JP2002048595A - Calibration method for sensor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a calibration method for sensor capable of improving the precision of the detection value obtained from a sensor without requiring a high cost. SOLUTION: A CPU 30 measures, every time when controlling an actuator 10 to set the rotating angle θ to be given to a potentiometer 20 to angles corresponding to a first to third inflection points P1, P2 and P3, a lower limit point Pmin and an upper limit point Pmax, the output P (voltage value) inputted from the potentiometer 20 in this state, and determines the error of the measured output P. The specified five points are connected through four straight parts to generate an approximate error characteristic line. The CPU 30 corrects the rotating angle θ on the basis of the data of the approximate error characteristic line Le in the determination of the rotating angle θ on the basis of the output P inputted from the potentiometer 20.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a sensor calibration method.

[0002]

2. Description of the Related Art There is a sensor which outputs a detection amount, for example, a voltage signal corresponding to a physical amount of a scientific phenomenon, for example, given a movement distance or a rotation angle given a physical amount. When such a sensor is used without calibration, the accuracy of the sensor itself must be improved because of the linearity of the input and output of the sensor and the variation of the output value, which increases the cost. In addition, when the absolute value of the physical quantity given to the sensor and the absolute value of the detected quantity output is required, the mounting accuracy of the sensor must be improved,
Extra parts and labor are required. Therefore, it is preferable to use the sensor after calibration.

[0003]

When the calibration of the sensor is performed, for example, when the calibration is performed at one point of the detected amount, it is possible to correct the variation in the absolute value of the sensor. It is not possible to correct the deviation of linearity. Therefore, it is difficult to improve the accuracy of the linearity of the sensor and the cost of the sensor. On the other hand, if the sensor is calibrated at many points of the detection amount, the apparent accuracy is improved, but it takes time to calibrate at many points, and the calibration of many points. In order to hold the application data, it is necessary to secure the capacity of the memory, which increases costs. The present invention seeks to solve such problems of the prior art, and the purpose thereof is to reduce the cost,
An object of the present invention is to provide a sensor calibration method capable of improving the accuracy of a detection value obtained from a sensor.

[0004]

According to the present invention, a physical quantity of a scientific phenomenon is given to output a detected quantity corresponding to the physical quantity, and the given physical quantity is included in the output detected quantity. An error characteristic line indicating a relationship with an error is a calibration method for a sensor having one or more inflection points, and is output when a physical quantity corresponding to an inflection point of the error characteristic line is given to the sensor. A first error measuring step of measuring an error included in the detected amount as an inflection point error, and the error characteristic line based on the inflection point error measured in the first error measuring step. An approximation error characteristic line generating step of generating an approximation error characteristic line approximated by a combination of curve portions;
And performing a calculation process to correct the error indicated by the approximate error characteristic line with respect to the detection value output from the sensor. Therefore, an inflection point error corresponding to the inflection point of the error characteristic line is measured, and based on the inflection point error, an approximation error characteristic line approximated by a combination of a plurality of linear portions or curved portions is generated. Since the error indicated by the error characteristic line is corrected, it is possible to correct the error with a minimum amount of measurement. The number of steps and cost required for the measurement can be reduced, and the accuracy of the detection value obtained from the sensor can be improved.

Further, according to the present invention, when a physical quantity of a scientific phenomenon is given, a detected quantity that is directly proportional to the physical quantity is output, and a difference between the given physical quantity and an error included in the output detected quantity is output. A method of calibrating a sensor having an error characteristic line indicating a relationship having one or more inflection points, wherein the detection amount is output when a physical quantity corresponding to an inflection point of the error characteristic line is given to the sensor. A first error measuring step of measuring an error contained in the inflection point error, and the error characteristic line is approximated by a combination of a plurality of linear portions based on the inflection point error measured in the first error measuring step. An approximate error characteristic line generating step of generating an approximate error characteristic line, and correcting an error indicated by the approximate error characteristic line with respect to the detection value output from the sensor. Characterized in that it comprises a correction step of performing arithmetic processing. Therefore, an inflection point error corresponding to the inflection point of the error characteristic line is measured, and based on the inflection point error, an approximation error characteristic line approximated by a combination of a plurality of straight line portions is generated. Since the error indicated by is corrected, it is possible to correct the error with the minimum measurement, without using a sensor that is more accurate and costly than before, and the man-hour required for calibration Cost can be reduced, and the accuracy of the detection value obtained from the sensor can be improved.

