JPH0257110A - Automatic steering control apparatus of farm working machine - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention enables agricultural machinery such as a rice transplanter to run approximately in parallel along so-called rows of crops that have already been planted in a field and lined up in rows. The present invention relates to the structure of an automatic steering control device.

[Conventional technology]

Conventionally, when planting seedlings in a field using a rice transplanter, for example, the rice transplanter is equipped with planting mechanisms at appropriate intervals on the left and right in the direction of movement.
The planting mechanism, which moves up and down as the rice transplanter advances, divides the seedling mat on the seedling platform into appropriate numbers of plants and plants them on the field, so the rice seedlings are placed on the field in an appropriate manner along the direction of movement of the rice transplanter. It is well known that the seedling planting interval is such that the seedlings are planted in a plurality of rows at appropriate intervals in the left and right directions with respect to the direction of travel.

As a prior art for an automatic steering device that allows a rice transplanter to run approximately parallel to rows of planted seedlings (hereinafter referred to as crop rows) that have already been planted in the field, Japanese Patent Laid-Open No. 62-61509
In the publication, a color video camera mounted on a rice transplanter that moves forward is used to image an appropriate range of the adjacent crop rows, and this image information is binarized and each planting is done at the crop location. After extracting the corresponding region, Hough
) Approximately calculate a straight line from the row of the plurality of regions by processing such as conversion, and calculate the difference between the calculated virtual straight line and the lateral shift and tilt shift with respect to arbitrary reference lines and reference points such as the vertical and horizontal center lines of the imaging screen. It is proposed that the aircraft's steering control be performed so that the

[Problem to be solved by the invention]

As in the prior art, when planting is performed using only an imaging means mounted on an agricultural machine such as a rice transplanter while detecting a row of crops and moving forward along the row of crops, the tracking performance for the row of crops is good. , it is possible to follow fine horizontal curvatures of the crop rows well, but it is not good at following the state where the crop rows are largely curved left and right, and furthermore, once the crop rows are If the detection is lost, it becomes impossible to determine in which direction the steering should be corrected.

In addition, for example, if an operator creates a row of crops that curves left and right during the first sample drive prior to unmanned automatic steering control, it will also cause problems during subsequent unmanned tracking control.
There was a problem in that the steering was performed in a meandering manner in accordance with the sample curved row of crops, making it almost impossible to form a straight row of crops thereafter.

On the other hand, if the steering is corrected based on the direction detected using only the geomagnetic sensor, it is possible to travel in a nearly straight line, but the relative positional relationship of the traveling aircraft to the adjacent row of crops is unknown, so deviations from the predetermined standard may occur. There was a problem in that it was not possible to perform tracing control that would fall within the permissible range.

Similarly, a management machine that applies fertilizer or sprays chemicals at an appropriate time after planting seedlings performs the work while moving along the rows of crops that have already been planted. Since the harvesting and threshing work is carried out without moving the harvester forward along the approximately linear boundary line of The steering control made it impossible to drive in a straight line.

SUMMARY OF THE INVENTION An object of the present invention is to provide an automatic steering device that can reliably realize a tracing-type steering correction that cannot be solved by only detecting crop rows using an imaging means or detecting the direction of the earth's magnetic field in an agricultural machine as described above. It is something to do.

[Means to solve the problem]

SUMMARY OF THE INVENTION The present invention provides an apparatus for automatically steering an agricultural machine such as a rice transplanter so that it runs approximately parallel to rows of crops that have already been planted in a field. The machine is equipped with an imaging means for taking an image of the row, and a geomagnetic sensor that detects the direction of the traveling machine using the earth's magnetism, and initially runs automatically based on the direction obtained by the geomagnetic sensor to form a standard crop row. On the sides of the reference crop row, etc., each image information obtained by the imaging means is
Each image is processed by a feature extraction means for extracting features such as value conversion, and the deviation of the characteristics of the crop row with respect to a standard such as straight forward movement of the agricultural machine and the azimuth deviation with respect to the standard obtained by the geomagnetic sensor are respectively calculated. , the automatic steering is performed by a determining means that determines the value of the steering output signal depending on the magnitude of the two deviations.

