JPH11325976A - Displacement input device for autonomous calibration of sensor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To obtain a displacement input device for autonomous calibration of sensor that integrally installs calibration function of average sensitivity in an autonomous calibration jig and enables to simultaneously calibrate the average sensitivity during working of linear error calibration. SOLUTION: A jig 5 is installed through a piezoelectric actuator 3 providing a displacement becoming a calibration input of sensor 4 on a base and a laser interferometer 20 is integrally installed under the jig 5. The laser interferometer 20 is constituted of a laser light source 14 placed on the base 1, a moving mirror 10 placed at the back side of the jig 5, a reference mirror 11 with a reflecting face orthogonal to the moving mirror 10 that is placed near the moving mirror 5 on the base 1, a light-receiving diode 15 detecting an interference fringe by reflecting light from the moving mirror 10 and the reference mirror 11.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a displacement input device for autonomous calibration which is used for sampling sensor input / output data and performing autonomous calibration such as linear error of the sensor by numerical calculation.

[0002]

2. Description of the Related Art In the development of displacement sensors and angle sensors aiming at nanometers and nanoradians, it has become difficult to obtain calibration standards for guaranteeing the accuracy as well as to improve the accuracy. Conventionally, X was used to calibrate the interpolation error between wavelengths of the interferometer.
Method using a line interferometer (DKBowen et al .: Subnanometr
e transducer characterization by X-ray interferome
try, Precision Engineering, 12, 3 (1990) 165)
A method of performing non-linear error correction using the linear drive range of PZT (W. How and G. Wilking: Investigation and compen
sation of the nonlinearity of heterodyne intergero
meters, Precision Engineering, 14, 2 (1992) 91).

However, the method of using an X-ray interferometer for calibration is as follows.
It is difficult for general displacement meter users to use. In the method using PZT, it is difficult to confirm whether calibration has been performed correctly. Furthermore, many of these high-definition sensors require delicate adjustment, and if possible, they often want to perform in-situ calibration frequently and while still attached to the device. The above-mentioned conventional method cannot meet this demand.

On the other hand, the present inventors have previously proposed an autonomous calibration method for linear errors in displacement sensors and angle sensors. Normally, to obtain calibration data for a sensor, a sensor system with higher precision than the sensor to be calibrated is required, whereas the autonomous calibration method requires the calibration data without using such a high-precision sensor system. Is a way to get The autonomous calibration method proposed by the present inventors is
Kei, Ken Morishima, Toru Sugibuchi: Autonomous Calibration of Linear Error of Displacement Meter, Journal of the Japan Society of Precision Engineering 59,12 (1993) 2043, Kei Seino, Suzuto Kuzu, Yoichi Nishino: Improvement of Interferometer Accuracy by Autonomous Calibration of Interpolation Error Journal of Precision Engineering, 62, 2 (1996) 279, Kei Seino, Zhang Shizhu: Research on high-precision autonomous calibration of angle sensors, Journal of Precision Engineering, 60, 11 (1994) 1591.

In this autonomous calibration method, for example, in the case of a displacement sensor, a reference sensor of the same type as the sensor to be calibrated is prepared and combined with a lever system so that the reference sensor detects n times the displacement of the sensor to be calibrated. Perform calibration measurement. Thus, when the calibration of the sensor to be calibrated is performed by the reference sensor, the linear error included in the calibration result is reduced to 1 / n by the lever system. It can converge to a state close to zero. In the autonomous calibration method proposed earlier, a lever system that expands the calibration input,
A sensor equivalent to the sensor to be calibrated is additionally required. For this reason, there is a disadvantage that the restriction on realizing in-situ calibration of the sensor incorporated in the device is too strong. To realize in-situ calibration, a method that does not use an extra sensor or lever system is desired.

