Method and system for testing precision of multi-axis loaded five-dimensional force sensor

Abstract

The invention provides a multi-axis loading five-dimensional force sensor precision test method and system, and relates to the technical field of sensor testing, and the method comprises the steps: simulating a five-dimensional force sensor through a multi-axis loading platform, obtaining simulation loading data, carrying out the data collection of the five-dimensional force sensor, and carrying out the deviation precision calculation, performing analysis and calibration based on the measurement deviation value, performing correction compensation on the five-dimensional force sensor, performing cyclic loading test based on compensation parameters, performing verification optimization on the compensation parameters, and backtracking compensation optimization data to the cyclic loading test for data iteration update to form a test optimization closed loop of the five-dimensional force sensor. The technical problems that in the prior art, a precision testing method of a five-dimensional force sensor in a multi-axis loading environment has limitation, and the performance of the sensor cannot be comprehensively evaluated are solved. The technical effects of comprehensively evaluating the performance of the five-dimensional force sensor under the multi-dimensional load and improving the measurement precision and reliability of the five-dimensional force sensor are achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of sensor testing, in particular to a multi-axis loading 5-dimensional force sensor precision testing method and system. BACKGROUND

[0002] With the development of high-end equipment such as mechanical arms, spacecraft, and precision medical devices, higher requirements are placed on the measurement accuracy of forces and torques under motion. Traditional single-axis or three-dimensional force sensors cannot simultaneously meet the precise measurement requirements under multi-dimensional force and torque coupling conditions, while 5-dimensional force sensors can simultaneously measure forces and torques acting on the sensor in five independent dimensions, providing a key role for mechanical analysis and control under various working conditions. However, in actual applications, 5-dimensional force sensors are easily affected by multi-dimensional coupling effects, nonlinear responses, and dynamic disturbances under multi-axis loading, leading to a decrease in measurement accuracy. Existing testing methods mainly rely on single-dimensional or simple combined loading, making it difficult to comprehensively evaluate the performance of sensors under multi-dimensional load, and lacking complete calibration and compensation means, which cannot form a closed-loop optimization mechanism.

[0003] In the prior art, the precision testing method of 5-dimensional force sensors under multi-axis loading has limitations, which cannot comprehensively evaluate the performance of the sensors, resulting in the technical problem that the measurement accuracy and reliability are difficult to guarantee. SUMMARY

[0004] The purpose of the present application is to provide a multi-axis loading 5-dimensional force sensor precision testing method and system to solve the technical problem in the prior art that the precision testing method of 5-dimensional force sensors under multi-axis loading has limitations, which cannot comprehensively evaluate the performance of the sensors, resulting in the measurement accuracy and reliability being difficult to guarantee.

[0005] In view of the above problems, the present application provides a multi-axis loading 5-dimensional force sensor precision testing method and system.

[0006] The first aspect of the present application provides a multi-axis loading 5-dimensional force sensor precision testing method, which comprises: simultaneously applying forces and torques in multiple directions to simulate the 5-dimensional force sensor through a multi-axis loading platform to obtain simulation loading data; synchronously collecting data of the 5-dimensional force sensor according to the simulation loading data to obtain a sensor output data set for deviation accuracy calculation and to obtain a measurement deviation value; analyzing and calibrating the simulation loading data based on the measurement deviation value, correcting and compensating the 5-dimensional force sensor according to the calibration data to generate compensation parameters; performing cyclic loading testing on the 5-dimensional force sensor based on the compensation parameters, verifying and optimizing the compensation parameters according to the test results, and backtracking the compensation optimization data to the cyclic loading testing for data iterative updating to form a testing optimization closed loop of the 5-dimensional force sensor.

[0007] Optionally, a multi-axis loading platform is configured, the multi-axis loading platform having five independently controlled loading units, each of the five independently controlled loading units having five dimensions of force and torque loading directions; a test working condition of a 5-dimensional force sensor is introduced to combine the five dimensions of force and torque loading directions for multi-dimensional loading analysis, to generate a loading instruction sequence, the loading instruction sequence including combined parameters of five dimensions of force components and torque components; based on the loading instruction sequence, real-time monitoring of the combined parameters of five dimensions of force components and torque components is simulated to be performed on the 5-dimensional force sensor, to obtain the simulated loading data.

[0008] Optionally, a synchronous trigger signal is constructed based on five dimensions of the 5-dimensional force sensor, the simulated loading data is sent to the 5-dimensional force sensor for synchronous collection according to the synchronous trigger signal, to obtain a plurality of data original signals; the plurality of data original signals are classified and stored according to five dimensions, to construct a sensor output data set; based on the sensor output data set, effective analysis is performed to extract data effective features, the sensor output data set is filtered according to the data effective features, to generate an effective data set; based on the 5-dimensional force sensor, standardized analysis is performed to generate a standard reference value, deviation accuracy calculation is performed between the effective data set and the standard reference value, to obtain five dimensions of measurement deviation values.

[0009] Optionally, based on the sensor output data set, working condition recognition is performed according to five dimensions, to determine a plurality of loading working condition information; based on the plurality of loading working condition information, stability analysis is performed, to set a stable output interval; the stable output interval is used as a constraint to perform feature calculation on the sensor output data set, to obtain data effective features; the data effective features are used as indexes to perform matching retrieval on the sensor output data set, to generate a plurality of data matching degrees; data with a data matching degree greater than a preset matching threshold value in the plurality of data matching degrees is extracted as effective data for integration, to generate the effective data set.

[0010] Optionally, deviation calculation is performed between the effective data set and the standard reference value according to five dimensions, to generate an absolute deviation value; based on the absolute deviation value, relative error analysis is performed according to five dimensions, to obtain a relative error value; based on the relative error value, distribution calculation is performed on five dimensions, to obtain error distribution features of five dimensions; measurement uncertainty analysis is performed on the absolute deviation value according to the error distribution features of five dimensions, to obtain data reliability of the absolute deviation value; according to the data reliability of the absolute deviation value, classification is performed according to data accuracy, to construct an accuracy level evaluation report; data matching is performed on the accuracy level evaluation report according to five dimensions, to determine the measurement deviation values of five dimensions.

[0011] Optionally, the simulated loading data is analyzed according to the measurement deviation value in five dimensions to identify a deviation trend graph; the simulated loading data is calculated based on the deviation trend graph to obtain a plurality of deviation component contributions; the simulated loading data is calibrated according to the plurality of deviation component contributions for multiple rounds to generate calibration data; the calibration data is synchronized to the 5-dimensional force sensor for influence analysis in five dimensions to obtain an influence coefficient of five dimensions for cross-compensation of the 5-dimensional force sensor, and a compensation offset of five dimensions is determined as the compensation parameter.

[0012] Optionally, a plurality of groups of working condition data are demarcated according to the test working conditions of the 5-dimensional force sensor, and a cyclic loading test sequence is set according to the plurality of groups of working condition data; the compensation parameter is applied to the programmable gain amplification signal processing of the 5-dimensional force sensor according to the cyclic loading test sequence, and a plurality of test data sets are obtained; the compensation parameter is compared and verified according to the data distance value calculated in five dimensions according to the plurality of test data sets; the compensation parameter with a data distance value greater than a preset distance threshold value is taken as key optimization data for data optimization to generate the compensation optimization data.