[Action/effect of the invention]

In the present invention, the creation of the first reference crop row is executed based on the direction detection of the geomagnetic sensor, which makes it more reliable to travel in a straight line than when the operator manually operates the row, and it is possible to The creation of curved reference crop rows is extremely easy.

Then, in the subsequent automatic steering, when farming is carried out along the sides of already planted crop rows as described above,
- Turn around approximately 180 times at the end of the first crop row, and carry out the agricultural work of one crop row along the side of the second crop row, which is what is called a straight round trip. Normally, when traveling in even-numbered rows after turning from the reference crop row, the imaging means follows the sides of the reference crop row, and the geomagnetic sensor detects the direction while automatically steering the vehicle. When traveling in odd-numbered rows after turning from row to row, it is sufficient to repeat automatic steering while following the sides of the row of crops using the imaging means and detecting the direction using the geomagnetic sensor as described above.

In this case, if the value of the output signal for the steering operation is determined from the image information provided by the imaging means and the deviations such as the lateral distance of the traveling machine from the crop row and the inclination in the direction of travel, Although it is possible to follow the rows of crops without losing sight of the rows of crops, if the output signal is determined by taking into account deviations in direction from the geomagnetic sensor, the straightness of the agricultural machine will also be improved. This eliminates the meandering motion that occurs when steering is performed by considering only deviations in crop rows.

Moreover, when there is a difference between the magnitude of the deviation of the characteristics of the crop row with respect to a standard such as straight movement of the agricultural machine obtained by the imaging means and the magnitude of the azimuth deviation with respect to the standard obtained by the geomagnetic sensor, When the azimuth deviation is larger, it can be determined that the azimuth of the traveling aircraft is significantly incorrect with respect to the azimuth of the crop rows, so control is mainly performed to correct the steering in the direction of eliminating the azimuth deviation detected by the geomagnetic sensor. By doing so, the steering can be corrected to eliminate large bends, and within the range of smaller deviations, it can be determined that there is no major deviation in the direction of the traveling aircraft, so deviations with respect to the crop row can be eliminated mainly from the detection results by the imaging means. By executing the following control to correct the steering, it is possible to reliably perform automatic steering for running in parallel, such as going straight, at an appropriate distance to the side of the row of crops.

Note that the determining means for determining the value of the output signal for the steering operation according to the magnitude of the difference between the two parts from the results of the deviation for the crop row obtained by the imaging means and the azimuth deviation obtained by the geomagnetic sensor is a normal one. CRISP control in feedbank control (calculating the value of the output signal by multiplying the deviation by a proportional constant, expressing the value of the output signal as a linear function relationship using two or more deviations as variables, etc.) )
or fuzzy control using fuzzy inference.

〔Example〕

An example applied to a rice transplanter will be described below. In the figure, 1 is a running body supported by front wheels 3, 3 on both left and right sides of the front part and rear wheels 4.4 on both left and right sides of the rear part. At the rear, a multi-row seedling planting device 7 consisting of a seedling platform 5 and a plurality of planting mechanisms 6 is mounted via a link mechanism 8 so as to be movable up and down.

The power of the engine 9 mounted on the top surface of the traveling body 1 is transmitted to the front and rear wheels 3.4 via a clutch 10 and a transmission case 11, while the power is transmitted to the front and rear wheels 3.4 via a PTO shaft 12 protruding from the transmission case 11. Power is transmitted to device 7. In addition, the code 13 indicates ON/OFF of the clutch 10.
14 is an actuator for traveling speed change, 15 is an actuator for
is an actuator for PTO shaft variable shaft peripheral speed.

The steering mechanism 18 is rotated via a steering handle 17 provided in front of the control seat 16 in the traveling aircraft 1, and the direction of the front wheels 3.3 is changed from side to side. The steering arm 20 is horizontally rotatably mounted on the rotation fulcrum shaft 19 of the steering mechanism 18, a pair of left and right tie rods 21, 21 connected to the steering arm 20, a hydraulic cylinder 22, and a manual It consists of a control valve 23 for steering, a steering gear box 24 that operates the control valve 23, and a pitman arm 25 that is swingable back and forth.