Accordingly, the present inventors further seek to accurately determine the linear error of the sensor using two data samplings and an approximate calculation and a convergence operation without using extra space or additional equipment. Proposed a new sensor autonomous calibration method (Japanese Patent Application No. 9-142638).
No., Kei Kiyono, Takawei, Ichiro Ogura: In-situ Autonomous Calibration Method for Geometric Sensors, Journal of the Japan Society of Precision Engineering 63-10 (1997) 1417-14
twenty one). That is, the method proposed by the present inventors is that the input quantity to be measured is x, the output is v, the average sensitivity is Sm, and the linear error is g.
As (x), the calibration curve is given by f (x) = v = Sm × x +
An autonomous calibration method of a sensor represented by g (x), wherein each output at a plurality of sampling points obtained by measuring a predetermined calibration range by the sensor is defined as vi, and an input value at each sampling point is defined as vi. A step of obtaining a 0th order approximate value x0i ≒ vi / Sm;
Assuming that the output at the sampling point given x is vi +, using the difference Δvi = vi + −vi between the outputs of the two points, a zeroth-order approximation g′0 (x0) of the derivative of the linear error g (x)
i) a step of calculating ≒ Δvi / Δx-Sm, and a zero-order approximation g′0 of the derivative obtained in this step.
(X0i) being numerically integrated to obtain a zero-order approximation g0 (x) = Σg'0 (x0i) Δx of the linear error g (x), and the linear error g (x )
, The approximate value of the input value at each of the sampling points is corrected using the zeroth-order approximate value g0 (xoi), the sampling point of the derivative is corrected using the corrected value, and the corrected result is further corrected. Repeating the process of correcting the approximate value of the linear error by numerical integration a required number of times.

[0007]

The calibration curve of the sensor is
An average straight line corresponding to a linear input / output characteristic and a linear error that is a deviation from the straight line are included. The purpose of the calibration is to measure the linear error, and in the new autonomous calibration method described above, if the small input change given for the calibration is not known, the slope of the average straight line of the calibration curve described above (that is, the average slope) Sensitivity) must be measured with high precision in advance.
For that purpose, it was necessary to measure the average sensitivity using a guaranteed step sample or the like separately from a device for performing data sampling for autonomous calibration. Accordingly, the present invention provides an input device for autonomous calibration of a sensor, in which the function of average sensitivity calibration is integrated into a jig for autonomous calibration so that the average sensitivity calibration can be performed simultaneously during the work of linear error calibration. The purpose is.

[0008]

A displacement input device for autonomous calibration of a sensor according to the present invention is mounted on a base via a piezoelectric actuator for giving a displacement serving as a sensor calibration input on the base. A jig, a laser light source attached to the base, a movable mirror attached to the back surface of the jig so as to be displaced with the jig, and an output light beam from the laser light source is vertically incident on the jig. A reference mirror attached to the base near the movable mirror with a reflecting surface orthogonal to the movable mirror, and a part of an output light beam from the laser light source is branched and vertically incident; And a light receiving element for detecting interference fringes due to light reflected from the reference mirror.

According to the present invention, a laser interferometer having a movable mirror, a reference mirror, a laser light source, a light receiving element, and the like is integrated under a jig with a piezoelectric actuator for providing a displacement input for autonomous calibration of a sensor. By incorporating the sensor, the average sensitivity of the sensor required for the autonomous calibration can be accurately calibrated simultaneously with the autonomous calibration of the sensor without using a step sample or the like separately. In particular, in a laser interferometer, the wavelength of a laser can be guaranteed with a high precision of 1/10000, and a higher precision sensitivity can be guaranteed than in the case of using a step sample that is guaranteed by a standard station in the world.

[0010]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a plan view showing a main part of an input device for autonomous calibration according to an embodiment of the present invention, and FIG. 2 is a front view. A support frame 2 is provided on a base 1, and a hollow cylindrical piezoelectric actuator 3 is mounted on the support frame 2. The piezoelectric actuator 3 can expand and contract in the axial direction of the hollow cylinder. A disc-shaped jig 5 for applying a displacement input to a sensor 4 to be calibrated is attached to an upper end of the piezoelectric actuator 3.