[0006]

Next, a first embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a schematic block diagram showing a configuration of a calibration device for performing a sensor calibration method according to the first embodiment;
FIG. 2 is a characteristic diagram showing a characteristic of a rotation angle versus an output value in a potentiometer as a sensor, FIG. 3 is a characteristic diagram showing an error characteristic showing a characteristic of a rotation angle versus an error output in a potentiometer, and FIG. 4 is a first embodiment. FIG. 13 is a characteristic diagram for describing generation of an approximate error characteristic line in the calibration method according to the embodiment.

Referring to FIG. 1, calibration apparatus 1
00 will be described. The calibration device 100 includes an actuator 10, a potentiometer 20
(Corresponding to a sensor in the claims), a CPU 30, a memory 40, and the like. The actuator 10 and the potentiometer 20 are provided in, for example, one device, for example, a robot device, and have a configuration in which the rotation angle of a member driven by the actuator 10 is detected by the potentiometer 20.

The actuator 10 is composed of, for example, a motor having a rotary drive shaft, and the CPU 30 controls the rotation of the rotary drive shaft. The potentiometer 20 has a rotation shaft (not shown) rotatably provided, an input terminal to which a predetermined voltage is input, and an output voltage obtained by dividing the predetermined voltage at a voltage division ratio directly proportional to the rotation angle of the rotation shaft. And a resistance application potentiometer having an output terminal. The rotation shaft is connected to a rotation drive shaft of the actuator 10 via a mechanism (not shown), and is configured to rotate by transmitting a rotation force from the rotation drive shaft. The CPU 30 is configured to input an output voltage from the output terminal of the potentiometer 20, measure the voltage value, and provide a control signal to the actuator 10 to control the rotation of the rotary drive shaft. The memory 40 stores data of an error characteristic line and an approximate error characteristic line described later generated by the CPU 30.

Next, referring to FIG.
The characteristic of 0 will be described. 2, the horizontal axis indicates the rotation angle θ of the rotation axis of the potentiometer 20. The vertical axis is the output value P indicating the output voltage of the potentiometer 20 in the form of percentage. That is, the rotation angle θ corresponds to a physical quantity in the claims, and the output value P corresponds to a detected value in the claims. In the figure, the symbol L indicates the theoretical value of the potentiometer 20, and represents a straight line in which the output value P is directly proportional to the rotation angle θ. Symbols E1, E
2 indicates a positive error (+ 3%) and a negative error (-3%) with respect to the theoretical value L, respectively. The symbol Lt is a representative characteristic of the actual potentiometer 20 and has a gentle S-shaped curve. This characteristic Lt is included in the range of the errors E1 and E2.

FIG. 3 shows an error characteristic line in which the error of the representative characteristic Lt with respect to the theoretical value L shown in FIG. 2 is plotted with respect to the rotation angle θ. In the figure, Le is an error characteristic line, E1,
E2 indicates a positive error (+ 3%) and a negative error (-3%). As shown in FIG. 3, the error characteristic line Le has three inflection points. That is, the rotation angle θ
Have inflection points corresponding to three angles of -50 degrees, 0 degrees and +50 degrees. These three inflection points will be referred to as first to third inflection points P1, P2, and P3 in ascending rotation angle order. Ideally, based on this error characteristic line Le, the output P
However, it is not realistic to measure and obtain all points of the error characteristic line Le because it takes too many steps. Therefore, in the present invention, the first to third inflection points P1, P1
2, an approximate error characteristic line of the error characteristic line Le is generated by P3 and four straight lines connecting five points of a lower limit value point Pmin and an upper limit value point Pmax corresponding to the lower limit value and the upper limit value of the rotation angle θ. I am trying to do it.

FIG. 4 shows the same error characteristic line Le as shown in FIG. 3, an approximate error characteristic line Lc obtained by approximating the error characteristic line with four straight lines, and an approximate error characteristic line Lc with respect to the error characteristic line Le. And a difference ΔL (= Lc−Le). In the figure, the approximate error characteristic line Lc is
A first straight line portion L1 connecting the lower limit point Pmin and the first inflection point P1, a second straight line portion L2 connecting the first inflection point P1 and the second inflection point P2, and a second inflection point P2 And the third inflection point P
3 and a fourth straight line portion L4 connecting the third inflection point P3 and the upper limit point Pmax. That is, in order to generate the approximate error characteristic line Lc, the error of the lower limit point Pmin (lower limit error), the error of the first inflection point P1 (first inflection point error), and the second
Error of inflection point P2 (second inflection point error), third inflection point P3
(The third inflection point error) and the error of the upper limit point Pmax (the upper limit error) may be measured, and the measured five errors may be sequentially connected by a straight line portion.

