RU2811341C2 - Device, systems and methods of particle counting - Google Patents

Abstract

FIELD: measurement technology.

SUBSTANCE: invention concerns a method for counting particles passing through a channel. When implementing the method, a light plane is formed across the channel with a matrix of LED emitters. As the particles pass through the light plane, the original output values generated in each light channel are recorded and the output values are stored in data cells. The original values are normalized so that each cell has a normalized output signal (NOS) value. The NOS values are analysed to identify data cells that are clustered together by at least one NOS value, with each cluster defining a separate event corresponding to at least one particle. For each detected event, the NOS values are summed up and the group velocity and volume of the event are determined. The number of particles of each identified first plane event is determined by classifying the volume of the event.

EFFECT: increase of accuracy of particle counting.

15 cl, 21 dwg

Description

State of the art

When operating air seeders, especially when applying high rate products such as small seeds (e.g. canola, flax, millet, oats, wheat, rye, barley, etc.), the product typically flows through the distribution lines to each row unit with a virtually continuous flow. It is therefore understood that to ensure accurate seed counting, a sensor system is required that is capable of detecting and counting the number of seeds passing through the distribution lines at one time, while distinguishing the seeds from debris that may pass through the distribution lines along with the seeds.

In fig. 1A is a perspective view of an embodiment of a single array particle counter assembly located in a channel.

In fig. 1B is a perspective view of an embodiment of a dual array particle counter assembly located in a channel.

In fig. 2A is a cross-sectional view of a single-die particle counter assembly as viewed along lines 2A-2A of FIG. 1A.

In fig. 2B is a cross-sectional view of a dual matrix particle counter assembly as viewed along lines 2B-2B of FIG. 1B.

In fig. 3A is a view of the raw voltage signals from the seeds shown in FIG. 2A or 2B passing through the first light plane of the particle counter assembly.

In fig. 3B shows the output scan of the Normalized Output Signal (NOS) after inverting and normalizing the original voltage signals shown in FIG. 3 A.

In fig. 4 is a flowchart of the process for identifying the start and end states of the NOS scan shown in FIG. 3B, to identify and mark Major Events.

In fig. 5 is a logic flow diagram of the process of identifying valley cells in the NOS scan shown in FIG. 3B.

In fig. 6 – logical flow diagram of the process of separating Main Events and marking the separated events.

In fig. 7 is a representation of the NOS scan shown in FIG. 3B showing the Major Event markings.

In fig. 8A is a representation of splitting the Main Event and marking the split Events after identifying the first trough column.

In fig. 8B is a representation of further division of the Main Event and further marking of the divided Events after identifying the second trough column.

In fig. 8C is a representation of further division of the Main Event and further marking of the divided Events after identifying the second trough column.

In fig. 9 – logical flow diagram of the process of classification of Major Events.

In fig. 10A is a flow chart summarizing the process of analyzing a Group obtained from a light plane of a single matrix to arrive at a final seed count and seed rate for output and display.

In fig. 10B is a flow diagram summarizing the process of comparing Groups obtained from each of the light planes of the two-matrix embodiment to arrive at a final seed count and seed rate for output and display.

In fig. 11 – side view of a variant of a conventional pneumatic seeder.

In fig. 12 is a schematic representation of the pneumatic seeder shown in FIG. 11, with particle counter assemblies located on the air seeder distribution lines.

In fig. 13 is a side view of the seeding section of a precision seeder with a particle counter assembly located on the distribution line of the seeding section.

In fig. 14 is a schematic diagram of an embodiment of the control system showing communication with the particle counter assembly and other components that may be used with the air seeder shown in FIG. 11, or the precision seeder shown in FIG. 13.

In fig. 15 – an embodiment of the process of setting up a control system, controlling a pneumatic seeder or precision seeder, as well as storing and displaying operating data.

Description of the invention

Referring to the drawings, in which like reference numerals denote the same or corresponding components in multiple views, FIG. 1A is an enlarged perspective view of one embodiment of a single-die particle counter assembly 200A located in a portion of a pipe or channel 202 through which particles pass. In fig. 2A is a cross-sectional view of particle counter 200A and channel 202 as viewed along lines 2A-2A shown in FIG. 1A.

In fig. 1B shows an enlarged perspective view of one embodiment of a dual-die particle counter assembly 200B located in the portion of pipe or channel 202 through which particles pass. In fig. 2B is a cross-sectional view of particle counter 200B and channel 202 as viewed along lines 2B-2B shown in FIG. 1B.

Throughout this description, reference numeral 200 is used to generally refer to particle counter assembly 200A or 200B, unless the specific description applies to only one of the embodiments, in which case reference numeral 200A will be used when considering only the single array embodiment, and Reference numeral 200B will be used when considering only the dual matrix embodiment.

It should be understood that particles may be carried by air flow through the channel 202 or may fall through the channel 202 under the influence of gravity. For example, as shown in FIG. 11 and 12, the particle counter assembly 200 may be located on the distribution line 58 of the air seeder 10 (discussed below) or, as shown in FIG. 13, on the seed tube 332 of the seed unit of the precision planter 300 (discussed below). In such applications, particle counter assembly 200 is used to detect and count passing seeds. In some air seeders, seeds and fertilizers (or other agricultural materials) pass through distribution lines 58 together, and the particle counter assembly 200 can distinguish seeds from other agricultural materials so that only the seeds are counted. The particle counter assembly can also distinguish between seeds and dirt or other debris passing through the distribution lines 58 of the air seeder 10 or the seed line 332 of the precision planters 300 so that only the seeds are counted. Additionally, in such applications, particle counter assembly 200 may be used to classify particles to distinguish between “good seeds” and seed fragments, thereby counting only “good seeds” that are expected to germinate. Although pneumatic seeders and precision planters are two typical applications for particle counter assembly 200, it should be understood that particle counter assembly 200 may be used in any other agricultural, industrial, or commercial application where it is desired to count particles passing through a pipe or channel to detect and counting only certain types of particles passing through a pipe or channel.

In the embodiment of single matrix particle counter assembly 200A shown in FIG. 1A and 2A, one light plane 210-1 is created across the channel 202. A single light plane 210-1 is created by an array of LED emitters 212-1 mounted within an emitter housing 214-1 located on one side of the channel 202. On the opposite side of the channel 202 and directly opposite the array of LED emitters 212-1 is a corresponding receiving array 216-1 , installed inside the receiver housing 218-1.

In the embodiment of dual matrix particle counter assembly 200B shown in FIG. 1B and 2B, two separate light planes 210-1, 210-2 are formed across the channel 202, the light planes 210-1, 210-2 being offset from each other in the longitudinal direction of the channel 202 by a distance "D" such that the particle is consistently pass through the first light plane 210-1 and then through the second light plane 210-2. Each light plane 210-1, 210-2 is generated by a corresponding matrix of LED emitters 212-1, 212-2, each matrix mounted in a corresponding emitter housing 214-1, 214-2 located on two sides of the channel 202 oriented at an angle of 90 degrees relative to each other about the longitudinal axis 204 of the channel 202 to detect the passage of particles in three dimensions. On opposite sides of channel 202 and directly opposite each of the respective LED arrays 212-1, 212-2 is a corresponding receiver array 216-1, 216-2, each receiver array 216-1, 216-2 mounted within a respective housing 218-1 , 218-2 receivers.

In any of the array embodiments 200A, 200B, the LED arrays 212-1, 212-2, respectively, comprise a number of closely spaced LED emitters 220, each of which produces a light beam or light channel 222 across the width or diameter of the channel 202. Corresponding receiving arrays 216-1, 216-2, as appropriate, comprise a corresponding number of closely spaced photodiode receivers 224 that receive light beams or light channels 222 from an opposing LED emitter 220. In one embodiment, the photodiode receivers 224 and LED emitters 220 are directly aligned with each other. each other such that a line between them is perpendicular to each of the photodiode receivers 224 and LED emitters 220. In other embodiments, the lines are not perpendicular.

In both embodiments 200A, 200B, the first LED array 212-1 is shown as having 12 LED emitters 220 producing light beams or light channels 222 (labeled C1-Cl2) extending across the width or diameter of the channel 202. In the dual array embodiment 200B , the second LED array 212-2 is also shown as having 12 LED emitters producing light beams or light channels 222 (labeled C13-C24) extending across the width of the channel 202 perpendicular to the light channels Cl-C12 of the first LED array 212-1. Each respective receiver array 216-1, 216-2 is, as appropriate, shown as having 12 corresponding photodiodes 224. It should be understood that more or fewer LED emitters and photodiode receivers can be used, spaced at larger or smaller intervals. It should also be understood that LEDs with wider or narrower beam angles and different light intensities may be used depending on the specific application, including depending on the width or diameter of the channel 202 and the size of particles passing through the channel that must be detected, and from various other factors obvious to those skilled in the art. By way of example only, in applications for detecting seeds passing through the distribution lines 58 of the planter 10, suitable LED emitters may be the SM1206NHC-IL LED emitters available from Bivar, Inc., having a beam angle of 30 degrees, spaced 0.08 inches apart. .2 cm), and suitable photodiodes 224 may be TEMD7000X01 photodiodes from Vishay Intertechnology, Inc., spaced 0.08 inches (0.2 cm). In the dual array embodiment 200B, a suitable longitudinal offset distance "D" between the light planes 210-1, 210-2 may be 0.125 inches (0.32 cm), 0.05 to 1 inch (0.13 to 2.5 cm) or 0.1 to 0.5 inches (0.2 cm to 1.3 cm).

