JPH0434386A - Method for correcting change in detection quantity of light of photoelectric sensor - Google Patents

Abstract

PURPOSE:To properly compensate the change in the detection quantity of light caused by the temp. characteristics of a light emitting element or the surface contamination of a lens, in a method for correcting the detection quantity of light of a photoelectric sensor, by determining the correction value of the detection quantity of light with due regard to a change in the distribution of measured values in addition to the mean value of the measured values of the detection quantity of light. CONSTITUTION:When the detection quantity of light of a photodetector is lowered by the surface contamination of the light lenses of a light emitting element 3 and a photodetector 5 in order to correct a change in the detection quantity of light of the photodetector 5, the gain of a variable gain amplifier 11 is made large. That is, the main value of the measured values of the detection quantity of light and the distribution range of said values become input and one of outputs after data processing becomes the gain for the amplification of the signal. When the distribution range of the measured values becomes larger than initial data or the previous one, the measuring number of times of the calculation of the mean value and the distribution range is increased in order to enhance the reliability as data and to prevent the generation of an oscillating phenomenon. Contrarily, when the distribution range of the measured values becomes smaller than the previous one, the measuring number of times of the calculation of the mean value and the distribution range is reduced.

Description

Detailed Description of the Invention (Industrial Field of Application) The present invention relates to a method for correcting changes in the amount of light received by a photoelectric sensor such as a photocoupler, and particularly to a method for correcting changes in light emitting efficiency of a light emitting element, clouding or dirt on a lens surface, etc. The present invention relates to a method for correcting changes in the amount of received light that are caused by changes in the amount of received light and prevents malfunctions.

(Prior art) A light emitting element such as a light emitting diode (LED) and a light receiving element such as a photodiode (PD) are used, and the light from the light emitting element is sensed by the light receiving element, thereby allowing goods to be transported in a non-contact manner. Photoelectric sensors configured to detect the movement or presence of objects such as objects are well known and are used in a wide variety of fields.

Problems with photoelectric sensors include changes in the luminous efficiency of the light emitting element due to temperature changes, clouding or dust on the lens surfaces of the light emitting element and light receiving element, and even vibrations. Due to misalignment of the optical axes between the light emitting element and the light receiving element, the amount of light detected by the light receiving element, that is, the amount of light received, may change, which may change the magnitude of the light receiving signal input to the light receiving circuit. , this may cause malfunction.

In view of the above-mentioned problems, it has already been proposed to incorporate an APC circuit using a monitor photodiode in addition to the light-receiving element, thereby compensating for changes in the luminous efficiency of the light-emitting element due to temperature changes, etc. In addition, the amount of light received by the light receiving element is measured an arbitrary number of times to find the average value, and by comparing this with the previous or initial data, a correction value for the amount of received light is found, and from this, the amount of light emitted or the gain on the light receiving side is determined. Feedback compensation has been proposed.

(Problem to be solved by the invention) An APC circuit using a monitor photodiode is effective in compensating the temperature of the light emitting element, but it is effective against changes in the amount of received light caused by fogging, dirt, etc. on the lens surface. We are unable to respond and cannot provide any compensation for this.

In devices that perform feedback control based on the average value of the amount of light received, in addition to temperature compensation of the light emitting element, fogging of the lens surface,
Changes in the amount of received light due to dirt, etc. will be compensated for, but if the change in received light amount is corrected by simply comparing average values, it will not be sufficient to compensate for variations in the measured value of the amount of received light or sudden, one-off changes. Therefore, appropriate correction for changes in the amount of received light is not necessarily performed.

Further, depending on the number of measurements to obtain the average value, the control system may oscillate due to variations in the measured values, and it may be difficult to perform stable correction.

The present invention was made by focusing on the above-mentioned problems in the conventional correction of changes in the amount of received light. Changes in the amount of received light can be compensated for accurately and reliably, and the control system will not oscillate due to variations in measured values, making stable corrections possible.Furthermore, even if there are sudden, one-off changes,
It is an object of the present invention to provide a method for correcting changes in the amount of received light, which performs highly reliable correction of changes in the amount of received light while minimizing interference with correction of the original changes according to the frequency of such changes.