As shown in FIG. 1, the difference ΔL between the approximation error characteristic line Lc and the error characteristic line Le thus generated falls within the range of ± 0.7% in error output. I have.

The calibration operation of the calibration device 100 will be described with reference to FIGS. First, the CPU 30 controls the actuator 10 to change the rotation angle θ given to the potentiometer 20 to the lower limit point P
At this angle, the voltage value input from the potentiometer 20 is measured, and an error in the measured voltage value (detection value) is obtained. The error is an error with respect to the theoretical value L described in FIG. Similarly, the CPU 30
The rotation angle θ to be given to the potentiometer 20 by controlling the actuator 10 is set to the first to third inflection points P1, P2, P3
And an angle corresponding to the upper limit point Pmax, the output P input from the potentiometer 20 in that state is set.
(Voltage value) is measured, and an error of the measured output P is obtained. As a result, the five points (Pmin, P
1, P2, P3, Pmax) are specified. The processing for determining the error of these five points is the first and second claims.
This corresponds to an error measurement step. Then, the specified five points are connected by four straight line portions (L1 to L4) to generate the approximate error characteristic line Lc. This processing corresponds to the approximation error characteristic line step in the claims. The data of the approximate error characteristic line Lc is stored in the memory 40.

When the potentiometer 20 is actually used, the value of the output P input from the potentiometer 20 to the CPU 30 is corrected by performing an arithmetic process for correcting the value of the output P based on the data of the approximate error characteristic line Lc. It is possible to obtain a large rotation angle θ. This correction processing corresponds to a correction step in the claims.

According to the sensor calibration method of the present embodiment described above, the inflection point errors corresponding to the three inflection points of the error characteristic line, and the lower limit corresponding to the lower limit value and the upper limit value of the rotation angle θ. An error of five points between the value point and the upper limit point is measured, and an approximate error characteristic line approximated by a combination of four linear portions is generated based on the five errors, and an error indicated by the approximate error characteristic line is generated. Was corrected, so 5
The error can be corrected with a small number of measurements, eliminating the need for high-precision and costly sensors compared to the past, and requiring no memory for storing large amounts of data. The required man-hour and cost can be reduced, and the accuracy of the detection value obtained from the sensor can be improved.

FIG. 5 is a characteristic diagram for explaining generation of an approximate error characteristic line in the calibration method according to the second embodiment. The difference between the second embodiment and the first embodiment is that both the upper limit value error and the lower limit value error are assumed to be zero, and the upper limit value point Pmax and the lower limit value point Pmi
This is a point that two measurements of n are omitted. The other operations are substantially the same as those of the first embodiment, and the description thereof is omitted.

Therefore, in the calibration method of the second embodiment, the approximate error characteristic line Lc is the lower limit point Pmi assumed to be zero.
n and a first straight line portion L1 connecting the first inflection point P1, a second straight line portion L2 connecting the first inflection point P1 and the second inflection point P2, a second inflection point P2 and a third It is configured to include a third straight line portion L3 connecting the inflection point P3 and a fourth straight line portion L4 connecting the third inflection point P3 and the upper limit point Pmax assumed to be zero. That is, in order to generate the approximate error characteristic line Lc, the error at the first inflection point P1 (first inflection point error), the error at the second inflection point P2 (second inflection point error), and the third
What is necessary is just to measure three errors of the error of the inflection point P3 (third inflection point error).

Here, a description will be given of the fact that the offset amount is taken into account when generating the approximate error characteristic line Lc in order to correct an output error (offset error) caused by a sensor mounting error. FIG. 6 is a characteristic diagram showing a characteristic of a rotation angle versus an output value for explaining offset adjustment. FIG. 6 shows a representative characteristic Lt and a characteristic diagram Loff in a state where the sensor is actually attached (a state in which the rotation angle is 180 degrees is offset with respect to the output of 50%). First, the output P is measured at the position of the second inflection point P2 (corresponding to a rotation angle of 180 degrees), which is the central inflection point, and the measured value at this time is used as an offset reference of the mounting position of the sensor. The offset amount of the approximate error characteristic line Lc is determined so that the characteristic diagram Loff matches the representative characteristic Lt, that is, the output corresponding to the rotation angle of 180 degrees matches 50%. Next, measurement is performed at the positions of the first and third inflection points P1 and P3, and as described above, the approximate error characteristic line L
Generate c. That is, the approximate error characteristic line Lc is generated in consideration of the positive or negative offset amount in order to correct the output error caused by the sensor mounting error.