In fig. 2A and 2B show a snapshot of a group of seeds S1-S5 passing through a channel 202, such as the distribution line 58 of the air seeder 10. It should be understood that when using an air seeder, the seeds are carried by the air flow along the distribution lines 58 at a relatively high speed and, essentially, , continuous flow (especially when planting small seeds with high rates of application, such as canola, flax, millet, oats, wheat, rye, barley, etc.), so the snapshot of a group of seeds in FIG. 2A or 2B is provided for ease of explanation of the signal scan shown in FIG. 3A generated by this group of seeds passing through a single light plane 210-1 or two light planes 210-1, 210-2, and other processes performed to analyze the signals for the purpose of counting and classifying seeds, as discussed below. Although the following description refers to seeds for the purpose of explaining the process that occurs when using an air seeder or precision planter, it should be understood that the term “seed” is used as an example only, and therefore the term “seed” should be understood interchangeably with the term particle.

Operation

As is known in the art, a photodiode uses light energy to generate an electric current, wherein the voltage increases in proportion to the intensity of the light. Thus, it should be understood that when a particle passes through the light channel 222 generated by the LED emitter 220, the associated photodiode 224 is momentarily “blacked out,” causing the signal output voltage to drop. After the particle passes through the light channel 222, so that the light channel is completely unblocked, the output signal of the photodiode 224 will return to its maximum intensity.

The output signals of the photodiodes 224 of one light plane 210-1 or two light planes 210-1, 210-2 are received by the controller 110 and stored in memory for further processing. The controller 110 may be part of a seed control system 100 (discussed below). The controller 110 may be located in the cab of a tractor pulling an air seeder 10, or a precision planter 300, as described below.

In fig. 3A shows a snapshot of the “raw” voltage signals (V _raw ) generated by the photodiodes 224 and detected by the controller 110 as a group of seeds S1-S5 passes through the first light plane 210-1 (ie, through channels C1-C12). The controller 110 registers the outputs of each of the photodiodes 224 associated with each light channel 222 into successive data cells over a series of data scans, with each of the successive data cells for each of the light channels 222 arranged in parallel (i.e., in rows or columns) . In the example shown in FIG. 3A, a snapshot contains a series of 50 acquisition cycles or scans, with each row corresponding to one acquisition cycle or one scan for each of the light channels 222 (C1-C12) organized into columns. It should be understood that the output signals may alternatively be organized such that the data scans are organized into sequential columns and the light channels 222 are organized into rows. By way of example only, data may be collected or scanned by controller 110 at a frequency of 10 kHz, but may be collected at higher or lower frequencies depending on the size and speed of particles to be detected and counted. Although not shown, it should be understood that for the dual matrix embodiment 200B, a similar set of raw voltage data will be generated and collected by the controller 110 as the seeds S1-S5 pass through the second light plane 210-2 (i.e., through the channels C13-C24) oriented at 90 degrees to the first light plane 210-1.

Data normalization process

Once the V _raw signals are collected and stored in the memory of the controller 110, the controller 110 runs a program 101 to invert and “normalize” the V _raw signals to values between 0 and 1 (no units). In fig. 3B shows the same snapshot as in FIG. 3A, after the V _raw signals have been inverted and normalized using the process described below, and are hereinafter referred to as "normalized output signal" or "NOS". The V _raw signals are inverted and normalized such that when light channel 222 has maximum light intensity (i.e. not blocked by passing seed), the signal value is 0. When light channel 222 has lowest light intensity (i.e. completely blocked by passing seed) seed), the signal value is 1. I.e. intermediate signal values from 0 to 1 indicate that light channel 222 is partially blocked by passing seed.

To invert and normalize the voltage of the original signal (V _raw ) shown in FIG. 3A for channels C1-C12 to obtain the NOS data shown in FIG. 3B for channels C1–C12, the following equation is used:

NOS=1 – (V _raw – V _b )/(V _u – V _b ),

where NOS is the normalized output signal,

V _raw – initial detected voltage of light channel 222,

V _b – average voltage when light channel 222 is blocked,

V _u is the average voltage when light channel 222 is unlocked.

The value of V _b may be set as a constant value (for example, 0.025 volts) or V _b may be the average output voltage of each photodiode 224 when the corresponding LED emitter 216 is turned off.

The V _u value may be a serial buffer operating on a first-in-first-out (FIFO) basis for 100 scans of the V _raw data of each photodiode 224 when there is no particle passing through the corresponding light channel 222 (e.g. C1-C12), and when the voltage V _raw has a change of less than ±0.01 V from the previous scan to the current scan.

It should be appreciated that for dual matrix embodiment 200B, the NOS data generation process for channels C13-C24 is the same as for channels C1-C12.

Referring to FIG. 2, the seeds are numbered S1-S5 from the very first (lowest) to the last (topmost) seed. When the very first single seed S1 passes through the first light plane 210-1, NOS values are generated as shown in FIG. 3B, as a small cluster of merged cells having values from 0.1 to 1 in a series of columns (channels C2-C4 in this example) and in a series of scan rows (rows 3-17 in this example). It should be noted that the very first scan (i.e. line 1 of the scan) is at the bottom of Fig. 3A and 3B. Still referring to FIG. 2, when a group of seeds S2-S5 (which are in close proximity to each other, with seeds S2, S3, seeds S2, S4 and seeds S4, 25 overlapping or touching each other) passes through the first light plane 210-1, NOS values are generated as shown in FIG. 3B, as a large cluster of merged cells having values from 0.2 to 1 in a series of columns (channels C4-C10 in this example) and in a series of scan rows (rows 15-48 in this example). Each cluster of NOS values from 0.2 to 1 is called a "Main Event". Thus, it is understood that the Main Event may be one discrete seed (such as seed S1) or the Main Event may be multiple seeds (such as seeds S2-S5) passing through the light planes 210-1 in close proximity to each other. to a friend. For the dual matrix embodiment 200B, a similar set of main events will be generated for seeds sequentially passing through channels C13-C24 of the second light plane 210-2 (not shown).

Identification of Major Events

After the NOS data is generated, the controller 110 uses the NOS data analysis program 101 to identify clusters of cells and associate each cluster of cells with a unique identifier for further processing discussed below. To distinguish between clusters of cells and associate each cluster of cells as a separate Major Event, NOS data is analyzed to determine when the first cell appears in each channel C1-C12 having a NOS value greater than a predefined minimum value (e.g., 0.15), which indicates that the seed begins to pass through this channel (that is, the "initial state" occurs), and when each channel C1-C12 remains the last NOS cell having a value greater than a predetermined minimum value (for example, 0.15) , which indicates that the seed has passed through the channel (i.e. there is a "final state"). All directly connected cells in all columns between the initial and final states constitute a single cluster of cells and define one Main Event. It will be appreciated that for dual matrix embodiment 200B, the process of distinguishing cell clusters and associating each cell cluster into a separate Main Event for channels C13-C24 is the same as described above for channels C1-C12.

Start/End State Identification Process

To identify the initial state for each channel column and the final state for each channel column, the program 101 of the controller 110 performs the initial/final state identification process 400 illustrated in FIG. 4. For the purposes of this document, only the NOS values of channels C1-C12 are considered. For the 200B dual matrix option, the same process is performed for channels C13-C24.

Referring to FIG. 4, at an initial step 402, the NOS data cell of a particular channel is analyzed, i.e. The value of the current NOS cell (NOS-c) is read and stored in memory and is initially assumed to be in the final state until otherwise specified as described herein. At step 404, the value of the NOS cell (NOS-p) of the same channel immediately preceding the NOS-c cell is read and stored in memory. At step 406, the value of the NOS cell (NOS-f) of the same channel following the NOS-c cell is read and stored in memory. At step 408, the change in signal relative to the previous cell (Look Behind Delta, Delta-lb) of this channel is calculated by subtracting the value of the NOS-p cell from the value of the NOS-c cell. The calculated Delta-lb change is stored in memory.

At step 410, the change in signal relative to the next cell (Look Ahead Delta, Delta-la) of that channel is calculated by subtracting the value of cell NOS-c from the value of cell NOS-f. The Delta-la value is stored in memory. At step 412, a memory search is performed to identify the corresponding state (ie, start or end state) for the NOS-p cell from the previous cycle. If the NOS-p cell is in the final state (i.e., the answer to step 412 is Yes), the program proceeds to step 414.

At step 414, if the Delta-lb value is greater than 0.1 (ie, the condition is met), the NOS-c cell is associated at step 420 as being in the initial state, and the associated state is written to memory. The program then proceeds to step 430 to cycle to the next NOS cell in the next column in the same scan line. It should be understood that when the program moves to the next NOS cell in the next column of the same row, the cell that was the NOS-f cell in the previous cycle now becomes the NOS-c cell in the new cycle. Similarly, a cell that was a NOS-c cell in the previous cycle now becomes a NOS-p cell in the new cycle. Returning to step 414, if the condition at step 414 is not met, the NOS-c signal is further analyzed at step 418.

