Dynamic error compensation method for scanning probe calibration

Abstract

The invention relates to the technical field of three-coordinate measuring machines, and particularly discloses a dynamic error compensation method for scanning probe calibration, which comprises the following steps: calibrating a probe deflection dynamic model based on a three-axis linear displacement sensor signal and delay time to obtain a calibrated probe deflection dynamic model; in the process of measuring the measured object by the three-coordinate measuring machine, recording the theoretical measurement position of the scanning probe and the signal of the three-axis linear displacement sensor at each sampling moment; performing interpolation processing on the three-axis linear displacement sensor signal at each sampling moment according to the delay time to obtain a three-axis linear displacement sensor signal at the delay moment; and inputting the signals of the three-axis linear displacement sensor at the delay moment into the calibrated measuring head deflection dynamic model for deflection calculation so as to obtain the actual deflection of the scanning measuring head at the current sampling moment. According to the dynamic error compensation method for scanning probe calibration provided by the invention, the calibration time is shortened, and meanwhile, the measurement precision is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of three-coordinate measuring machines, and more particularly to a dynamic error compensation method for scanning probe calibration. BACKGROUND

[0002] Contact three-coordinate measuring machines (CMMs) are one of the most widely used industrial measuring instruments due to their high accuracy, robustness, and versatility. Probe systems are the key components of three-coordinate measuring machines, which determine the measurement mode of the machine. Contact probe systems can be divided into two categories: touch trigger probes and scanning probes. Touch trigger probes can only output an on or off signal when they contact the measured surface, resulting in pre-travel errors due to machine motion. Scanning probes can output continuous analog signals based on the deflection of the probe stylus, so they can eliminate the inherent pre-travel errors of touch trigger probes. However, since scanning probes require more parameters to characterize, the calibration of scanning probes is more complex than that of touch trigger probes. Renishaw provides a calibration program that requires scanning multiple latitudes and longitudes forward and backward on a sphere, but the calibration method of scanning probes is usually very time-consuming, with calibration time being more than 6 times that of touch trigger probes.

[0003] At the same time, traditional three-coordinate measuring machines can achieve micron-level measurement accuracy within their measurement range, regardless of whether they use touch probes or scanning probes, meeting the accuracy requirements of most industrial tests. However, existing research has not considered the dynamic effects of probes, mainly focusing on the dynamic errors of three-coordinate measuring machine axes, and the impact of dynamic errors on probe systems has not been given enough attention. Therefore, for scanning probes on three-coordinate measuring machines, it is very important to design a model and calibration method that can accurately and effectively characterize probe signals while keeping parameters simple and easy to identify. SUMMARY

[0004] The present application aims to overcome the above-mentioned defects and provide a dynamic error compensation method for scanning probe calibration, which can shorten the calibration time while improving the measurement accuracy.

[0005] As a first aspect of the present application, a dynamic error compensation method for scanning probe calibration is provided, wherein the scanning probe is installed at the end of the Z-axis column of a three-coordinate measuring machine, and a three-axis linear displacement sensor is provided in the scanning probe, which includes an X-axis linear displacement sensor, a Y-axis linear displacement sensor, and a Z-axis linear displacement sensor. When the scanning probe contacts the measured object, a certain deflection occurs between the actual measurement position of the scanning probe and the theoretical measurement position. The theoretical measurement position of the scanning probe is represented as [X p , Y p , Z p ] T, the actual measurement position of the scanning probe is represented as [x p , y p , z p ] T , the actual deflection of the scanning probe is represented as [d x , d y , d z ] T ;

[0006] The dynamic error compensation method for the scanning probe calibration comprises the following steps:

[0007] Step S1: constructing a probe deflection dynamic model based on three-axis linear displacement sensor signals and delay time;

[0008] Step S2: calibrating the probe deflection dynamic model based on three-axis linear displacement sensor signals and delay time by a 25-point calibration method to obtain a calibrated probe deflection dynamic model;

[0009] Step S3: recording the scanning probe theoretical measurement position and three-axis linear displacement sensor signals at each sampling time during the measurement of the measured object by the three-coordinate measuring machine; wherein the three-axis linear displacement sensor signals are X-axis linear displacement sensor signal, Y-axis linear displacement sensor signal and Z-axis linear displacement sensor signal respectively;

[0010] Step S4: interpolating the three-axis linear displacement sensor signals at each sampling time according to the delay time to obtain the three-axis linear displacement sensor signals at the delay time; wherein the delay time is the current sampling time plus the delay time;

[0011] Step S5: inputting the three-axis linear displacement sensor signals at the delay time into the calibrated probe deflection dynamic model for deflection calculation to obtain the actual deflection of the scanning probe at the current sampling time.