Returning to FIG. 5, the difference ΔL between the approximation error characteristic line Lc and the error characteristic line Le generated in this way falls within a range of ± 0.8% in error output. Thus, a practically satisfactory result is obtained without much difference from the case of the first embodiment. Therefore, according to the second embodiment, in addition to the operation and effect of the first embodiment, there is an advantage that the number of measurements can be reduced from five to three.

In each of the above-described embodiments, a potentiometer is used as an example of a sensor. Therefore, the physical quantity given to the sensor is a rotation angle, and the detection quantity output from the sensor is a voltage. However, the present invention is not limited to these embodiments, but can be applied to other types of sensors. For example, the physical quantity of the scientific phenomenon given to the sensor may be a physical or chemical physical quantity, for example, and the form of the detection quantity output from the sensor is not limited to voltage.

The relationship between the physical quantity given to the sensor and the detected quantity output from the sensor does not necessarily need to be directly proportional. Further, in the present embodiment, the approximation error characteristic line is configured by a combination of linear portions, but may be configured by a combination of curved portions.

[0022]

As described above, according to the present invention, when a physical quantity of a scientific phenomenon is given, a detected quantity corresponding to the physical quantity is output, and for the given physical quantity, the detected quantity output is An error characteristic line indicating a relationship with an included error is a method for calibrating a sensor having one or more inflection points, and is output when a physical quantity corresponding to an inflection point of the error characteristic line is given to the sensor. A first error measuring step of measuring an error included in the detected amount as an inflection point error, and dividing the error characteristic line into a plurality of linear portions based on the inflection point error measured in the first error measuring step. Or an approximation error characteristic line generating step of generating an approximation error characteristic line approximated by a combination of curved portions, and the approximation error characteristic line for the detection value output from the sensor. The error that is has a configuration comprising a correction step of performing arithmetic processing so as to correct. Therefore, an inflection point error corresponding to the inflection point of the error characteristic line is measured, and based on the inflection point error, an approximation error characteristic line approximated by a combination of a plurality of linear portions or curved portions is generated. Since the error indicated by the error characteristic line is corrected, it is possible to correct the error with a minimum amount of measurement. The number of steps and cost required for the measurement can be reduced, and the accuracy of the detection value obtained from the sensor can be improved.

Further, the present invention outputs a detected quantity which is directly proportional to the physical quantity by giving a physical quantity of a scientific phenomenon, and calculates a difference between the given physical quantity and an error included in the output detected quantity. A method of calibrating a sensor having an error characteristic line indicating a relationship having one or more inflection points, wherein the detection amount is output when a physical quantity corresponding to an inflection point of the error characteristic line is given to the sensor. A first error measuring step of measuring an error contained in the inflection point error, and the error characteristic line is approximated by a combination of a plurality of linear portions based on the inflection point error measured in the first error measuring step. An approximate error characteristic line generating step of generating an approximate error characteristic line, and correcting an error indicated by the approximate error characteristic line with respect to the detection value output from the sensor. And configured to include a correction step of performing arithmetic processing. Therefore, an inflection point error corresponding to the inflection point of the error characteristic line is measured, and based on the inflection point error, an approximation error characteristic line approximated by a combination of a plurality of straight line portions is generated. Since the error indicated by is corrected, it is possible to correct the error with the minimum measurement, without using a sensor that is more accurate and costly than before, and the man-hour required for calibration Cost can be reduced, and the accuracy of the detection value obtained from the sensor can be improved.

[Brief description of the drawings]

FIG. 1 is a schematic block diagram illustrating a configuration of a calibration device for performing a sensor calibration method according to a first embodiment.

FIG. 2 is a characteristic diagram showing a characteristic of a rotation angle versus an output value in a potentiometer as a sensor.

FIG. 3 is a characteristic diagram showing an error characteristic indicating a characteristic of a rotation angle versus an error output in the potentiometer.