[0013] Optionally, the load is parsed according to the cyclic loading test sequence to determine a load expected value, and signal analysis is performed according to the load expected value to traverse a plurality of loading stages to generate a signal amplification multiple; a load-gain relationship list is constructed, the load expected value is mapped to the load-gain relationship list, and a gain control parameter is called; the gain control parameter and the signal amplification multiple are matched according to the load expected value to generate a gain control instruction; the compensation parameter is applied to the 5-dimensional force sensor to execute the gain control instruction for preliminary signal conditioning to generate a standardized signal; a high-precision ADC module is used to synchronously sample the standardized signal to generate a plurality of sampling data for digital filtering processing to generate a preliminary test data set; the preliminary test data set is verified through a plurality of cyclic loading tests to generate the plurality of test data sets.

[0014] Optionally, the full range of the 5-dimensional force sensor is extracted, load analysis is performed based on the full range, and load distribution characteristics are obtained; the full range is analyzed according to the load distribution characteristics, and a linear response region and a nonlinear response region are obtained; the linear response region is divided into a plurality of first load intervals at equal intervals, and the nonlinear response region is divided into a plurality of second load intervals at unequal intervals; the force component and the torque component are coupled and analyzed, and the plurality of first load intervals and the plurality of second load intervals are cross-identified according to the coupling analysis result to determine a cross-load interval; the plurality of first load intervals, the plurality of second load intervals, and the cross-load interval are used to apply loads to the 5-dimensional force sensor through the multi-axis loading platform, and a plurality of gain output signals are obtained; gain scoring is performed based on the plurality of gain output signals according to the plurality of first load intervals, the plurality of second load intervals, and the cross-load interval, and a target gain value is determined according to the gain scoring; the plurality of first load intervals, the plurality of second load intervals, and the cross-load interval are mapped and associated with the target gain value, and a load-gain relationship list is constructed.

[0015] In a second aspect of the present application, a multi-axis loaded 5-dimensional force sensor precision test system is provided, the system comprising: a sensor simulation module for simultaneously applying forces and torques in multiple directions to a 5-dimensional force sensor through a multi-axis loading platform to simulate and obtain simulated loading data; a data acquisition module for synchronously acquiring data of the 5-dimensional force sensor according to the simulated loading data, obtaining a sensor output data set for deviation precision calculation, and obtaining a measurement deviation value; a data analysis module for analyzing and calibrating the simulated loading data based on the measurement deviation value, modifying and compensating the 5-dimensional force sensor according to the calibration data, and generating compensation parameters; and a data updating module for performing cyclic loading test on the 5-dimensional force sensor based on the compensation parameters, verifying and optimizing the compensation parameters according to the test results, backtracking the compensation optimization data to the cyclic loading test for data iterative updating, and forming a test optimization closed loop of the 5-dimensional force sensor.

[0016] The one or more technical solutions provided in the present application have at least the following technical effects or advantages: The method provided by the embodiment of the application obtains simulation loading data by simultaneously applying forces and torques in multiple directions to a 5-dimensional force sensor through a multi-axis loading platform; obtains sensor output data sets by performing data synchronization collection on the 5-dimensional force sensor according to the simulation loading data, and obtains a measurement deviation value by performing deviation precision calculation; analyzes and calibrates the simulation loading data based on the measurement deviation value, corrects and compensates the 5-dimensional force sensor according to the calibration data, and generates compensation parameters; performs cyclic loading test on the 5-dimensional force sensor based on the compensation parameters, verifies and optimizes the compensation parameters according to the test results, traces back the compensation optimization data to the cyclic loading test to perform data iteration and update, and forms a test optimization closed loop of the 5-dimensional force sensor. The actual working conditions are accurately simulated, the performance of the sensor under multi-dimensional load is comprehensively evaluated, high-precision measurement is realized, and the measurement precision and reliability of the 5-dimensional force sensor are improved.

[0017] The above description is only a summary of the technical solutions of the application. In order to more clearly understand the technical means of the application, the specific embodiments of the application can be implemented according to the content of the description, and in order to make the above and other purposes, characteristics and advantages of the application more obvious and easy to understand, the following specific embodiments of the application are described. It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the application, nor is it intended to limit the scope of the application. Other features of the application will become apparent from the following description. BRIEF DESCRIPTION OF DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description are only exemplary, and other drawings can be obtained by the provided drawings without creative labor for those skilled in the art.

[0019] Figure 1 The flowchart of the multi-axis loading 5-dimensional force sensor precision test method provided by the application.

[0020] Figure 2 The structural schematic diagram of the multi-axis loading 5-dimensional force sensor precision test system provided by the application.

[0021] Explanation of reference signs: sensor simulation module 11, data acquisition module 12, data analysis module 13, data update module 14. DETAILED DESCRIPTION

[0022] This application provides a method and system for testing the accuracy of a 5D force sensor under multi-axis loading. It addresses the limitations of existing methods for testing the accuracy of 5D force sensors under multi-axis loading environments, which fail to comprehensively evaluate sensor performance, leading to difficulties in guaranteeing measurement accuracy and reliability. The method achieves accurate simulation of actual working conditions, comprehensively evaluates sensor performance under multi-dimensional loads, realizes high-precision measurement, and improves the measurement accuracy and reliability of the 5D force sensor.

[0023] The technical solutions of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. It should be understood that the present invention is not limited to the exemplary embodiments described herein. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention. It should also be noted that, for ease of description, only the parts related to the present invention are shown in the accompanying drawings, not all of them.

[0024] Example 1, as Figure 1 As shown, this application provides a method for testing the accuracy of a 5D force sensor under multi-axis loading, the method comprising: Simulated loading data was obtained by applying forces and torques in multiple directions simultaneously to a 5D force sensor using a multi-axis loading platform.

[0025] Furthermore, a multi-axis loading platform is used to simulate the simultaneous application of forces and torques in multiple directions to a 5D force sensor, obtaining simulated loading data. The method includes: configuring a multi-axis loading platform with five independently controlled loading units, each with five force and torque loading directions; introducing the test conditions of the 5D force sensor and combining them with the five force and torque loading directions to perform multi-dimensional loading analysis, generating a loading command sequence containing combined parameters of the five force and torque components; and real-time monitoring of the combined parameters of the five force and torque components in the 5D force sensor simulation based on the loading command sequence to obtain the simulated loading data.

[0026] Specifically, the 5-dimensional force sensor is a high-precision measuring device that can simultaneously measure forces and torques in five directions, for example, measuring forces in three directions and torques in two directions, the three directions of force including force Fx along the X-axis direction, force Fy along the Y-axis direction and force Fz along the Z-axis direction, and the two directions of torque including rotational torque Tx around the X-axis and rotational torque Ty around the Y-axis. According to the 5-dimensional force sensor, a multi-axis loading platform is configured, which refers to an experimental platform that independently controls multiple loading units to apply forces and torques in different directions. Each loading unit is composed of a motor, a transmission mechanism, a force sensor and a control unit, wherein the motor is used to provide power and control the size and direction of the force, the transmission mechanism, such as gears, belts, etc., transmits the power of the motor to the loading point, the force sensor is used to monitor the size of the applied force in real time to ensure the accuracy of the loading, and the control unit is used to receive the loading instruction sequence and control the speed and torque of the motor to control the applied force and torque. The multi-axis loading platform is equipped with five independently controlled loading units, each corresponding to a dimension of force and torque loading direction, and can accurately apply forces and torques in different directions and sizes. The loading units are controlled by high-precision motors and drive devices to ensure the stability and repeatability of the loading process.

[0027] After the platform configuration is completed, multi-dimensional loading analysis is performed in combination with the test working conditions of the 5-dimensional force sensor. The test working conditions include the magnitude, direction, and loading frequency of the force and torque that the 5-dimensional force sensor may encounter in actual application. For example, according to the test working conditions of the 5-dimensional force sensor, the range, direction, and typical working state of the force and torque that each dimension may withstand are determined, and a load combination matrix is defined in combination with the actual application scenario. Each row represents the five-dimensional force (Fx, Fy, Fz) and torque (Tx, Ty) components of a loading stage. Using the multi-dimensional loading analysis method, each combination matrix is simulated or subjected to finite element analysis to evaluate the response characteristics and feasibility of the 5-dimensional force sensor under different force and torque combinations, and to eliminate load combinations that exceed the range or are unstable. According to the analysis results, each feasible load combination is converted into specific control parameters, including the force amplitude, direction, loading sequence, and duration of each loading unit, forming a complete loading instruction sequence. Each loading instruction sequence contains the combined parameters of the force components and torque components of the five dimensions, thereby ensuring that the loading process can realistically simulate the actual force conditions of the sensor. The multi-axis loading platform applies the combined parameters of the force components and torque components of the five dimensions according to the generated loading instruction sequence. For example, according to the test working conditions, a loading instruction sequence is generated, including applying a combination load of X-axis positive force Fx=50N, Y-axis negative force Fy=-30N, Z-axis positive force Fz=20N, X-axis torque Tx=15Nm, and Y-axis negative torque Ty=-10Nm to the 5-dimensional force sensor. The five independent loading units of the multi-axis loading platform work cooperatively according to the instruction sequence. Loading unit 1 applies a 50N positive thrust along the X-axis, loading unit 2 generates a 30N tensile force in the negative Y-axis direction through a lever mechanism, loading unit 3 vertically presses downward by 20N in the positive Z-axis direction, loading unit 4 and loading unit 5 are arranged at the edges of the sensor, such as 0.1m from the center, loading unit 4 applies a 100N force along the positive Z-axis, generating a Ty=-10Nm torque, and loading unit 5 applies a 150N force along the negative Y-axis, generating a Tx=15Nm torque. The force arm is 0.1m, and all loading units are synchronously controlled through force and displacement closed-loop control.

[0028] During the loading process, the multi-axis loading platform monitors the magnitude and direction of the applied force and torque in real time, and outputs the monitoring data to form simulated loading data. The simulated loading data records the force and torque values applied by each loading unit at each time point and forms complete five-dimensional combination data. The independent loading units of the multi-axis loading platform and the real-time monitoring mechanism ensure the accuracy and stability of the loading process, improving the reliability and repeatability of the test.

[0029] According to the simulated loading data, the 5-dimensional force sensor is subjected to data synchronization acquisition, and a sensor output data set is obtained for deviation accuracy calculation to obtain a measurement deviation value.

[0030] Further, according to the analog loading data, the 5-dimensional force sensor is synchronously collected to obtain the sensor output data set for deviation accuracy calculation, and the measurement deviation value is obtained. The method comprises the following steps: a synchronous trigger signal is constructed based on the five dimensions of the 5-dimensional force sensor, the analog loading data is sent to the 5-dimensional force sensor for synchronous collection according to the synchronous trigger signal, and a plurality of data original signals are obtained; the plurality of data original signals are classified and stored according to the five dimensions to construct a sensor output data set; based on the sensor output data set, effective analysis is performed to extract data effective features, the sensor output data set is filtered according to the data effective features, and an effective data set is generated; based on the 5-dimensional force sensor, a standard analysis is performed to generate a standard reference value, and the effective data set and the standard reference value are subjected to deviation accuracy calculation to obtain the measurement deviation value of the five dimensions.

[0031] Specifically, a synchronous trigger signal is constructed for the five dimensions of the 5-dimensional force sensor through a high-speed controller or a synchronous trigger module, such as an FPGA, to ensure that the 5-dimensional force sensor can be aligned with the output action time of the multi-axis loading platform at each analog loading, and the data is synchronously collected. According to the synchronous trigger signal, the analog loading data applied by the multi-axis loading platform is sent to the 5-dimensional force sensor for synchronous collection to obtain corresponding data original signals, which are the force and torque signals directly output by the 5-dimensional force sensor during the loading process of the multi-axis loading platform, reflecting the instantaneous response of the 5-dimensional force sensor under different loading conditions. The collected plurality of data original signals are classified and stored according to the five dimensions, such as X-direction force, Y-direction force, Z-direction force, X-axis torque, and Y-axis torque, to construct a plurality of dimensional sensor output data sets, each of which contains original data of the corresponding dimension. The data in the sensor output data set is subjected to effective analysis to extract data effective features, and the sensor output data set is filtered according to the data effective features to generate an effective data set.

[0032] The 5-dimensional force sensor is subjected to standardization analysis through a 5-dimensional force sensor calibration standard or a theoretical model to generate a standard reference value of each dimension, which represents the target output of the 5-dimensional force sensor under ideal working conditions and is used for deviation comparison. The effective data set and the standard reference value are compared and analyzed to calculate the measurement deviation value of the five dimensions, which is the deviation degree of each dimensional output relative to the standard reference, including absolute deviation, relative error, and error distribution characteristics. By accurately matching the real analog loading data with the 5-dimensional force sensor standard output, deviation analysis is performed to improve the effectiveness and reliability of the 5-dimensional force sensor precision test.

[0033] Further, based on the sensor output data set, effective analysis is carried out, data effective features are extracted, the sensor output data set is screened according to the data effective features, and an effective data set is generated. The method comprises: based on the sensor output data set, working condition recognition is carried out according to five dimensions, and a plurality of loading working condition information is determined; stable analysis is carried out according to the plurality of loading working condition information, and a stable output interval is set; the stable output interval is used as a constraint to carry out feature calculation on the sensor output data set, and data effective features are obtained; the data effective features are used as indexes to carry out matching retrieval on the sensor output data set, and a plurality of data matching degrees are generated; data with a data matching degree greater than a preset matching threshold in the plurality of data matching degrees are extracted as effective data for integration, and the effective data set is generated.