(Means for Solving the Problems) The purpose of the present invention is to measure the amount of light received by the light-receiving element of the photoelectric sensor, and to measure the average value and distribution based on a predetermined number of measurements. The detection is performed every number of times, and the average value of the current measurement value is compared with the average value of the previous measurement value, and the distribution of the current measurement value and the distribution of the measurement value described above are compared, and the change in the amount of received light is determined by comparing both. This is achieved by a method for correcting changes in the amount of light received by a photoelectric sensor, which is characterized by determining a correction value for the amount of light received to compensate for the change in the amount of light received by a photoelectric sensor.

Furthermore, in the method for correcting changes in the amount of received light according to the present invention, when the measured value of the amount of light received by the light receiving element is significantly different from the previous measured value, a total of 791 (a) is included in the calculation and distribution of the average value. Or, reduce the weight given to the calculation and distribution frequency of the average value of the measured value, so that as the number of times a measured value that is significantly different from the previous measured value increases, the measured value becomes the average value. The weight given to the calculation and distribution frequency may be increased.

Further, in the method for correcting changes in the amount of light received by a photoelectric sensor according to the present invention, when the distribution range of the measured values is expanded compared to before, the number of measurements to detect the average value and distribution is increased, and the distribution range is expanded. When the size is reduced compared to before, the number of measurements for detecting the average value and distribution may be reduced.

(Function) According to the method for correcting changes in the amount of light received by a photoelectric sensor according to the present invention, the correction value for the amount of light received is determined by taking into consideration changes in the distribution of the measured values in addition to changes in the average value of the measured values of the amount of light received. Therefore, changes in the amount of received light due to temperature characteristics of the light projecting element, dirt on the lens surface, etc. can be appropriately compensated for.

In addition, when the measured value of the amount of light received differs greatly from the previous measured value, that is, when a sudden and one-off change occurs,
The distribution frequency of the average value and measured value is updated according to the frequency of this, and even if a sudden and one-off change occurs, the hindrance to correction for the original change is minimized, and this is also reliable. This allows for highly accurate received light amount correction.

In addition, the number of measurements to calculate the average value is determined according to changes in the distribution range of the measured values, which prevents the control system from oscillating due to variations in the measured values and ensures stable correction. Become.

(Example) The present invention will be described in detail below with reference to the accompanying drawings.

FIG. 1 is a circuit block diagram of a photoelectric sensor and a received light amount change control device used to implement a photoelectric amount change correction method for a photoelectric sensor according to the present invention.

In FIG. 1, reference numeral 1 indicates a photoelectric sensor, and the photoelectric sensor 1 is composed of a light projecting element 3 and a light receiving element 5. The light projecting element 3 is composed of a pulse modulation type LED or the like, and is driven by a light projecting circuit 7 to emit light. The light receiving element 5 is constituted by a photodiode or the like, receives light from the light projecting element 3, converts the light into current, and outputs a current signal.

The current signal from the light receiving element 5 is given to the I/D converter 9 and converted into a voltage value. Since this voltage signal is generally weak, it is amplified by the amplifier 11. The amplifier 11 is configured as a variable gain amplifier whose gain can be variably set for gain adjustment for correcting changes in the amount of received light.

Since the light that the light-receiving element 5 is sensitive to is a pulse-modulated wave from the light-emitting element 3, the voltage signal from the variable gain setter 11 is applied to a peak-hold sample-hold circuit 13, which converts it into a DC component. The signal is then converted into a digital signal by the A/D converter 15 and input to the CPU 17.

Further, the voltage signal output from the variable gain amplifier 11 is given to a signal processing circuit 19. The signal processing circuit 19 may be similar to a signal processing circuit in a general photoelectric sensor, and performs operations such as determining whether light is incident or not using a comparator and removing noise components using a noise filter. It outputs a signal. This operation signal 19 is also input manually to the CPU 17 .

The CPU 17 may have a general structure, and may have an EEP
The operation system stored in ROM21,
Changes in the amount of light received by the photoelectric sensor 1 are corrected according to the program. In addition, the CPU 17 is the A/D converter 1
5 and an EEPROM 21 may be used.

The CPU 17 outputs a gain setting signal to the variable gain setter 11 to correct changes in the amount of light received by the photoelectric sensor 1, and also outputs an alarm signal and manually inputs a trigger signal.