If at step 418 the NOS-c value is greater than 0.15 V (ie, the condition is met), then at step 420 the NOS-c cell is associated as being in the initial state and the associated state is written to memory. The program then proceeds to step 430 to move to the next NOS cell of the next column in the same scan line. However, if the condition of step 418 is not met, the program proceeds to step 430 to cycle to the next NOS cell of the next column in the same scan line.

Returning to step 412, if the NOS-p cell is not in the final state (i.e., the answer to step 412 is No), the program proceeds to step 422. At step 422, if the NOS-p cell is in the initial state (i.e., i.e. the answer to step 422 is Yes), the program proceeds to step 424. If at step 424 the value of Delta-lb is less than -0.2 and the value of Delta-la is less than -0.012 (i.e., the conditions are met), then at step 428, the NOS-c cell is associated as being in a final state, and its associated state is written to memory. The program then proceeds to step 430 to move to the next NOS cell of the next column in the same scan line. If, however, the conditions of step 424 are not met, the NOS-c cell is further analyzed at step 426.

If, at step 426, the value of the NOS-c cell is less than 0.15 V (ie, the condition is met), then at step 428, the NOS-c cell is associated as being in a final state, and its associated state is written to memory. The program then proceeds to step 430 to move to the next NOS cell of the next column in the same scan line. If, however, the condition of step 426 is not met, the program proceeds to step 430 to move to the next NOS cell of the next column in the same scan line.

The process is repeated for each of the cells in the row across all column channels (C1-C12). When a scan line is completed, the program moves to the next scan line and repeats the process for each of the cells in that row through all columns (C1-C12), etc., thereby identifying each initial state and each final state.

Once an initial state cell and a final state cell are identified for each channel of a column, all cells in that column from the initial state cell to the final state cell are associated into a column cluster. Additionally, all column clusters that are directly connected by at least one cell are then associated as part of the Main Event. All cells that make up this Main Event are then given a unique Main Event label, generated by the program and stored in memory for later use, discussed below. For example, the unique label of a Main Event may be an integer, and each subsequent Main Event may be assigned the next sequential integer such that each Main Event has a unique identifier in increments of one integer.

For example, as shown in FIG. 3B, the cluster of immediately adjacent cells after the very first starting cell (ie, the lowest starting cell) defines the first Major Event in the scan. As shown in FIG. 7, all contiguous cells composing this Main Event are labeled with the integer 999. The next cluster of immediately adjacent cells after the next very first starting cell defines the next Main Event, and each of the contiguous cells composing this Main Event is labeled in FIG. 7 by the next consecutive integer 1000.

For the dual array embodiment 200B, the same process is repeated for the NOS data scans associated with channels C13-C24 of the second light plane 210-2. As discussed below, it should be understood that the second light plane 210-2 (which is oriented at 90 degrees to the first light plane 210-1) may produce a NOS data scan that has differences in identified seed clusters due to differences in proximity and overlap seeds in a two-dimensional scan of the second light plane 210-2. Thus, these differences in seed clusters may result in differences in the numbering of Major Events due to differences between the start and end states identified in the start/end identification process 400 of seed clusters as viewed in the orientation of the second light plane 210-2. These differences are taken into account in the 900A, 900B Group Comparison process discussed below.

Main Event Analysis

To determine the exact number of seeds, the NOS data scan is further analyzed so that through the Major Event analysis process 500 shown in FIG. 5 determine whether the Main Event corresponds to a single seed or whether the Main Event includes one or more seeds. The Main Event analysis process 500 includes two sub-processes: (A) identifying “valley columns” to determine whether the Main Event needs to be further divided or broken down into individual Events; and (B) marking each individual Event with a unique identifier for later use. The Major Event analysis process 500 (and each of its subprocesses) is performed by the controller 110 of the monitoring system 100 (discussed below) through the software program 101. A description of the Main Event analysis process 500 (and each of its subprocesses) is provided for the Cl-C12 channel scan only. However, it should be noted that for dual matrix embodiment 200B, a similar process 500 is performed for channels C13-C24.

A. Trough Column Identification Process

A "valley column" is a column in which the NOS cell value is less than the values of the NOS cells immediately to the left and right of the current NOS cell being analyzed. In general, if a cell with a lower NOS value is detected between two high NOS cells in adjacent columns, this suggests that two seeds are passing through the light plane 210-1 side by side in close proximity to each other or partially overlapping each other. . For example, if the voltage of the cluster signal generated by the passage of two seeds is considered as mountains or plateaus, the valley columns represent the valleys between the mountains or plateaus that demarcate the two seeds. The trough column includes only those cells in the NOS data scan that connect the cluster of cells that make up the Main Event to the left and right of the currently analyzed NOS cell (NOS-c). Referring to the NOS data scan shown in FIG. 3B, in Main Event 999 there are no trough columns, but in Main Event 1000 there are three trough columns 1000A, 1000B, 1000C. Due to the presence of trough columns in the Main Event 1000, it needs to be divided into separate Events, as explained below in the section “Main Event Dividing and Labeling Process 600”, but first the process of identifying the trough columns is explained.

Referring to FIG. 3B and the logic block diagram of FIG. 5, valley columns are identified by analyzing each cell in each row of the channel column scan (C1-C12) before moving to the next row of the scan. At an initial step 502, it is first determined whether the NOS cell being analyzed (i.e., the current NOS cell (NOS-c)), the NOS cell to its left (NOS-1), and the NOS cell to its right (NOS-r) are associated with the Main Event (as described above). If the NOS-c cell was not previously associated with the Main Event, the program proceeds to step 530 to analyze the next NOS cell in the next column in the same scan line. If the answer at step 502 is Yes, the program proceeds to step 504 and analyzes the NOS data scans to identify the first starting cell after the ending cell in that column to determine whether there are any NOS cells after the identified starting cell in that column up to NOS-c cells labeled as trough cell (as explained below). If the answer to step 504 is No, the program proceeds to step 530. If the answer to step 504 is Yes, the program proceeds to step 506.

For example, as shown in FIG. 3B, assume that the cell in scan row 23 in channel or column 5 (hereinafter referred to as r26:c5) is the current cell being analyzed (that is, the NOS-c cell). At step 502, cell NOS-c r26:c5 was previously associated with the Main Event labeled 1000, and the cell to its right (ie, cell NOS-r r26:c6) was also previously associated with the Main Event labeled 1000. However, the cell on the left (that is, cell NOS-1 r26:c4) is not associated with any Primary Event, so the answer at step 502 is No, and the program proceeds to step 530 and moves to the next cell in the same row in the next column. It should be understood that if the cells to the right or left of a NOS-c cell are not associated with the same Primary Event, the NOS-c cell cannot be a trough cell. Moving on to cell r26:c7 as a NOS-c cell and repeating step 502, cell r26:c7 was previously associated with the Main Event labeled 1000, and the cell to its right (i.e., NOS-r cell r26:c8) is also was previously associated with the Main Event labeled 1000. Likewise, the cell to its left (that is, cell r26:c6) was also associated with the Main Event 1000, so the answer at step 502 is Yes, and the program proceeds to step 504. At step 504 the program looks through previous scans in the same column (i.e. column 7) up to the first starting cell in that column (which in this case is NOS-c cell r26:c7) and determines that there are no cells there before r26:c7 , which were previously labeled as valley cells because the earliest valley cell defines the valley column. Since the answer at step 504 regarding cell r26:c7 is Yes, the program proceeds to step 506.

At step 506, the NOS data scan is analyzed to determine whether a NOS-c cell is in a column immediately adjacent to a NOS cell that was previously labeled as a valley cell (explained below). If the answer to step 506 is Yes, the program proceeds to step 530. If the answer to step 506 is No, the program proceeds to step 510. Referring to the previous example regarding cell r26:c7, the program confirms that the cell at r26:c7 does not border with any cells previously labeled as valley cells, since two valley cells or two valley columns cannot be located side by side with each other. Since the answer at step 506 regarding cell r26:c7 is Yes, the program proceeds to step 510.

At step 510, the program reads and stores the value of cell NOS-c. At step 512, the program reads and stores the value of the NOS cell to its left (NOS-1). At step 514, the program reads and stores the value of the NOS cell to its right (NOS-r). At step 516, the program calculates a "Delta-L value" which is the amount of difference between the value of the NOS-1 cell and the NOS-c cell. At step 518, the program calculates a "Delta-R value" which is the difference between the value of cell NOS-r and the value of cell NOS-c. Referring to the previous example regarding cell NOS-c r26:c7, its value in FIG. 3B is equal to 0.2; the NOS-1 value of cell r26:c6 is 0.8; the NOS-r value of cell r26:c8 is 0.2 and therefore the Delta-L value is 0.6 (i.e. 0.8-02) and the Delta-R value is 0 (i.e. 0 ,2-0.2).

If, at step 520, the Delta-L value and the Delta-R value are greater than or equal to 0.03 (ie, the condition is met), then at step 522, the program marks the NOS-c cell as a “valley cell” and stores it in memory. The program then proceeds to step 530 to cycle to the next NOS cell in the next column in the same scan line. If the Delta-L or Delta-R value is less than 0.03 (i.e., the condition of step 520 is not met), the program proceeds to step 524. Referring to the previous example regarding the cell at r26:c7, although the Delta-L value calculated above is 0 ,6 is greater than 0.03, the Delta-R 0 value is less than 0.03, so the program proceeds to step 524.