The sensor 4 to be calibrated is, for example, a displacement sensor such as a capacitance type displacement sensor. The sensor 4 is the support frame 2
It is held at a predetermined position in the Z-axis direction by a Z-axis sliding mechanism 7 provided on a support column 6 fixed to (or the base 1).

Below the jig 5, a laser interferometer 20 for controlling the displacement of the jig 5 by the pitch of the interference fringes is incorporated. The laser light source 14 of the laser interferometer 20 is a semiconductor laser and is mounted on a base. In order to guide the output laser light from the laser light source 14 to the back surface of the jig 5, the plate beam splitter 16 and the 波長 wavelength plate 1
7 and a mirror 13 are attached to the base 1, and a cube beam splitter 12 is attached to the support frame 2.

On the back surface of the jig 5 is mounted a movable mirror 10 which is displaced together with the jig 5 and into which laser light is vertically incident. A reference mirror 11 is attached to the support frame 2 integrally with the cube beam splitter 12 and has a reflecting surface orthogonal to the movable mirror 10 and into which the light split by the cube beam splitter 12 is vertically incident. The optical path length difference between the movable mirror 10 and the reference mirror 11 is 1
It is set to about 0 wavelength or less. Thereby, the light reflected by the movable mirror 10 and the reference mirror 11 is multiplexed by the cube beam splitter 12 to form interference fringes. The interference light is extracted via the mirror 13, the 波長 wavelength plate 16 and the plate beam splitter 16. In order to detect the extracted interference light, a light receiving diode 15 is attached to the base 1.

The output of the light receiving diode 15 is detected by a detection circuit 21 including a current / voltage conversion circuit,
Sent to 2. The controller 22 performs zero cross detection of interference fringes and the like, and according to the detection result, the drive circuit 23
To the control signal. The drive circuit 23 supplies a drive voltage to the piezoelectric actuator 3 according to the control signal. When the input displacement is given to the jig 5 and the jig 5 needs to be temporarily fixedly held at a certain input displacement, the controller 22 performs, for example, zero cross point detection of interference fringes,
A function of locking the actuator 3 based on the detection signal is provided. Alternatively, the controller 22 vibrates the jig 5 in a predetermined range, and outputs a zero-cross detection signal.
It can also be used as a trigger signal for taking in calibration data.

A method for autonomously calibrating the displacement sensor 4 using such a displacement input device will now be specifically described. FIG. 4 shows the input / output relationship of the displacement sensor, that is, the calibration curve. In most cases, a line representing the average sensitivity (calibration line)
Is guaranteed to fall within the range of ± a% of its output. Assuming that the input displacement of the sensor 4 is x and the output at that time is v, the function v = f (x) of the calibration curve is represented by a straight line having a gradient given by the average sensitivity Sm, as shown in the following equation 1, and Is represented by a linear error g (x) indicating the deviation of

[0016]

## EQU1 ## v = f (x) = Sm.x + g (x)

In this embodiment, the output data obtained at a plurality of first sampling points in a predetermined calibration range by the displacement sensor 4 to be calibrated and the output data obtained by applying a minute change to each of the first sampling points. The output data obtained at the second sampling point is used, and the linear error is calibrated only by the numerical operation of the digitized data. FIG. 3 shows the flow of the autonomous calibration process.

As shown in FIG. 3, output data is sampled by applying a displacement to the displacement sensor 4 at predetermined intervals within a calibration range (step S1). The sampling points xi (i = 1, 2,..., N) in step S1 do not necessarily have to be at equal intervals, but are arranged as evenly as possible in the calibration range.