In order to correct changes in the amount of light received by the light receiving element 5, it is necessary to sequentially capture the measured value of the amount of light received, and also for correcting the amount of received light,
In other words, a measured value of the amount of light received in a state that can serve as a reference is required as a measured value of the amount of light received that serves as initial data for sensitivity correction. Depending on whether the photoelectric sensor is of a reflected light detection type or a transmitted light detection type, the conditions that can be used as a reference include the light-on condition in which the amount of received light is stable when light enters, as shown in Figure 2. As shown in Fig. 3, there are two ways of receiving light: the amount of light received when light is blocked, or the stable dark-on state, so the amount of light received in this on state is determined as = 1 measurement, assuming that the state where light is stably received is the on state. I'll do what I do. To be more precise, for example, in the case of light-on, the maximum amount of light received when it is on is taken, and in the case of dark-on, it is processed by the hardware configuration or software to take the minimum amount of light received when it is off. For example, changes in the amount of light during transition from off to on or from on to off will not be erroneously taken in as data for sensitivity correction. Furthermore, if the timing of taking in the measured value of the amount of received light cannot be applied by an external gate, if it is difficult, a tri-input signal may be input to the CPU 17, and the gate can be applied accordingly.

In order to see historical changes in the amount of light received by the light receiving element 5 of the photoelectric sensor 1, it is necessary to determine the changes in the average value of the measured values a predetermined number of times as well as the differences in changes due to variations in the measured values. . An example of data processing required for this purpose will be described next.

First, it is necessary to import initial data to serve as a comparison standard, so for example, immediately after setting the photoelectric sensor 1,
Alternatively, after inputting some trigger, the measured value of the amount of received light is captured a predetermined number of times. That is, for example, the measured value of the amount of light received is acquired when the article is passed through the conveyance line until the photoelectric sensor 1 is turned on n times.

Then, the CPU 17 calculates the average value and the variation width, that is, the distribution range, of the measured values of the amount of light received n times, and stores them in the internal memory of the CPU 17 as initial data.

Next, the actual operation begins, and similarly, the measured values of the amount of received light are taken sequentially n times, and when the measured values of the amount of received light are taken n times, the same as the initial data described above, the received light amount measurements are performed n times. From the value, calculate the average value and distribution range, compare this with the initial data, determine the received light amount correction value as described later, and output the gain setting signal to the variable gain amplifier 11 based on this. do. Then, the received light amount measurement values are sequentially acquired over the next n times, the average value and distribution range of these are calculated, and this is compared with the average value and distribution range calculated - times before. Calculating the received light amount correction value is repeated.

Next, a processing algorithm of a method for correcting changes in the amount of light received by the present invention for correcting changes in the amount of light received by the photoelectric sensor will be described with reference to a first embodiment.

The light emitting element 3 is used to correct changes in the amount of light received by the light receiving element 5.
When the amount of light received decreases due to dirt on the lens surface of the light receiving element 5, the gain of the variable gain amplifier 3 shown in FIG. 1 is increased. In other words, the average value and distribution range of the measured value of the amount of received light are determined manually, and one of the outputs after data processing becomes a gain for signal amplification.

In addition, if the distribution range of measured values is larger than the initial data or previous data, the reliability of the data will be increased,
In order to prevent the occurrence of an oscillation phenomenon, a situation arises in which it is preferable to increase the number of measurements for calculating the average value and distribution range. On the other hand, if the distribution range of measured values is smaller than before, it is better to reduce the number of measurements to calculate the average value and distribution range to speed up the response speed for correcting the amount of received light and increase efficiency. There is also. In such a case, the output after data processing is the average value and the number of measurements for calculating the distribution range.

Here, we have described the average value and dispersion of the measured value of the amount of received light as human data, but some of the data here also includes, for example, a foreign object that accidentally passes momentarily along the sensor's detection axis and causes light to emit light. As in the case where the axis is interrupted, there are cases where data that should originally be removed is included in the data for average value calculation, etc. However, as the number of occurrences (number of appearances) of these data increases, it becomes necessary to separately prepare processing algorithms for them in order to increase the reliability of the data as processed data. Hereinafter, this processing will be referred to as exception processing.