The control valve 23 is connected rearward to the steering arm 20 via a ball-and-socket joint, and the spool at the rear end of the control valve 23 is connected to the pitman arm 25.

Further, the rear end of a hydraulic cylinder 22, whose front end is rotatably connected to the traveling body 1, is rotatably connected to the steering arm 20, and swings in accordance with the rotation angle of the control handle 17. The pitman arm 25 moves the spool of the control valve 23 forward and backward, sends hydraulic pressure from the hydraulic pump 26 driven by the engine 9, moves the piston rod in the hydraulic cylinder 22 in and out, and responds to the rotation of the steering arm 20. Then, change the direction of both the left and right front wheels by 3°3.

This hydraulic cylinder 22 is also driven by an electromagnetic solenoid type automatic steering control valve 27, and in this case, the steering angle of the front wheels 3 is determined by a potentiometer 28 attached to the rotation fulcrum shaft 23.
This can be determined by detecting the rotation angle of the steering arm 20 at .

The automatic steering control valve 27 is operated by a steering controller 3 driven by a central control device 30 for automatic steering and traveling.
1, and is actuated by the ON-OFF actuator 13 of the clutch 10, the travel shift actuator 14, the PTO shaft circumferential speed change actuator 15, and the travel controller 32 driven by the central control device 30. Operate.

The detection device 33 in the automatic steering control device of the present invention consists of a tracing imaging means 34 that images the crop row, and a geomagnetic sensor 35 that is mounted on the agricultural machine and detects its absolute direction. , a feature extracting means 36 for extracting features such as binarizing the image information captured by the image capturing means 34; and a feature extraction means 36 for extracting features such as binarization of the image information captured by the imaging means 34; It includes a comparison section 37 that calculates a deviation, and an image processing controller 38 that processes the features obtained by the feature extraction means 36 and exchanges necessary information (data) with the central control device 30.

The geomagnetic sensor 35 detects the azimuth angle of the reference line of the traveling direction of the traveling aircraft 1 with respect to the geomagnetism (for example, north), and uses a combination of a Hall element and a magnetoresistive element to detect the magnetism in a specific direction (geomagnetic field). It can be configured so that it has good sensitivity in certain directions and is insensitive in other directions.

The geomagnetic sensor 35 is preferably provided at the upper end of a support that projects upward from the traveling body 1 in order to reduce the influence of magnetism due to metal parts such as steel in the traveling body 1 as much as possible.

The imaging means 34 is rotatably attached to an arm projecting laterally of the traveling body 1 or an upright support via a posture maintaining means (not shown) such as a steering arm.

Considering the case where the traveling body 1 moves forward along the side of the crop row, turns around 180 degrees at the end of the crop row, and travels, the scanning imaging means 34 is provided on both left and right sides of the traveling body 1. It is preferable to provide one each, or to provide one so that the left and right installation positions can be changed.

The imaging means 34 follows the reference line Ko of the imaging travel to the traveling aircraft 1.
Set it so that it is parallel to the direction of travel on the side, facing forward and diagonally downward.

The imaging means 34 is an area sensor that detects an object, and has a two-dimensional plane with an imaging screen extending vertically and horizontally like an x-y plane, such as a so-called video camera, and is, for example, a two-dimensional MO3 sensor. In devices with a built-in image sensor or two-dimensional CCD image sensor, the image formed through the lens is sensed by each image sensor (photoelectric device) arranged in a two-dimensional array on the image plane, and then displayed on the imaging screen. 40
This information can be output as an electrical signal.

Further, the imaging means 34 may be either color or black and white, but by using the copying crop row 34 for imaging the crop row in color, the field scene and the crop row can be distinguished and their characteristics can be further enhanced. Can be clearly recognized.

