JP3623592B2 - Temperature compensation method for force sensor - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a temperature correction method in a force sensor used in the field of measuring force or torque using a strain gauge. The present invention is applied to, for example, a force sensor used in a robot or the like for use in stress measurement of building components or force control.
[0002]
[Prior art]
A force sensor combines a mechanical elastic body using metal with a detector of a type that converts a stress such as a strain gauge into a change in electrical resistance value (hereinafter simply referred to as “resistance value”). It is a sensor that detects force or torque acting on the body as a change in resistance value. In general, in a detector using a conversion element that converts a measurement amount into a minute change in resistance value, the conversion element is used to extract a change in resistance value corresponding to the measurement amount (stress in a strain gauge) as a voltage value. A method using a Wheatstone bridge that incorporates is widely adopted.
[0003]
FIG. 1 is a diagram showing an outline of a circuit configuration of a force sensor according to this method. As shown in the figure, the force sensor has eight gauges G1 to G8. The output voltages of the gauges G1 to G8 are respectively amplified by programmable amplifiers (variable gain amplifiers) 11 to 18 to provide output LPFs composed of outputs LPF1 to LPF8 for each gauge.
[0004]
The gauges G1 to G8 have an equivalent circuit configuration. Therefore, the circuit configuration of the gauge will be described with reference to FIG. 2 in which one gauge Gn (n = 1, 2, 3,... 8) is extracted. One gauge Gn has a bridge configuration composed of two strain gauges A and B and two fixed bridge resistors RA 1 and RB. By applying an excitation voltage Ve to this bridge, the electrical resistance change of the strain gauge is changed. Is output as a voltage value Vn.
[0005]
However, as shown in FIG. 1, the balance resistors RA 1 and RB are shared by all the gauges G1 to G8. An initial error correction circuit including a variable resistor VRn and a fixed resistor Rn is provided to correct an error at the time of initial setting. This variable resistor VRn is adjusted so that the output LPFn (n = 1, 2, 3... 8) of each gauge Gn in the standard state at the time of installation (for example, room temperature 20 ° C .; no load) is all zero. Is done.
[0006]
Each strain gauge A, B is oriented and fixed in a predetermined direction on the force sensor body. The voltage appearing at each terminal C of the gauges G1 to G8 is applied to IN ⁺ of the corresponding programmable amplifiers 11 to 18, while the voltage appearing at the terminal D between the shared balance resistors RA and RB is the corresponding programmable amplifier 11. Applied to IN ⁻ of ˜18. The programmable amplifiers 11 to 18 amplify the differential voltage and generate eight gauge outputs LPF1 to LPF8.
[0007]
The gauge outputs LPF1 to LPF8 (LPF) are sent to an appropriate signal processing unit and converted to a six-axis force. Usually, the conversion to 6-axis force is performed through an operation (linear conversion) by multiplying LPF1 to LPF8 by a 6 × 8 conversion matrix. A configuration example of the signal processing unit will be described later.
[0008]
Now, it is assumed that a force sensor that has completed initial error correction under an accurate standard state is installed on the object W. In this state, when any force / moment is applied to the object W, some or all of the gauge outputs LPF1 to LPF8 are not zero, and a six-axis force in which at least any component is not zero is detected. The problem here is when the measurement environment deviates from the standard state.
[0009]
Due to the temperature characteristics of the resistors, programmable amplifiers, etc. constituting each gauge G1 to G8, each gauge output shifts under the influence of temperature change. That is, even if the six-axis force applied to the object W is unchanged, the gauge output is shifted. The influence of temperature change on each gauge output is described by a temperature coefficient. If the temperature coefficient is known, accurate measurement can be performed by correcting each gauge output according to the environmental temperature and then converting it to a six-axis force.
[0010]
Since the temperature coefficient is often different for each individual force sensor, an operation for obtaining the temperature coefficient for each individual force sensor is required. Conventionally, assuming that this temperature coefficient is not constant within the usable range (for example, 0 ° C. to 50 ° C.) of the force sensor, each gauge output is detected at many points in the usable range, From the transition, the temperature coefficient is obtained as a value that varies depending on the temperature by, for example, broken line approximation.
[0011]
[Problems to be solved by the invention]
However, in order to obtain a gauge output that correctly reflects each set temperature environment, the temperature inside the force sensor is different for each set temperature (for example, 5 ° C, 10 ° C,... 50 ° C). It is necessary to wait until it converges to a stable state (note that it does not necessarily match the set temperature due to heat generation of the circuit). This standby time is, for example, 2 hours for one set temperature, which is a heavy burden.
[0012]
Accordingly, an object of the present invention is to overcome the above-described problems, easily obtain a temperature coefficient for temperature correction of the gauge output of the force sensor, and correct the temperature using the temperature coefficient. The purpose is to provide a correction method and to reduce the work burden when using the force sensor.