At step 524, the program locates the value of the NOS cell two columns to the right (NOS-2r) of the NOS-c cell and determines a "Delta-2R value" which is the amount of the difference between the value of the NOS-2r cell and the value of the NOS-r cell. If the Delta-2R value is greater than or equal to 0.03 and the Delta-L value is greater than or equal to 0.03 (i.e., both conditions are met), the program proceeds to step 526. However, if any condition of step 524 is not met, the program proceeds to step 530. Referring to the previous example regarding the cell at r26:c7, the value of cell NOS-2r (ie, cell r26:c9) is 1.0. Thus, since the Delta-L value of 0.6 is greater than or equal to 0.03, and the Delta-2R value of 0.8 (i.e. 1.0-0.2) is greater than or equal to 0.03, the conditions are met, and the program proceeds to step 526.

At step 526, the program determines whether the condition that the value of cell NOS-c is less than or equal to the value of cell NOS-r is met. If so, then at step 522, the NOS-c cell is marked and stored in memory as a valley cell. Referring to the previous example regarding the cell at r26:c7, since the value of cell r26:c7 is equal to the value of cell r26:c8, the condition is satisfied and therefore cell r26:c7 is marked as a trough cell. The program then moves to cell r23:cl0. If the answer at step 526 is No, then at step 528 the NOS-r cell is marked and stored in memory as a valley cell.

The program then proceeds to step 530 of cycling to the next column to the right in the same scan line. After moving to the next column, you should understand that the cell that was a NOS-r cell in the previous cycle will become a NOS-c cell in the new cycle, and the cell that was a NOS-c cell in the previous cycle will become a NOS- cell in the new cycle. 1.

The above process is repeated for each of the cells in the row across all channel columns (C1-C12). When a scan line is completed, the program moves to the next scan line and repeats the process for each of the cells in that line across all columns (C1-C12), and so on, thereby identifying all the cells in the valley. All cells immediately adjacent to valley cells make up a valley column.

For dual array embodiment 200B, the same process is repeated for NOS data scans with respect to channels C13-C24 of the second light plane 210-2. As discussed previously, it should be understood that the second light plane 210-2 (which is oriented at 90 degrees to the first light plane 210-1) may produce a NOS data scan that has differences in identified seed clusters due to differences in proximity and overlap seeds in a two-dimensional scan of the second light plane 210-2. Thus, these differences in seed clusters may result in differences in the numbering of Major Events due to differences between the start and end states identified during the seed cluster start/end identification process 400 when viewed in the orientation of the second light plane 210-2 . In addition, due to the 90 degree difference in the orientation of the first and second light planes 210-1, 210-2, different notch columns may be associated with respective Key Events. For example, as shown in FIG. 2B, the second light plane 210-2 may have a gap or notch column between seeds not detected in the two-dimensional "view" of the first light plane 210-1 during the notch column identification process 500 for channels C13-C24.

B. The Process of Dividing and Labeling the Main Event

In fig. 7 shows the original Main Event labels 999 and 1000 assigned by the Main Event identification process 400, with columns 1000A, 1000B, 1000C identified in accordance with the valley column identification process 400 described above. In fig. 8A shows the Main Event 1000 of FIG. 7, initially divided into two separate Events labeled 1000 and 1001 in accordance with the Main Event splitting and labeling process 600 (discussed below). In fig. 8B shows the Main Event 1000 further divided into three Events 1000, 1001, 1002 in accordance with the Main Event splitting and marking process 600 (discussed below). And in fig. 8C shows the Main Event 1000 further divided into four Events 1000, 1001, 1002, 1003 in accordance with the Main Event division and marking process 600 (discussed below).

When the valley columns (if any) in the Main Event are identified, the program performs the Main Event splitting process 600 according to the logic flow diagram shown in FIG. 6. If there are no valley columns in the Main Event, the Main Event splitting and labeling process 600 is skipped for that Main Event and the program proceeds to defining the Event metrics (discussed below).

Referring to FIG. 6, at an initial step 602, the NOS data scans are analyzed to identify the earliest Major Event having a valley column, which according to FIG. 7 is a Main Event labeled 1000. At step 604, the program identifies the earliest (ie, lowest) trough cell in this Main Event, which in FIG. 8A is valley cell 1000A in scan row 26 in column C7 (i.e. r26:c7).

At step 606, the program splits the Main Event into two separate Events (ie, a left Event and a right Event). All cells in the Main Event to the right of the trough column, along with all cells in the lower half of the trough column, are associated with the right Event. All cells to the left of the trough column and above the trough column, along with the cells in the top half of the trough column, are associated with the left Event.

At step 608, all cells associated with the right Event are assigned the last used label. At step 610, all cells associated with the left Event are assigned a new label, incremented to the next sequential integer. In the example shown in FIG. 8A, it can be seen that all cells to the right of the first valley column 1000A, and the cells making up the lower half of the valley column 1000A, retain their original label (ie, label 1000). All cells to the left of the valley column 1000A and above the valley column 1000A, as well as the top half of the valley column 1000A, are labeled with the next consecutive integer (that is, 1001).

The program then proceeds to step 612 to determine whether the Main Event includes other trough columns. If there is no other valley column in the Main Event (ie, the answer at step 612 is No), the process of splitting and marking the Main Event ends and the program proceeds to the process of determining Event metrics discussed below. If there are other bottom columns in the Main Event (ie, the answer at step 612 is Yes), the program proceeds to step 616 and moves to the next earliest bottom column (ie, the bottom column with the next lowest bottom cell). Referring to FIG. 8A, the next earliest valley column is valley column 1000B.

After moving to the next valley column, at step 618 the program determines whether the label associated with the cell to the right of the earliest valley cell is different from the label of the earliest valley cell. If the answer at step 618 is Yes, the process of steps 606-612 described above is repeated. If the answer at step 618 is No, the program proceeds to step 620. In the example shown in FIG. 8A, the cells to the right of the earliest valley cell (ie, cell number r31:c9) in column 1000B have the same label (ie, 1000) as the earliest valley cell. Accordingly, the answer at step 618 is No, and so the program proceeds to step 620.

At step 620, the program again splits the Main Event into two separate Events (ie, a left event and a right event). All cells in the Main Event to the left of the trough column, below the trough column, and in the lower half of the trough column are associated with the left Event. All cells in the Main Event to the right of the trough column, above the trough column, and in the top half of the trough column are associated with the right Event.

At step 622, all cells associated with the left Event are assigned the last used label. At step 624, all cells associated with the right Event are assigned a new label, incremented to the next sequential integer. Referring now to FIG. 8B, it can be seen that all cells to the left of valley column 1000B, and the cells making up the bottom half of valley column 1000B, retain their original labels (ie, labels 1000 and 1001). All cells to the right of valley column 1000B and above valley column 1000B, as well as the top half of valley column 1000B, are labeled with the next consecutive integer (ie, 1002).

The program then proceeds to step 612 to determine whether the Main Event contains other trough columns. If there are no other valley columns in the Main Event (ie, the answer at step 612 is No), the process of splitting and marking the Main Event ends and the program proceeds to the process of determining Event metrics, which is discussed below. If there are other bottom columns in the Main Event (ie, the answer at step 612 is Yes), the program proceeds to step 616 and moves to the next earliest bottom column (ie, the bottom column with the next lowest bottom cell). Referring to FIG. 8B, the next earliest valley column is valley column 1000C.

After moving to the next valley column, the program repeats step 618 and the corresponding steps thereafter. In the example shown in FIG. 8B, the cells to the right of the earliest valley cell (ie, cell number r41:c7) in valley column 1000C have a label different from the label of the earliest valley cell (ie, 1000). Accordingly, the answer at step 618 is Yes, so the program proceeds to step 606.

At step 606, the program splits the Main Event into two separate Events (ie, a left Event and a right Event). All cells in the Main Event to the right of the trough column, along with all cells in the lower half of the trough column, are associated with the right Event. All cells to the left of the trough column and above the trough column, along with cells in the top half of the trough column, are associated with the left Event.

At step 608, all cells associated with the right Event are assigned the last used label. At step 610, all cells associated with the left Event are assigned a new label, incremented to the next sequential integer. Referring now to FIG. 8C, it can be seen that all the cells to the right of the valley column 1000C and the cells making up the bottom half of the valley column 1000C are assigned the last used label (ie, 1002). All cells to the left of the 1000C valley column and above the 1000C valley column, as well as the top half of the 1000C valley column, are labeled as the next sequential integer (i.e., 1003).

The program then proceeds to step 612 to determine whether the Main Event contains other trough columns. If there are no other valley columns in the Main Event (ie, the answer at step 612 is No), the process of splitting and marking the Main Event ends and the program proceeds to the process of determining Event metrics discussed below. If there are other bottom columns in the Main Event (ie, the answer at step 612 is Yes), the program proceeds to step 616 and moves to the next earliest bottom column (ie, the bottom column with the next lowest bottom cell). In fig. 8C there are no other valley columns, so the process of separating and labeling the Main Event is completed, and the program proceeds to the process of determining the Event metrics described below.