Next, the traveling body 1 is placed in a field, and after executing one-way travel from the edge of the field, it is turned around approximately 180 times at the terminal end to execute two-way travel, and then turns at the end to make three-way travel. The main entrance runs, and the four exits run in the same way, and so on.
A schematic flowchart of the processing executed by the detection device 33 and the central control device 30 when performing agricultural work such as rice planting while moving forward in a straight line and then turning back and traveling back and forth.
4a and 4-b) will be explained.

First, in step S1 following the start, after setting initial values, the operator of the agricultural machine drives the traveling machine 1 to position it at the edge of the field. Next, in step S2, before starting the single-head run, the geomagnetic sensor 35 is activated, and in step S3, the operator detects the direction of travel of the traveling aircraft 1 [reference orientation (θ0)] while the operator moves the traveling aircraft 1. A standard crop row is created by moving forward and carrying out agricultural work such as seedling planting. At this time, the direction of the traveling aircraft 1 should not change too much, but since there may be slight changes in the orientation due to changes in the orientation of the traveling aircraft, the reference orientation (θ0) should be determined at appropriate time intervals or by adjusting the traveling distance as appropriate. It may be determined based on the average value of the detected values of the direction detected a plurality of times at each interval.

The value of this reference direction (θ0) is used as the reference direction for the so-called steering correction control in which the vehicle travels while following the sides of the crop row. Then, in step S5, the -
At the end of the main run, it turns approximately 180 times and enters a two-way run along the sides of the reference crop row created during the second run, and thereafter executes the steering correction control subroutine shown in step S6. This involves repeating the agricultural work of creating rows of crops.

In addition, when working with a combine harvester or a lawn mower, the crop row refers to the boundary line between the cut area and the uncut area.

Fig. 4-b shows the content of the steering correction control subroutine, in which the imaging means 34 and the geomagnetic sensor 35 are operated simultaneously in parallel (step R1, step R
2), the steering correction control is performed using the detection results of both, and in step R3, when driving after the second exit, the geomagnetic sensor 35 detects the direction θ2 at appropriate time intervals (Δt), and The azimuth deviation θ from the reference azimuth (θ0) of
-θ〇-θ2 is calculated in step R4. When the azimuth deviation θ obtained in this manner is calculated a plurality of times, the average value thereof is calculated and determined in step R5. At that time, when the latest data is detected and calculated, by removing the oldest data and adding the latest data to calculate a so-called moving average value, it is possible to grasp trends due to new data that is being obtained one after another. .

In parallel with these processes, in step R6, the imaging means 34
The image data obtained in is captured at appropriate time intervals (Δt), and then in step R7, the planting is performed using binarization, etc., which extracts and distinguishes the crop area (NAE) from other field scenes for each screen. Perform feature extraction. This feature is stored in the image processing controller 38, etc.

FIG. 5 shows a state in which a planted crop location (NAE) is imaged on the imaging screen 40, and FIG. 6 shows a 2Jl@ screen.

In this embodiment, when the imaging means 34 is used for color, the RGB color system [red (R), green (G), blue (
The characteristics of the field scene are extracted from the signals of the red, green, and blue components (using the color light of B) as the primary color light and white is obtained by adding light, and the signal output of these three color components is When the signal output ratio of the green (G) component to the total sum (R+G+B=1) is greater than or equal to a predetermined value, it is determined to be a seedling (crop), and the area (N) is divided (S) from other parts of the imaging screen 40.
(egmentation) Execute the binarization process to specify the

Note that binarization may be performed using color difference processing in which a portion of the color line segment in which the color difference image data (C; -B) obtained by curving the green component to the blue component has an output of a certain level or more is determined to be a seedling.

Each time one image data is captured by the imaging means 34,
For the plurality of plantings obtained by binarizing the data, the crop area (Nl), (N2), (N3), etc. are simultaneously displayed on - number of screens 40, so in step R8, two After performing calculations to determine the coordinates of the position of each crop point, each valued planting area (Nl),
(N2), (N3)... is calculated for the virtual straight line as the closest linear approximation, and the declination angle Δρ and lateral deviation deviation ρ of the virtual line KN with respect to the reference line KO on the imaging screen 40 are calculated. (See Figure 6).

Calculation of β on this virtual line is performed using software pre-installed in the image processing controller 3B, such as the well-known least squares error method or the Hough (HOUGII) transform method. When a certain point (Xi, Yi) in the orthogonal coordinate system is given, all line segments passing through it are P i =Xicos 0 in the polar coordinate system.
It can be expressed as i4-Yisin θi.

Therefore, each of the areas (Nl) on the imaging screen 40.

For each (N2), (N3)...
Execute the Hough processing expressed in i), count the frequency of taking the same ρ and θ as a dimensional histogram, find the maximum values of ρ0 and θ0, and calculate the virtual straight line of the crop row for each of the - screens. This specifies β.

β is the reference line Ko for each virtual straight line obtained in this way.
The reference point 0 (on the imaging screen 40 7! From the center position, etc.) to a virtual straight line along the X-axis direction! The lateral deviation deviation ρ, which is the distance up to this point, is calculated for each virtual straight line Kffi (step R9).

In addition, when it is on the right side of the reference line KO in the imaging screen 40, ρ>0, and when it is on the left side, ρ<0.

Furthermore, the case where the virtual line tilts to the right from the reference line Ko as it goes upward on the imaging screen 40 is assumed to be Δρ>0, and the opposite case is assumed to be Δρ<0.

Then, in step RIO, the plurality of virtual straight lines H
The average value of the values of Δρ and ρ calculated for each x is calculated, but at this time, in order to take into account the trend based on the latest data, as in the case of θ, a so-called moving average calculation is performed.

Next, in step R11, it is determined whether the azimuth deviation θ is larger than the dead zone β, and similarly, in step R12, it is determined whether the calculated declination angle Δρ is within the range of the dead zones α and β. The determination result determines how to weight the output signals of subsequent steering control.

In order to ensure the stability of steering correction control, when the calculated azimuth deviation θ and yaw angle Δρ are within a certain range (within a dead zone), the control valve of the steering drive means, such as a hydraulic cylinder, is The drive electromagnetic solenoid does not operate, and the electromagnetic solenoid operates only when the range is exceeded (the dead zone is exceeded).

Therefore, by using the left/right deviation angle Δρ of the crop row with respect to the reference line Ko from the image information obtained by the imaging means 34, fine steering correction is made so as to follow the crop row in a substantially parallel manner along the sides of the crop row. The dead zone (angle to the left and right with respect to the reference line) α for this deflection angle Δρ should be set small.

On the other hand, the azimuth deviation θ obtained by the geomagnetic sensor 35 is used to determine whether or not the direction of travel is significantly tilted to the left or right with respect to the reference crop row, so the dead zone in this case (Left and right angles with respect to the reference line) β is the above α
(See Figure 7).

In Fig. 4-b, the part surrounded by the -dotted chain line is the means for determining which deviation (factor) to focus on when determining the output signal for steering, and it uses the inference of fuzzy control. The magnitude of the output signal may be determined using The magnitude of the output signal may be determined as a linear functional relationship between the steering controller 31 and the automatic steering control valve 27 in accordance with the magnitude of the output signal.
to drive the electromagnetic solenoid.

Then, in step R11, when it is determined that the azimuth deviation θ detected by the geomagnetic sensor is larger than the large dead zone β, the traveling direction of the traveling body 1 is largely deviated from the direction in which the crop rows are lined up. Therefore, in order to preferentially execute steering correction in the direction in which the azimuth deviation θ is smaller, steering control is executed with increased weighting so as to place emphasis on the detection result of the azimuth deviation θ by this geomagnetic sensor (step R13).

When it is determined in step R12 that the deflection angle Δρ by the imaging means 34 is smaller than the value of the smaller dead zone α, the output signal for corrective steering is not outputted and the deviation angle Δρ is maintained as it is. In other words, there is no change in the direction of the wheels (step R14).

In the state of α〈Δρ〈β, the azimuth deviation is not so far apart from the reference direction, and the deviation angle Δρ is rather large from the detection result of the imaging means 34 for the crop row, so the image Weighting is performed to place emphasis on the detection result by means 34 (step R15).