Therefore, there are three processing algorithms for correcting changes in the amount of received light: those related to sensitivity correction, those related to adjustment of the number of measurements for calculating the average value and distribution range, and those related to exception handling. Each processing algorithm will be explained below.

◎Processing related to sensitivity correction As shown in Figure 4, if the fluctuations in the measured values of the amount of light received over n times are normally distributed, the data of the current measured value and the data of the previous measured value are The relationship is as shown in Figure 5 when the current measured value does not fall within the distribution range of the previous measured value, and when it does fall within the distribution range as shown in Figure 6. be.

As shown in Figure 5 (if the distribution range of the current =I measurement value does not completely belong to the previous measurement value, the average value of the current measurement value will be at the level of the average value of the previous measurement value). It is only necessary to adjust the gain so that the gain returns to .The calculation formula at this time is G for the current gain, G for the gain after correction.

, A is the average value of the previous measurement values. -2, the average value of the current measurement value is A. Then, it is shown by the following formula.

G, = (A,,-□/A,)Gn-1-(1) As shown in Figure 6, the distribution range of the current measurement value completely belongs to the distribution range of the previous measurement value. In this case, even if the average value itself of the measured values changes, it can be considered that this is simply a variation within the allowable variation. This is done to make the size smaller than in the case where there is. For example, the fifth
When the gain adjustment amount is set to approximately half that of the case shown in the figure, the calculation formula for calculating the gain is shown by the following formula.

G, =0.5 (A, -t /A-) G--t
・”(2> Therefore, it is determined whether the change in the distribution range of the received light amount measurement value at a predetermined number of times is a pattern as shown in FIG. 5 or a pattern as shown in FIG. 6. The gain correction amount is set, and the amount of received light is corrected in a feedback manner.

Furthermore, if further correction is not possible due to limitations such as the variable gain setting width of the variable gain setter 11, the CP
U17 only needs to output an alarm signal, which allows it to have a diagnostic function.

◎Processing algorithm for adjusting the number of measurements for average value calculation and distribution range calculation For example, suppose that the number of measurements can be variably set from a minimum of 20 times to a maximum of 50 times, and at the time of initial data acquisition, this is set to 35 times. It is assumed that

If the distribution range of the current measurement value is the same as the previous one, the open side number of 35 times will not be changed, and the closer the distribution range is to zero, the closer the measurement number will be to the lowest value of 20 times. conduct. On the other hand, if the distribution range becomes larger, for example, if the distribution range of measured values is three times that of the previous one, set the number of measurements to be the maximum of 50.
Processing is performed so that the larger the distribution range, the greater the number of measurements. The calculation formula for determining the number of measurements is given by the following formula, where the distribution range of the previous measurement value is Rn-1, the distribution range of the current measurement value is R1, and the number of measurements after adjustment is n.

If R,<R,-1, n = (R,,/R,,-1) 15+20
-(3) When R7≧R,1, n = (R,, /Rn-+) 7.5 +27.5
．． -(4) However, since the number of measurements n is only a positive integer, in the above calculation formula, if n is a rational number,
It is sufficient if the decimal point is rounded off.

Furthermore, if the distribution variation of the received light amount needle #JM increases beyond a predetermined value, an alarm may be output, so that the user can be notified of large fluctuations in the amount of light received by the photoelectric sensor 1. become.

Exception Processing Algorithm Here, as an example of exception processing, a measurement value whose distribution range R, , of the current measurement value exceeds three times the previous distribution range R, , -1 is subjected to exception processing. This exception handling is ignored as the number of occurrences of the measurement value subject to the above-mentioned exception handling is small, and the larger the number of occurrences of the measurement value, the greater the weighting as correction data.

In this case, in detail, when the measured value of the amount of light received by the light receiving element 5 is significantly different from the previous measured value, the measured value is not included in the calculation and distribution of the average value, or the measured value is The weight given to the calculation of the average value and the distribution frequency is reduced, and as the number of times a measurement value that is significantly different from the previous measurement value increases, the weight given to the calculation of the average value and the distribution frequency is increased. Increase weight.