[0034] Specifically, the collected multi-dimensional data is segmented according to time sequence or sampling sequence, each segment containing simultaneously collected multi-dimensional signals, and the statistical characteristics of each dimension are calculated for each time segment or sample segment, including mean, variance, peak-to-peak value and signal change rate, for reflecting the stress state of the segment data. Through clustering algorithms such as K-means and DBSCAN, sample segments with similar five-dimensional signal characteristics are classified into the same category, and each category corresponds to a loading condition information, i.e. the typical response state of the sensor under a specific force and torque combination. The data of the 5-dimensional force sensor under different stress states is divided into multiple loading condition information for distinguishing different force and torque combination states. Moreover, by calculating the mean, variance and standard deviation of the signal, the stability of the multiple loading condition information is analyzed, and the time period for the 5-dimensional force sensor multi-dimensional output signal to reach stability is obtained, and a stable data interval is set to exclude transient response and disturbance phase data. The stable output interval is used as a constraint condition to calculate the features of the data in the sensor output data set, such as mean, variance and standard deviation, to obtain data effective features, which are statistical indicators reflecting data quality and reliability, used to evaluate the stability and consistency of the data. The data effective features are used as indexes for matching and searching in the sensor output data set, and the matching degree between the data effective features and the original signals in the sensor output data set is calculated through similarity calculation methods such as Euclidean distance, cosine similarity or standardized error score. According to the application scenarios and precision requirements of the 5-dimensional force sensor, a preset matching threshold is set to distinguish between valid data and invalid data, ensuring that the selected data can accurately reflect the performance of the 5-dimensional force sensor under actual working conditions. The multiple data matching degrees are compared with the preset matching threshold, and the data with a matching degree greater than the preset matching threshold are integrated to form an effective data set, and each data in the effective data set meets the stability, representativeness and multi-dimensional consistency. Through effective analysis and screening of the sensor output data, the quality and reliability of the data are improved, and the reliability and effectiveness of the multi-axis loading 5-dimensional force sensor test scheme are improved.

[0035] Further, based on the 5-dimensional force sensor, a standard reference value is generated by standardization analysis, deviation accuracy calculation is performed between the effective data set and the standard reference value, and the measurement deviation values of the five dimensions are obtained. The method comprises: performing deviation calculation between the effective data set and the standard reference value according to the five dimensions, and generating absolute deviation values; performing relative error analysis according to the five dimensions based on the absolute deviation values, and obtaining relative error values; performing distribution calculation on the five dimensions based on the relative error values, and obtaining error distribution characteristics of the five dimensions; performing measurement uncertainty analysis on the absolute deviation values according to the error distribution characteristics of the five dimensions, and obtaining data reliability of the absolute deviation values; classifying according to data accuracy according to the data reliability of the absolute deviation values, and constructing an accuracy level evaluation report; and performing data matching on the accuracy level evaluation report according to the five dimensions, and determining the measurement deviation values of the five dimensions.

[0036] Specifically, the effective data set and the standard reference value of the five dimensions are compared dimension by dimension. For each dimension, the difference between the actual output data of the sensor in the effective data set and the standard reference value is calculated, and the absolute value of the difference is taken to generate the absolute deviation value. Based on the absolute deviation value, the absolute deviation value is divided by the standard reference value to perform relative error analysis, and the relative error value of each data point in each dimension is obtained, which is used to quantify the relative size of the error of each data point. According to the relative error value of each dimension, distribution calculation is performed, and the distribution characteristics of the error in different working conditions and time periods are counted, including mean, variance, standard deviation and other statistical indicators, to obtain the error distribution characteristics of the five dimensions. The error distribution characteristics reflect the distribution of the deviation in different dimensions. Based on the calculated error distribution characteristics, the absolute deviation value is analyzed by calculating the confidence interval, and the data reliability of the absolute deviation value of each dimension is obtained. Then, according to the data reliability of the absolute deviation value, the data is classified into different levels according to the data accuracy, for example, data reliability greater than or equal to 90% belongs to high accuracy, data reliability greater than or equal to 70% and less than 90% belongs to medium accuracy, and data reliability less than 70% is low accuracy, to construct an accuracy level evaluation report. The higher the accuracy, the higher the data reliability. Finally, the accuracy level evaluation report is matched with the data of the five dimensions. For each dimension of the five dimensions, each data point in the effective data set is traversed. For each data point, the data reliability value corresponding to it in the accuracy level evaluation report is found, the accuracy level to which the measurement deviation value of each dimension belongs is determined according to the set threshold, and the measurement deviation values of the five dimensions are obtained.

[0037] Based on the measurement deviation value, the simulation loading data is analyzed and calibrated, and the 5-dimensional force sensor is corrected and compensated according to the calibration data to generate compensation parameters.

[0038] Further, based on the measurement deviation value, the simulated loading data is analyzed and calibrated, and the 5-dimensional force sensor is corrected and compensated according to the calibration data to generate compensation parameters. The method comprises: according to the measurement deviation value, the simulated loading data is analyzed for deviation in five dimensions, and a deviation trend graph is identified; based on the deviation trend graph, the simulated loading data is matched and calculated to obtain a plurality of deviation component contribution degrees; according to the plurality of deviation component contribution degrees, the simulated loading data is calibrated in multiple rounds to generate calibration data; the calibration data is synchronized to the 5-dimensional force sensor for influence analysis in five dimensions to obtain influence coefficients of the five dimensions, the 5-dimensional force sensor is cross-compensated, and compensation offsets of the five dimensions are determined as the compensation parameters.

[0039] Specifically, for the simulated loading data of each dimension of the five dimensions, the corresponding dimension measurement deviation value is subtracted point by point to obtain the deviation analysis result of the corresponding dimension under each loading condition. Batch calculation can be performed through a numerical calculation tool MATLAB. According to the deviation analysis result, through visualization tools such as Matplotlib, Seaborn and Plotly, the deviation of each dimension with the load is plotted into a curve graph, the horizontal axis is the simulated loading value, and the vertical axis is the deviation value. The deviation changes with the load through the broken line or scatter point to form a deviation trend graph, which reflects the dynamic law of the output deviation of the 5-dimensional force sensor under different force and torque combinations.

[0040] The deviation trend graph is used to match and calculate the simulated loading data. Through the least square method, the proportion of each dimension deviation in the total deviation is calculated to obtain a plurality of deviation component contribution degrees. According to the contribution degree, the deviation component is distributed to the correction value of each dimension. In each iteration, the correction factor of the simulated loading data is adjusted, and the correction factor is used to correct the deviation in the simulated loading data. The corrected deviation is calculated, and the contribution degree of each dimension is re-evaluated according to the new deviation value. This process is repeated continuously, and the correction factor is adjusted based on the latest deviation and contribution degree in each iteration until the deviation converges below a preset threshold to generate calibration data. The iterative process uses iterative algorithms such as gradient descent and Newton method for multiple rounds of calibration. The calibration data is synchronized to the 5-dimensional force sensor, and the output data of each dimension after calibration is obtained through 5-dimensional force sensor simulation. The influence amplitude of the calibration data on the output of each dimension is calculated through linear regression or sensitivity analysis to obtain the influence coefficients of the five dimensions for comprehensive evaluation of the sensor performance. The influence coefficient refers to the influence degree quantitative index of the calibration data on the sensor output, which is used to evaluate the calibration effect.

[0041] Based on the influence coefficient of the five dimensions, the deviation of each dimension of the 5-dimensional force sensor is distributed to the corresponding dimension according to the influence coefficient for cross compensation, the compensation deviation offset is obtained, and the compensation deviation offset is taken as the compensation parameter of the 5-dimensional force sensor, for example, wherein, the cross compensation refers to that in the multi-dimensional force sensor, due to the coupling effect between the dimensions, for example, the force applied in the X direction will cause the response of the Y direction or the torque dimension, and the deviation of a certain dimension cannot be completely eliminated by separate correction. Therefore, the deviation of each dimension is distributed and corrected according to the influence degree on other dimensions, for example, according to the influence analysis, a 5*5 matrix is established, each element represents the influence proportion of the calibration data of a certain dimension on the output of other dimensions, the (i, j) element in the matrix represents the influence coefficient of the deviation of the i-th dimension on the output of the j-th dimension, the measured deviation of each dimension is multiplied by the corresponding influence coefficient to obtain the contribution value of the deviation to the adjustment of each dimension, and the contribution values of all dimensions are superimposed on the corresponding dimensions to obtain the cross compensation deviation offset that needs to be adjusted for each dimension. Through deviation analysis and cross compensation, the compensation parameter reflecting the true characteristics of the 5-dimensional force sensor is generated, thereby significantly reducing the measurement error and improving the overall accuracy and reliability of the 5-dimensional force sensor under multi-axis loading conditions.

