Double-probe pose calibration method based on gauge block rotation

Abstract

In the high-end manufacturing industry, a high-precision rotary body is widely applied, and the measuring efficiency and precision can be improved by adopting a double-optical-probe measuring machine. The invention provides a calibration method for deflection, pitching and zero point relative positions of double probes. The contour of the gauge block is used as a reference, and the two probes are aligned with two surfaces of the gauge block and scan in X, Y and Z directions respectively to form a linear equation. And a plurality of groups of linear equations are obtained by rotating the gauge block for multiple times to form a linear equation set, and the installation deviation and the relative position of the two probes can be finally solved through a least square method. Compared with a traditional ball calibration method, the method has the advantages that due to a linear measurement model, the influence of a nonlinear error of the probe and a linear error of a displacement table can be inhibited, and a more accurate measurement result can be obtained.

Description

Technical Field

[0001] This invention primarily relates to a method for calibrating installation deviations of optical probes. By measuring the profile of a gauge block in the X, Y, and Z directions using dual probes and repeatedly changing the gauge block's position, the zero-point position offset and installation deviations, such as yaw and pitch angles, of the dual probes in these directions are determined. The calibration results can effectively guide the adjustment and compensation of installation deviations, and the method belongs to the field of precision instrument manufacturing and precision testing and metrology. Background Art

[0002] High-precision thin-walled rotating parts and films are used in a variety of fields, including aerospace, navigation, and optics. For example, hemispherical resonators, high-precision aerospace bearings, and various optical films all require strict control of thickness uniformity and internal and external contour deviations. Due to the recent increase in production of these parts, efficient quality control methods have become indispensable. Using dual optical point measurement probes, aligning them with the upper and lower surfaces of the film and the internal and external contours of the rotating part, respectively, effectively measures the thickness and contour of these films and rotating parts.

[0003] In industry, a six-axis adjustment frame is often used to adjust the yaw angle, pitch angle, and zero point relative position of the two probes in the X, Y, and Z directions. This ensures that the light from the two probes is aligned with each other, and that when measuring a rotating body, the extension line of the light direction can pass through the axis of rotation. This adjustment is generally achieved by measuring a standard gauge block, adjusting the probe posture, and finding the maximum output value of the probe. This ignores the situation where multiple installation deviations of the probes couple to produce measurement errors, resulting in a lack of effective data guidance. To address this problem, some scholars have proposed using two probes with installation deviations to measure a standard sphere, and calculating the installation deviation of the probe by comparing the ideal value without installation deviation obtained by mathematical calculation with the measured value with installation deviation. Although this method can effectively calibrate multiple installation deviations such as zero point offset and yaw and pitch of the probe, it also produces noise amplification problems due to the nonlinear contour of the sphere itself. As a result, in the presence of measurement noise, all installation deviations cannot be accurately calibrated, which in turn leads to a decrease in measurement accuracy. Summary of the Invention

[0004] The present invention is aimed at the situation where the yaw angle, pitch angle, and zero offset in the three directions of X, Y, and Z of the dual optical alignment probe are unknown under non-ideal conditions. This requires accurate acquisition and further compensation or adjustment. This method uses a standard gauge block as a measurement reference. The two probes scan the gauge block contour in the three equations of X, Y, and Z respectively, establish a polynomial, and construct multiple linear equations through multiple postures of the gauge block to obtain the yaw angle, pitch angle, and offset in the three directions of X, Y, and Z of the two probes. This guides the error compensation and probe adjustment of the measurement system.

[0005] The calibration method employed in this invention involves first placing a gauge block in an arbitrary position, with its two faces aligned with two probes. A translation stage is then used to drive the gauge block or dual probes to scan the block's contour in the X, Y, and Z directions, obtaining output data from both probes. The gauge block's position is then randomly altered three or four times using a tilt stage, and a system of linear equations is constructed. The yaw and pitch angles of the two gauge blocks, as well as their zero offsets in the X, Y, and Z directions, are then solved using least-squares fitting.

[0006] The method for calibrating the installation deviation and relative position of the dual optical probes based on the rotation of the gauge block includes the following steps:

[0007] (1) First, fix the gauge block vertically on the tilting table, and fix the tilting table on the four-axis motion table by hot melt glue or tooling. The four-axis motion table includes a linear motion table in the X, Y, and Z directions and a turntable. The turntable plane is parallel to the XY plane.

[0008] (2) Adjust the tilt table to a small angle, taking care not to exceed the probe's inclined measurement limit angle, and align the two sides of the gauge block with the two probes. Select an appropriate gauge block size so that the measurement points on both sides of the gauge block do not exceed the probe's range. After selecting the appropriate size, secure the gauge block with a fixture and hot melt adhesive to prevent it from loosening during measurement.

[0009] (3) Determine a measurement zero point, start the X-direction translation stage, move the gauge block in the X-direction, and evenly collect 20 points of the output of each of the two probes.

[0010] (4) Return the X-axis to the zero position, start the Y-direction translation stage, move the gauge block in the X-direction, and evenly collect 20 points of the output of each of the two probes.

[0011] (5) Return the Y axis to the zero position, start the Z-direction translation stage, move the gauge block in the Z direction, and evenly collect 20 points of output from each of the two probes.

[0012] (6) Use the turntable and tilt table to change the position of the gauge block, while being careful not to exceed the measuring slope limit of the probe. Repeat steps (3) to (5).

[0013] (7) According to the solution method, the corresponding linear equations are listed, and the measurement results of (3) to (5) are calculated using the least squares method to obtain the zero point offset, yaw angle and pitch angle of the two probe installations.