[0013]
[Means for Solving the Problems]
The present invention measures each gauge output and internal temperature of the force sensor only at two points near the lower limit and the upper limit of the usable range of the force sensor under a constant six-axis force load condition. The above-mentioned technical problem is solved by determining the temperature coefficient for each gauge from the measurement results at the points and using this to perform temperature correction by software during actual measurement.
[0014]
The temperature correction is expressed by the following equation (1), for example.
Gcomp-i = Gi- {si * (T-T0) + Gt0-i} (1)
Here, the meaning of each code is as follows.
Gcomp-i; i-th gauge output after correction Gi; i-th gauge output before correction; temperature coefficient T related to i-th gauge; force sensor internal temperature output T0 at measurement; reference environmental temperature (for example, Force sensor internal temperature output Gt0-i under 0 ° C); i-th gauge output under reference ambient temperature (for example, 0 ° C.)
DETAILED DESCRIPTION OF THE INVENTION
FIG. 3 exemplifies a system configuration for performing temperature correction according to the present invention using gauge outputs LPF (LPF1 to LPF8) obtained from the circuits shown in FIGS.
The entire force sensor system is a central processing unit (hereinafter referred to as “CPU”) comprising a sensing unit 10 having the circuit configuration shown in FIGS. 1 and 2, an analog multiplexer 21, an A / D converter 22, and a microprocessor. 20), a volatile memory 23, a nonvolatile memory 24, and a temperature sensor 26. The analog multiplexer 21, the A / D converter 22, and the memories 23 and 24 are connected to the CPU 20 via the bus line 25.
[0016]
The analog multiplexer 21 incorporates an interface circuit with the CPU 20 and a driver circuit that drives the analog multiplexer 21. The analog multiplexer 21 is driven based on a command sent from the CPU 20, and the gauge outputs LPF 1 to LPF 1 from the sensing unit 10 are driven. The LPF 8 and the temperature sensor output are sequentially sent to the A / D converter 22. The A / D converter 22 converts each gauge output and temperature sensor output into digital values and outputs them to the bus line 25.
[0017]
The CPU 20 performs processing for overall control of the entire system, control commands for the analog multiplexer 21, storage of gauge output (before correction) and temperature sensor output output from the A / D converter 22, temperature correction processing, and six-axis force The conversion process is performed according to various programs stored in the nonvolatile memory 24. In order to implement the present invention, the memory 23 or 24 stores data necessary for the calculation of the equation (1), calculates a temperature coefficient based on the data, temporarily stores various data, and stores the data of the equation (1). A program for executing calculation processing and the like is stored.
[0018]
Hereinafter, a processing procedure for acquiring data necessary for temperature correction using such a force sensor system and a processing procedure for obtaining a temperature-corrected detection output using the data will be described with reference to the flowcharts of FIGS. I will explain. In the present embodiment, the reference temperature Tst for calculating the equation (1) is specified as the lower limit usable temperature Tst = Tmin specified in the specification, and another point for measuring the gauge output in order to obtain the temperature coefficient si. Useable upper limit temperature Tmax specified in the certificate. Note that the six-axis force load for obtaining the temperature coefficient si is zero. In addition, it is assumed that the initial error adjustment of each of the gauges G1 to G8 under the reference temperature Tst (= Tmim) has been completed.
[0019]
First, the excitation voltage Ve is applied under the reference temperature Tst (= Tmin) while being applied to the sensing unit 1.
[0020]
Next, a program for measuring the temperature coefficient is started, and repeated detection of the internal temperature of the force sensor is started (step S1). This process is continued until it is determined that the internal temperature of the force sensor is stable (step S2). That is, it is determined that the internal temperature of the force sensor is stable when the temperature output difference ΔT at the time of detection before and after falls below an appropriate small value ε, and the latest internal temperature output is the temperature output T0 under the reference temperature. (Step S3). Further, in the next step 4, the gauge outputs G01 to G08 (digital values) under the reference temperature are sequentially captured and stored.
[0021]
Next, with the excitation voltage Ve applied to the sensing unit 1, the force sensor is placed under the second gauge output measurement temperature Tmax.
[0022]
The program for measuring the temperature coefficient is started again, the repeated detection of the internal temperature of the force sensor is started (step S5), and the stability of the internal temperature of the force sensor is determined by the same process as step S3 (step S5). S6). Then, after storing the internal temperature output T1 at that time (step S7), the gauge outputs G11 to G18 (digital values) at the second temperature are sequentially taken in and stored (step S8). Finally, the temperature coefficients si of the gauges G1 to G8 are calculated and stored from the two sets of gauge outputs G01 to G08 and G11 to G18 (step S9), and the process is terminated. The temperature outputs T0 and Gt0-i under the reference temperature are stored in the nonvolatile memory 24 together with the temperature coefficient si.
[0023]
The process for detecting the six-axis force using the force sensor siltem after the above process is as follows.
A program for 6-axis force measurement is started, the internal temperature of the force sensor is detected, and stored as a temperature output T at the time of measurement (step M1). Further, each gauge output (digital value) is sequentially fetched and stored as gauge outputs G1 to G8 before correction (step M2).