When θ<β and Δρ>β, the imaging means 34
The detection results of the geomagnetic sensor 35 and the geomagnetic sensor 35 are weighted to be considered approximately equally (step R).
16).

Thereafter, in the central control unit 30 having a data storage section, a calculation section, a comparison section, etc., a steering instruction amount S is calculated based on the plurality of deviations ρΔρ, θ according to a control theory such as fuzzy control, and the steering instruction amount S is A case will be described in which an output signal for steering correction is outputted to the steering controller 31, for example, the automatic steering control valve 27 for rotationally driving the steering wheel, in response to the instruction amount S.

In this case, in the fuzzy control, data from the imaging means 34 and the geomagnetic sensor 35 are processed at high speed;
In order to compensate for the loss of accuracy in crop row detection and azimuth deviation detection, it is possible to obtain practical control outputs with less accurate detection values, that is, ambiguous inputs.
In addition, the relationship between the conditions of the detection target and the controlled amount in control,
In other words, it is suitable for control when it is difficult to strictly model and describe the relationship between input and output.

Furthermore, fuzzy control is also suitable when the detection means is an imaging means and a geomagnetic sensor, as in the present invention, in which a plurality of unmistakable elements are combined as a control condition part.

Next, fuzzy control using fuzzy inference will be explained.

Generally, in control using fuzzy inference, a control algorithm is described as an ambiguous relationship between input variables of multiple pieces of information for control, for example, two input variables (x, y) and an output (operated amount) 2 to a control device. It is something to do.

for example, If X is small and y is large, 2 should be in the middle.

If X is large and y is medium, make 2 large.

The control algorithm is expressed by what is called a fuzzy control rule of the form (if-then), such as: rule i
The part f... is called the antecedent part, and the part then... is called the consequent part.

Now, in applying this fuzzy control rule to automatic steering control, in this embodiment, a set of data of lateral deviation ρ and deviation angle Δρ of crop rows is used as an antecedent part, and automatic steering operation instruction amount S is the consequent part, nine fuzzy control rules 1-1 to 1-9 are set, the data set of the lateral deviation ρ and the azimuth deviation θ is the antecedent part, and the automatic steering operation instruction amount S is The nine fuzzy control rules from 2-1 to 2-9, for a total of 18 rules, that are used as the consequent part are listed in Table 1 (Part 1 and Part 2).
). Here, rule 1-1 indicates that if the label of ρ is O and the label of Δρ is 0, S should be set to O (straight ahead).

In Tables 2 to 4, the fuzzy variables taken by the input variables of the deviations ρ, Δρ, and θ and the output variable of the operation instruction amount S are shown as discrete fuzzy variables that are discretized into an integer domain. It is something.

The area of fuzzy variables, which is an ambiguous area, is defined by giving the degree (grade) to which the elements (members) of the entire set of input variables are included in the area (domain). The function that gives is called the Menharsisop function. For example, the domain of fuzzy variables of ρ is the element (member) of the entire set of strike-slip deviations ρ.
It is defined by giving the degree (grade) that is included in the domain (domain).

The (value) at the top of each table indicates the range of each variable. For example, -5 to 5 in Table 2 is the input variable ρ
is the value taken by

In the example, the label of the fuzzy variable in each table is “
"Big right", "Small right j", "0", "Small left", "
The degree to which the value of the variable is included in the set of these labels (degree of fit = membership value)
is expressed in stages as an integer from 0 to 10.

1 “Fushii 1’1 membership of X (110,67) pixels △ membership of Bu ×( 1/20) rad.

Membership of θ Bu ×( 1/20) rad.

Below the member zip steering operation instruction amount of S, ρ-3 (unit: 110.67 pixels), Δρ-1(
Unit 1/20rad, ), θ--3 (Unit 1/2
0rad, ), the fuzzy inference control method in this embodiment will be explained.