Next, at each sampling point x in step S1
A sampling point xi ₊ = where a small displacement Δx is given to i
Data sampling at xi + Δx is performed (step S
2). The minute displacement Δx may be given as a known value, but actually, a value obtained by averaging calculation of output data is used. From the data obtained by the two samplings described above, the output at the sampling point xi is defined as vi, first, the linear error g (x) is ignored, and x is obtained based on the following equation (2).
Calculate the 0th order approximate value x0i of i (step S3).

[0020]

[Equation 2] x01 ≒ vi / Sm

On the other hand, assuming that the output at the sampling point xi ₊ obtained in step S2 is vi ₊ , the difference value Δ of the output of the corresponding two sampling points xi and xi ₊ separated by Δx.
vi is obtained by the following equation (3).

[0022]

## EQU3 ## Δvi = vi ₊ −vi

Using the obtained output difference value Δvi,
A zero-order approximation g′0 (x0i) of the derivative of the linear error g (x) is obtained by numerical calculation based on the following equation 4 (step S4).

[0024]

G′0 (x0i) = Δg (x) / Δx ≒ Δvi / Δx−Sm

The approximate value g'0 (x0i) of the derivative indicates the slope of the linear error g (x) at x0i as shown in FIG. Subsequently, the obtained approximate value g′0 (x0i) of the derivative is numerically integrated to obtain a zeroth-order approximate value g0 of the linear error g (x).
(X) = Σg′0 (x0i) Δx is calculated (step S
5). The number of times of correction processing j is initialized (step S6), and the 0th-order approximation g of the linear error obtained in step S5
Using 0 (x), an iterative convergence operation for converging the linear error g (x) to a predetermined range is performed. That is, the 0th-order approximate value x0i of the sampling point xi is corrected using g0 (x) to obtain a first-order approximate value x1i (step S7). FIG.
, Each sampling point x using g0 (x)
When the 0th-order approximate value x0i of i is corrected, a first-order approximate value x1i as shown in the following equation 5 is obtained.

[0026]

X1i = {vi-g0 (x0i)} / Sm

Then, using the first-order approximate value x1i, a first-order approximate value g'1 (x1i) of the derivative is obtained according to Equation 4 (step S8). The first approximation g1 (x1i) of the linear error g (x) is obtained by numerically integrating the obtained approximation (step S8). Thereafter, for example, a convergence determination is made as to whether or not the residual error has converged to a predetermined range (step S10).
If the convergence condition is not satisfied, the number of times of the correction processing j is increased (step S11), and the same correction processing is repeated thereafter. This improves the approximation accuracy of the derivative g '(x) and the linear error g (x). In general, a derivative g′j (xji) expressed by using the j-th approximate value xji and a linear error gj (xji) obtained by integrating the derivative g′j (xji) are obtained as shown in Expression 6.

[0028]

G'j (xji) ≒ Δvi / Δx-Sm gj (xji) = ∫g'j (xji) dx

By using this, j obtained by modifying the approximate value xji
The + 1st approximate value xj + 1i is obtained as in the following equation (7).

[0029]

Xj + 1i = {vi-gj (xji)} / Sm

When the number j of the correction processing is increased, the degree of approximation of xi and the linear error function g (x) is gradually improved, and it is possible to converge on an accurate calibration curve. In the above description, the procedure for directly obtaining the linear error function has been described. However, there is also a method for obtaining the inverse function of the linear error function first, and in this case, the repetition of the correction is not necessary (see the above cited documents). Generally, the relationship between the average sensitivity Sm and the small displacement Δx is expressed by the following equation (8).

[0031]

Sm · Δx = ΣΔvi / n

Therefore, if one of Sm and Δx is known accurately, the other can be obtained by calculation. For example, if the average sensitivity Sm is known in advance, the minute displacement Δx can be accurately obtained by calculation based on Equation 8. Using the obtained small displacement Δx, the calculation of Equation 4 or less becomes possible. At this time, the sampling points xi only need to be arranged substantially uniformly, and it is not always necessary to know the exact value.