Next, as the number of occurrences of the measurement value subject to exception processing increases, calculations are performed using that value, and as the number of occurrences decreases, it is processed to be smaller. A calculation example of how to change the weighting when the (Y) average value of the measurement value of the exception is later added to the overall measurement data is shown below.In addition, in this calculation example, Assume that there are measured values E of exceptional objects.

[In the case of E > A-, -1+3 Rn-1] Ae=
[(E+ +E2 +-z*)/k(A,-+ +3
R, -t )l (k/n)1+ (A,, +3R,
i ) ”'(5) ΣE = A e dispute k Φ
(k/n) −(f3),', An=
(ΣE+ΣM)/(k (k/n)+(n-k)!
...(7) However, ΣE: Sum of measured values other than exceptions In this case, the minimum value of measured values within +R=A e -R=3R, [If E<A, -13Rn-+] Ae=(An-+ 3Rn-t) [((A7-
1-3Rn-! )-(E□+E2+-Ek)/k)
(k/n)] ...(8)ΣE=Ae
-k ・(k/n) -(9),', An
= (ΣE+ΣM)/fk (k/n)+(n-k)
)...(10) At this time, +R=3R
, the maximum value of the measured value within -R=Ae (5) and (8), the average value of the deviation of the measured value E of the exception from the previous measured value distribution range is multiplied by (k/a). As a result, as the number of occurrences k of the measurement value of the exception object increases, the calculation is performed so that the value approaches the value of the measurement value of the exception.

Note that if it is desired to change the degree of weighting on the number of occurrences k of the measurement value E of the exception, processing such as raising (k/n) to a power may be performed.

Equations (6) and (9) calculate the total sum of the measured values E of the exception, but by multiplying the number k by (k/n), the measured value E of the exception The larger the number k of data, the closer it becomes to the actual measured value, and the smaller the number, the higher the probability that it will be ignored (disregarded).

Equations (7) and (0) represent the weighted average values of all measured values after processing the measured value E of the exception in the above processing.

It should be noted that if the measured value of the exception object appears more than a predetermined value, an alarm may be output, thereby making it possible to notify the user of a change in the external environment.

Next, a case will be described in which the process of correcting the change in the amount of received light as described above is performed using fuzzy inference.

By using fuzzy inference, the processing algorithm system becomes simple and easy to understand, it becomes possible to universally incorporate processing suitable for the situation, and responsiveness can be improved.

The processing algorithm for correcting changes in the amount of received light using fuzzy inference will also be explained separately into sensitivity correction, measurement frequency adjustment, and exception handling, similar to the above-described embodiment.

◎Fuzzy inference algorithm related to sensitivity correction Examples of fuzzy inference rules related to sensitivity correction are shown in (10) to (15).
) is shown in the formula.

The input requirement for this fuzzy inference is the average value A of the measured values of the amount of received light. This is data for determining whether or not all of the measured values of the amount of light received this time are all included within the dispersion set (distribution range) of the previous measured values, and the output is the gain change rate ΔG.

IfA,=NL and An±R,<A,-,±R,,-.

then ΔG=PL ・(11) If A,=
NL and An ±R, (:A,, l-1±Rn-1t
hen ΔG=PM ・(12) If A,=
ZR then ΔG−ZR・ (13) If A,
,=PL nd An ±R, Ren A, -t ±R, -1the
n ΔG=NM ・ (14) If A,,=
PL and An ±R,,l<A,,-,±R0-1t
hen ΔG=NL ・(15) (11) and ([
5) In the fuzzy inference rule shown in equation 5, since the current measured value of the received light amount exceeds the variation range of the previous measured value of the received light amount, the gain is set so that the average value returns to the previous value. Feedback control is performed by adjustment, and (12
In the fuzzy inference rules shown in equations ) and (14), the rate of change in gain is made smaller because the measured value of the amount of received light is all within the variation range of the measured value of the amount of received light last time. , performs a little feedback control.

For example, in the case of the fuzzy inference rule of equation (11),
[If the average value has decreased considerably and the current measurement data exceeds the range of dispersion of the previous measurement data, increase the gain considerably.''

FIG. 7 shows the manual membership function of the average value A1 of the measured values of the amount of received light, and FIG. 8 shows the output membership function of the gain change rate ΔG. For example, when processing is performed using an 8-bit CPU, the minimum value (N
L) is 1, the maximum value (PL) is 256, and the value when there is almost no change in the average value (ZR) is the previous average value A. -1.