For dual array embodiment 200B, the same process is repeated for NOS data scans with respect to channels C13-C24 of the second light plane 210-2. As discussed previously, it should be understood that the second light plane 210-2 (which is oriented at 90 degrees relative to the first light plane 210-1) may produce a NOS data scan that has differences in identified seed clusters due to differences in proximity and overlap seeds during two-dimensional scanning of the second light plane 210-2. Thus, these differences in seed clusters may not only result in differences in the numbering of Major Events due to differences between the start and end states identified during the start/end identification process 400 of seed clusters when “looked at” in the orientation of the second light plane 210 -2, but the 90-degree different orientation of the first and second light planes 210-1, 210-2 may cause the valley column identification process 500 for channels C13-C24 to identify different valley columns for the respective Main Events. As a result of the differences in the valley columns, differences in the division and labeling of Events may occur during the Main Event division and Event labeling process 600. These differences are taken into account in the Group Comparison Process 900A, 900B, which is discussed below.

Defining Event Metrics

After the Main Event separation and labeling process 600 is completed, the program determines several metrics for each Event that are used during the Group Comparison Process 900A, 900B discussed below, including Event Duration, Event Width, Event Breaks, Event Stress Amount, and Event Length.

Event Duration is the total duration of the Main Event from the scanning time of the last line of the Main Event minus the scanning time of the first line of the Main Event. Referring to FIG. 8C, the Event Duration for Main Event 999 is indicated by reference numeral 702, and the Event Duration for Main Event 1000 is indicated by reference numeral 704.

The Event Width is the total width of the Main Event bounded by the channels 222 (C1-C12) having a value in the NOS cells multiplied by the distance between the channels 222. Referring to FIG. 8C, the Event Width for Main Event 999 is indicated by reference numeral 706, and the Event Width for original Main Event 1000 (now divided into individual Events 1000, 1001, 1002, 1003) is indicated by reference numeral 708.

Event Breaks are the total number of times any channel in any Main Event transitions from the start state to the end state and back to the start state. Referring to FIG. 8C, there are no Event Breaks in Main Event 999. However, in the original Main Event 1000 (now divided into individual Events 1000, 1001, 1002, 1003), Event Breaks are indicated by reference numeral 710.

The Event Voltage Volume is the sum of the Main Event NOS cell values multiplied by the Group Velocity (discussed in more detail below).

Event Length is the true distance equal to the Event Duration multiplied by the Group Velocity (discussed in more detail below).

Grouping Major Events

Major Events are grouped for the purposes of: (a) comparing the number of seeds in a given Scan Window and (b) calculating Group Velocity. The Scan Window begins when the earliest Main Event begins, and the Scan Window is open until a specified number of scans without detecting seeds are detected (for example, no NOS value greater than 0.15 is detected) after the last final state for a unified cluster of NOS cells defining this Primary Event. For example, if the specified number of scans since the last end state of the earliest Main Event in which no seeds are detected is, say, 10 scans, the Scan Window will close after 10 scans in which no seeds are detected. In other words, using the proposed 10 scan gap, if the scan rate is 10 kHz, the Scan Window will close 1 ms after the last end state of the earliest Main Event that opens the Scan Window. All Major Events covered by this Scan Window are called a "Main Event Group" or "Group" for short.

The Group Speed ("Group Speed") is determined differently depending on whether the single die option 200A or the dual die option 200B is used. For the dual array embodiment 200B, since there are two light planes 210-1, 210-2, the Group Velocity can be accurately calculated by determining the change in time between the first light plane 210-1 and the second light plane 210-2 detecting the first Main Event. in the Group, divided by the distance between the light planes 210-1, 210-2 (eg, equal to 0.125 inches in the embodiment described above). The average seed throughput rate can be calculated using a FIFO buffer for a predetermined number of Groups (eg 3000).

For the single array embodiment 200A, since there is only one light plane 210-1, the Group Velocity is calculated using the characteristic Event Width divided by the characteristic Event Duration. As the name suggests, the characteristic Event Width is the most common Event Width for a Group during a predefined moving number of Major Events (i.e., the most common number of channels 222 with values in NOS cells occurring during a predefined number of Events multiplied by distance between channels 222). Likewise, the characteristic Event Duration, as the name suggests, is the most common Event Duration that occurs during a predetermined number of Major Events. The purpose of using a characteristic value is to obtain an average measurement of the width and duration of a single Main Event, under the assumption that the Main Event will most often consist of a single particle.

As stated above, Event Voltage Volume and Event Length are calculated using Group Velocity (calculated by any of the methods described above depending on whether single matrix option 200A or dual matrix option 200B is used). In the two-matrix variant 200B, which outputs the speed of each cluster of cells (consisting of a single seed or a group of seeds), an additional filter can be applied that makes an exception for Event Stress Volume and Event Length if the calculated Group Speed is greater than the average speed of 2 ,5 times. In such cases, the cluster of cells defining the Main Event is most likely garbage. For such cases, the average speed is used for all events other than the first Main Event in the Group.

Metric histograms

To classify the product flowing through the sensor assembly 200, the most frequently occurring characteristic values of Event Voltage Volume (EVV), Event Extent, and Event Width are determined and stored using histograms for one light plane 210-1 (for the 200A single light plane). matrix) or for each light plane 210-1, 210-2 (for option 200B with two matrices). This is done using a predefined FIFO buffer (eg 3000 Main Events).

The EVV Histogram has 30 bars or intervals that include only the lowest 90% of Major Events. The characteristic Voltage Volume (VVM) value is found by averaging the values of the 4 highest bars or intervals.

The Event Length Histogram has 20 bars or intervals that include only the bottom 90% of Major Events. The characteristic value of Extent is found by averaging the values of the 4 highest bars or intervals.

An Event Width histogram has a range of bars or bins from 0 to 13. The characteristic Width value is determined by averaging the values of the two highest bars or bins.

Classification of Major Events

In fig. 9 is a logic flow diagram of a Classification Process 800 for classifying a Main Event as a "five" (i.e., five seeds), a "four" (i.e., four seeds), a "three" (i.e., three seeds), "two" (i.e. two seeds), "one" (i.e. one seed) or garbage. According to the logic block diagram shown in FIG. 9, the program first checks whether the Main Event meets the classification criteria as a five, then as a four, then as a three, then as a two, then as a one, and then as garbage. Once a Major Event is classified as meeting the criteria of one of these gradations, the program exits the loop and moves on to the next Major Event. In the example calculations below to classify a Main Event as a five, four, three, two, one or trash, certain EVV values were empirically selected to distinguish the number of seeds in the Main Event. Other EVV values may be more accurate for estimating the number of seeds in a Main Event, and therefore the EVVs given in the example calculations should be considered non-limiting EVVs for purposes of categorizing a Main Event when estimating the actual number of seeds composing a Main Event.

At block 802, if the EVV is greater than 6.8 times the VVM such that the condition of block 802 is met, the program classifies the Major Event at block 804 as Five and then proceeds to block 840 to classify the next Major Event. If the condition of step 802 is not met, the program proceeds to step 806. At step 806, if EVV is more than 5 times VVM, i.e. the condition of step 806 is met, the program classifies the Main Event at step 808 as a Quad, and then proceeds to step 840 to classify the next Main Event. If the condition of step 806 is not met, the program proceeds to step 810.

At step 810, if EVV is more than 3.9 times VVM, i.e. the condition of step 810 is satisfied, the program classifies the Main Event at step 818 as Triple, and then the program proceeds to step 840 to classify the next Main Event. If the condition of step 810 is not met, the program proceeds to step 814 to check whether the Main Event includes a break (ie, a start-end-start transition). If the Major Event includes a gap, the program proceeds to step 816 to check if the EVV value is more than 2.2 times the VVM value, and if the conditions of both step 814 and step 816 are met, the program classifies the Major Event in step 818 as Triple and then proceeds to step 840 to classify the next Main Event. If the condition of either step 814 or step 816 is not met, the program proceeds to step 820.

At block 820, if the EVV is more than 2.9 times the VVM such that the condition of block 820 is met, the program classifies the Major Event at block 826 as a Double, and proceeds to block 840 to classify the next Major Event. If the condition of step 820 is not met, the program proceeds to step 822 to check whether the Main Event includes a break (start-end-start transition). If the Main Event includes a break, the program proceeds to step 824 to check if EVV is 1.2 times the value of VVM, and if the condition of both step 822 and step 824 is met, the program classifies the Main Event at step 826 as Double (two) and then proceeds to step 840 to classify the next Main Event. If the condition of either step 822 or step 824 is not met, the program proceeds to step 828.

If at step 828 EVV is greater than 0.85 VVM, i.e. the condition of step 828 is satisfied, the program classifies the Main Event at step 832 as a Single, and then proceeds to step 840 to classify the next Main Event. If the condition of step 828 is not met, the program proceeds to step 830 to check if the Event Length is greater than 0.6 times the characteristic Length value. If the condition of step 830 is met, the program classifies the Main Event at step 832 as a Single and then proceeds to step 840 to classify the next Main Event. If the condition of step 830 is not met, the program classifies the Main Event at step 834 as garbage, and then proceeds to step 840 to classify the next Main Event.