[0024]
Next, using the existing data Gt0-i, si, and T0, the right side of the equation (1) is calculated to obtain corrected gauge outputs Gcomp-1 to Gcomp-8 (step M3). Further, the temperature-corrected 6-axis force is obtained using the 6-axis force conversion matrix data (step M4). The obtained 6-axis force is used in accordance with the application as in the prior art.
[0025]
As described above, according to the present invention, the temperature coefficient of each gauge is obtained through the gauge output measurement of only two points. However, with such a method, temperature correction with considerably high accuracy can be executed. This is because the temperature coefficient actually takes a substantially constant value with considerably good accuracy in the normal usable range of the force sensor. FIG. 6 and FIG. 7 are graphs of actual measurement data confirming this.
[0026]
First, the graph of FIG. 6 shows four points in 15 ° C increments between an external temperature of 0 ° C (reference temperature) and 45 ° C for a commonly used force sensor. And gauge outputs G1 to G8 are actually measured and shown in a line graph. The horizontal axis represents the internal temperature output, and the vertical axis represents the gauge outputs G1 to G8 (the unit used bits for convenience). In this graph, for any gauge, the bending angle of the broken line is not so large, indicating that the temperature coefficient has a substantially constant value.
[0027]
In the graph of FIG. 7, the temperature coefficient for each gauge is obtained in advance from the gauge output when the external temperature is 0 ° C. (reference temperature) and 45 ° C., and is applied to the temperature correction of the data shown in the graph of FIG. This is shown by a line graph. As in FIG. 6, the horizontal axis represents the internal temperature output, and the vertical axis represents the gauge outputs G1 to G8.
[0028]
In this graph, the gradient of temperature-gauge output is significantly reduced for all gauges, and the temperature coefficient in the practical sense is greatly corrected downward. From these graphs, it is understood that the temperature characteristic of the force sensor is greatly improved by the temperature correction using the temperature coefficient obtained from the gauge output of only two points.
[0029]
Note that the gauge output measurement for obtaining the temperature coefficient is normally performed under a 6-axis force no-load condition, but it goes without saying that it may be performed under a constant 6-axis force load condition instead of no load. .
[0030]
【The invention's effect】
According to the present invention, since the temperature coefficient for performing temperature correction at the time of measurement in the usable range of the force sensor is obtained from the gauge output outputs of only two points, the burden associated with the use of the force sensor is reduced. Further, by using the temperature coefficient obtained in this way, temperature correction that is sufficiently practical can be achieved, and the force sensor can have a small temperature coefficient in an effective sense.
[Brief description of the drawings]
FIG. 1 is a diagram illustrating an outline of a circuit configuration of a force sensor.
FIG. 2 is a diagram for explaining by extracting one gauge Gn (n = 1, 2, 3,... 8) in the circuit configuration shown in FIG.
FIG. 3 illustrates a system configuration for performing temperature correction according to the present invention using gauge outputs LPF (LPF1 to LPF8) obtained from the circuits shown in FIGS. 1 and 2;
FIG. 4 is a flowchart illustrating a process for obtaining a temperature coefficient.
FIG. 5 is a flowchart for explaining processing at the time of six-axis force measurement.
FIG. 6 shows a force sensor that is generally used, and takes four points in 15 ° C. between an external temperature of 0 ° C. (reference temperature) and 45 ° C., and the sensor internal temperature and gauge output G 1 ˜G8 was measured and indicated by a line graph.
7 is obtained in advance from the gauge output at the external temperature of 0 ° C. (reference temperature) and 45 ° C., and applied to the temperature correction of the data represented by the graph of FIG. This is shown by a line graph.
[Explanation of symbols]
10 sensing units 11 to 18 programmable amplifier 20 CPU
21 Analog multiplexer 22 A / D converter 23 Volatile memory 24 Non-volatile memory 25 Bus line 26 Temperature sensor G1-G8 Strain gauge

Claims (3)

Under constant six-axis force load conditions, each gauge output and internal temperature of the force sensor is measured at only two points near the lower limit and upper limit of the usable range of the force sensor, and the measurement results at these two points A temperature correction method for a force sensor, in which a temperature coefficient for each gauge is determined and software is used to perform temperature correction during actual measurement. Under constant six-axis force load conditions, each gauge output and internal temperature of the force sensor is measured only at two points near the lower limit and upper limit of the usable range of the force sensor, and the measurement results at these two points The temperature coefficient for each gauge is determined from
Gcomp-i = Gi- {si * (T-T0) + Gt0-i}
(However,
Gcomp-i; i-th gauge output after correction Gi; i-th gauge output before correction; temperature coefficient T relating to the i-th gauge; force sensor internal temperature output T0 at the time of measurement; force under reference ambient temperature Temperature sensor internal temperature output Gt0-i; i-th gauge output under reference environmental temperature)
A temperature correction method in the force sensor, wherein temperature correction is performed by software including calculation represented by The temperature correction method for a force sensor according to claim 1 or 2, wherein the constant six-axis force load condition is no six-axis force load.

Images