[0042] Based on the compensation parameter, the 5-dimensional force sensor is subjected to cyclic loading test, the compensation parameter is verified and optimized according to the test result, the compensation optimization data is traced back to the cyclic loading test for data iteration update, and a test optimization closed loop of the 5-dimensional force sensor is formed.

[0043] Further, based on the compensation parameter, the 5-dimensional force sensor is subjected to cyclic loading test, the compensation parameter is verified and optimized according to the test result, and the method comprises the following steps: a plurality of groups of working condition data are divided according to the test working condition of the 5-dimensional force sensor, and a cyclic loading test sequence is set according to the plurality of groups of working condition data; the compensation parameter is applied to the programmable gain amplification signal processing of the 5-dimensional force sensor according to the cyclic loading test sequence, and a plurality of test data sets are obtained; the compensation parameter is compared and verified according to the data distance value calculated according to the five dimensions according to the plurality of test data sets; the compensation parameter with a data distance value greater than a preset distance threshold value is taken as key optimization data for data optimization, and the compensation optimization data is generated.

[0044] Specifically, according to the test working condition of the 5-dimensional force sensor, different force and torque combinations that the 5-dimensional force sensor can withstand are divided into multiple groups of working condition data, and a cyclic loading test sequence is set according to the multiple groups of working condition data, which refers to multiple loading tests according to the set working condition data, for covering the response of the 5-dimensional force sensor under typical use scenarios and ensuring that the test process can comprehensively evaluate the performance of the 5-dimensional force sensor under various conditions. According to the cyclic loading test sequence, the 5-dimensional force sensor is subjected to programmable gain amplification signal processing by applying compensation parameters, and the gain is dynamically adjusted by a programmable gain amplifier (PGA) according to the compensation parameters to adapt to different test working conditions. Through cyclic loading tests, multiple rounds of test data sets are obtained, each round of test data set including the output data of the 5-dimensional force sensor under different working conditions. Based on the multiple rounds of test data sets, the compensation parameters are calculated for data distance values in five dimensions, specifically by calculating the difference between the output data of the 5-dimensional force sensor and the standard reference value to obtain the data distance value, which is used to evaluate the effectiveness of the compensation parameters. According to the multiple data distance values calculated, the compensation parameters are compared and verified, and if the data distance value is greater than a preset distance threshold, the corresponding compensation parameter is marked as a key optimization data, and an optimization algorithm such as genetic algorithm, particle swarm optimization, etc. is used to optimize the marked compensation parameter to generate compensation optimization data. The preset distance threshold is a standard value set according to expert experience and actual requirements, which is used to determine whether the compensation parameter needs to be further optimized.

[0045] The optimized compensation parameters, i.e. compensation optimization data, are traced back to the cyclic loading test and re-applied to the 5-dimensional force sensor for testing. By tracing back the optimized parameters to the cyclic loading test, the process of repeatedly applying the loading sequence, collecting data, and calculating the deviation is repeated to realize data iteration and update, forming a test optimization closed loop for the 5-dimensional force sensor, which is used to ensure that the compensation parameters can achieve optimal performance under different working conditions and improve the measurement accuracy and reliability of the 5-dimensional force sensor. Through data iteration and closed-loop test optimization, the compensation parameters are continuously optimized to form a dynamic and adaptive test and optimization process, improving the high precision, reliability and repeatability of the 5-dimensional force sensor under multi-axis loading conditions. Moreover, through the test of multiple groups of working condition data, the adaptability and reliability of the 5-dimensional force sensor under various actual application conditions are improved.

[0046] Further, according to the cyclic loading test sequence, the compensation parameters are applied to the programmable gain amplification signal processing of the 5-dimensional force sensor to obtain a plurality of test data sets, and the method comprises: performing load analysis according to the cyclic loading test sequence, determining a load expected value, performing signal analysis according to the load expected value in a plurality of loading stages to generate a signal amplification multiple; constructing a load-gain relationship list, mapping the load expected value to the load-gain relationship list, and calling a gain control parameter; matching the gain control parameter with the signal amplification multiple according to the load expected value to generate a gain control instruction; applying the compensation parameters to the 5-dimensional force sensor to execute the gain control instruction for preliminary signal conditioning to generate a standardized signal; using a high-precision ADC module to synchronously sample the standardized signal to generate a plurality of sampling data for digital filtering processing to generate a preliminary test data set; and verifying the preliminary test data set through a plurality of cyclic loading tests to generate the plurality of test data sets.

[0047] Specifically, according to the load analysis of each loading stage in the cyclic loading test sequence, the load expected value of force and torque corresponding to each stage is calculated. According to the load expected value of each stage, combined with the load amplitude and direction of each stage, the signal is analyzed using signal analysis tools such as fast Fourier transform to generate the signal amplification factor of each stage, which refers to the amplification factor required to make the 5-dimensional force sensor output signal reach the optimal measurement range. Through the load analysis of the 5-dimensional force sensor, a load-gain relationship list is constructed, and the load expected value of each stage is mapped to the load-gain relationship list for matching analysis to obtain the corresponding gain control parameter, which refers to the parameter used to adjust the programmable gain amplifier to achieve precise control of the signal amplification factor. For example, according to the load analysis and signal characteristic test, a table or database is established to record the gain amplification factor and control parameter of the programmable gain amplifier corresponding to the load value of each dimension. Then, the load expected value of each stage in the cyclic loading test sequence is looked up or matched to the load-gain relationship list through interpolation method, so as to obtain the corresponding gain control parameter. If the load value is between the table intervals, linear or polynomial interpolation is used to calculate the accurate gain control parameter to ensure precise and controllable signal amplification factor. The gain control parameter and signal amplification factor are compared and corrected according to the load expected value, and the output value is adjusted according to the deviation between the two to generate specific instruction set for controlling the programmable gain amplifier, i.e. gain control instruction. The gain control instruction contains specific amplification factor, duration and signal channel information, ensuring that the amplitude of the sensor output signal in each stage meets the expectation. In each loading stage, the gain control instruction is applied through the programmable gain amplifier to perform preliminary signal conditioning on the 5-dimensional force sensor to generate a standardized signal. Specifically, the preliminary signal of each dimension is first superimposed with a compensation deviation value to eliminate measurement deviation, and then the signal amplitude is adjusted according to the gain control instruction to match the target amplification factor, so as to amplify the signal to the ideal range while maintaining the proportional relationship and coupling characteristics of the output of each dimension. The conditioned output is the standardized signal, which ensures that the output amplitude is within the ideal range and is compensated for deviation. A high-precision ADC module is used to synchronously sample the standardized signal to obtain multiple sampling data, and a digital filtering method such as FIR filter and IIR filter is used to digitally filter the multiple sampling data to remove noise and transient interference of the sampling data, generating a preliminary test data set. Through multiple rounds of cyclic loading tests, the preliminary test data sets of each round are compared and verified to form a multi-round test data set including five dimensions and each loading stage. Through multiple rounds of cyclic loading tests, the effectiveness and stability of the compensation parameter under different working conditions are ensured, and the measurement accuracy and reliability of the 5-dimensional force sensor are improved.

[0048] Further, the construction process of the load-gain relationship list includes: extracting a full range of the 5-dimensional force sensor, performing load analysis based on the full range to obtain load distribution characteristics; analyzing the full range according to the load distribution characteristics to obtain a linear response region and a nonlinear response region; performing equal-interval division based on the linear response region to obtain a plurality of first load intervals, and performing non-equal-interval division based on the nonlinear response region to obtain a plurality of second load intervals; coupling the force components and the torque components for coupling analysis, and performing interval cross identification on the plurality of first load intervals and the plurality of second load intervals according to the coupling analysis result to determine cross load intervals; applying loads to the 5-dimensional force sensor according to the plurality of first load intervals, the plurality of second load intervals, and the cross load intervals through the multi-axis loading platform to obtain a plurality of gain output signals; performing gain scoring based on the plurality of gain output signals according to the plurality of first load intervals, the plurality of second load intervals, and the cross load intervals, and determining a target gain value according to the gain scoring; and mapping and associating the plurality of first load intervals, the plurality of second load intervals, and the cross load intervals with the target gain value to construct the load-gain relationship list.

[0049] Specifically, in the process of constructing the load-gain relationship list, first, the full range of the 5-dimensional force sensor in each dimension is obtained, and the full range refers to the maximum and minimum force and torque values that can be measured by the 5-dimensional force sensor in each dimension. Based on the full range, load analysis is performed on different load levels, and multiple sampling points are divided on each dimension according to the full range of the sensor. The corresponding force or torque is applied to each sampling point, the output signal is collected through a high-precision ADC module, and the output amplitude, mean, variance and other statistical characteristics are calculated to obtain the response characteristics of each sampling point. The signal of all sampling points is distributed analyzed to obtain the load distribution characteristics of each dimension, which refers to the distribution of the sensor output signal under different load conditions. Then, the full range is divided into a linear response region and a nonlinear response region according to the load distribution characteristics, where the linear response region refers to the interval where the sensor output is approximately linearly related to the input force or torque, and the nonlinear response region refers to the interval where the output and input have obvious nonlinear deviation. Specifically, the load-output curve of each dimension is fitted by linear fitting, quadratic or polynomial fitting, and the fitting residual is calculated. When the residual is lower than a preset threshold, it is determined as a linear response region, and the interval with larger residual or obvious deviation from the linear trend is determined as a nonlinear response region. According to the linear response region, multiple first load intervals are obtained by equal interval division, and according to the nonlinear response region, multiple second load intervals are obtained by non-equal interval division. The force components and torque components of each dimension are retrieved for coupling analysis to obtain coupling analysis results and evaluate the interaction between different dimensions. According to the coupling analysis results, the first load intervals and the second load intervals are cross-identified to form cross-load intervals for representing the coupling effect under multi-dimensional loading conditions. For example, a coupling matrix is constructed according to the force components and torque components, which reflects the degree of mutual influence between different dimensions. The correlation coefficient or coupling strength in the coupling matrix is calculated to quantify the interaction effect between different dimensions. According to the coupling analysis results, the first load intervals and the second load intervals are cross-identified to form cross-load intervals for representing the coupling effect under multi-dimensional loading conditions. Through a multi-axis loading platform, actual forces and torques are applied to the 5-dimensional force sensor according to the first load intervals, the second load intervals and the cross-load intervals. For each loading interval, the gain setting is adjusted through a programmable gain amplifier, and the sensor output signal is synchronously sampled using a high-precision ADC module to record the amplitude, waveform and dynamic response of each dimension, thereby obtaining the gain output signal under each gain setting. Based on multiple gain output signals, gain scoring is performed according to the first load intervals, the second load intervals and the cross-load intervals, and the scoring method includes calculating the linearity, normalized error, signal-to-noise ratio and fluctuation range of the steady-state output of the signal, and generating a comprehensive score by combining the indicators.By comparing the comprehensive scores of various gain settings within the same load range, the gain with the highest score is selected as the optimal gain for that range, and this optimal gain is used as the target gain value. Finally, a mapping relationship is established between multiple first load ranges, multiple second load ranges, and cross-load ranges and the target gain value, constructing a load-gain relationship list. This list is a mapping table that associates different load ranges with corresponding gain control parameters, enabling precise control of signal amplification under different load conditions. Furthermore, the load-gain relationship list is a lookup table or database, containing range numbers, load ranges, and corresponding gain parameters, facilitating quick retrieval during cyclic loading tests and signal conditioning. By constructing the load-gain relationship list, precisely matched control parameters are provided for signal amplification and compensation of the sensor under different load conditions. This allows the sensor output signal in each load range to undergo precise gain adjustment, achieving high-precision, multi-dimensional measurement and signal standardization, thereby improving the testing accuracy, compensation capability, and adaptability of the 5D force sensor.

[0050] Example 2, based on the same inventive concept as the multi-axis loading 5D force sensor accuracy testing method in the aforementioned examples, such as... Figure 2 As shown, this application provides a 5D force sensor accuracy testing system with multi-axis loading, wherein the 5D force sensor accuracy testing system with multi-axis loading includes: The sensor simulation module 11 is used to simulate the simultaneous application of forces and torques in multiple directions to a 5D force sensor through a multi-axis loading platform to obtain simulated loading data. The data acquisition module 12 is used to synchronously acquire data from the 5D force sensor based on the simulated loading data, obtain the sensor output dataset, calculate the deviation accuracy, and obtain the measurement deviation value. The data analysis module 13 is used to analyze and calibrate the simulated loading data based on the measurement deviation value, correct and compensate the 5D force sensor based on the calibration data, and generate compensation parameters. The data update module 14 performs cyclic loading tests on the 5D force sensor based on the compensation parameters, verifies and optimizes the compensation parameters based on the test results, and backtracks the compensation optimization data to the cyclic loading test for iterative data updates, forming a closed loop for the test optimization of the 5D force sensor.

[0051] Further, the sensor simulation module 11 is further configured to: configure a multi-axis loading platform, the multi-axis loading platform having five independently controlled loading units, the five independently controlled loading units respectively having five dimensions of force and torque loading directions; introduce a 5-dimensional force sensor test working condition combined with the five dimensions of force and torque loading directions for multi-dimensional loading analysis, generate a loading instruction sequence, the loading instruction sequence containing combined parameters of five dimensions of force components and torque components; based on the loading instruction sequence, perform real-time monitoring of the combined parameters of five dimensions of force components and torque components on the 5-dimensional force sensor simulation, and obtain the simulation loading data.

[0052] Further, the data acquisition module 12 is further configured to: based on the five dimensions of the 5-dimensional force sensor, construct a synchronous trigger signal, send the simulation loading data to the 5-dimensional force sensor for synchronous acquisition according to the synchronous trigger signal, and obtain a plurality of data original signals; store the plurality of data original signals according to five dimensions, construct a sensor output data set; based on the sensor output data set, perform effective analysis, extract data effective features, filter the sensor output data set according to the data effective features, and generate an effective data set; based on the 5-dimensional force sensor, perform standardized analysis, generate a standard reference value, calculate the deviation accuracy of the effective data set and the standard reference value, and obtain five dimensions of measurement deviation values.