[0012] Further, the scanning probe is also provided with parallel leaf springs, which are X-axis parallel leaf springs, Y-axis parallel leaf springs and Z-axis parallel leaf springs respectively, and the X-axis linear displacement sensor, the Y-axis linear displacement sensor and the Z-axis linear displacement sensor are all provided with magnetic cores, the X-axis parallel leaf spring is connected with the magnetic core on the X-axis linear displacement sensor, the Y-axis parallel leaf spring is connected with the magnetic core on the Y-axis linear displacement sensor, and the Z-axis parallel leaf spring is connected with the magnetic core on the Z-axis linear displacement sensor, the scanning probe moves after contacting the measured object, the scanning probe can drive the X-axis parallel leaf spring, the Y-axis parallel leaf spring and the Z-axis parallel leaf spring to move, the movement of the X-axis parallel leaf spring, the Y-axis parallel leaf spring and the Z-axis parallel leaf spring can drive the magnetic cores connected therewith to move, at this time, the X-axis linear displacement sensor, the Y-axis linear displacement sensor and the Z-axis linear displacement sensor output displacement signals, and the displacement signals output by the X-axis linear displacement sensor, the Y-axis linear displacement sensor and the Z-axis linear displacement sensor are proportional to the displacement of the respective magnetic cores.

[0013] Further, the construction of the probe deflection dynamic model based on the three-axis linear displacement sensor signals and the delay time also includes:

[0014] The probe deflection dynamic model based on the three-axis linear displacement sensor signals and the delay time is expressed as:

[0015]

[0016] wherein, is the X-axis linear displacement sensor signal corresponding to the i+t1 moment, is the Y-axis linear displacement sensor signal corresponding to the i+t2 moment, is the Z-axis linear displacement sensor signal corresponding to the i+t3 moment, i represents the current moment, t1, t2 and t3 all represent the delay time, k px , k qy and k rz are linear terms, h p , h q and h r are quadratic terms, k qx , k rx , k py , k ry , k pz and k qz are cross-coupling terms.

[0017] Further, the three-axis linear displacement sensor signal and the delay time based probe deflection dynamic model is calibrated by the 25-point calibration method to obtain a calibrated probe deflection dynamic model, and the calibration method further includes:

[0018] Determine the parameter k in the probe deflection dynamic model based on the three-axis linear displacement sensor signal and the delay time through the 25-point calibration method px , k qy , k rz , h p , h q , h r , k qx , k rx , k py , k ry , k pz , k qz , t1, t2 and t3, to obtain the calibrated probe deflection dynamic model

[0019] The process of the 25-point calibration method is as follows:

[0020] Select 25 points on the standard sphere, each point has a different normal direction, measure the 25 points on the standard sphere by the scanning probe, the deflection generated by the scanning probe when measuring the 25 points is different; when measuring each point by the scanning probe, it needs to go through the approach, stable and retract three stages, in the calibration process, the data of the approach, stable and retract three stages are used, 10 samples are selected from the approach, stable and retract three stages, 750 sample points are generated in the measurement of 25 points, the 750 sample points are located on a sphere with a radius equal to the radius of the standard sphere plus the radius of the probe tip, when the scanning probe measures the ith sample point, the theoretical measurement position P i =[X i , Y i , Z i ] of the scanning probe is obtained by the three-coordinate measuring machine, at this time, the X-axis linear displacement sensor signal p, the Y-axis linear displacement sensor signal q and the Z-axis linear displacement sensor signal r are collected, formula (1) is solved through the LM algorithm, to obtain the parameter k in the probe deflection dynamic model based on the three-axis linear displacement sensor signal and the delay time px , k qy , k rz , h p , h q , h r , k qx , k rx , k py , k ry , k pz , k qz , t1, t2 and t3

[0021]

[0022] Wherein, [X c , Y c , Zc ] is the coordinate of the center of the standard sphere, R is the radius of the standard sphere, r is the radius of the scanning probe tip, [X i , Y i , Z i ] is the theoretical measurement position of the scanning probe when measuring the i-th sample point.

[0023] Furthermore, the three-axis linear displacement sensor signal signal at the delay time i+t Expressed as:

[0024]

[0025] Among them, signal i+t is the X-axis linear displacement sensor signal p at the delay time i+t , Y-axis linear displacement sensor signal q at the delay time i+t Or the Z-axis linear displacement sensor signal r at the delayed time i+t , i is the current sampling time, t is the delay time, i+t is the delay time, and t is t1, t2 or t3.

[0026] Furthermore, it also includes:

[0027] According to the theoretical measurement position of the scanning probe at the current sampling moment [X p , Y p , Z p ] T and the actual deflection of the scanning probe at the current sampling moment [d x , d y , d z ] T , calculate the actual measurement position of the scanning probe at the current sampling moment [x p ,y p , z p ] T , the calculation formula is as follows:

[0028]

[0029] The present invention provides a method for dynamic error compensation in scanning probe calibration, which has the following beneficial effects: a simple time delay constant is used to compensate for the dynamic error in the probe signal, a corresponding calibration model and method are established to identify the constant time delay, and calibration is completed by probing 25 points on a sphere, thereby shortening the calibration time and improving the measurement accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. Together with the following specific embodiments, they are used to explain the present invention, but do not constitute a limitation of the present invention.

[0031] Figure 1 A flow chart of a dynamic error compensation method for scanning probe head calibration provided by the present application.

[0032] Figure 2 A structural schematic diagram of a three-coordinate measuring machine provided by the present application.

[0033] Figure 3 A working principle diagram of a scanning probe head provided by the present application.

[0034] Figure 4 A structural schematic diagram of a scanning probe head provided by the present application.

[0035] Figure 5 A scanning probe head deflection schematic diagram after a scanning probe head tip contacts a surface of a measured object provided by the present application.

[0036] Figure 6 A scanning probe head theoretical deflection curve and a scanning probe head actual deflection curve comparison schematic diagram in a proximity phase provided by the present application.

[0037] Figure 7 A 25-point distribution diagram of a standard sphere provided by the present application.

[0038] Figure 8 A probe signal change diagram with machine motion provided by the present application. DETAILED DESCRIPTION

[0039] In order to further clarify the technical means and effects of the present application for achieving the predetermined inventive objectives, the following describes in detail the specific implementation, structure, features and effects of a dynamic error compensation method for scanning probe head calibration according to the present application, with reference to the accompanying drawings and preferred embodiments. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the protection scope of the present application.

[0040] It should be noted that the terms "first", "second", and the like in the specification and claims of the present application and the above-described accompanying drawings are used to distinguish similar objects, and do not necessarily indicate a specific order or sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the present application described herein can be implemented. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device including a series of steps or units does not necessarily limit to those steps or units clearly listed, but can include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0041] A dynamic error compensation method for scanning probe calibration is provided in the embodiment, as shown in the accompanying drawings Figures 2-4 The scanning probe 2 is installed at the end of the Z-axis column of the coordinate measuring machine 1, and a three-axis linear displacement sensor 7 is arranged in the scanning probe 2, which is an X-axis linear displacement sensor, a Y-axis linear displacement sensor and a Z-axis linear displacement sensor respectively. When the scanning probe 2 contacts the measured object 3, a certain deflection 6 is generated between the actual measurement position 5 of the scanning probe and the theoretical measurement position 4. The theoretical measurement position 4 of the scanning probe is represented as [X p , Y p , Z p ] T The theoretical measurement position of the scanning probe is obtained by the coordinate measuring machine, and the actual measurement position 5 of the scanning probe is represented as [x p , y p , z p ] T The actual deflection 6 of the scanning probe is represented as [d x , d y , d z ] T ; wherein the actual measurement position [x p , y p , z p ] T of the scanning probe is represented as the superposition of the theoretical measurement position [X p , Y p , Z p ] T of the scanning probe and the actual deflection [d x , d y , d z ] T of the scanning probe:

[0042]

[0043] It should be noted that the controller of the coordinate measuring machine obtains the theoretical measurement position [X p , Y p , Z p ] T of the scanning probe according to the movement values of the X-axis, Y-axis and Z-axis motors, and therefore, in order to obtain the final actual measurement position [x p , y p , z p ] T , the actual deflection [d x , d y , d z ] T of the scanning probe needs to be accurately obtained.

[0044] It should be understood that, as shown in Figure 4 The structure of the entire scanning probe can be regarded as a mechanism composed of three independent structures of X-axis, Y-axis and Z-axis, wherein the linear displacement sensors LVDT are installed along the axes, and the signals output by the X-axis linear displacement sensor, Y-axis linear displacement sensor and Z-axis linear displacement sensor are respectively denoted as p, q and r. The calibration is to find the relationship between the linear displacement sensor output signals p, q and r and the probe deflection [d x , d y , d z ] T .

[0045] As shown in Figure 4 The scanning probe 2 is also provided with parallel leaf springs 8, which can perform approximate linear motion within a relatively short distance, and the linear displacement sensor can accurately detect the change in motion during the precise displacement of the parallel leaf springs 8. The X-axis parallel leaf spring, Y-axis parallel leaf spring and Z-axis parallel leaf spring are provided with magnetic cores 9 on the X-axis linear displacement sensor, Y-axis linear displacement sensor and Z-axis linear displacement sensor, respectively. The X-axis parallel leaf spring is connected with the magnetic core 9 on the X-axis linear displacement sensor, the Y-axis parallel leaf spring is connected with the magnetic core 9 on the Y-axis linear displacement sensor, and the Z-axis parallel leaf spring is connected with the magnetic core 9 on the Z-axis linear displacement sensor. The scanning probe 2 will move after contacting the measured object 3. The scanning probe 2 can drive the X-axis parallel leaf spring, Y-axis parallel leaf spring and Z-axis parallel leaf spring to move. The movement of the X-axis parallel leaf spring, Y-axis parallel leaf spring and Z-axis parallel leaf spring can drive the magnetic cores 9 connected thereto to move. At this time, the X-axis linear displacement sensor, Y-axis linear displacement sensor and Z-axis linear displacement sensor will output displacement signals. The displacement signals output by the X-axis linear displacement sensor, Y-axis linear displacement sensor and Z-axis linear displacement sensor are proportional to the displacement of the respective magnetic cores 9.

[0046] As shown in Figure 1 The dynamic error compensation method for calibrating the scanning probe includes the following steps:

[0047] Step S1: constructing a probe deflection dynamic model based on three-axis linear displacement sensor signals and delay time;

[0048] Preferably, the construction of the probe deflection dynamic model based on three-axis linear displacement sensor signals and delay time further includes:

[0049] Mathematical kinematic relationship between LVDT and stylus tip displacement: Both the parallel leaf spring and the LVDT exhibit good linear characteristics in a small displacement range, the relationship between them should be represented as a linear model. Due to manufacturing and assembly errors, the three LVDTs cannot be perfectly aligned with the X, Y and Z axes, introducing the corresponding cross-coupling terms, and adding a quadratic term to improve the robustness of the model, and then to cope with the nonlinearity of the stylus and large displacement. In previous work, the mathematical model of the probe under quasi-static conditions was established as:

[0050]

[0051] Considering that the parallel leaf spring and the LVDT both exhibit strong linear characteristics within a limited deflection range, the probe mathematical model is expressed as a semi-linear model. However, this semi-linear model does not take into account the time delay. When the detection speed increases, the measurement accuracy decreases significantly. Therefore, the probe calibration model is optimized for dynamic conditions.

[0052] In actual detection measurement, it can be observed that there is a difference between the voltage signal and the actual deflection of the scanning stylus. Currently, there is no direct method to obtain the accurate deflection of the scanning stylus tip. Therefore, the motion of the coordinate measuring machine after the stylus tip contacts the measured surface is taken as the theoretical stylus deflection, as shown in Figure 5 At the same time, the actual stylus deflection can be calculated by the LVDT signal of the stylus model. When the coordinate measuring machine is controlled to move along the x-axis perpendicular to the fixed plane at a speed of 2 mm / s, the motion signal of the coordinate measuring machine and the LVDT signal are recorded simultaneously, and then the stylus theoretical deflection and the stylus actual deflection are calculated, as shown in Figure 6 When the stylus tip maintains contact with the measured surface, the stylus estimated deflection and the stylus theoretical deflection should be the same. However, in actual measurement, the stylus estimated deflection shows a significant delay relative to the stylus theoretical deflection. When approaching the measured surface using different speeds, a roughly consistent time delay can be observed. The above phenomenon illustrates that the existing quasi-static stylus model cannot display the dynamic characteristics of the stylus during detection. The delay of the detection signal will seriously affect the accuracy of the detection, especially at high detection speeds. Assuming that the signal delay is 20 ms and the detection speed is 1 mm / s, the single-point measurement error will reach 0.02 mm, which is much larger than the motion error of the coordinate measuring machine.

[0053] The parallel leaf spring mechanism used in the probe structure is a common flexible parallelogram module used for linear displacement guidance in nanopositioning. This system is usually identified as a second-order model. The typical detection process includes approach, stabilization, and retraction stages. In the approach stage, the probe contacts the surface to be measured with a uniform linear motion. From the perspective of control system theory, the response of the detection signal should be the response of a second-order system under ramp excitation. The actual response of the probe signal is as follows: Figure 6 As shown in the figure, during the probe retraction phase, the probe moves away from the surface being measured in a uniform linear motion, and the probe signal response is similar to that during the approach phase. Experimental experience shows that in a second-order model, at low frequencies, there is an approximately linear relationship between phase and frequency. In this case, the effect of phase lag manifests as a nearly constant time delay in the time domain. By compensating for this time delay during probe calibration and measurement, dynamic errors during the probe phase can be significantly reduced.

[0054] The probe detection process is similar to the response of a second-order system excited by a ramp. Each channel detection signal, namely p, q, and r, needs to be compensated by a constant delay vector t = [t1, t2, t3]. By adding a time delay element to the detection signal, the quasi-static model is transformed into a comprehensive dynamic model that includes time delay. That is, the dynamic model of the probe deflection based on the three-axis linear displacement sensor signal and the delay time is expressed as:

[0055]

[0056] in, is the X-axis linear displacement sensor signal corresponding to the i+t1 moment, is the Y-axis linear displacement sensor signal corresponding to the i+t2 moment, is the Z-axis linear displacement sensor signal corresponding to the i+t3 moment, i represents the current moment, t1, t2, and t3 are the delay times of the p, q, and r signals respectively, and k px 、k qy and k rz are all linear terms, h p 、h q and h r are all quadratic terms, k qx 、k rx 、k py 、k ry 、k pz and k qz The semi-linear model of probe deflection simplifies the model parameters, avoids overfitting problems, and reduces the need for a large amount of calibration data.

[0057] In actual operation, the sampling rate of the three-coordinate measuring machine controller is 1000hz, that is, one point is collected every 1ms, but the time delay of the signal may not be an accurate integer, and interpolation needs to be performed on the signal occurring in the sub-sampling period.

[0058] For example, the three-axis linear displacement sensor signal signal i+t is expressed as:

[0059]

[0060] signal i+t is the X-axis linear displacement sensor signal p i+t , the Y-axis linear displacement sensor signal q i+t , or the Z-axis linear displacement sensor signal r i+t at the delay time, i is the current sampling time, t is the delay time, i+t is the delay time, and t is t1, t2, or t3.

[0061] Step S2: calibrating the probe deflection dynamic model based on the three-axis linear displacement sensor signal and the delay time by a 25-point calibration method to obtain a calibrated probe deflection dynamic model;

[0062] Preferably, the calibration of the probe deflection dynamic model based on the three-axis linear displacement sensor signal and the delay time by the 25-point calibration method to obtain the calibrated probe deflection dynamic model further comprises:

[0063] The parameters k px , k qy , k rz , h p , h q , h r , k qx , k rx , k py , k ry , k pz , k qz , t1, t2, and t3 in the probe deflection dynamic model based on the three-axis linear displacement sensor signal and the delay time are determined by the 25-point calibration method to obtain the calibrated probe deflection dynamic model.

[0064] The process of the 25-point calibration method is as follows:

[0065] In order to determine the above model parameters k px , k qy , k rz , h p , h q , h r , kqx , k rx , k py , k ry , k pz , k qz , t1, t2 and t3, 25 points on the standard sphere are selected for calibration, as shown in Figure 7 , the distribution of 25 points refers to the ISO 10360-5 standard, and each point has a different normal direction, resulting in different deflections of the probe at each point. The 25 points on the standard sphere are measured by the three-coordinate scanning probe, and the deflection of the three-coordinate scanning probe when measuring the 25 points is different; the three-coordinate scanning probe measures each point and experiences three stages of approach (S1), stabilization (S2) and retraction (S3), as shown in Figure 8 , the relationship between the X-axis motor movement and the p signal is obvious, and the movement-signal relationship of the Y-axis and the Z-axis is similar to that of the X-axis. Each detection starts from the approach point, so in the first half of S1-1, the probe tip does not contact the spherical surface, and the detection signal remains unchanged. After the probe tip contacts the spherical surface, the detection signal gradually increases with the increase of the probe deflection. It should be noted that due to the existence of time delay, the voltage signal is not synchronized with the movement signal. Specifically, at the junction of S1-1 and S1-2, i.e. when the scanning probe just contacts the measured surface, the probe signal will not change immediately. The retraction stage is similar, S3-1 represents the stage when the probe tip still contacts the surface of the standard sphere during retraction, and S3-2 represents the stage when the probe tip leaves the surface of the standard sphere; during the calibration process, the data of S1-2, S2 and S3-1 stages are used, and in these three stages, when the three-coordinate scanning probe measures the i-th sample point, the controller of the three-coordinate measuring machine acquires the theoretical measurement position P i = [X i , Y i , Z i ] of the three-coordinate scanning probe at a sampling rate of 1000HZ, at this time, the X-axis linear displacement sensor signal p i , the Y-axis linear displacement sensor signal q i , and the Z-axis linear displacement sensor signal r i are collected, more sample points are sampled in S1-2 and S3-1 stages because the signal delay has the most significant effect in these two stages. All these sample points should be located on a sphere with a radius equal to R+r, where R is the radius of the standard sphere and r is the radius of the scanning probe tip, and the parameters k px , k qy , k rz , h in the probe deflection dynamic model based on the three-axis linear displacement sensor signals and the delay time are solved by the LM algorithm (Levenberg-Marquardt).p , h q , h r , k qx , k rx , k py , k ry , k pz , k qz , t1, t2 and t3

[0066]

[0067] where [X c , Y c , Z c ] is the center coordinate of the standard sphere, and [X i , Y i , Z i ] is the theoretical measurement position of the scanning probe when measuring the ith sample point.

[0068] It should be noted that the initial value of the parameter is set, the initial value of the delay time t can be set to zero, the initial value of the radius is obtained from the theoretical standard sphere radius and the theoretical tip radius, the voltage displacement curve in the approach stage is analyzed to obtain the initial value of the linear term. Specifically, it is assumed that the tip center remains unchanged after contacting the spherical surface, and the displacement generated during the approach of the machine is equal to the displacement generated by the probe itself. Therefore, the initial value of the linear term can be expressed as the ratio of displacement to voltage difference. For example, the initial value of k px can be expressed as dx / dp, and k py and k pz can be obtained by analyzing the voltage-displacement curves of the y-axis and z-axis. Because the time delay of the probe signal only affects the phase of the signal, but not the amplitude of the signal, the linear term will not be affected by the time delay. For the cross-coupling term and the quadratic term, they are initially set to zero. Finally, the center coordinates of the sphere are determined by manual probe point fitting.

[0069] Step S3: During the measurement of the measured object by the three-coordinate measuring machine, the theoretical measurement position of the scanning probe and the three-axis linear displacement sensor signals at each sampling time are recorded; wherein the three-axis linear displacement sensor signals are respectively the X-axis linear displacement sensor signal, the Y-axis linear displacement sensor signal and the Z-axis linear displacement sensor signal;

[0070] Step S4: The three-axis linear displacement sensor signals at each sampling time are interpolated according to the delay time to obtain the three-axis linear displacement sensor signals at the delay time; wherein the delay time is the current sampling time plus the delay time;

[0071] Step S5: inputting the three-axis linear displacement sensor signal of the delay moment into the calibrated probe deflection dynamic model for deflection calculation to obtain the actual deflection of the scanning probe at the current sampling moment [d x , d y , d z ] T .

[0072] Specifically, it further comprises:

[0073] According to the theoretical measurement position [X p , Y p , Z p ] T of the scanning probe at the current sampling moment and the actual deflection of the scanning probe at the current sampling moment [d x , d y , d z ] T , the actual measurement position [x p , y p , z p ] T of the scanning probe at the current sampling moment is calculated, and the calculation formula is as follows:

[0074]

[0075] Measurement accuracy method verification: in a constant temperature room with a temperature of 20°, a same coordinate measuring machine is used, a scanning probe (here, a self-designed MWU probe is used) and a Renishaw SP25 probe are sequentially arranged, and precision comparison verification is sequentially performed, as shown in Table 1, four groups of verification objects are named as MWU_s, MWU_f, SP_r and MWU_p, MWU_s represents that the MWU probe 25-point method is used for calibration, and points are collected at a slow speed in the S2 stable stage; MWU_f represents that the MWU probe 25-point method is used for calibration, and points are collected at a fast speed in the S2 stable stage; SP_r represents that the Renishaw SP25 probe Renishaw method is used for calibration; MWU_p represents that the MWU probe 25-point method is used for calibration, and points are collected at a fast speed in the S3-1 back-off stage; and verification is performed on the four objects in the order of the four experiments. Specifically, the signals collected in the stable stage are not affected by dynamic errors, while the signals collected in the shrinkage stage are affected by dynamic errors. The simulation results verify the effectiveness of the model under dynamic conditions.

[0076] First, the standard ball P From is compared at different dotting speeds. As shown in Table 2, 25 points are measured on the other half of the standard ball, the deviation of each point relative to the center of the least square sphere is calculated, and the maximum deviation is subtracted from the minimum deviation (referred to as “P FormThe probe deflection is 0.18mm, and the stable stage sampling and the back-off stage sampling are respectively used, and different probe speeds are used for measurement, so as to verify the dynamic performance of the probe under the calibration method.

[0077] Secondly, the standard ball error under different scanning speeds is compared. As shown in Table 3, the probe deflection is 0.18mm, the standard ball with a diameter of 44.9613mm and a shape error of 0.08um is measured by using different speeds to perform the upper hemisphere spiral scanning, and the standard ball radius error and P From error are evaluated, and the dynamic performance of the probe under the calibration method is verified.

[0078] Thirdly, the length measurement precision is compared. The probe deflection is 0.18mm, and the length measurement is performed on the length block with a length of 29.9998mm along the X-axis, the Y-axis and the XY diagonal direction axis of the coordinate measuring machine. In the measurement, as shown in Table 4, the signal in the stable stage S2 is used as the data for evaluating the length, and in another measurement, as shown in Table 5, the signal in the back-off stage S3-1 is used as the data for evaluating the length, and the dynamic performance of the probe under the calibration method is verified.

[0079] Table 1 comparison of experimental objects

[0080] Test subjects MWU_s MWU_f SP_r MWU_p Calibration method 25-point calibration 25-point calibration Ranscho calibration 25-point calibration Sampling phase S2 S2 S2 S3-1 Calibration deflection range (mm) 0.18 0.18 0.2-0.5 0.18 Calibration speed (mm / s) 0.05 1 3 1 Calibration time (s) 215 102 587 102

[0081] Table 2 comparison of P From of the ball standard ball under different dotting speeds (um)

[0082]

[0083] Table 3 comparison of standard ball error (um) under different scanning speeds

[0084]

[0085] Table 4 comparison of length measurement precision of the probe (stable stage data acquisition)

[0086]

[0087] Table 5 comparison of length measurement precision of the probe (back-off stage data acquisition)

[0088]

[0089] The dynamic error compensation method for the scanning probe calibration provided by the application proposes a probe model containing a delay time, and the model can compensate the dynamic error by using a simple time delay constant. Compared with the traditional calibration method, the probe model containing the delay time proposed in the application has higher precision in the measurement test, and the calibration time is only 17.4% of that of the traditional calibration method.

[0090] The above merely describes preferred embodiments of the present application, and is not intended to limit the present application in any form. Although the present application has been disclosed with the preferred embodiments as above, it is not intended to limit the present application, and any person skilled in the art can make some changes or modifications to the above disclosed technical contents to obtain equivalent embodiments with equivalent changes, as long as the changes or modifications do not depart from the technical solution of the present application. Any simple modification, equivalent change and modification made to the above embodiments according to the technical essence of the present application still falls within the scope of the technical solution of the present application.

Claims

1. A dynamic error compensation method for scanning probe head calibration, characterized by, The scanning probe (2) is installed at the end of the Z-axis column of the three-coordinate measuring machine (1), a three-axis linear displacement sensor (7) is arranged in the scanning probe (2), the three-axis linear displacement sensor (7) is respectively an X-axis linear displacement sensor, a Y-axis linear displacement sensor and a Z-axis linear displacement sensor, when the scanning probe (2) contacts the measured object (3), a certain deflection (6) of the actual measurement position (5) of the scanning probe relative to the theoretical measurement position (4) is generated, the theoretical measurement position (4) of the scanning probe is represented as [X p , Y p , Z p ] T , the actual measurement position (5) of the scanning probe is represented as [x p , y p , z p ] T , and the actual deflection (6) of the scanning probe is represented as [d x , d y , d z ] T ; The dynamic error compensation method for calibrating the scanning probe comprises the following steps: Step S1: constructing a probe deflection dynamic model based on three-axis linear displacement sensor signals and delay time; Step S2: calibrating the probe deflection dynamic model based on three-axis linear displacement sensor signals and delay time by a 25-point calibration method to obtain a calibrated probe deflection dynamic model; Step S3: recording the scanning probe theoretical measurement position and three-axis linear displacement sensor signals at each sampling time during the measurement of the measured object by the three-coordinate measuring machine; wherein the three-axis linear displacement sensor signals are X-axis linear displacement sensor signal, Y-axis linear displacement sensor signal and Z-axis linear displacement sensor signal respectively; Step S4: interpolating the three-axis linear displacement sensor signals at each sampling time according to the delay time to obtain three-axis linear displacement sensor signals at the delay time; wherein the delay time is the current sampling time plus the delay time; Step S5: inputting the three-axis linear displacement sensor signals at the delay time into the calibrated probe deflection dynamic model for deflection calculation to obtain the actual deflection of the scanning probe at the current sampling time.

2. The dynamic error compensation method for scanning probe head calibration of claim 1, wherein, The scanning probe (2) is also provided with parallel leaf springs (8), which are X-axis parallel leaf springs, Y-axis parallel leaf springs and Z-axis parallel leaf springs respectively, and each of the X-axis linear displacement sensor, Y-axis linear displacement sensor and Z-axis linear displacement sensor is provided with a magnetic core (9), the X-axis parallel leaf spring is connected with the magnetic core (9) on the X-axis linear displacement sensor, the Y-axis parallel leaf spring is connected with the magnetic core (9) on the Y-axis linear displacement sensor, and the Z-axis parallel leaf spring is connected with the magnetic core (9) on the Z-axis linear displacement sensor, the scanning probe (2) moves after contacting the measured object (3), the scanning probe (2) can drive the X-axis parallel leaf spring, Y-axis parallel leaf spring and Z-axis parallel leaf spring to move, the movement of the X-axis parallel leaf spring, Y-axis parallel leaf spring and Z-axis parallel leaf spring can drive the magnetic core (9) connected therewith to move, at this time, the X-axis linear displacement sensor, Y-axis linear displacement sensor and Z-axis linear displacement sensor output displacement signals, and the displacement signals output by the X-axis linear displacement sensor, Y-axis linear displacement sensor and Z-axis linear displacement sensor are proportional to the displacement of the respective magnetic cores (9).

3. The dynamic error compensation method for scanning probe head calibration of claim 1, wherein, The construction of the probe deflection dynamic model based on three-axis linear displacement sensor signals and delay time also comprises: The probe deflection dynamic model based on three-axis linear displacement sensor signals and delay time is expressed as: wherein, is the X-axis linear displacement sensor signal corresponding to the i+t1 moment, is the Y-axis linear displacement sensor signal corresponding to the i+t2 moment, is the Z-axis linear displacement sensor signal corresponding to the i+t3 moment, i represents the current moment, t1, t2 and t3 all represent delay time, k px , k qy and k rz are linear terms, h p , h q and h r are quadratic terms, k qx , k rx , k py , k ry , k pz and k qz are cross-coupling terms.

4. The dynamic error compensation method for scanning probe head calibration of claim 3, wherein, The calibration of the probe deflection dynamic model based on three-axis linear displacement sensor signals and delay time by the 25-point calibration method to obtain the calibrated probe deflection dynamic model also comprises: The parameter k in the dynamic model of the probe deflection based on the triaxial linear displacement sensor signal and the delay time is determined by a 25-point calibration method px , k qy , k rz , h p , h q , h r , k qx , k rx , k py , k ry , k pz , k qz , t1, t2 and t3, to obtain the calibrated dynamic model of the probe deflection The process of the 25-point calibration method is as follows: 25 points on the standard sphere, each point has a different normal direction, the standard sphere is measured by a scanning probe, the deflection generated by the scanning probe when measuring the 25 points is different; the scanning probe measures each point, and each point measurement experiences three stages of approaching, stabilizing and retracting; in the calibration process, data of the three stages of approaching, stabilizing and retracting are used, 10 samples are selected from the three stages of approaching, stabilizing and retracting, 750 sample points are generated in the measurement of 25 points, and the 750 sample points are located on a sphere with a radius equal to the radius of the standard sphere plus the radius of the probe tip; when the scanning probe measures the ith sample point, the theoretical measurement position P i of the scanning probe is obtained by using the three-coordinate measuring machine i , Y i , Z i ] at this time, the X-axis linear displacement sensor signal p, the Y-axis linear displacement sensor signal q and the Z-axis linear displacement sensor signal r are collected, formula (1) is solved by using the LM algorithm, so as to obtain the parameters k px , k qy , k rz , h p , h q , h r , k qx , k rx , k py , k ry , k pz , k qz , t1, t2 and t3 in the probe deflection dynamic model based on the three-axis linear displacement sensor signal and the delay time. where [X c , Y c , Z c ] are the spherical coordinates of the center of the standard sphere, R is the radius of the standard sphere, r is the radius of the scanning probe tip, [X i , Y i , Z i ] are the theoretical measurement positions of the scanning probe when measuring the i-th sample point.

5. The method of claim 1, wherein, the three-axis linear displacement sensor signal signal at the delay instant i+t is represented as: wherein signal i+t is the X-axis linear displacement sensor signal p at the delayed time i+t , Y-axis linear displacement sensor signal q at the delayed time i+t , or Z-axis linear displacement sensor signal r at the delayed time i+t , i is the current sampling time, t is the delay time, i+t is the delayed time, and t is t1, t2, or t3.

6. The method of claim 1, wherein, It also comprises: According to the theoretical measurement position of the scanning probe at the current sampling moment [X p , Y p , Z p ] T and the actual deflection of the scanning probe at the current sampling moment [d x , d y , d z ] T , calculate the actual measurement position of the scanning probe at the current sampling moment [x p ,y p , z p ] T , the calculation formula is as follows:

Images