JPH0441368B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a guidance system for an unmanned vehicle that guides the vehicle while detecting guidelines on the floor using an optical sensor.

The method of detecting guidelines on the floor using the optical sensor and steering and guiding the vehicle based on the detected values is generally called an optical guidance method, in which light emitted from a light source is reflected by the floor surface, The guideline position is detected based on the difference in the amount of reflected light received by each light receiving element, but in the conventional method, the guideline position is detected by
Since it is simply detected as being the same as the position of the light receiving element that has received reflected light above a certain amount of received light (threshold value), the reflection from the guideline may temporarily drop below the threshold value in a place where the guideline is dirty. In addition, the optical sensor may not be able to accurately detect the guideline due to the difference in the material of the floor surface depending on the location, which may increase or decrease as the vehicle travels. Ta.

In addition, even if the position is outside the guideline, strong reflected light may unexpectedly enter the light-receiving element outside the guideline position due to the influence of local areas with high reflectance on the floor or foreign objects with high reflectance.
There was an inconvenience that this locally high amount of received light was judged as a guideline.

Therefore, the present invention smoothes the amount of light received by each light receiving element of an optical sensor by averaging it with the amount of light received by adjacent light receiving elements, and also smoothes the amount of light received by each light receiving element based on the fluctuation in the total amount of light received by all the light receiving elements. hand,
The above-mentioned disadvantages are resolved by changing the height of the threshold for guideline detection.
Examples will be described below based on the drawings.

That is, FIG. 1 is a schematic plan view of a three-wheeled unmanned vehicle 1 as an example of an unmanned vehicle, where 2 is a drive wheel that is directly connected to a travel motor 3 and steers around a vertical axis 4, and 5 is a drive wheel that rotates around a vertical axis 4. Sprocket 6 fixed to sprocket 4 and chain 7 attached to sprocket 6
9 is a driven wheel, and in this unmanned vehicle 1, an optical sensor 11 is particularly installed at a position biased toward the front left side of the drive wheel 2. . In this unmanned vehicle 1, the optical sensor 11 is installed at a side position away from the drive wheel 2 position as described above, and is laid on the floor by pasting tape or applying white paint. Since the vehicle runs while viewing the guide line 12 at a position offset from the center of the vehicle body 1, the wheels 2 do not move on the guide line 12 and the guide line 12 can be prevented from becoming dirty.

Next, to explain the structure of the optical sensor 11 of this example, as shown in FIGS. 2 to 5, the sensor of this example has a substantially rectangular base 13.
16 holes 14 and 15 are bored in two rows in the longitudinal direction,
In the holes 14 and 15, infrared light emitting diodes 16 as a light source are inserted in one row, and phototransistors 17 as light receiving elements are inserted in the other row. 18 is a mounting block to the vehicle body 1, 19 is a printed circuit board, 21 is a red filter, and the infrared light emitting diode 16 has a wide directional characteristic (about 60
The phototransistor 17 is loaded with a narrow directivity (approximately 20 degrees) at an angle to the board 13, and the phototransistor 17 is loaded perpendicularly to the board 13.
Directly reflected light emitted from the infrared light emitting diode 16 and reflected by the floor surface F or the guideline 12
L'1 and L'2 are designed not to enter the linear phototransistor 17 to eliminate the influence of gloss on the floor (Figs. 3 and 4).

FIG. 5 shows the circuit of the infrared light emitting diode 16 and phototransistor 17, and shows the voltage (Vp)
By applying this, light L is emitted, and reflected light from the floor surface or guideline is incident on the phototransistor 17, and a voltage proportional to the amount of incident light is output (OUT).

In this way, the output voltage from each phototransistor 17 changes in proportion to the amount of light reflected from the floor directly below the phototransistor 17, and the magnitude of the output voltage indirectly affects the light reflectance of the floor. difference,
In other words, it is designed to detect whether it is a floor surface or a guideline.Next, the method for detecting the guideline position will be explained.

That is, as shown in FIG. 6, each phototransistor 17 is connected to a computer 25 mounted on the vehicle body 1 via an analog multiplexer 22, an amplifier 23, and a comparator 24, and the computer 25 is connected to each interface 26, 2
The computer 25 calculates the guideline position by calculating and analyzing the output signal of the phototransistor 17 using a method described in detail later, and calculates the guideline position based on the calculated value. The driving motor 3 or the steering motor 8 is driven as appropriate to guide the vehicle. 28 is a D/A compada, 3
1, 32, and 33 are in the computer 25, respectively.
RAM, CPU and ROM.

That is, the comparator 24 and the D/A converter 28 are circuits for A/D conversion, and the multiplexer 22 is appropriately switched by a phototransistor selection command 29 from the computer 26, and the output voltage from the phototransistor 17 at one end of the sensor is changed. The output voltages are sequentially input to the A/D converter circuit, and the output voltages are converted into corresponding digital values by a known sequential comparison method and then sent to the computer 25.
The data is stored in the RAM 31 inside.