[0053] Further, the data acquisition module 12 is further configured to: based on the sensor output data set, identify working conditions according to five dimensions, determine a plurality of loading working condition information; based on the plurality of loading working condition information, perform stability analysis, and set a stable output interval; use the stable output interval as a constraint to calculate features of the sensor output data set, and obtain data effective features; use the data effective features as an index to match and search the sensor output data set, and generate a plurality of data matching degrees; extract data with a data matching degree greater than a preset matching threshold value from the plurality of data matching degrees as effective data for integration, and generate the effective data set.

[0054] Further, the data acquisition module 12 is further configured to: calculate the deviation of the effective data set and the standard reference value according to five dimensions, generate an absolute deviation value; based on the absolute deviation value, perform relative error analysis according to five dimensions, obtain a relative error value; based on the relative error value, perform distribution calculation on five dimensions, obtain error distribution features of five dimensions; according to the error distribution features of five dimensions, perform measurement uncertainty analysis on the absolute deviation value, obtain data reliability of the absolute deviation value; according to the data reliability of the absolute deviation value, perform hierarchical classification according to data accuracy, and construct an accuracy level evaluation report; match data according to five dimensions on the accuracy level evaluation report, and determine the measurement deviation values of five dimensions.

[0055] Further, the data analysis module 13 is further configured to: perform bias analysis on the simulation loading data in five dimensions according to the measurement bias value, identify a bias trend graph; perform matching calculation on the simulation loading data based on the bias trend graph, obtain a plurality of bias component contribution degrees; perform multi-round calibration on the simulation loading data according to the plurality of bias component contribution degrees, and generate calibration data; synchronize the calibration data to the 5-dimensional force sensor for influence analysis in five dimensions, obtain an influence coefficient of five dimensions, cross-compensate the 5-dimensional force sensor, and determine a compensation offset of five dimensions as the compensation parameter.

[0056] Further, the data updating module 14 is further configured to: divide a plurality of groups of working condition data according to the test working condition of the 5-dimensional force sensor, set a cyclic loading test sequence according to the plurality of groups of working condition data; perform programmable gain amplification signal processing on the 5-dimensional force sensor according to the cyclic loading test sequence and the compensation parameter, and obtain a plurality of test data sets; compare and verify the compensation parameter according to the plurality of test data sets and a data distance value calculated in five dimensions; perform data optimization on the compensation parameter with a data distance value greater than a preset distance threshold as key optimization data, and generate the compensation optimization data.

[0057] Further, the data updating module 14 is further configured to: perform load analysis according to the cyclic loading test sequence, determine a load expected value, perform signal analysis according to the load expected value in a plurality of loading stages, and generate a signal amplification multiple; construct a load-gain relationship list, map the load expected value to the load-gain relationship list, and call a gain control parameter; match the gain control parameter and the signal amplification multiple according to the load expected value, and generate a gain control instruction; perform the gain control instruction on the 5-dimensional force sensor using the compensation parameter to generate a standardized signal; perform synchronous sampling on the standardized signal using a high-precision ADC module, generate a plurality of sampling data, perform digital filtering processing, and generate a preliminary test data set; verify the preliminary test data set through a plurality of cyclic loading tests, and generate the plurality of test data sets.

[0058] Further, the data updating module 14 is further configured to: extract a full range of the 5D force sensor, perform load analysis based on the full range to obtain load distribution characteristics; analyze the full range according to the load distribution characteristics to obtain a linear response region and a nonlinear response region; perform equal-interval division based on the linear response region to obtain a plurality of first load intervals, perform non-equal-interval division based on the nonlinear response region to obtain a plurality of second load intervals; perform coupling analysis on the force component and the torque component, and perform interval cross identification on the plurality of first load intervals and the plurality of second load intervals according to the coupling analysis result to determine cross load intervals; apply loads to the 5D force sensor according to the plurality of first load intervals, the plurality of second load intervals, and the cross load intervals through the multi-axis loading platform to obtain a plurality of gain output signals; perform gain scoring on the plurality of gain output signals according to the plurality of first load intervals, the plurality of second load intervals, and the cross load intervals, and determine a target gain value according to the gain scoring; and map and associate the plurality of first load intervals, the plurality of second load intervals, and the cross load intervals with the target gain value to construct the load-gain relationship list.

[0059] The various embodiments in the specification are described in a progressive manner, and each embodiment focuses on the difference from other embodiments. The foregoing Figure 1 The multi-axis loaded 5D force sensor precision testing method and specific examples in Embodiment One are also applicable to the multi-axis loaded 5D force sensor precision testing system of the present embodiment. Based on the foregoing detailed description of the multi-axis loaded 5D force sensor precision testing method, those skilled in the art can clearly understand the multi-axis loaded 5D force sensor precision testing system of the present embodiment. Therefore, for the sake of brevity of the specification, the multi-axis loaded 5D force sensor precision testing system of the present embodiment will not be described in detail here.

[0060] The above description of the disclosed embodiments enables a person skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

[0061] Obviously, for those skilled in the art, without departing from the principles of the present application, the present application can be improved and modified in several ways, and these improvements and modifications also fall within the protection scope of the present application.

Claims

1. A method of testing the accuracy of a multi-axially loaded 5-dimensional force sensor, characterized in that, The method comprises: Simulate the application of forces and torques in multiple directions to the 5-dimensional force sensor through a multi-axis loading platform to obtain simulation loading data; According to the simulation loading data, perform data synchronous collection on the 5-dimensional force sensor to obtain a sensor output data set for deviation accuracy calculation, and obtain a measurement deviation value; Based on the measurement deviation value, analyze and calibrate the simulation loading data, correct and compensate the 5-dimensional force sensor according to the calibration data, and generate compensation parameters; Based on the compensation parameters, perform cyclic loading test on the 5-dimensional force sensor, verify and optimize the compensation parameters according to the test results, backtrack the compensation optimization data to the cyclic loading test for data iterative update, and form a test optimization closed loop of the 5-dimensional force sensor.

2. The multi-axial loaded 5-dimensional force sensor precision testing method of claim 1, wherein, Simulate the application of forces and torques in multiple directions to the 5-dimensional force sensor through a multi-axis loading platform to obtain simulation loading data, the method comprising: Configure a multi-axis loading platform, the multi-axis loading platform having five independently controlled loading units, each of the five independently controlled loading units having five dimensions of force and torque loading directions; Introduce the test working conditions of the 5-dimensional force sensor to perform multi-dimensional loading analysis in combination with the five dimensions of force and torque loading directions, generate a loading instruction sequence, and the loading instruction sequence contains combined parameters of force components and torque components in five dimensions; Based on the loading instruction sequence, real-time monitor the combined parameters of force components and torque components in five dimensions simulated on the 5-dimensional force sensor to obtain the simulation loading data.