When the displacement input device of this embodiment is used, the average sensitivity Sm required for the linear error calibration of the displacement sensor described above is obtained.
Can be calibrated on-the-fly during a linear error calibration operation. That is, the minute displacement given to the jig 5 is
By averaging the obtained output data while controlling at an interference fringe interval of 0, a precise average sensitivity Sm can be obtained. In the case where the average sensitivity is unknown using a displacement sensor other than the interferometer, it was conventionally necessary to measure the average sensitivity separately using a step sample, etc., but according to this embodiment, the average sensitivity is integrated with the autonomous calibration jig. With the laser interferometer 20 incorporated in the above, the average sensitivity calibration can be performed simultaneously with the linear error calibration. Moreover, since the average sensitivity is determined based on the wavelength of the interferometer, it is possible to obtain an average sensitivity with extremely high accuracy. In the case of a step sample or the like, there is a possibility that the reliability may be lost due to contamination, but there is no such concern in the laser interferometer. Further, the laser interferometer 2
Since 0 is only required to detect a displacement within the maximum displacement of the actuator 3, the configuration is simple and compact. If the optical path difference between the interference arms is at most about 10 wavelengths, the drift of the interference fringes can be suppressed to a small value.

The present invention is not limited to the above embodiment.
For example, in the embodiment, a laser interferometer for space propagation is incorporated, but using an optical fiber or a glass rod as a waveguide medium,
Interferometers having mirrors fixed to their end surfaces may be used.
Further, for calibration of a sensor having a wide measurement range, the entire input device of the embodiment can be installed on a movable base having a large movable range to be an input device. In this case, by fixing the reading unit of the encoder to a device mounted on the X-axis stage, autonomous calibration including the average pitch of the encoder can be performed. Further, the input device according to the present invention can be applied not only to the autonomous calibration of the displacement sensor but also to the autonomous calibration of the angle sensor.

[0035]

As described above, according to the present invention, the laser interferometer is integrated under a jig with a piezoelectric actuator for giving a displacement input for autonomous calibration of the sensor. At the same time as the calibration, the average sensitivity of the sensor required for the autonomous calibration can be accurately calibrated. In particular, in a laser interferometer, the wavelength of the laser can be guaranteed with high accuracy, and the sensitivity can be guaranteed with higher accuracy than when a step sample is used.

[Brief description of the drawings]

FIG. 1 is a plan view showing main part calibration of a displacement input device according to an embodiment of the present invention.

FIG. 2 is a front view of the displacement input device of the embodiment.

FIG. 3 is a diagram showing a flow of a process of autonomous calibration of a sensor using the apparatus of the embodiment.

FIG. 4 is a diagram for explaining difference value calculation and input approximate value calculation on a calibration curve of a sensor.

FIG. 5 is a diagram for explaining calculation of an approximate value of a derivative of a linear error.

FIG. 6 is a diagram for explaining calculation of an approximate value of a linear error.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Base, 2 ... Support frame, 3 ... Piezoelectric actuator, 4 ...
Displacement sensor, 5 ... jig, 10 ... movable mirror, 11 ... reference mirror, 12 ... cube beam splitter, 12 ... mirror, 14 ... laser light source, 15 ... light receiving diode, 16 ...
Plate beam splitter, 17 ... 1/4 wavelength plate, 21
... Detection circuit, 22 ... Controller, 23 ... Drive circuit.

Claims (1)

[Claims] A base mounted on the base, a jig mounted on the base via a piezoelectric actuator for applying a displacement serving as a sensor calibration input, a laser light source mounted on the base, A movable mirror which is attached to the back surface of the jig so as to be displaced together with the jig and into which the output light beam from the laser light source is vertically incident; and a reflection orthogonal to the movable mirror on the base near the movable mirror. A reference mirror attached with a surface, a part of the output light beam from the laser light source is branched and vertically incident, and a light receiving element for detecting interference fringes due to reflected light from the movable mirror and the reference mirror. A displacement input device for autonomous calibration of a sensor, comprising: