CN112325808B - Flatness real-time calibration compensation measurement method based on multiple PSDs - Google Patents

Abstract

The invention discloses a multi-PSD-based flatness implementation calibration compensation measurement method. The method includes: acquisition of target and reference data: synchronously acquiring the measurement data of the target PSD and multiple reference PSDs ₁ and ₂ in real time, respectively, and the synchronization error : 200ms; establishment of error model: according to the terrain environment where the test is located, analyze the influence of flatness accuracy and have a specific form of error mainly ground vibration, and establish an error model in this environment; station error separation: next Multiple sets of test measurements, adding data samples, analyzing the data, and separating errors; real-time flatness calibration: real-time compensation for flatness through the compensation system to ensure the flatness accuracy of the scanning test device, compensation accuracy (resolution less than 0.005mm) ). It can effectively remove the vibration error of the laser plane generator itself, and effectively improve the plane compensation accuracy.

Description

A real-time calibration compensation measurement method for flatness based on multi-PSD

technical field

The invention relates to a method for real-time flatness detection and compensation and calibration of a plane motion device, in particular to a multi-PSD-based flatness real-time calibration and compensation method, which is especially suitable for the flatness real-time calibration of large-size and high-precision electromagnetic wave plane scanning devices with compensation.

Background technique

Large-scale high-precision plane scanning device is one of the key equipments for electromagnetic wave testing such as detection of amplitude and phase characteristics of tight field quiet zone and near-field testing. At the same time, in the near-field test, the transceiver probe is always installed on the plane scanning device. By moving the probe in the plane and at the same time according to a certain sampling interval, the test target is sent and received. , wave absorption characteristics and other electromagnetic characteristics to detect. Among them, in order to ensure the detection and test accuracy, according to the RUZE theory, the root mean square error of the flatness of the scanning device is required to be less than 1% of the working wavelength. The flatness of the plane scanning device is difficult to meet the test requirements through hardware, and the flatness accuracy requirements of the scanning detection device are usually ensured by means of compensation.

At present, there are two ways to compensate the large-scale plane scanning detection device. One is to measure the flatness of the plane scanning device and interpolate to calculate the error of each position through a high-precision large-scale measuring device such as a laser tracker, and write the error compensation amount into the motion. The control system performs off-line compensation. This method cannot detect the change of the flatness error of the scanning device in real time during the detection process. It requires irregular error detection and calibration, and the real-time performance is poor; the second is to fix the position of the scanning device probe by The sensitive detector PSD can detect the flatness error of the scanning device in real time by receiving the laser signal, and feedback the error to the motion system in real time, so as to realize the dynamic real-time compensation of the flatness error. The method is to directly compare the position information received by the PSD with the initial information to obtain the displacement error, and directly feed the displacement error as the flatness error to the motion control system for compensation. In fact, the displacement error detected by the PSD includes the flatness error of the scanning device and the station error such as the vibration of the PSD sensor and the laser light source. The station error is also an important source of affecting the flatness compensation accuracy of the scanning measurement device. It is seldom reported in the existing technical documents to separate out the error and improve the real-time compensation accuracy of the flatness.

SUMMARY OF THE INVENTION

Based on the above problems, a real-time flatness calibration compensation measurement method based on multi-PSD is proposed, which is used to separate the station error from the displacement error detected by the PSD and improve the plane compensation accuracy of the plane scanning device.

The object of the present invention is achieved through the following technical solutions, and the method includes:

Step 1) PSD position information acquisition: synchronously acquire the position data of the target PSD and multiple reference PSDs ₁ , PSD ₂ ... PSD _n (n≥2) in real time, wherein the synchronization error between each PSD is not greater than: 20ms;

Step 2) Three-factor error model establishment: establish a three-factor error model of actual flatness error, station error, and external disturbance error;

Step 3) Separation of station errors: on the basis of the error model, multiple sets of test measurements are performed, data samples are added, and the station errors are separated by analyzing the data;

Step 4) Flatness real-time calibration compensation: the flatness is compensated in real time by the compensation system to ensure the flatness accuracy of the scanning test device, and the resolution of the compensation motion system is better than 0.005mm.

It can be seen from the technical solutions provided by the present invention that the real-time calibration compensation measurement method for flatness based on multi-PSD provided by the embodiment of the present invention analyzes the error source and establishes a mathematical model, adopts the multi-probe measurement method to collect data, The data analysis finds out the error relationship of the station at different positions, and the simultaneous function equation solves the error amount for compensation and calibration. The implementation scheme has low cost and wide applicability.

Description of drawings

1 is a schematic flowchart of a real-time calibration compensation measurement method for flatness based on multiple PSDs provided by an embodiment of the present invention;

Fig. 2a, Fig. 2b, Fig. 2c, Fig. 2d, Fig. 2e are respectively schematic diagrams of the test measurement model and its components according to the embodiment of the present invention;

Figure 3a and Figure 3b are respectively schematic diagrams of the accuracy error of the flatness surface before and after the test compensation according to the embodiment of the present invention;

Detailed ways

The embodiments of the present invention will be described in further detail below. Contents that are not described in detail in the embodiments of the present invention belong to the prior art known to those skilled in the art. For convenience of description, this embodiment adopts one target PSD and two reference PSDs.

The real-time calibration compensation measurement method for flatness based on multiple PSDs of the present invention, its preferred specific embodiment is:

Include steps:

PSD position information acquisition: acquire the measurement data of the target PSD and multiple reference PSD ₁ , PSD ₂ ... PSD _n (n≥2) in real time, wherein the synchronization error between each PSD is not greater than: 20ms;

Three-factor error model establishment: Establish three-factor error models of actual flatness error, station error, and external disturbance error;

Station error separation: On the basis of the error model, multiple sets of test measurements are carried out, data samples are added, and the station error is separated by analyzing the data;

Flatness real-time calibration compensation: The flatness is compensated in real time through the compensation system to ensure the flatness accuracy of the scanning test device. The resolution of the compensation motion system is better than 0.005mm.

The device model built for this method includes XY axis shift stage (1), compensated Z axis (2), target PSD (3), reference PSD ₁ (4), reference PSD ₂ (5), laser emission a device (6), wherein the compensation Z-axis (2) and the target PSD (3) constitute a scanning test device, and the target PSD (3) is fixed on the compensation Z-axis (2), wherein the compensation Z-axis contains a control driver inside;

On the XY-axis shift stage (1), the sliding stage can move along the X-axis or the Y-axis in the XY plane along with the drive motor, and the laser plane signal emitted by the laser transmitter (6) can be mapped on the XY plane. The photosensitive surface of the target PSD is moved by the slide table and received by the sensor, and the reference PSD ₁ and PSD ₂ are placed in the laser plane coverage area to receive the signal.

In the establishment of the three-factor error model, it is analyzed that the error effects of the flatness calibration compensation in the test site environment mainly include the actual flatness error of the slide guide rail, the station error and the external interference error. The overall error model is established below.

Building an overall error model includes:

Based on the analysis of the station error model, the station error in the test site environment is mainly the error caused by the environmental vibration, then the vibration model of a single position has the following form:

Where A, B, C are acceleration, velocity and displacement coefficients.

Similarly, the external interference at a single location is mainly embodied in the form of electromagnetic interference. First, the device can be equivalent to a combination of resistance, inductance and capacitance. The weak error caused by external electromagnetic interference has the following forms:

Among them, K ₁ , K ₂ and K ₃ are the interference coefficients of capacitance, inductance and resistance, and C, L and R are equivalent inductance, capacitance and resistance of the circuit.

The corresponding displacement relationship can be expressed as:

ψ(t)=Kφ(t), where K is the scaling factor.

The data Z(n) recorded by the target PSD contains the reference flatness information and the error amount information caused by the three factors. The overall three-factor error model and the mathematical model corresponding to the data collected by the target PSD and the reference PSD are established below.

Z(n)=S(n)+E(n)+ψ(n)

Among them, S(n) is the actual flatness error of the slide rail, E(n) is the error caused by the vibration of the station, and the Z-axis offset due to external electromagnetic interference is ψ(n).

Model for target PSD data:

Z _{0, i} (N)=S _{0, i} (N)+E _{0, i} (N)+ψ _{0, i} (N), i≤N

Among them: i is the i-th measurement position during the operation of the target PSD, S _{0, i} (N) is the actual flatness error of the slide rail;

The data recorded by reference PSD ₁ and reference PSD ₂ at different measurement positions are G _{1, i} (n), G _{2, i} (n), respectively, and the measurement data includes the amount of vibration error of the station and the amount of error caused by external electromagnetic interference errors. Information. In theory, m measurement probes are required, and the data are G _{m, i} (n), (m=1, 2, . Taking the data of the first reference PSD at the first measurement position as an example, the following relationship is obtained:

G _1,1 (n)=E _1,1 (n)+ψ _1,1 (n)

where E _1,1 (n) is the error caused by the vibration of the station at this point, and ψ _1,1 (n) is the Z-axis offset caused by the external electromagnetic interference at this point.

The set of first reference PSD measurements at i points is as follows:

同样G₂、G₃、…、G_M。

Likewise G ₂ , G ₃ , ..., G _M .

For the target PSD measurement data, there is G ₀ =(G _0,1 G _0,2 G _0,3 ...G _0,i ) ^T , where G _0,1 =E _0,1 (n)+ψ _0,1 ( n).

By synthesizing the error model equations at all measurement points, considering the error characteristics caused by the vibration of the station and external electromagnetic interference, the data processing transformation method is used to find the relationship between different reference PSDs at different measurement positions. Taking two reference PSD measurement data as an example, the equation system is obtained:

G ₂ =J ₁ G ₁

Among them, J ₁ is the relationship coefficient between the measured data of the first reference PSD and the second reference PSD, G ₁ is the measured data of the first reference PSD, and G ₂ is the measured data of the second reference PSD.

Synthesizing all equations, there are:

G _(m-1)×1 =J _1×(m-1) ·G′ _(m-1)×1 (m≥2)

in,

G _(m-1)×1 =(G ₂ G ₃ … G _m ) ^T , J _1×(m-1) =(J ₁ J ₂ … J _m-1 ) ^T , G′ _{(m-1)× 1} = (G ₁ G ₂ … G _m-1 ) ^T

Taking the measurement data of the first reference PSD and the second reference PSD as an example to obtain J ₁ , G ₁ and G ₂ are respectively subjected to empirical modal decomposition, and the Hilbert Huang data analysis method is used to obtain J ₁ . The above method is repeated to obtain the coefficient matrix J _1×(m-1) .

In the field environment, the flatness error caused by the vibration of the station and the external electromagnetic interference always exists in the data measured by the target PSD. According to the calculated coefficient matrix J _1×(m-1) , there is still the following relationship:

G ₀ =J ^-1 ·G _k

Among them, G _k is the signal data measured by the kth reference PSD, and G ₀ is the station error contained in the target PSD and the Z-axis offset caused by external electromagnetic interference.

The real-time measurement data information of the target PSD during the moving process includes the actual flatness position, the station error, and the error amount of the Z-axis offset caused by external electromagnetic interference. Then the Z-axis flatness that needs to be compensated is Δz:

Δz=Z(n)-J ^-1 ·G _k =S(n)

where J ^-1 ·G _k is the sum of errors caused by station errors and external electromagnetic interference.

The calculated and analyzed Δz is compensated in real time by the compensation control system to ensure the flatness accuracy of the scanning test device. At the same time, the compensated data surface is checked to verify the effectiveness of the detection method.

The multi-PSD-based flatness implementation calibration compensation measurement method of the present invention can effectively reduce the vibration error of the laser plane generator itself and effectively improve the plane compensation accuracy.

Specifically include:

First, obtain the data of the target PSD and the reference PSD ₁ and PSD ₂ respectively;

The schematic diagram of the test model includes an XY axis shift stage, a scanning measurement device (the target PSD is fixed on the compensation Z axis), the reference PSD ₁ and PSD ₂ , and a laser transmitter. The XY axis shift stage can make the scanning test device in the The XY plane moves unidirectionally along the X direction or the Y direction. At the same time, the two sets of reference PSDs can receive the laser signal in real time in the laser coverage area, and the sensor data is recorded in real time during the entire measurement process.

According to the environment in which the test is located, analyze the error that affects the flatness accuracy and has a specific form, and establish a three-factor error model in this environment;

Among them, the main factor affecting the flatness of the scanning measurement device in the test site environment is the vibration of the laser reference plane caused by the vibration of the station, and there is also the phenomenon of external electromagnetic interference. The characteristic of the real-time flatness measurement and calibration method of the present invention is that it can separate the errors caused by the vibration of the station and the interference errors, and according to the established mathematical model, adopt the corresponding data analysis method, and finally compensate Δz in real time through the compensation control system to reduce the The influence of vibration and interference on the flatness ensures the flatness accuracy of the scanning test device.

The relevant models established here for the main errors are as follows:

The data Z(n) recorded by the target PSD contains the reference flatness information and the error amount information caused by the three factors. The overall three-factor error model and the mathematical model corresponding to the data collected by the target PSD and the reference PSD are established below.

Z(n)=S(n)+E(n)+ψ(n)

Among them, S(n) is the actual flatness error of the slide rail, E(n) is the error caused by the vibration of the station, and the Z-axis offset due to external electromagnetic interference is ψ(n).

In the following, multiple sets of test measurements are performed, data samples are added, and sensor data are analyzed to separate station errors.

To simulate large-scale measurement conditions, you can change the distance between the sliding table and the laser source, and then collect the measurement data of the reference PSD at different points after fixing the distance. The purpose is to find the vibration of the test site and the error of electromagnetic signal interference at different positions. has a specific variation pattern. For example, the distance between the laser source and the slide table can be set as Ni ( _i =1,2,...,n), and then data acquisition is performed, the position measurement of the two sets of reference PSDs is changed, and the data is collected; then change the laser source The distance from the sliding table is repeated for multiple sets of tests.

Data analysis is carried out on the collected data, and the coefficient matrix J between the station error between different positions and the Z-axis offset caused by external electromagnetic interference is solved, and then the compensation amount Δz is obtained. The error at any point position is compensated in real time to ensure that the scanning measurement device maintains a high flatness accuracy during movement. This step method is to perform compensation in real time at the same time during the measurement process.

At present, the data before compensation is obtained through sensor scanning and fitted to a curved surface. The Z direction range of the plane is -0.1mm~+0.4mm, and the Z direction range of the compensated data fitting plane is -0.02mm~+0.03mm (as shown in the appendix). Figure), the flatness accuracy is effectively improved.

The multi-PSD-based flatness real-time calibration measurement method disclosed in the present invention utilizes multiple groups of PSD sensors to measure the entire test at the same time. The proportional relationship is obtained by analyzing the measurement data, separated by the error separation method, and compensated in real time during the scanning process. Based on the multi-PSD measurement, the invention separates the vibration of the station from the interference error, and controls and compensates the Z axis in real time through the compensation control system, so as to keep the operation with high flatness accuracy. Compared with the prior art, the method proposed in this patent does not depend on any special hardware. By establishing the mathematical model of the corresponding error source, using the multi-probe error separation method to solve the size of the error, and then compensating and calibrating, the economic cost of the whole scheme is low and the applicability is wide.

Specific examples:

The present invention is based on multi-PSD flatness real-time calibration compensation measurement, and the basic idea of the invention is to establish an error factor model by utilizing the characteristic that the error has a deterministic variation law in different space states. Select multiple sets of photoelectric sensors with higher accuracy to measure at the same time, analyze and process the data to obtain the compensation amount Δz, and compensate in real time through the compensation system to ensure the flatness accuracy of the measuring device. The schematic flowchart is shown in Figure 1.

According to the characteristics of the error law, the test and measurement models of Fig. 2a, Fig. 2b, Fig. 2c, Fig. 2d, Fig. 2e are established, including XY axis shift stage, scanning measurement device (target PSD and compensation Z axis, compensation Z axis includes control driver ), reference PSD ₁ and reference PSD ₂ , laser transmitter. The laser transmitter is fixed on the bracket to generate the laser plane.

The function of the XY axis shift stage is to simulate a large-scale scanning frame, in which the slide table moves along the X or Y axis direction driven by the driving stepper motor. The scanning measurement device here is composed of the target PSD and the compensation Z axis, and the target PSD is fixed. On the Z axis, the function is to detect the laser signal, and accept the control signal for real-time compensation, so as to ensure that the Z axis is always in the reference plane.

The reference PSD ₁ and PSD ₂ are always in the laser coverage range, and the laser plane signal emitted by the laser transmitter can be mapped on the photosensitive surface of the target PSD moving with the slide table and received and controlled.

A three-factor error model was established.

The next step is to carry out test measurements. _Multiple groups of tests can be established, such as groups _{A 1} , _{A 2} , . . . , start the slide table to move along the X axis on the XY axis shift table, and at the same time, the two sets of reference PSDs measure and collect data at two points a ₁ and b ₁ until the target PSD runs to the end point on the X axis, at which time the two sets of reference PSDs are changed. The measurement position of the reference PSD is a ₂ , b ₂ and then continue to collect data, and at the same time start the slide again to run along the X-axis towards the opposite end of the other side. The positions where the reference PSD was added during the experiment of the A1 group were (a ₃ , b ₃ ), (a ₄ , b ₄ ), _. . . , ( _ai , b _i ). After the measurement of group A ₁ is completed, change the distance to N ₂ , and use the same steps to perform the test of group A ₂ , and so on.

The Δz is obtained by analyzing the measurement data, and the compensation control system compensates the error Δz on the Z axis in real time, so as to ensure that the control scanning measurement device maintains a high-precision flatness during the movement.

The above description is only a preferred embodiment of the present invention, but the protection scope of the present invention is not limited to this. Substitutions should be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims (3)

1. a flatness real-time calibration compensation measurement method based on multiple PSD, is applicable to multiple groups of reference PSD compensation, is characterized in that, comprises the steps: Step 1) PSD location information acquisition: synchronously acquire the location data of the target PSD and multiple reference PSDs ₁ , PSD ₂ . . . PSD _n in real time, where n≥2, and the synchronization error between each PSD is not greater than: 20ms; Step 2) Three-factor error model establishment: establish a three-factor error model of actual flatness error, station error, and external disturbance error; Step 3) Separation of station errors: on the basis of the error model, multiple sets of test measurements are performed, data samples are added, and the station errors are separated by analyzing the data; Step 4) Flatness real-time calibration compensation: the flatness is compensated in real time through a compensation system, and the resolution of the compensation system is better than 0.005mm; The device model built for this method includes an XY axis shift stage (1), a compensated Z axis ( ₂ ), a target PSD ( ₃ ), a reference _PSD1 , PSD2...PSDn, a laser transmitter (6) , wherein the compensation Z-axis (2) and the target PSD (3) constitute a scanning test device, and the target PSD (3) is fixed on the compensation Z-axis (2), wherein the compensation Z-axis contains a control driver inside; On the XY-axis shift stage (1), the sliding stage can move along the X-axis or the Y-axis in the XY plane along with the drive motor, and the laser plane signal emitted by the laser transmitter (6) can be mapped on the XY plane. The photosensitive surface of the target PSD on which the sliding stage moves is received by the sensor, and the reference PSD ₁ , PSD ₂ ... PSD _n is placed in the coverage area of the laser plane to receive the signal; Obtain analog or digital error signals through related processing modules; For the establishment of the three-factor error model, it is firstly analyzed that the error effects of the flatness calibration compensation under the test site environment include the actual flatness error of the slide guide rail, the station error and the external interference error; The overall three-factor error model and the mathematical model corresponding to the data collected by the target PSD and multiple reference PSD ₁ , PSD ₂ ... PSD _n are established below: Z(n)=S(n)+E(n)+ψ(n) Among them, S(n) is the actual flatness error of the slide rail, E(n) is the error caused by the vibration of the station, and the Z-axis offset due to external electromagnetic interference is ψ(n); Model for target PSD data: Z _0,i (N)=S _0,i (N)+E _0,i (N)+ψ _0,i (N), i≤N Among them: i is the i-th measurement position during the operation of the target PSD, and S _0,i (N) is the actual flatness error of the slide rail; The data recorded by reference PSD ₁ and reference PSD ₂ at different measurement positions are G _1,i (n) and G _2,i (n), respectively, where the measurement data includes station vibration errors and errors caused by external electromagnetic interference errors. In theory, m measurement probes need to be used, and the data is G _m,i (n) represents the data recorded by the mth reference PSD at the ith measurement point, m=1,2,...,n, Taking the data of the first reference PSD ₁ at the first measurement position as an example, the following relationship is obtained: G _1,1 (n)=E _1,1 (n)+ψ _1,1 (n) where E _1,1 (n) is the error caused by the vibration of the station at this point, and ψ _1,1 (n) is the Z-axis offset caused by external electromagnetic interference at this point; The first set of reference PSD ₁ measurements at i points is as follows:

Likewise G ₂ , G ₃ , ..., G _M ; For the target PSD measurement data, there is G ₀ =(G _0,1 G _0,2 G _0,3 ... G _0,i ) ^T , where G _0,1 =E _0,1 (n)+ψ _0,1 ( n); Synthesizing the error model equations at all measurement points, considering the error characteristics caused by both the vibration of the station and the external electromagnetic interference, the data processing transformation method is used to find the relationship between different reference PSDs at different measurement positions. Take the PSD measurement data as an example to get the equation system: G ₂ =J ₁ G ₁ Wherein J ₁ is the relationship coefficient between the measured data of the first reference PSD ₁ and the second reference PSD ₂ , G ₁ is the measured data of the first reference PSD ₁ , and G ₂ is the measured data of the second reference PSD ₂ ; Synthesizing all equations, there are: G _(m-1)×1 =J _1×(m-1) ·G′ _(m-1)×1 , m≥2 in G _(m-1)×1 =(G ₂ G ₃ ···G _m ) ^T ,J _1×(m-1) =(J ₁ J ₂ ···J _m-1 ) ^T ,G′ _{(m-1)× 1} = (G ₁ G ₂ …G _m-1 ) ^T Take the measurement data of the first reference PSD ₁ and the second reference PSD ₂ as an example to obtain J ₁ , perform empirical mode decomposition on G ₁ and G ₂ respectively, and use the data analysis method of Hilbert-Huang transform to obtain J ₁ , repeat the above method and then obtain the coefficient matrix J _1×(m-1) ; In the field environment, the flatness error caused by the vibration of the station and the external electromagnetic interference always exists in the data measured by the target PSD. According to the calculated coefficient matrix J _1×(m-1) , there is still the following relationship: G ₀ =J ^-1 ·G _k Among them, G _k is the signal data measured by the kth reference PSD _k , and G ₀ is the station error contained in the target PSD and the Z-axis offset caused by external electromagnetic interference. 2. the flatness real-time calibration compensation measurement method based on multiple PSDs according to claim 1, is characterized in that, carrying out multiple groups of tests to increase data samples, after analyzing the data and solving Δz, comprising: The real-time measurement data information of the target PSD during the moving process includes the actual flatness position, the station error, and the error amount of the Z-axis offset caused by external electromagnetic interference. The Z-axis flatness that needs to be compensated is Δz: Δz=Z(n)-J ^-1 ·G _k =S(n) where J ^-1 ·G _k is the sum of errors caused by station errors and external electromagnetic interference. 3. the flatness real-time calibration compensation measurement method based on multi-PSD according to claim 2, is characterized in that, calculates the flatness error Δz of scanning test device at any position, and carries out real-time compensation by compensating system, guarantees scanning Test the flatness accuracy of the device.

Images