[0014] This paper proposes a tilt adjustment method based on the gauge block rotation method. Combining computational geometry with optimization techniques, this method leverages the linear characteristics of the gauge block to suppress noise. A linear equation is formulated, and a least-squares fit is used to measure the yaw and pitch angles and the relative zero point position of the dual-probe alignment system. Compared to traditional spherical calibration methods, the linear model sampled by this method suppresses noise, achieving higher calibration accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 Schematic diagram of the measuring device

[0016] In the figure: 1, Z-axis motion stage 2, optical probe 3, rotation stage 4, Y-axis motion stage 5, X-axis motion stage

[0017] Figure 2 Schematic diagram of the gauge block rotation method

[0018] In the figure: 1. Optical probe A 2. Gauge block 3. Optical probe B 4. Yaw angle of optical probe A 5. Pitch angle of optical probe B 6. Y-axis offset of optical probes A and B 7. Pitch angle of probe A 8. Pitch angle of probe B 9. X / Y / Z distance between the measurement zero points of probes A and B (m) x ,m y ,m z ), corresponding to 6, 9, and 10 in the figure. The thickness d of the gauge block corresponds to 11.

[0019] Figure 3 Measurement flow chart DETAILED DESCRIPTION

[0020] The following describes in detail the method and device for adjusting the inclination of an inclined hole axis based on geometric optimization and minimum area theory in conjunction with the accompanying drawings and specific embodiments:

[0021] The theory and method proposed in this article of the present invention are as follows Figure 1 The measurement device shown is implemented and calibrated to Figure 1 The optical probe shown in FIG. The X-axis mechanism 5, the Y-axis mechanism 4, the Z-axis mechanism 1, and the rotation axis 3 are combined to verify the measurement method in this paper. The workbench 4 can move along the X and Y axes, the Z axis is equipped with a probe sensor 2, and the workbench 4 is equipped with a two-dimensional rotation stage 3. The installation position deviation of the probe is shown in FIG. Figure 2 , calibrated by a gauge block with a thickness of d, as shown in Figure 11. What needs to be calibrated are the roll angles (α1, α2) of probes A and B corresponding to 5 and 6 in the figure, as well as the pitch angles (β1, β2) 8 and 9, and the relative position of the zero points of the two probes (m x ,m y ,m z ) corresponds to 6, 9, and 10 in the figure.

[0022] The measuring principle and process of the present invention are:

[0023] (1) First, install the gauge block on the fixture, and align the two sides with probe A and probe B respectively. At this time, the plane equations of the two sides of the gauge block are recorded as Ax+By+Cz+D1=0 and Ax+By+Cz+D 2` = 0. The installation deviation of the probe zero point is recorded as (m x ,m y ,m z ), the yaw and pitch angles of the two probes are (α1, β1) and (α2, β2).

[0024] (2) Let the probe scan in the X\Y\Z directions respectively, and obtain the measurement array as (L1 (1) ,L1 (2) ,L1 (3) ,...,L1 (n) ) and (L2 (1) ,L2 (2) ,L2 (3) ,...,L2 (n) ).

[0025] (3) Rotate the gauge block, change the plane equation, and re-obtain the scanning results. Multiple measurements constitute an equation group, and the least squares method is used to calculate the installation deviation and zero point relative position of the two probes. The measurement process is as follows: Figure 3 shown.

[0026] formula:

[0027]

[0028] Therefore, a method based on the gauge block rotation method is used to calculate the yaw and pitch angles of the two probes and the relative zero point position. Its unique feature is that it replaces the traditional nonlinear model with a linear model, resulting in higher accuracy than traditional algorithms. By scanning the front and back of the gauge block separately, a linear equation relationship is obtained. Multiple pose transformations are used to generate a linear system of equations, which are then solved to ultimately determine the installation deviation and relative zero point position of the two probes. This method can be applied to high-precision internal and external contour measurement of rotating bodies and high-precision thin film measurement. It can provide guidance for probe pose adjustment and error compensation.

Claims

1. A dual-probe posture calibration method based on gauge block rotation, characterized by: For high-precision measurement of rotating objects and thin films, a dual-probe position calibration method based on the gauge block rotation method is proposed to calibrate the installation error and zero-point relative position of dual-aligned optical probes. The two surfaces of the gauge block are aligned with the two optical probes, which are then scanned in the X, Y, and Z directions. The resulting contour data is used to construct the corresponding linear equations. To fully solve all parameters, the gauge block is transformed, that is, its plane equation is changed. Using multiple arbitrary positions to construct a set of linear equations, the yaw angle, pitch angle, and zero-point relative position relationship of the two optical probes can be obtained. The measuring principle and process of the present invention are: (1) First, install the gauge block on the fixture, and align the two sides with probe A and probe B respectively. At this time, the plane equations of the two sides of the gauge block are recorded as Ax+By+Cz+D1=0 and Ax+By+Cz+D 2` = 0. The installation deviation of the probe zero point is recorded as (m x ,m y ,m z ), the yaw and pitch angles of the two probes are (α1, β1) and (α2, β2). (2) Move the translation stage in the xyz direction to drive the two probes to move. Let the probes scan in the X\Y\Z directions respectively, and obtain the measurement array of the two probes as (L1 (1) ,L1 (2) ,L1 (3) ,...,L1 (n) ) and (L2 (1) ,L2 (2) ,L2 (3) ,...,L2 (n) ). (3) Rotate the gauge block, change the plane equation, and repeat (1) to 2 to obtain the scanning results. Establish the equation group based on the geometric relationship established by the rotation coordinate transformation. (4) Multiple measurements are performed to obtain different plane equations of the gauge block, forming a set of equations. When the number of equations exceeds the number of unknowns, the installation deviation of the two probes and the relative position of the zero point are calculated using the least squares method. Mode:

Images