That is, for example, in order from the end of the sensor 11, a voltage of 3 volts is output from the first phototransistor 17 through the amplifier 23 and A/D conversion, a voltage of 3.5 volts is output from the second phototransistor 17, and a voltage of 3.5 volts is output from the second phototransistor 17.
Each voltage value such as 4 volts from the phototransistor 17 is temporarily stored in the memory 31, and the stored voltage values are processed as follows.

That is, if the stored voltage value, that is, the amount of light received, is expressed in a bar graph, it becomes a bar line with white circles in Figure 8, and the numbers on the horizontal axis of the graph indicate the position of each phototransistor (see Figure 7), and the numbers on the vertical axis indicate the position of each phototransistor (see Figure 7). The axis shows the amount of light received, but the value of this bar line with white circles is based on the actual amount of light received, so it faithfully reflects the effects of locally bright areas on the floor, highly reflective dust, etc. Even in areas other than the guideline, a locally high value may be shown (value 14 in Figure 8). To eliminate this effect, the voltage value obtained as above is first Smooth it like this.

That is, as shown in FIG.
The actual voltage value from the phototransistor No. 1 and the phototransistor No. 16, that is, the transistors across the sensor, are each detected to be 1.5 volts.
For each value, take the average with the actual voltage value from the phototransistor (No. 2 and No. 15) one inside, and use that value as the smoothed voltage value at each end. . In other words, if the actual value of the second phototransistor is, for example, 2.3 volts, then
(1.5 + 2.3) / 2 = 1.9 gives a smoothing value of 1.9 for the 1st phototransistor, and if the actual value of the 15th phototransistor is, for example, 2.1 volts, then (2.1 + 1.5) / 2 = 1.8, the smoothing value of the 16th phototransistor is 1.8.

In addition, each phototransistor 17 of No. 2 to No. 15
Regarding the smoothing of the voltage value, the average value of the values on both sides is taken.

In other words, for example, the 14th phototransistor 1
If the actual voltage value from 7 is 3.4 volts, and the values of 13th and 15th are 2.7 volts and 2.0 volts, respectively, then (3.4 + 2.7 + 2.0) / 3 = 2.7.
As the smoothing value of phototransistor 17 No. 14
2.7 is obtained, and the above calculation is performed for each phototransistor to smooth all values.

The smoothed values are shown as bars with black circles in FIG.

Next, an average value M is further calculated from all the smoothed values for each phototransistor 17 obtained by the above calculation, and a constant value (α) is set to this calculated average value M (2.5 volts in FIG. 8). The resulting value is set as the threshold T (3.0 volts in FIG. 8).

As is clear from Figures 7 and 8, the actual voltage value (white circle) before smoothing exceeds the threshold T even at the phototransistor position (No. 14) other than the guideline position (Nos. 5 to 8). Although there are values, after smoothing, only the values of the phototransistors (Nos. 5 to 8) at the guideline position exceed the threshold T, so that accurate detection of the guideline position can be performed.

Note that the constant value (α) to be added to the average value M may not be a fixed value, but may be a value obtained by multiplying the average value M by a constant coefficient, or a constant value of the average value M (for example, by smoothing It may be a value obtained by multiplying the difference from the later maximum value (value No. 7 in FIG. 8) by a series number.

Then, the center of the guideline is determined based on the threshold value T determined by the above calculation and each value after smoothing.

That is, based on each of the above-mentioned smoothing values and threshold T, for each phototransistor (Nos. 1 to 16), those whose smoothed values exceed the threshold T are set as "1", and those whose smoothed values do not exceed the threshold T. ``0''
For example, in the above example, the 1st to 4th and 9th to 16th phototransistors 17 are "0", and the 5th to 8th phototransistors 17 are "0". is given "1".

Then, to the value (V) "0" or "1" given by the above binarization, the weight (W) regarding the position of each phototransistor (No. 1 to No. 16) is added.
(In other words, for example, a number that increases from the leftmost side to the rightmost side, in this case 0 to 30) is multiplied, and the product value is divided by the sum of all values after binarization. Therefore, a central guideline is required.

That is, in the case of the above example, as shown in FIG. 9, the first phototransistor 17 has "0".
However, the second phototransistor 17 has "2".
However, ``30'' is given as the position weight (W) to the 16th phototransistor 17, and the center position of the guideline is calculated as the position weight in the following manner.

〓(position weight)×(value after binarization)/〓(value after binarization) =(0×0)+(2×0)+…+(8×1)+(10×
1)+(12×1)+(14×1)+…+(30×0)/0+
0+...+1+1+1+1+0+...+0=11 Therefore, in the above example, the position weight is "11", that is, the 6th and 7th in the sensor 11.
It is detected that the center of the guideline is located directly below the midpoint of the phototransistor number.

