Error correction method and system for force sensor

Abstract

The present invention relates to the field of mechanical measurement, and in particular to an error correction method and system for a force sensor. The method comprises the following steps: using a sensor to collect a first stress value at each position on a structure surface, and using finite element analysis software to fit a second stress value at each position on the structure surface; fitting a first fitting value according to a variation rule of the first stress value in a preset range at each position, and calculating the reliability of the first fitting value; fitting a second fitting value according to a second stress value in a preset range at each position, and calculating the reliability of the second fitting value; taking the reliability as a weight to perform a weighted summation of the first fitting stress value and the second fitting stress value to obtain a comprehensive fitting value at each position; and realizing error correction based on the comprehensive fitting value at each position. Effectively improve the accuracy of error correction.

Description

Technical Field

[0001] The present invention relates to the field of mechanical measurement, and in particular to an error correction method and system for a force sensor. Background Art

[0002] Wear and tear or improper management can cause measurement errors in stress sensors, which can lead to the inability to collect accurate stress data, thus affecting subsequent stress analysis. Therefore, the stress sensor needs to be calibrated at regular intervals. Before calibrating the stress sensor, it is necessary to obtain relatively accurate stress values, so that the sensor can be calibrated and adjusted based on the relatively accurate stress values.

[0003] When obtaining relatively accurate stress values, relatively accurate stress values ​​are often obtained by fitting and analyzing the collected data. This method has a good effect on removing individual errors. However, sometimes the collected stress values ​​have a large error content. It is easy to fit the error information into the variation law by simply fitting the measured values. Therefore, at this time, such errors cannot be removed by simply fitting the measured values. Therefore, how to accurately fit relatively accurate stress values ​​when the error content is large has become the research focus of this invention.

[0004] The patent application document with publication number CN112414532A discloses a batch error detection device for gravity sensors. The method in the patent application document mainly realizes error detection by mechanical means, and does not involve the problem that a relatively accurate value cannot be fitted when the error content is large. Therefore, the method in the patent application document cannot solve the technical problem in the present invention. Summary of the invention

[0005] In order to solve the problem of how to accurately fit a relatively accurate stress value when the error content is large, the present invention provides an error correction method and system for a force sensor.

[0006] In a first aspect, the present invention provides a method for error correction of a force sensor, which adopts the following technical solution:

[0007] A method for error correction of a force sensor comprises the steps of:

[0008] The first stress value of each position on the surface of the structure is collected by using a sensor, and the second stress value of each position on the surface of the structure is fitted by using finite element analysis software;

[0009] According to the variation law of the first stress value in the preset range of each position, the first fitting value is fitted, and the reliability K1 of the first fitting value is calculated. , S represents a sequence of structural change sharpness of all positions within the preset range of the position, L1 represents a sequence of first stress values ​​of all positions within the preset range of the position, represents the correlation between sequence S and sequence L1, represents the similarity of the variation law of the first stress value of the position and the kth other position, W represents the number of positions on the structure surface, and max() represents the maximum value function; the second fitting value is fitted according to the second stress value within the preset range of each position, and the reliability of the second fitting value is calculated; the first fitting stress value and the second fitting stress value are weighted and summed using the reliability as the weight to obtain the comprehensive fitting value of each position;

[0010] Error correction is achieved based on the comprehensive fitting value at each position.

[0011] The present invention combines the variation law information of stress data collected by the sensor and the variation law information of stress data obtained by finite element analysis to fit accurate stress data, thereby realizing error correction. This correction method can effectively prevent the problem that accurate stress data cannot be fitted by using sensor data with errors, and effectively improves the accuracy of the fitted stress data. Further, when fitting accurate stress data by combining the variation law information of stress data collected by the sensor and the variation law information of stress data obtained by finite element analysis, the reliability of the fitted data obtained by the sensor stress data and the reliability of the fitted data obtained by the finite element analysis are analyzed, and the reliability of each stress data is set according to the reliability. The reference degree of the fitting data obtained by various methods is improved, thereby improving the accuracy of the fitting stress data; further, when analyzing the accuracy of the fitting data obtained by various methods, taking into account that the more accurate the data used in the fitting law is, the more accurate the obtained fitting law is, and thus the accuracy of the data used to obtain the fitting law is analyzed to accurately describe the accuracy of the fitting data obtained by various methods; further, when analyzing the accuracy of the data used in the fitting law, taking into account that the stress value is related to the structural sharpness, the accuracy of the data used in the fitting law is accurately evaluated by analyzing the relationship between the stress value collected by each method and the structural sharpness, and then the accuracy of the data fitted by various methods is accurately evaluated.