3. The multi-axial loaded 5-dimensional force sensor precision testing method of claim 1, wherein, According to the simulation loading data, perform data synchronous collection on the 5-dimensional force sensor to obtain a sensor output data set for deviation accuracy calculation, and obtain a measurement deviation value, the method comprising: Based on the five dimensions of the 5-dimensional force sensor, construct a synchronous trigger signal, send the simulation loading data to the 5-dimensional force sensor for synchronous collection according to the synchronous trigger signal, and obtain a plurality of data original signals; Classify and store the plurality of data original signals according to five dimensions to construct a sensor output data set; Based on the sensor output data set, perform effective analysis, extract data effective features, filter the sensor output data set according to the data effective features, and generate an effective data set; Based on the 5-dimensional force sensor, perform standardized analysis to generate a standard reference value, perform deviation accuracy calculation on the effective data set and the standard reference value, and obtain measurement deviation values in five dimensions.

4. The multi-axial loaded 5-dimensional force sensor precision testing method of claim 3, wherein, Based on the sensor output data set, perform effective analysis, extract data effective features, filter the sensor output data set according to the data effective features, and generate an effective data set, the method comprising: According to the five dimensions, perform working condition recognition on the sensor output data set to determine a plurality of loading working condition information; According to the plurality of loading working condition information, perform stability analysis and set a stable output interval; Use the stable output interval as a constraint to perform feature calculation on the sensor output data set to obtain data effective features; Use the data effective features as an index to perform matching retrieval on the sensor output data set to generate a plurality of data matching degrees; Extract the data matching degree greater than the preset matching threshold value in the plurality of data matching degrees as valid data for integration, and generate the valid data set.

5. The multi-axial loaded 5-dimensional force sensor precision testing method of claim 3, wherein, Based on the standardized analysis of the 5-dimensional force sensor, the standard reference value is generated, the deviation accuracy calculation is performed between the valid data set and the standard reference value, and the measurement deviation values of the five dimensions are obtained, the method comprising: According to the five dimensions, the deviation calculation is performed between the valid data set and the standard reference value, and the absolute deviation value is generated; Based on the absolute deviation value, the relative error analysis is performed according to the five dimensions, and the relative error value is obtained; Based on the relative error value, the distribution calculation is performed on the five dimensions, and the error distribution characteristics of the five dimensions are obtained; According to the error distribution characteristics of the five dimensions, the measurement uncertainty analysis is performed on the absolute deviation value, and the data reliability of the absolute deviation value is obtained; According to the data accuracy of the absolute deviation value, the accuracy level evaluation report is constructed; According to the accuracy level evaluation report, the data matching is performed according to the five dimensions, and the measurement deviation values of the five dimensions are determined.

6. The multi-axial loaded 5-dimensional force sensor precision testing method of claim 1, wherein, Based on the measurement deviation value, the analog loading data is analyzed and calibrated, the 5-dimensional force sensor is modified and compensated according to the calibration data, the compensation parameter is generated, and the method comprises: According to the measurement deviation value, the deviation analysis is performed on the analog loading data according to the five dimensions, and the deviation trend graph is identified; Based on the deviation trend graph, the matching calculation is performed on the analog loading data, and the plurality of deviation component contribution degrees are obtained; According to the plurality of deviation component contribution degrees, the analog loading data is calibrated for multiple rounds, and the calibration data is generated; The calibration data is synchronized to the 5-dimensional force sensor for influence analysis according to the five dimensions, the influence coefficients of the five dimensions are obtained, the 5-dimensional force sensor is cross-compensated, and the compensation offset of the five dimensions is determined as the compensation parameter.

7. The multi-axial loaded 5-dimensional force sensor precision testing method of claim 2, wherein, Based on the compensation parameter, the 5-dimensional force sensor is subjected to cyclic loading test, and the compensation parameter is verified and optimized according to the test result, the method comprising: According to the test working condition of the 5-dimensional force sensor, a plurality of working condition data is divided, and a cyclic loading test sequence is set according to the plurality of working condition data; According to the cyclic loading test sequence, the programmable gain amplification signal processing is applied to the 5-dimensional force sensor by using the compensation parameter, and a plurality of test data sets are obtained; According to the plurality of test data sets, the compensation parameter is calculated according to the five dimensions to compare and verify the data distance value; The compensation parameter with the data distance value greater than the preset distance threshold value is taken as the key optimization data for data optimization, and the compensation optimization data is generated.

8. The multi-axial loaded 5-dimensional force sensor precision testing method of claim 7, wherein, According to the cyclic loading test sequence, the programmable gain amplification signal processing is applied to the 5-dimensional force sensor by using the compensation parameter, and a plurality of test data sets are obtained, the method comprising: According to the cyclic loading test sequence, the load is resolved, the load expected value is determined, the signal analysis is performed according to the load expected value in a plurality of loading stages, and the signal amplification multiple is generated; A load-gain relationship list is constructed, the load expected value is mapped to the load-gain relationship list, and the gain control parameter is called; The gain control parameter is matched with the signal amplification multiple according to a load expected value to generate a gain control instruction; The compensation parameter is applied to perform the gain control instruction on the 5-dimensional force sensor to generate a standardized signal; A high-precision ADC module is adopted to synchronously sample the standardized signal to generate a plurality of sampling data for digital filtering processing to generate a preliminary test data set; The preliminary test data set is verified through a plurality of cycles of cyclic loading tests to generate a plurality of test data sets.

9. The multi-axial loaded 5-dimensional force sensor precision testing method of claim 8, wherein, The process of constructing a load-gain relationship list includes: The full-range range of the 5-dimensional force sensor is extracted, and load analysis is performed based on the full-range range to obtain load distribution characteristics; The full-range range is analyzed according to the load distribution characteristics to obtain a linear response region and a nonlinear response region; The linear response region is divided at equal intervals to obtain a plurality of first load intervals, and the nonlinear response region is divided at unequal intervals to obtain a plurality of second load intervals; The force component and the torque component are coupled and analyzed, and the plurality of first load intervals and the plurality of second load intervals are cross-identified according to the coupling analysis result to determine cross-load intervals; The 5-dimensional force sensor is loaded according to the plurality of first load intervals, the plurality of second load intervals, and the cross-load intervals through the multi-axis loading platform to obtain a plurality of gain output signals; Gain scores are determined according to the plurality of gain output signals, the plurality of first load intervals, the plurality of second load intervals, and the cross-load intervals, and a target gain value is determined according to the gain scores; The plurality of first load intervals, the plurality of second load intervals, and the cross-load intervals are mapped and associated with the target gain value to construct the load-gain relationship list.

10. A multi-axial loaded 5-dimensional force transducer precision test system characterized by, Steps for implementing the multi-axis loading 5-dimensional force sensor precision test method according to any one of claims 1 to 9 include: A sensor simulation module is configured to simulate the 5-dimensional force sensor by simultaneously applying forces and torques in multiple directions through the multi-axis loading platform to obtain simulation loading data; A data acquisition module is configured to synchronously acquire data of the 5-dimensional force sensor according to the simulation loading data to obtain a sensor output data set for deviation precision calculation to obtain a measurement deviation value; A data analysis module is configured to analyze and calibrate the simulation loading data based on the measurement deviation value, correct and compensate the 5-dimensional force sensor according to the calibration data, and generate a compensation parameter; A data update module is configured to cyclically load test the 5-dimensional force sensor based on the compensation parameter, verify and optimize the compensation parameter according to the test result, backtrack the compensation optimization data to the cyclic loading test for data iterative updating, and form a test optimization closed loop of the 5-dimensional force sensor.

Images