The membership function of the gain change rate ΔG is defined as NL, ZR as the value of the change rate with respect to the current gain so that even if the average value of the measured values changes, it will reliably return to the average value of the previous measured value after correction. , PL.

◎Fuzzy inference algorithm for adjusting the number of measurements An example of a fuzzy inference rule for adjusting the number of measurements to calculate the average value and distribution range of the measured value is shown in (16) to (■
8) It is shown in the formula.

If R1, /R,,,=NL then Δn = NL ・= (16) I
f R, /R,, = ZR then Δn = Z R = (17) If R,
/R, 1, = PL then Δn = PL = (18) The input requirements for this fuzzy inference rule are the distribution range R of the current measurement value relative to the distribution range R1-1 of the previous measurement value,
The output is the amount of change Δn as the adjustment frequency of the number of measurements. This indicates how many times to increase or decrease the number of measurements from the current number of measurements.

For example, in the case of the fuzzy inference rule shown in equation (16), the command is ``If the distribution range of the current measurement data is considerably smaller than that of the previous one, reduce the number of measurements considerably'', and (17 Equation (18) means that the number of measurements should be increased considerably because the change in the distribution range is quite large. In other words, the fuzzy inference rule is set so that the number of measurements increases as the distribution range becomes larger.

FIG. 9 shows the input membership functions of the distribution range ratios R, /R, . The minimum value (NL) of the distribution range ratio becomes 0 from R7=0, and there is almost no change (Z R)
becomes 1. Further, the maximum value (PL) is set to, for example, twice the previous distribution range Rn-1. Note that this maximum value may be set to an optimum value depending on the application rate. Figure 10 shows the adjustment degree Δn of the number of measurements.
shows an output membership function that shows . Regarding the adjustment degree Δn, in this case, NL=-15, ZR value 0,
PL is set to 15.

◎Fuzzy inference algorithm for exception handling An example of fuzzy inference rules for sensitivity correction in exception handling (19
) to (23).

The input requirements for this fuzzy inference are the average value Ae of measured values that are considered to be exceptions, such as the average value Ae of measured values that exceed twice the distribution range of the previous measured value, and the occurrence of the measured value that is the exception. The output is the gain change rate ΔG2. This gain change rate ΔG2 is determined after the gain adjustment by the sensitivity correction described above. Therefore, in this case, it is assumed that the gain change rate ΔG of the sensitivity correction described above does not include a correction component due to the measured value of this exception.

If Ae=PL and k=PLthen
ΔG2=NL − (19) If Ae=PL
and k=NLthen ΔG2=ZR・(
20) If Ae=ZR then ΔG2=ZR・(21,) If Ae=
NL and k=NL then ΔG2=ZR
−(22) If Ae=NL and k=PLthen
ΔG2=PL ・(23) In this fuzzy inference rule, the larger the average value Ae of the measured values of the exception, the smaller the gain (corrected to be), but at that time,
The smaller the number k of appearances of the measurement value of this exception, the smaller the magnitude of the correction. In other words, the smaller the number of appearances k of the measurement value of the exception, the smaller the influence of the measurement value of the exception. For example, (19)
In the case of the fuzzy inference rule shown in the formula, the command is ``If the average value of the measured values of the exception is quite large and the number of occurrences thereof is also quite large, reduce the gain by a considerable amount.''

This fuzzy reasoning makes it possible to reliably correct the amount of light even if there is a sudden abnormal measurement value.

Fig. 11 shows the input membership function of the average value Ae of the measured values of the exception, Fig. 12 shows the manual membership function of the number of occurrences of the measured value of the exception, and Fig. 13 shows the output of the gain correction factor ΔG2. The membership functions are shown respectively. Regarding the input membership function shown in FIG. 11 and the output membership function shown in FIG. 14,
This is the same as the input membership function and output membership function in the fuzzy inference algorithm for sensitivity correction described above. The minimum value (NL) is the case where the measured value of the exception does not appear at all in the number of occurrences k,
This is zero. The maximum value (PL) is the maximum number of measurements at that time, and its intermediate value is ZR. If it is desired to change the weighting regarding this number of times, the ZRO value may be manipulated.