As above, (position weight) × (value after binarization)
By dividing the sum of (values after binarization) by the sum of (values after binarization), the sum of (values after binarization) corresponds to the width of the guideline, so regardless of the fluctuation of the guideline width, It is possible to accurately detect the center position of the guideline, and to reduce the influence of the appearance of locally high values that cannot be eliminated even by the smoothing process described above.

That is, for example, in FIG. 8, it is assumed that the smoothed value of the 14th phototransistor 17 other than the guideline position exceeds the threshold value T, and even after the conversion to two values or more, the value of 14th phototransistor 17 is "1". Assuming that is given, the operation to calculate the center of the guideline above is (8 x 1) + (10 x 1) + (12 x 1) + (
14 × 1) + (26 × 1) / 1 + 1 + 1 + 1 + 1 = 14, and the detected value “14” corresponds to the 8th phototransistor position in the sensor 11, so the detected value is the width of the guideline (5th to 8th phototransistor position). phototransistor position).
The local influence of the 14th phototransistor position is reduced.

Once the center position of the guideline has been calculated in the computer 25 as described above, the output to the steering motor 8 is
The steering motor 8 is turned off, and depending on the direction of deviation of the guideline position with respect to the sensor 11, the steering motor 8 is rotationally driven to the left or right in a direction to correct the deviation.

That is, for example, in the above example, the detected value is "11" and the position weight of the center of the sensor 11 is "15", which means that the sensor 11, that is, the vehicle body 1, is shifted to the right in Fig. 7. , the steering motor 8 is rotated in a direction to move the vehicle body 1 to the left. If the vehicle body 1 is shifted to the left,
Of course, the steering motor 8 is rotated in the opposite direction.

The above process is shown in a flowchart as shown in Figure 10, and in this example, from "Start" to "Turn the steering wheel to the left" or "Turn the steering wheel to the right" and then return to "Storing the amount of received light in memory". One cycle time is approximately 30 milliseconds,
By repeating this process while driving, the unmanned vehicle can smoothly travel along the guideline even if there are obstacles such as dirt on the guideline or different reflections from the floor depending on the location.

In addition, at the point where the guideline is interrupted, the onboard computer 25 determines the interruption and stops the unmanned vehicle.
If the ROM 33 stores driving information after the guideline was interrupted (information on driving speed, steering angle, etc. stored as a function of the distance traveled), the unmanned vehicle will be able to move around the interrupted part of the guideline. The vehicle can also drive automatically based on the above driving information.

As is clear from the above explanation, the present invention changes the threshold value for guideline detection based on the increase/decrease in the total amount of light received by all the light receiving elements, so that the reflected light from the guideline is localized. Even if the contrast between the guideline and the guideline decreases due to temporary degradation due to dirt, etc., or the contrast with the guideline decreases due to changes in the light reflected from the floor surface due to differences in the material of the floor surface depending on the location, etc., the guideline cannot be detected accurately. This allows the unmanned vehicle to be guided well without deviating from the guideline. In addition, since the amount of light received by each light receiving element of the optical sensor is smoothed by averaging the amount of light received by adjacent light receiving elements, it is possible to smooth the amount of light received by each light receiving element of the optical sensor with the amount of light received by the adjacent light receiving element. Even if a strong reflected beam unexpectedly enters the light receiving element at a position other than the guideline due to the influence of a foreign object with a high can be induced well.

[Brief explanation of the drawing]

Fig. 1 is a schematic plan view showing an example of an unmanned vehicle, Fig. 2 is a plan view of an optical sensor, and Fig. 3 and 4
The figures are respectively a cross-sectional view taken along line A-A and line B-B in Figure 2, Figure 5 is a circuit diagram, and Figure 6 is a block diagram showing the connections of the computer, optical sensor, etc. on the unmanned vehicle. Fig. 7 is a schematic diagram showing the positional relationship between the phototransistor of the optical sensor and the guideline, Fig. 8 is a bar graph showing the actual amount of light received by each phototransistor and the value after smoothing, and Fig. 9 is a schematic diagram showing the positional relationship between the phototransistor of the optical sensor and the guideline. FIG. 10 is a schematic diagram showing position weights and binarized values for phototransistors, and is a flowchart of the processing steps. 1...Unmanned vehicle, 5...Steering device,
11...Optical sensor, 12...Guideline, 17...Phototransistor, F...Floor surface,
M...average value, T...threshold value.

Claims (1)

[Claims] 1 A guidance method for an unmanned vehicle that uses an optical sensor having multiple light-receiving elements to guide the vehicle while detecting guidelines on the floor, in which the amount of light received by each light-receiving element of the optical sensor is calculated by comparing the amount of light received by an adjacent light-receiving element. A guidance method for an unmanned vehicle characterized by smoothing by averaging the amount of light received by all light receiving elements, and changing the height of a threshold value for guideline detection based on fluctuations in the total amount of light received by all light receiving elements. .