The fuzzy control rules established from the set of ρ-3 and Δρ-1 are shown in Table 2, Nos. 1 to 1-1. No. 1-4.
There are three, No. 1-8, and the fuzzy control rules that are established from the set of ρ-3 and θ--3 are shown in Table 2, from No. 2 to No.
2-I No, 2-2 No, 2-4 No, 2-8-4
It is one.

In fuzzy control rule No. 1-1, the label of ρ as the antecedent part takes "0". Variable ρ-3 at this time
The fitness degree of the membership function at the label "0" that can be taken is "3" from Table 2. Similarly, the label of Δρ as the antecedent part takes "0". The degree of fitness of the membership function at the label "0" that the variable Δρ-1 can take at this time is "7" from Table 3.

The fitness of these two member zip functions is "3" and "7".
”, the smaller value “3” is set as N of the fuzzy control rule.
o, 1-1 is the degree of conformance to the condition of the entire antecedent part, ``If ρ is O and Δρ is 0.''

Furthermore, since the label of the operation instruction amount S given by the consequent part (inference result) of No. 1-1 of the fuzzy control rule "S is set to 0" is "0", in Table 5 The upper limit of the fitness of the discrete membership function is set to "3", which is the fitness of the entire antecedent part (Fig. 8 (
a)).

Thereafter, in the same way as above, the fuzzy control rule teeth 2-8 are set for a total of 7
After executing the process, the amount of operation instruction S obtained from each rule is combined. The combination is obtained by superimposing the respective operation instruction amounts S and adopting the maximum value.

The final value of the operation instruction amount S is determined by the value of the center of gravity of the set of membership functions of S obtained in this way (
3=-1,71) (see Figure 10).

These membership functions can be determined by referring to each table above.
Fuzzy control rule level 1-1 is shown in FIG. 8(a), similarly, fuzzy control rule level 1-4 is shown in FIG. 8(b), and fuzzy control rule level 18 is shown in FIG. 8(C).
), and Fig. 9(a) for fuzzy control rule interval 2-1.
The fuzzy control rule No. 2-2 is shown in No. 9.
FIG. 9(b), fuzzy control rule interval 2-4 is shown in FIG. 9(c), and fuzzy control rule No. 2-8 is shown in FIG. 9(d).

In this case, for example, the fuzzy control rule IVk11.1-1
The degree of suitability of the operation instruction amount S is the area below the height "3", which is surrounded by a thick line.

The central control device 3
0, the electromagnetic solenoid of the steering control valve 27 is activated, and the hydraulic cylinder 22 that changes the rotation angle of the steering arm 20 in the steering mechanism is driven to perform corrective steering, and on the side of a predetermined crop row, Automatic steering control is executed in which the traveling machine 1 moves forward in parallel along the crop row.

In this case, the potentiometer 28 that detects the rotation angle of the steering arm 20 determines whether or not the front wheels 3 already have a steering angle that tilts to the right or left with respect to the direction of travel. This can be useful for convergence control that progresses to a parallel state.

Furthermore, when executing steering correction control, the actual steering angle of the wheels can be detected by the potentiometer 28;
By comparing the amount of change in the angle of the potentiometer 28 at this time or the above-mentioned steering instruction amount S with the amount of change in the azimuth deviation θ of the geomagnetic sensor, it is determined whether the field is soft or hard. be able to.

In other words, when the amount of change in the azimuth deviation θ of the geomagnetic sensor is smaller than the actual angle at which the wheels change direction or the amount of wheel steering command (steering amount), a slipping phenomenon such as skidding of the wheels against the field surface occurs. It can be determined that the field is weak.

On the other hand, when the amount of change in the azimuth deviation measured by the geomagnetic sensor is large relative to the amount of steering, the wheels rarely skid relative to the field, and the direction change of the traveling aircraft follows the steering correction well. Therefore, it can be determined that the field is hard.

In addition, when farming in areas such as large-scale, graded fields, it is easy for agricultural machinery to move straight because the field is rectangular in plan view. By changing the so-called weighting settings so as to increase the fitness of the membership function of the azimuth deviation θ, it is possible to strongly exert automatic steering control that targets the azimuth.