[0012] Preferably, the method for obtaining the preset range includes:

[0013] The mean of the first stress value and the second stress value of each position is recorded as the comprehensive stress value, and all the comprehensive stress values ​​in the preset window of each position are fitted with a three-dimensional polynomial, which is recorded as the first analysis model of each position. The Euclidean distance of the coefficient vector of the first analysis model of every two adjacent positions is used as the clustering distance. All positions are clustered to obtain several categories, and the mean of the radius of the circumscribed circle of the positions of all categories is used as the preset radius; a circle with each position as the center and the preset radius as the radius is used as the preset range.

[0014] The present invention sets a preset range by analyzing the size range of data with similar variation patterns, thereby providing a basis for subsequently acquiring data with similar patterns around each position.

[0015] Preferably, the method for obtaining the coefficient vector includes:

[0016] Get the coefficients of each item in each dimension of the fitting model, arrange the coefficients of all items in ascending order according to the degree of the item, and record the obtained vectors as the local coefficient vectors of each dimension. The vector obtained by splicing the local coefficient vectors of all dimensions of the fitting model together is recorded as the coefficient vector.

[0017] The present invention describes the fitting model information through coefficient vectors, and the fitting model information described by this method is more comprehensive.

[0018] Preferably, the method for obtaining the sharpness of the structural change includes:

[0019] A preset number of directions are obtained through any position. For any direction, a position within a preset radius on one side of the position is recorded as a first reference position, and a position within a preset radius on the other side is recorded as a second reference position; a polynomial is fitted using the coordinates of all first reference positions, and a vector consisting of the coefficients of the polynomial is recorded as a first coefficient vector; a polynomial is fitted using the coordinates of all second reference positions, and a vector consisting of the coefficients of the polynomial is recorded as a second coefficient vector; the modulus of the difference between the first coefficient vector and the second coefficient vector is taken as the sharpness of the local structural change of the position in the direction; the average of the sharpness of the local structural change of the position in all directions is taken as the sharpness of the structural change of the position.

[0020] The present invention characterizes the sharpness of the structural change at each position by analyzing the difference in the change rules of the position coordinates on both sides of each position. This description method is relatively simple and has high accuracy.

[0021] Preferably, the method for obtaining the reliability of the second fitting value includes:

[0022] ;

[0023] Wherein, L2 represents a sequence of second stress values ​​of all positions within the preset range of the position, It indicates the similarity of the variation law of the second stress value of this position with the kth other position, and K2 indicates the reliability of the second fitting value of this position.

[0024] Preferably, the method for obtaining the similarity of the change rules includes:

[0025] The coefficient vector of the fitting model at a position is obtained, and the reciprocal of the Euclidean distance between the coefficient vector of the fitting model at the position and the coefficient vectors of the fitting models at other positions is recorded as the variation pattern similarity.

[0026] The present invention reflects the similarity of the variation rules of each position with other positions by analyzing the similarity of the fitting models of each position. This calculation method is simpler and has higher implementation efficiency.

[0027] Preferably, the error correction based on the comprehensive fitting value of each position includes:

[0028] The comprehensive fitting value at each position is taken as the stress value after error correction.

[0029] In a second aspect, the present invention provides an error correction system for a force sensor, which adopts the following technical solution:

[0030] A force sensor error correction system includes: a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned force sensor error correction method is implemented.

[0031] By adopting the above technical solution, the above-mentioned error correction method of a force sensor is generated into a computer program and stored in a memory so as to be loaded and executed by a processor, thereby manufacturing a terminal device based on the memory and the processor for easy use.