Group Comparison Process

Fig. 10A and 10B illustrate Group Comparison Process 900A, 900B, respectively. It should be understood that FIG. 10A illustrates the Group Compare Process 900A for the single array embodiment 200A as it illustrates the steps for only one light plane 210-1, whereas FIG. 10B illustrates the Group Comparison Process 900B for the two-array embodiment 200B as it illustrates the steps for both light planes 210-1, 210-2. The Group Comparison process 900A, 900B is performed by the program after completion of the processes 400, 500, 600, 800 and their respective intermediate processes described above with respect to the data generated by the light plane 210-1 (for the single-array variant 200A) or the data generated by each light planes 210-1, 210-2 (for version 200B with two matrices). As in fig. 10A and in FIG. 10B, at step 902, the number of individual Events in each Group (eg, events 999, 1000, 1001, 1002, 1003 in FIG. 8C) is summed to determine the number of seeds in the Group. It is assumed that the number of summed discrete Events of each Group is equal to the number of seeds that make up the Group.

For the dual array embodiment 200B, an additional step 904 is performed where the total number of Group seeds from each light plane 210-1, 210-2 is compared and the larger number is stored in memory as the number of seeds for that Group. For example, the Group for the first light plane 210-1 identified five separate Events (ie 999, 1000, 1001, 1002, 1003) corresponding to five seeds. However, the second light plane 210-2 may not separate the Seed Groups S2-S5 into four separate Events because the light plane 210-2 may not detect three notch channels if the seeds are positioned such that scanning the second light plane reveals only two notch channels. . Thus, in this example, the second light plane 210-2 can only detect four separate Events corresponding to the four seeds. At step 904, the seed count of the first light plane that identified a greater number of seeds (ie, five seeds) will then be used for the overall seed count of the Group.

At step 906, after the total number of seeds in the Group has been determined and after the Group Velocity has been determined (by any of the methods described above for the single matrix option 200A or for the dual matrix option 200B), the final count of seeds in the Group and the average seed speed is then output to device 130 display of the control system 100 for viewing by an operator.

It should be understood that use of the dual array option 200B may improve seed counting accuracy because the second light plane 210-2 is capable of distinguishing seeds in the third dimension that otherwise could not be differentiated and counted using the single array/single light plane option , which is only capable of detecting seeds in two-dimensional space. Thus, since the dual array embodiment produces a higher number of seeds detected between the two light planes, the dual array embodiment 200B will generally be more accurate than the single array embodiment 200A. However, using the second light plane 210-2 requires twice the processing power compared to using a single light plane (eg, 12 inputs compared to 24 inputs in the example shown). Based on testing, seed counting using the 200A single matrix option was shown to have a seed counting accuracy of 90% (i.e. the measured seed count was within 10% of the true seed count). The dual matrix embodiment 200B increased the seed count accuracy to 97% (ie, the measured seed count was within 3% of the true seed count). However, for air seeders, even 90% seed counting accuracy is extremely accurate. Accordingly, given the additional processing power required by the 200B dual die option, using the 200A single die option may be adequate for many, if not most, air seeder applications.

Pneumatic seeder

In fig. 11 is a side view of a conventional air seeder 10 of the type disclosed in US Patent Publication No. 20100264163 and US Patent No. 8695396, which are incorporated herein by reference in their entirety. In fig. 12 is a schematic illustration of the air seeder 10 shown in FIG. 11, with a particle counter assembly 200 located on the distribution line 58 of the air seeder 10. The air seeder system 10 is shown as consisting of an air cart 11 and a grouser implement 24. The air cart 11 has hoppers 12, 14 for storing seeds, fertilizers, or other granular products. for delivery to the soil. The hoppers 12, 14 are mounted on a frame 16 supported by support wheels 18 to be moved forward along the ground by a towing vehicle (not shown in the drawing) connected to the front coupling device 20. The grunt implement 24 has a frame 26 supported by support wheels 28 and connected to the rear of the pneumatic cart frame 16 by means of a hitch 30. The frame 26 carries a plurality of row units 60. In alternative arrangements, the grouser implement 24 may be located in front of the pneumatic cart 11, or the grouser implement 24 and the pneumatic cart 11 may be located on a common frame. Thus, all references to the air seeder system 10 or, for short, the “air seeder” 10 are intended to refer to each of the above-mentioned devices.

Operation of a pneumatic seeder

Referring to FIG. 12, the air seeder 10 includes an air distribution system 34 similar to that disclosed in US Pat. No. 6,213,690, which is incorporated herein by reference in its entirety. The air distribution system 34 includes a fan 36 to direct air through the main conduit 38. At the bottom of each hopper 12, 14 is a metering mechanism 40 for dispensing metered amounts of seed, fertilizer, or other granular products through the product channels 42, 44 from the respective hoppers 12. 14 into the main channels 38 (only one main channel 38 is shown in FIGS. 10 and 11). The product metered by the metering mechanism 40 into the main conduits 38 is carried by the air flow to a downstream distribution tower 50. Typically, each main conduit 38 has one tower 50. In addition, separate main conduits 38 may be provided for each of the respective hoppers, so that the various products contained in these respective bins 12, 14 can be separately dispensed into a respective tower 50 for delivery via separate distribution lines 58 to separate seeding units 60, as described below. Alternatively, products from respective bins 12, 14 may be combined in a common main line 38, as shown in FIG. 12, for joint distribution. Although two hoppers 12, 14 are shown with associated metering mechanisms 40 and main lines 38, it should be understood that any number of hoppers, metering mechanisms 40 and main lines 38 may be provided on the air seeder 10 as needed.

The dispensing mechanism 40 may be a volumetric dispensing mechanism, but may be any other suitable dispensing mechanism known in the art or developed in the future. As mentioned above, product lines 42, 44 direct product from the metering mechanism 40 to main channels 38, which carry the product in the air flow to downstream distribution towers 50. Each tower 50 has an upper distribution head 52 located at the uppermost end of the vertical distribution pipes 54. Head 52 evenly divides the product stream into multiple distribution lines 58. Each distribution line 58 delivers product to a downstream row unit 60. The seeding unit 60 opens a furrow 62 in the soil surface. The distribution line places the product in the furrow 62, and the trailed compaction or closing wheel 64 (Fig. 11) compacts the soil above the placed product. Although the row unit 60 in FIG. 11 and 12 show one post with a furrow opening point 62 and only one distribution line 58 for each row unit 60, it should be understood that the row unit 60 may be a single channel, dual row row unit using a cutting disc and boot with dual distribution lines 58 for delivering seeds and fertilizers to the soil as described in U.S. Patent Nos. 8,275,525, 9,826,667 and 9,968,030, each of which is incorporated herein by reference in its entirety, and which are commercially embodied in the Case 500 Series air seeder.

Still referring to FIG. 12, the metering mechanisms 40 include variable speed meter drives 72, 74 connected to respective product flow meters 76, 78 located at the bottom of the respective bins 12, 14. When the drives 72, 74 rotate the respective flow meters 76, 78, the product from the respective bins 12, 14 are fed through respective product lines 42, 44 into the main line 38, which in turn transports the product to the distribution tower 50. The feed rate controller 84 is connected to variable speed flow meter drives 72, 74.

The feed rate controller 84 is in signal communication with the controller 110 of the control system 100 (discussed later). In addition to the controller 110 executing a software program 101 to perform each of the processes 400, 500, 600, 800, 900 described above, the controller 110 is also coupled to a speed sensor 168 that determines the ground speed of the air seeder 10. The controller 110 controls the controller 84 feed speeds to adjust the speeds of the dispenser drives to maintain the selected product feed rate when the movement speed changes. In one alternative embodiment, the feed rate controller 84 may be coupled to be actuated based on the tractor's ground speed through a transmission with a gear ratio that is adjustable from the tractor cab. The controller 110 is also connected to the GPS receiver 166, 131. The controller 110 controls the feed rate controller 84 to adjust the metering rate depending on the location of the air seeder in the field. As discussed below, the control system 100 includes an input device, such as a graphical user interface 112, allowing the operator to set a desired product delivery rate, such as pounds per acre or seeds per acre, etc.

Precision seeder

In fig. 13 is a side view of one of a plurality of row units 310 of a conventional planter 300 with a particle counter system 200 located on the seed tube 332 of the row unit 310 to count seeds as they fall by gravity through the seed tube 332. Each row unit 310 is supported by an instrument beam 314 using a parallel linkage 316 that allows each row unit to move vertically independently of the toolbar and other spaced row units to accommodate changes in terrain or when the row unit encounters a rock or other obstruction as the planter moves across the field. Each row unit 310 may include a front mounting bracket 320 on which a hopper support beam 322 and a subframe 324 are mounted. The hopper support beam 322 supports the seed hopper 326 and the fertilizer hopper 328, and also operatively supports the seed meter 330 and seed tube. 332. The subframe 324 includes a downwardly extending shank 325 that, in operation, supports the furrow opening assembly 334. The furrow closing assembly 336 is supported in operation by the rear end of the subframe 324. The furrow opening assembly 334 may include one or more furrow opening disks 344, which are supported in operation by the shank 325. The furrow opening assembly 334 may include one or more gauge wheels 348, which during operation are supported on a subframe by 324 rotating arms of gauge wheels. The depth adjuster 368 may be selectively positioned to change the depth of the furrow opening discs 344 relative to the gauge wheels 348 to vary the depth of the furrow created by the furrow opening discs.