Furthermore, in order to enforce automatic steering control that targets the heading, the operator switches the fuzzy control switch so that only the antecedent part of the combination of ρ and θ is applied in the fuzzy control rule. Also good.

In this way, by strengthening automatic steering control that targets heading, the straight-line performance of agricultural machines can be greatly improved.

On the other hand, in places where the field edges are curved from side to side, ρ is
The fuzzy control switch is changed so that only the antecedent part of the combination of By changing this, it is possible to weaken the automatic steering control that targets this direction.

The weighting settings may be changed automatically or manually.

If fuzzy control is adopted in automatic steering control as in this example, it is possible to perform control using multiple different detection targets (deviations as inputs in this example) and various combinations of input conditions as condition parts. Moreover, the degree of input suitability can be changed extremely easily using control software.

If the azimuth deviation θ, declination angle Δρ, and lateral deviation deviation ρ from the reference line of the crop row are calculated using the moving average values of the results of multiple detections, it is possible to equalize the error caused by the detection results only - times and calculate the most recent one. It is possible to include the detection results and consider the trends.

Furthermore, due to the unevenness of the plowing platform in the field, the traveling machine tilts to the left and right.In other words, when the wheels enter the depressions in the tilling platform, the traveling machine tilts downward to that side, making it difficult to image the rows of crops. If the image capturing means is fixedly attached to the traveling machine body, the position of the crop row will be imaged as if it were shifted laterally due to the tilting of the traveling machine body as described above. Attach the tilt sensor
It is preferable to perform a correction that laterally moves the left and right reference lines in the imaging means in accordance with the left and right inclination angle of the traveling aircraft.

[Brief explanation of the drawing]

The drawings show an embodiment of the present invention, in which Fig. 1 is a plan view of a riding rice transplanter, Fig. 2 is a side view, and Fig. 3 is a block diagram of an automatic steering/travel control device and an explanatory diagram of its operation including a hydraulic circuit. , 4th-
Figures a and 4-b are control flowcharts, Figure 5 is a diagram of the imaging screen by the imaging means, Figure 6 is a diagram of the binarized imaging screen, Figure 7 is an explanatory diagram of the dead zone, and Figure 8 Figure (a),
(b) (C) is a diagram of the membership function in case of fuzzy control rule 1, (a), (b), (c) in Figure 9
(d) is a diagram of the membership function in the case of fuzzy control rule No. 2, and FIG. 10 is an explanatory diagram showing the method of determining the steering instruction amount. 1... Traveling body, 2... Frame, 3.4...
... Wheels, 5 ... Seedling stand, 6 ... Planting mechanism, 7
...Seedling planting device, 9...Engine, 11...Beaution case, 17...Steering handle, 2o...
... Control handle, NAE ... Planting at crop location, 1
9...Rotation fulcrum shaft, 20...Steering arm, 22...Hydraulic cylinder, 27...Automatic steering control valve, 3o...Central control unit, 31... Steering controller, 32... Traveling controller, 33...
...Detection device, 34...Imaging means, 35...
Geomagnetic sensor, 36...Feature extraction means, 37...
...Comparison section, 38... Image processing controller.

Claims (1)

[Claims] (1) In a device that automatically controls an agricultural machine such as a rice transplanter so that it runs approximately parallel to and along the rows of crops already planted in a field, the machine The machine is equipped with an imaging means for taking an image of the ground, and a geomagnetic sensor that detects the direction of the traveling machine using the earth's magnetism. Initially, the machine automatically travels based on the direction obtained by the geomagnetic sensor to form a reference crop row, and On the sides of the crop row, etc., each image information obtained by the imaging means is processed by a feature extracting means for extracting features such as binarization, and the crop row is determined based on a standard such as straight forward movement of the agricultural machine. The vehicle is characterized by automatic steering using a determining means that calculates the deviation of the characteristic of the steering wheel and the azimuth deviation with respect to the reference obtained by the geomagnetic sensor, and determines the value of the steering output signal according to the magnitude of both deviations. Automatic steering control device for agricultural machinery.