[0032] The present invention has the following technical effects:

[0033] The present invention combines the stress data variation law information collected by the sensor and the stress data variation law information obtained by finite element analysis to fit accurate stress data, thereby realizing error correction. This correction method can effectively prevent the problem that accurate stress data cannot be fitted using sensor data with errors, and effectively improves the accuracy of the fitted stress data.

[0034] Furthermore, when fitting accurate stress data by combining the variation law information of stress data collected by the sensor and the variation law information of stress data obtained by finite element analysis, the reliability of the fitting data obtained by the sensor stress data and the reliability of the fitting data obtained by finite element analysis are analyzed, and the reference degree of the fitting data obtained by each method is set according to the reliability, thereby improving the accuracy of the fitting stress data;

[0035] Furthermore, when analyzing the accuracy of the fitting data obtained by various methods, it is considered that the more accurate the data used for the fitting law is, the more accurate the fitting law is. Therefore, the accuracy of the data used for obtaining the fitting law is analyzed to accurately describe the accuracy of the fitting data obtained by various methods.

[0036] Furthermore, when analyzing the accuracy of the data used in the fitting rules, considering that the stress value is related to the structural sharpness, the accuracy of the data used in the fitting rules is accurately evaluated by analyzing the relationship between the stress values ​​and the structural sharpness collected by various methods, and then the accuracy of the data fitted by various methods is accurately evaluated. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] By reading the detailed description below with reference to the accompanying drawings, the above and other purposes, features and advantages of the exemplary embodiments of the present invention will become readily understood. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-restrictive manner, and the same or corresponding reference numerals represent the same or corresponding parts.

[0038] Figure 1 The present invention is a flowchart of a method for error correction of a force sensor according to an embodiment of the present invention. DETAILED DESCRIPTION

[0039] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0040] It should be understood that when the terms "first", "second", etc. are used in the claims, descriptions, and drawings of the present invention, they are only used to distinguish different objects, rather than to describe a specific order. The terms "include" and "comprise" used in the description and claims of the present invention indicate the presence of the described features, wholes, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or their collections.

[0041] The embodiment of the present invention discloses a method for correcting an error of a force sensor. Figure 1 , comprising steps S1 to S3:

[0042] S1: Use sensors to collect the first stress values ​​at various locations on the surface of the structure, and use finite element analysis software to fit the second stress values ​​at various locations on the surface of the structure.

[0043] Specifically, stress values ​​at various positions on the surface of the structure are collected using stress sensors and recorded as first stress values ​​at various positions; finite element analysis is performed on the structure to obtain a stress model of the structure, and stress values ​​at various positions on the surface of the structure are obtained in the stress model and recorded as second stress values ​​at various positions.

[0044] S2: Fitting the first fitting value according to the variation law of the first stress value within the preset range of each position, and calculating the reliability of the first fitting value; fitting the second fitting value according to the second stress value within the preset range of each position, and calculating the reliability of the second fitting value; taking the reliability as the weight, performing weighted summation on the first fitting stress value and the second fitting stress value to obtain the comprehensive fitting value of each position.

[0045] It should be noted that according to the relevant theory of stress mechanics, the stress on the surface of the structure has a certain correlation in space, so by fitting the correlation law of the surface of the structure, a relatively accurate stress value can be obtained. However, if the data collected by the sensor has a lot of errors, the correlation law fitted by the data collected by the sensor contains some error information, so it is impossible to fit a relatively accurate correlation law using only the data collected by the sensor, and thus it is impossible to fit a relatively accurate stress value relatively accurately.

[0046] It should be further explained that finite element analysis can fit a stress distribution model based on the physical characteristics of the structure. The stress distribution model contains stress information on the surface of the structure. Therefore, the stress information can be combined to compensate for the problem that the sensor data cannot be accurately fitted.

[0047] S20: fitting a first fitting value according to a variation rule of the first stress value in a preset range at each position, and calculating the reliability of the first fitting value.

[0048] It should be noted that, since the stress values ​​collected by the sensor have errors, the stress values ​​fitted using the stress values ​​with errors will also have certain errors; at the same time, the stress data obtained by the stress distribution model obtained by finite element analysis also have certain errors, and the stress values ​​fitted using the stress values ​​with errors will also have errors. The errors of the stress values ​​at each position obtained by each of these two methods are different. At some positions, the error value of one method is larger, and at some positions, the error value obtained by the other method is larger. For example, at some locations with large stress, the error of the stress value collected by the sensor is larger, and at some locations with complex structural changes, the error of the stress value obtained by finite element analysis is larger. Therefore, the fitting values ​​of these two methods can be combined for analysis to obtain relatively accurate stress values. First, the accuracy of the stress value fitting values ​​obtained by each method needs to be analyzed.

[0049] Preferably, as an example, fitting the first fitting value according to the variation rule of the first stress value in the preset range of each position includes:

[0050] The coordinates of each dimension of each position are taken as independent variables, and the first stress value of each position is taken as the dependent variable. A multidimensional polynomial is fitted to the first stress value within a preset range of any position, which is recorded as the first fitting model of each position. The stress value of the position fitted by the first fitting model is recorded as the first fitting value.

[0051] It should be added that the method for obtaining the preset range includes:

[0052] The mean of the first stress value and the second stress value of each position is recorded as the comprehensive stress value, and all the comprehensive stress values ​​in the preset window of each position are fitted with a three-dimensional polynomial, which is recorded as the first analysis model of each position. The Euclidean distance of the coefficient vector of the first analysis model of every two adjacent positions is used as the clustering distance. All positions are clustered to obtain several categories, and the mean of the radius of the circumscribed circle of the positions of all categories is used as the preset radius; a circle with each position as the center and the preset radius as the radius is used as the preset range.

[0053] It is understandable that when fitting data, data similar to the law is generally selected to fit the law, and then the data is fitted according to the law. The preset range obtained in the above process can reflect the distribution size of the data with similar laws, so the data obtained within the preset range of a position generally has a similar variation law to the data at that position.

[0054] It should be added that the method for obtaining the coefficient vector includes:

[0055] Get the coefficients of each item in each dimension of the fitting model, arrange the coefficients of all items in ascending order according to the degree of the item, and record the obtained vectors as the local coefficient vectors of each dimension. The vector obtained by splicing the local coefficient vectors of all dimensions of the fitting model together is recorded as the coefficient vector.

[0056] S21: Calculate the reliability of the first fitting value.

[0057] It should be noted that the first fitting value is obtained based on the data fitting within the preset range, so the more accurate the data within the preset range, the more accurate the fitting value obtained, so the accuracy of the fitting value can be evaluated by evaluating the accuracy of the data within the preset range; according to the relevant knowledge of stress mechanics, the sharper the structural change, the greater the stress value of the structure, and the sharpness of the structural change is correlated with the stress value, so the accuracy of the data within the structural range can be judged by the correlation; due to the poor reproducibility of the variation law of abnormal data, if there is a large similarity between the variation law of the preset range of a position and the variation law of other positions, it means that the amount of abnormal data in the preset range of the position is small, so the accuracy of the data within the preset range of the position can be evaluated by the similarity between the variation law of the preset range in the position and other positions.

[0058] Preferably, as an example, calculating the reliability of the first fitting value includes:

[0059] ;

[0060] Wherein, S represents a sequence of structural change sharpness of all positions within the preset range of the position, L1 represents a sequence of first stress values ​​of all positions within the preset range of the position, Indicates the similarity of the variation law of the first stress value at this position and the kth other position, W indicates the number of positions on the structure surface, max() indicates the maximum value function, represents the correlation between sequence S and sequence L1, and K1 represents the reliability of the first fitting value.

[0061] Understandably, It reflects the similarity of the variation law of the first stress value at this position and other positions. The larger the value, the more likely it is that there are other positions with similar variation laws to the first stress value at this position. The error is generally random. If there is an error in the first stress value within the preset range of the position, it is generally difficult to find other positions with similar variation laws. Therefore, the first stress value within the preset range of the position is less disturbed by the error, and the reliability of the data fitted based on the first stress value with a smaller degree of error interference is greater. It reflects the correlation between the sharpness of the structural change and the first stress value. The larger the value is, the greater the correlation between the sharpness of the structural change and the first stress value within the preset range of the position is. According to the relevant theory of stress mechanics, the sharpness of the structural change is correlated with the stress value. Therefore, if the calculated correlation between the sharpness of the structural change and the first stress value within the preset range of the position is greater, it means that the first stress value within the preset range of the position is not affected by the problem of reduced correlation caused by errors. Therefore, the possibility of errors in the first stress value within the preset range of the position is small, and the accuracy of the fitting value fitted based on the first stress value within the preset range of the position is greater.