A ride quality sensor 164, which may be an accelerometer, may be mounted on the row unit 310 to measure vertical speed and acceleration of the row unit 310. Speed sensors 168, such as radar speed sensors or GPS speed sensors, may be mounted on the instrument bar 314 or row unit 310. A down force actuator 318, such as an air cushion, hydraulic or pneumatic cylinder, or the like, acts on a parallel linkage 316 to apply a down force to the row unit 310. A down force valve 174, such as an electrically actuated servo valve , can control the amount of clamping force applied by the clamping force actuator 318 .

In operation, as the planter 300 moves in the forward direction indicated by arrow 311, the furrow opening assembly 334 cuts a furrow 338 in the soil surface. The seed hopper 326, which stores seeds for planting, provides a constant supply of seeds to the seed meter 330. In an alternative embodiment, the precision planter 300 may be a center-loading planter comprising a frame-mounted bulk solids hopper known in the art; in such embodiments, the seed hopper 326 may be a small auxiliary hopper in communication with the bulk product hopper. The seed meter 330 selectively interacts with the drive 172 through the clutch 170 so that individual seeds are metered and discharged into the seed tube 332 at regular intervals depending on the desired amount of seed and the speed at which the planter 300 is towed across the field. The drive 172 and clutch 170 may be of the type disclosed in US Pat. No. 8,307,771, which is incorporated herein by reference in its entirety. In other embodiments, there is no clutch 170 and the actuators 172 are electric actuators such as those disclosed in Applicant's International Publication WO2017011355, which is incorporated herein by reference in its entirety. The particle sensor assembly 200 is supported by the seed tube 332 and detects the passage of seeds through the seed tube 332. The seeds fall from the end of the seed tube 332 into the furrow 338 and are covered with soil by the cover wheel assembly 336. As in the air seeder embodiment, in the precision planter 300, the display device 130, communication module 120, and controller 110 may be installed in the cab of a tractor towing the precision planter 300 across a field. The tractor may also be equipped with one or more speed sensors 168, such as a Hall effect wheel speed sensor or a radar speed sensor.

Control system

In fig. 14 is a schematic diagram of a control system 100 for use with an air seeder 10 or a precision seeder 300. The control system 100 includes a controller 110 including a central processing unit 116 and a graphical user interface 112, a memory 114 that stores software executing processes 400, 500, 600, 800, 900, and other processes described above. The monitoring system 100 may also include a communication module 120 and a display device 130. The display device 130, communication module 120, and controller 110 may be installed in a tractor cab. The controller 110 is electrically coupled to the communication module 120 via a wire harness 150. The communication module 120 may include an authentication chip 122 and a memory 126. The communication module 120 is electrically coupled to the display device 130 via a wire harness 152. The display device 130 may include a graphical user interface 132, memory 134, a CPU 136, and a wireless Internet connection 154 for connecting to a cloud storage server 140. One such wireless Internet connection 154 may be a cellular modem 138. Alternatively, wireless Internet connection 154 may be a wireless adapter 139 for establishing an Internet connection through a wireless router.

Display device 130 may be a consumer computing device or other multi-function computing device. The display device 130 may contain general purpose software, including an Internet browser. The display device 130 may also include a motion sensor 137, such as a gyroscope or accelerometer, and may use the signal generated by the motion sensor 137 to determine a desired modification to the graphical user interface 132. Display device 130 may also include a digital camera 135, wherein images from camera 135 may be referenced to global positioning system (GPS) coordinates, stored in memory 134, and transmitted to cloud storage server 140. The display device 130 may also include a GPS receiver 131.

When operating as shown in FIG. 15, the monitoring system 100 may perform the process generally designated 1500. Referring to FIG. 15 in combination with FIG. 14, at step 1505, communication module 120 performs an optional authentication procedure in which communication module 120 receives a first set of authentication data 190 from monitoring device 110 and authentication chip 122 compares authentication data 190 with a key, token, or code stored in module memory 126 120 communication or transmitted from the display device 130. If the authentication data 190 is correct, the communication module 120 preferably transmits the second set of authentication data 191 to the display device 130 such that the display device 130 allows other data to be transmitted between the monitoring device 110 and the display device 130 through the communication module 120.

At step 1510, the monitoring device 110 receives configuration input specified by the user through the graphical user interface 112. In some embodiments, there may be no graphical user interface 112 and the user can enter configuration through the graphical user interface 132 of the display device 130. The configuration input may include parameters such as spatial offsets between the GPS receiver 166 and the row units 60, 300 of the corresponding air seeder 10 or precision planter 300, as well as operating parameters of the air seeder 10 or precision planter 300 (e.g., clutch parameters 170 , actuators 172 and valves 174 clamping force). The monitoring device 110 then transmits the final configuration data 188 to the display device 130 via the communication module 120.

At step 1512, display device 130 may access prescription data files 186 from cloud storage server 140. Prescription data files 186 may include a file (eg, a shape file) containing geographic boundaries (eg, field boundaries) and geographic locations (eg, GPS coordinates) associated with operating parameters (eg, seed rate). The display device 130 may allow a user to edit the prescription data file 186 using a graphical user interface 132 . The display device 130 may reconfigure the prescription data file 186 for use by the controller 110 and transmit the resulting prescription data 185 to the controller 110 via the communication module 120 .

At step 1514, as the air seeder 10 or precision planter 300 moves across the field, the controller 110 sends command signals 198 to the clutches, actuators, and down force controls. These command signals 198 may include signals to determine whether one or more clutches are engaged, signals to determine the speed at which the actuators are operated, and signals to determine the down force applied to the row units 60, 310.

At step 1515, as the planter 10 or precision planters 300 moves across the field, the controller 110 receives raw application data 181, including signals from the particle sensor assembly 200, down force sensors 162, ride quality sensors 164, GPS sensors, and speed sensors 168. The controller 110 processes the original application data 181 and stores said application data in the memory 134. The controller 110 may transmit the processed application data 182 to the display device 130 via the communication module 120. The processed application data 182 may be streaming, fragmented, or partial data. It should be understood that in method 1500, implement control and data storage are performed by controller 110 such that if display device 130 ceases to function, is removed from control system 100, or is used for other functions, storage of implement operations and critical data is not interrupted.

At step 1520, display device 130 receives and stores in memory 134 real-time processed application data 182. At step 1525, display device 130 may render a map of processed application data 182 (eg, a population map). At step 1530, display device 130 may display a numerical aggregate of application data (eg, the population planted by the row unit in the last 5 seconds). At step 1535, display device 130 may store in memory 134 the location, size, and other display characteristics of the application map images rendered at step 1525. At step 1538, after completion of seeding or planting operations, display device 130 may transmit processed application data file 183 to cloud storage server 140. The processed application data file 183 may be a complete file (eg, a data file). At step 1540, monitoring device 110 may store completed application data in memory 114 (eg, in a data file).

The method for mapping and displaying application data 182 may be the same or similar to the application data maps disclosed in US Pat. No. 9,699,958, which is incorporated herein by reference in its entirety.

The above description and drawings are for illustration purposes only and should not be construed as limiting the invention. Various modifications to the embodiments and general principles and characteristics of the system and methods described herein will be apparent to those skilled in the art. Accordingly, the invention is intended to fall within the broadest scope consistent with the appended claims and the full scope of equivalents covered by those claims.

Claims (73)