FIG. 4 is a characteristic diagram for explaining generation of an approximate error characteristic line in the calibration method according to the first embodiment;

FIG. 5 is a characteristic diagram for explaining generation of an approximate error characteristic line in the calibration method according to the second embodiment.

FIG. 6 is a characteristic diagram illustrating a characteristic of a rotation angle versus an output value for explaining offset adjustment.

[Explanation of symbols]

100 Calibration device, 10 Actuator, 20 Potentiometer, 30 CPU
40: memory, Le: error characteristic line, Lc approximation error characteristic line, Pe: error output, θ: rotation angle.

Claims (8)

[Claims] 1. A physical quantity of a scientific phenomenon is given to output a detection quantity corresponding to the physical quantity, and an error indicating a relationship between the given physical quantity and an error included in the output detection quantity. A calibration method for a sensor having a characteristic line having one or more inflection points, wherein an error included in the detection amount output when a physical quantity corresponding to an inflection point of the error characteristic line is given to the sensor. A first error measuring step of measuring the error characteristic line as an inflection point error, and an approximation in which the error characteristic line is approximated by a combination of a plurality of linear portions or curved portions based on the inflection point error measured in the first error measuring step. An approximate error characteristic line generating step of generating an error characteristic line; and correcting an error indicated by the approximate error characteristic line with respect to the detection value output from the sensor. A calibration method for a sensor, comprising: a correction step of performing an arithmetic process. 2. A physical quantity of a scientific phenomenon is given to output a detection quantity in direct proportion to the physical quantity, and an error indicating a relationship between the given physical quantity and an error included in the output detection quantity. A calibration method for a sensor having a characteristic line having one or more inflection points, wherein an error included in the detection amount output when a physical quantity corresponding to an inflection point of the error characteristic line is given to the sensor. A first error measuring step of measuring the error characteristic line as an inflection point error; and an approximation error characteristic line obtained by approximating the error characteristic line with a combination of a plurality of linear portions based on the inflection point error measured in the first error measuring step. Generating an approximate error characteristic line, and performing an arithmetic process on the detection value output from the sensor so as to correct an error indicated by the approximate error characteristic line. A calibration method for a sensor, comprising: 3. A second error for measuring an upper limit error and a lower limit error, which are errors included in a detection amount output when an upper limit value and a lower limit value of the physical quantity given to the sensor are given to the sensor. The method further comprises a measuring step, wherein the generation of the approximate error characteristic line by the approximate error characteristic line generation step is performed based on the inflection point error, the upper limit value error, and the lower limit value error. 2. The sensor calibration method according to 2. 4. An inflection point of the error characteristic line is composed of three inflection points of a first, second, and third inflection points in ascending order of the physical quantity, and the first, second, and third inflection points are provided. When the inflection point errors corresponding to the points are defined as first, second, and third inflection point errors, the approximation error characteristic line is a first connecting the lower limit value error and the first inflection point error. A straight line portion, a second straight line portion connecting the first inflection point error and the second inflection point, a third straight line portion connecting the second inflection point error and the third inflection point error, 4. The sensor calibration method according to claim 3, further comprising a third inflection point error and a fourth straight line portion connecting the upper limit error. 5. The execution of the second error measurement step by assuming that both the upper limit error and the lower limit error measured in the second error measurement step are zero. 3. The sensor calibration method according to 3. 6. The inflection point of the error characteristic line is composed of three inflection points of first, second, and third inflection points in ascending order of the physical quantity, and the first, second, and third inflection points are provided. When the inflection point errors corresponding to the points are defined as first, second, and third inflection point errors, the approximate error characteristic line includes the lower limit value error assumed to be zero and the first inflection point error. , A second straight line portion connecting the first inflection point error and the second inflection point, and a third straight line portion connecting the second inflection point error and the third inflection point error. The sensor calibration according to claim 5, further comprising a straight line portion and a fourth straight line portion connecting the third inflection point error and the upper limit error assumed to be zero. Method. 7. The sensor according to claim 1, wherein the sensor comprises a rotatable rotation shaft, an input terminal to which a predetermined voltage is input, and an output voltage obtained by dividing the predetermined voltage at a voltage division ratio directly proportional to a rotation angle of the rotation shaft. An output terminal to be output, the physical quantity is the rotation angle,
3. The sensor calibration method according to claim 2, wherein the detection amount is the output voltage. 8. The sensor calibration method according to claim 7, wherein said potentiometer is a resistance application type potentiometer.