By combining the fuzzy inference algorithm related to exception handling and the fuzzy inference algorithm related to sensitivity correction, fuzzy inference can be used to calculate, for example, ``There is a considerable change in the amount of light,
This may be done according to fuzzy reasoning such as "Since there is a slight possibility that the measured value is an exception, the sensitivity should be changed slightly larger."

As described above, by setting the fuzzy inference rules and membership functions, it becomes possible to reliably and stably correct light amount changes using a simple algorithm.

(Effects of the Invention) As can be understood from the above explanation, according to the method for correcting changes in the amount of light received by a photoelectric sensor according to the present invention, changes in the distribution of measured values are taken into account in addition to changes in the average value of the measured values of the amount of received light. Since the received light amount correction value is determined in this manner, changes in the received light amount due to temperature characteristics of the light projecting element, dirt on the lens surface, etc. can be appropriately compensated for.

In addition, when the measured value of the amount of received light differs greatly from the previous measured value, that is, when a sudden, one-off change occurs, the average value and the distribution frequency of the measured value are updated according to the frequency of this change. Even if a rapid, one-off change occurs, the interference with correction for the original change is minimized, and highly reliable correction of the amount of received light can be performed even in response to such a sudden change. In addition, the number of measurements to calculate the average value is determined according to changes in the distribution range of the measured values, which prevents the control system from oscillating due to variations in the measured values and ensures stable correction. Become.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing an embodiment of a photoelectric sensor and a control device for correcting changes in received light amount used to implement the method for correcting changes in received light amount of a photoelectric sensor according to the present invention, and FIG. 2 is a block diagram of a light-on type photoelectric sensor. Figure 3 is a graph showing changes in the amount of light received. Figure 3 is a graph showing changes in the amount of light received by a dark-on type photoelectric sensor. Figure 4 is a graph showing the average value of the measured value of the amount of light received a predetermined number of times and its distribution range. Figures 5 and 6 are graphs showing the difference between the distribution range of the current measurement value and the distribution range of the previous measurement value, respectively, and Figure 7 is the average of the measurement values, which is an input requirement in fuzzy inference regarding sensitivity correction. Figure 8 is a graph showing the output membership function of the gain change rate, which is its output, and Figure 9 is the input J requirement in fuzzy inference regarding adjustment of the number of measurements. A total of 11f1 (a graph showing the human membership function for the distribution range of m, Figure 10 is the output of the measurement frequency adjustment degree), a graph showing the j membership function, Figure 1.1 is related to exception handling. A graph showing the human membership function of the average value of the measured value of the exception object, which is the human factor in fuzzy inference. Figure 12 is a graph showing the input membership function of the number of occurrences of the exceptional measurement value. Figure 13. is a graph showing the output membership function of the gain change rate which is its output. 1... Photoelectric sensor 3... Light emitting element 5... Light receiving element 11... Variable gain amplifier 17... CPU 19... Signal processing circuit

Claims (1)

[Claims] 1. Measure the amount of light received by the light-receiving element of the photoelectric sensor, detect the average value and distribution based on a predetermined number of measurements, and compare the average value of the current measurement value and the previous measurement value. The method is characterized in that the average value of the measured values is compared, and the distribution of the current measured value is compared with the distribution of the measured values, and a received light amount correction value is determined based on these comparisons to compensate for changes in the received light amount. A method for correcting changes in the amount of light received by a photoelectric sensor. 2. When the measured value of the amount of light received by the light receiving element is significantly different from the previous measured value, the measured value is not included in the calculation and distribution of the average value, or the measured value is not included in the calculation and distribution of the average value. It is characterized by reducing the weight given to the frequency, and increasing the weight given to the calculation of the average value and the distribution frequency according to the increase in the number of times a measurement value that is significantly different from the previous measurement value is measured. 2. The method for correcting changes in the amount of light received by a photoelectric sensor according to claim 1. 3. When the distribution range of the measured values is expanded compared to before, increasing the number of measurements to detect the average value and distribution;
3. The method of correcting changes in the amount of light received by a photoelectric sensor according to claim 2, wherein when the distribution range is reduced compared to before, the number of measurements for detecting the average value and distribution is reduced.