1. A method for counting particles passing through a channel, the method comprising the steps of: (A) forming a first number of light channels with a certain spacing relative to each other across the channel perpendicular to the longitudinal axis of the channel, the first number of light channels jointly defining a first light plane, each channel of the first number of light channels is formed by a corresponding one LED emitter of the first number of LED emitters, located on the first side of the channel, each LED emitter of the first number of LED emitters is spaced at a known distance from the other emitters and creates a certain light intensity; (B) generating initial output signal values proportional to the light intensity of each light channel of the first number of light channels using a first number of photodiodes located on a second side of the channel, each photodiode of the first number of photodiodes located opposite one of the LED emitters of the first number of LEDs emitters; (C) as the particles pass through the first light plane, parallelly recording initial values of the output signal generated in each light channel of the first number of light channels during successive data scans, wherein the initial values of the output signal are stored in data cells in the first data set; (D) normalizing the original output signal values in each of the data cells of the first data set so that each data cell of the first data set has a Normalized Output Signal (NOS) value; (E) analyzing the NOS values of each of the data cells of the first data set to identify data cells that are clustered together by at least one NOS value within a given range of NOS values, wherein each of the identified clusters of the combined data cells of the first data set defines a separate event of the first data set. a plane corresponding to at least one particle passing through the first light plane; (F) marking each individual first plane event with a unique first plane identifier; (G) summing the NOS values of each identified first plane event; (H) determining the group velocity of each identified first plane event; (I) determining, for each said detected first plane event, an event volume by multiplying the sum of the NOS values of each said detected first plane event obtained in step (G) by the group velocity determined in step (H) for each said detected first plane event ; (J) determining the number of particles of each identified first plane event by classifying the volume of the event. 2. The method according to claim 1, additionally containing the steps of: (K) forming a second number of light channels at a given distance from each other across the channel perpendicular to the longitudinal axis of the channel, wherein the second number of light channels jointly defines a second light plane, wherein each light channel from the second number of light channels is generated by a corresponding LED emitter from the second number of LED emitters located on the third side of the channel, oriented at an angle of 90 degrees to the first side of the channel, wherein each LED emitter of the second number of LED emitters is located at a known distance from other emitters and creates a certain light intensity, and the second light plane is offset relative to the first light plane along the longitudinal axis of the channel to a known distance; (L) generating initial output signal values proportional to the light intensity in each light channel of a second number of light channels using a second number of photodiodes located on a fourth side of the channel, wherein each photodiode of the second number of photodiodes is located opposite one of the second number of LED emitters LED emitters; (M) parallelly recording the initial output signal values generated in each light channel of the second number of light channels as particles pass through the second light plane through successive data scans, the initial output signal values being stored in data cells in the second data set; (N) normalizing the original values of the original output signal in each of the data cells of the second data set so that each data cell of the second data set has a normalized output signal (NOS) value; (O) analyzing the NOS values of each of the data cells of the second data set to identify data cells that are clustered together by at least one NOS value within a given range of NOS values, wherein each of the identified clusters of the combined data cells of the second data set defines a separate event of the second data set. a plane corresponding to at least one particle passing through the second light plane; (P) marking each said individual second plane event with a unique second plane identifier; (Q) summing the NOS values of each said detected second plane event; (R) determining the group velocity of each identified second plane event; (S) determining, for each said detected second plane event, an event volume by multiplying the sum of the NOS values of each said detected second plane event determined in step (Q) by the group velocity of each said detected second plane event from step (R); (T) determining the number of particles of each detected second plane event by classifying the volume of the event for each detected second plane event; (U) associating each of said detected events of the first plane with a corresponding event of said detected events of the second plane; (V) determining which of said first and second plane associated events has a greater number of particles; (W) calculating the current number of particles for the period of operation by adding the larger number of particles determined in step (V) to the number of particles of the immediately preceding event from said associated detected events of the first and second plane. 3. The method according to claim 1, in which, at step (D) of normalizing the values of the original output signal in each data cell of the first data set: converting the original output signal values in each data cell of the first set of data into a NOS value between 0 and 1, wherein a NOS value of 0 corresponds to the maximum light intensity in the corresponding light channel of the first number of light channels and indicates that the light channel is not any of the particles passing through the light channel is blocked, and a NOS value of 1 corresponds to the minimum light intensity in the corresponding light channel of the first number of light channels and indicates that the light channel is completely blocked by particles passing through the light channel. 4. The method according to claim 2, in which at stage (N) of normalizing the values of the original output signal in each data cell of the second data set: converting the original output signal values in each data cell of the second set of data into a NOS value between 0 and 1, wherein a NOS value of 0 corresponds to the maximum light intensity in the corresponding light channel of the second number of light channels and indicates that the light channel is not any of the particles passing through the light channel is blocked, and a NOS value of 1 corresponds to the minimum light intensity in the corresponding light channel from the second number of light channels and indicates that the light channel is completely blocked by particles passing through the light channel. 5. The method according to claim 1, further comprising the step of: calculating the number of passing particles during the operating period by adding the number of particles determined in step (J) to the number of particles of the immediately preceding event from said detected events of the first plane. 6. The method according to claim 1, in which, at the stage of determining the group velocity of said detected first plane event: determining the characteristic width for a given number of events of the first plane; determine the characteristic duration for a given number of events of the first plane; divide the characteristic width by the characteristic duration; wherein the characteristic width is determined by finding the most common event width of the first plane events for a given number of events, wherein the event width of each of the first plane events is determined by counting the first number of channels having NOS values defining each first plane event and multiplying said number by known distance between LED emitters; A the characteristic duration is determined by finding the most common event duration for first plane events for a given number of first plane events, each event duration of each first plane event being determined by determining the duration between the first NOS value and the last NOS value defining each first plane event. 7. The method according to claim 2, in which, at the stage of determining the group velocity of said detected event of the second plane: determining a time difference between said detected first plane event and said detected second plane event associated with said first plane event, and dividing the time difference by a known offset distance of the second light plane relative to the first light plane. 8. The method of claim 2, wherein the known offset distance of the second light plane relative to the first light plane is 0.05 to 1 inch (0.13 to 2.5 cm). 9. The method according to claim 1, in which, at the stage of classifying each of the first plane event volumes identified at stage (J), to determine the number of particles in the first plane event: determining whether the detected volume of the first plane event satisfies the criteria for a given number of particles. 10. The method according to claim 2, wherein in the step of determining the number of particles in step (T) by classifying the volume of the event: determining whether the event volume of the detected second plane event satisfies the criteria for a predetermined number of particles. 11. The method of claim 9 or 10, wherein the predetermined number of particles is one of: (i) five particles, (ii) four particles, (iii) three particles, (iv) two particles, and (v) one particle. 12. The method according to claim 1 or 2, additionally containing the steps of: analyzing the NOS values of each of the data cells of the first or second data set to identify depressions in the combined clusters of NOS data cells defining said individual events of the first or second plane; subdividing each identified first or second plane event having valleys in the combined clusters of NOS data cells into additional individual first or second plane events. 13. The method according to claim 1 or 2, wherein the channel is a distribution pipe of a pneumatic seeder or a seed pipe of a precision seeder, and the particles are one of (i) seeds; (ii) fertilizers; or (iii) seeds and fertilizer. 14. Particle counting system containing: a first array of emitters having a first number of spaced LED emitters, each LED emitter of the first number of LED emitters creating a light channel having a certain light intensity, the first array of emitters located on a first side of the channel such that the first plurality of light channels extend across the channel perpendicular to the longitudinal axis a channel, wherein the first plurality of light channels defines a first light plane; a first array of receivers having a first number of spaced photodiodes, wherein the first array of receivers is located on a second side of the channel, and each photodiode of the first number of photodiodes is located opposite one of the LED emitters of the first number of LED emitters, each photodiode of the first number of photodiodes is configured to generate initial output signal values proportional to the light intensity in each corresponding light channel of the first number of light channels; a controller in signal communication with the first matrix of receivers and configured to: (i) receiving initial output signal values from each photodiode of the first number of photodiodes during a series of sequential data scans and storing the initial output signal values in data cells of the first data set; (ii) normalizing the original output signal values in each of the data cells of the first data set so that each data cell of the first data set has a normalized output signal (NOS) value; (iii) analyzing the NOS values of each of the data cells of the first data set to identify data cells that are clustered together by at least one NOS value within a specified range of NOS values, wherein each of the identified clusters of the combined data cells of the first data set defines a separate event of the first data set. a plane corresponding to at least one particle passing through the first light plane; (iv) marking each said individual first plane event with a unique first plane identifier; (v) determining, for each said first plane detected event, an event volume by multiplying the sum of the NOS values of each said first plane detected event by the group velocity of each said first plane detected event; (vi) determining the number of particles of each identified first plane event by classifying the volume of the event. 15. Particle counting system according to claim 14, additionally containing: a second array of emitters including a second number of spaced apart LED emitters, each LED emitter of the second number of LED emitters creates a light channel having a certain light intensity, the second array of emitters is located on a third side of the channel so that the second set of light channels extends perpendicularly across the channel the longitudinal axis of the channel and at an angle of 90 degrees relative to the first number of light channels, and the second number of light channels defines the second light plane, while the second light plane is offset relative to the first light plane along the longitudinal axis of the channel by a known distance; a second receiver array including a second number of spaced apart photodiodes, wherein the second receiver array is located on a fourth side of the channel, and each photodiode of the second number of photodiodes is located opposite one LED emitter of the second number of LED emitters, wherein each photodiode of the second number of photodiodes is configured to generating initial output signal values proportional to the light intensity in each corresponding light channel of the second number of light channels; wherein each photodiode from the second number of photodiodes of the second receiver array is in signal communication with the first receiver array; the controller is additionally configured to: (i) receiving initial output signal values from each photodiode of the second number of photodiodes during a series of sequential data scans and storing the initial output signal values in data cells of the second data set; (ii) normalizing the original output values in each of the data cells of the second data set so that each data cell of the second data set has a NOS value; (iii) analyzing the NOS values of each of the data cells of the second data set to identify data cells that are clustered together by at least one NOS value within a specified range of NOS values, wherein each of the identified clusters of the combined data cells of the second data set defines a separate event of the second data set. a plane corresponding to at least one particle passing through the second light plane; (iv) marking each said individual second plane event with a unique second plane identifier; (v) determining, for each said second plane detected event, an event volume by multiplying the sum of the NOS values of each said second plane detected event by the group velocity of each said second plane detected event; (vi) determining the number of particles of each detected second plane event by classifying the volume of the event for each detected second plane event; (vii) associating each of said detected events of the first plane with a corresponding event of said detected events of the second plane; (viii) determining which of said first and second plane associated events has a greater number of particles; (ix) calculating the current number of particles for the period of operation by adding the larger number of particles determined in step (viii) to the number of particles of the immediately preceding event from said associated detected events of the first and second plane.

Images