[0062] The above embodiments involve the similarity of change rules and the sharpness of structural changes. The following describes the method for obtaining the similarity of change rules and the sharpness of structural changes.

[0063] First, the method of obtaining the similarity of change rules is introduced.

[0064] Preferably, as an example, a method for obtaining the similarity of change rules includes:

[0065] The coefficient vector of the fitting model at a position is obtained, and the reciprocal of the Euclidean distance between the coefficient vector of the fitting model at the position and the coefficient vectors of the fitting models at other positions is recorded as the variation pattern similarity.

[0066] It can be understood that the coefficient vector reflects the law of change. The Euclidean distance between the coefficient vector of the fitting model at this position and the coefficient vectors of the fitting models at other positions can reflect the difference between the fitting model at this position and the fitting models at other positions. The greater the difference in the fitting models, the greater the difference in the law of change between this position and other positions.

[0067] Then the method of obtaining the sharpness of structural changes is introduced.

[0068] Preferably, as an example, the method for obtaining the sharpness of structural change includes:

[0069] A tangent plane is made through the position, and with the position as the center, a preset number of uniformly distributed rays are obtained on the tangent plane and recorded as directions, a plurality of views of the structure are obtained, a view at the position is obtained and recorded as a reference view, and the positions corresponding to each point in the reference view are projected in each direction to obtain a plurality of projection points; for one direction, a position corresponding to a projection point within a preset radius is obtained on one side of the position as the first reference position of the position, and a position corresponding to a projection point within a preset radius is obtained on the other side as the second reference position; a polynomial is fitted using the coordinates of all first reference positions, and a vector formed by the coefficients of the polynomial is recorded as a first coefficient vector; a polynomial is fitted using the coordinates of all second reference positions, and a vector formed by the coefficients of the polynomial is recorded as a second coefficient vector; the modulus of the difference between the first coefficient vector and the second coefficient vector is taken as the sharpness of the local structural change of the position in the direction; the average of the sharpness of the local structural change of the position in all directions is taken as the sharpness of the structural change of the position.

[0070] It can be understood that when the coordinate law on one side of a position is significantly different from that on the other side, it indicates that a sudden change in the coordinate law occurs at that position, and thus the structural change at that position is more acute.

[0071] S22: Fitting a second fitting value according to a second stress value within a preset range at each position.

[0072] Preferably, as an example, fitting the second fitting value according to the second stress value within the preset range of each position includes:

[0073] The coordinates of each dimension of each position are taken as independent variables, and the second stress value of each position is taken as the dependent variable. A multidimensional polynomial is fitted to the second stress value within a preset range of any position, which is recorded as the second fitting model of each position. The stress value of the position fitted by the second fitting model is recorded as the second fitting value.

[0074] S23: calculating the reliability of the second fitting value;

[0075] Preferably, as an example, calculating the reliability of the second fitting value includes:

[0076] ;

[0077] Wherein, L2 represents a sequence of second stress values ​​of all positions within the preset range of the position, It indicates the similarity of the variation law of the second stress value of this position with the kth other position, and K2 indicates the reliability of the second fitting value of this position.

[0078] S24: Using reliability as a weight, weighted summing is performed on the first fitting stress value and the second fitting stress value to obtain a comprehensive fitting value at each position.

[0079] It can be understood that the accuracy of the first fitting stress value and the second fitting stress value is evaluated by reliability, and then the weight is set according to the accuracy to adjust the reference degree of the first fitting stress value and the second fitting stress value when obtaining the accurate stress value.

[0080] S3: Error correction is achieved based on the comprehensive fitting values ​​at each position.

[0081] Preferably, as an example, error correction is implemented based on the comprehensive fitting value of each position, including:

[0082] The comprehensive fitting value at each position is taken as the stress value after error correction.

[0083] An embodiment of the present invention further discloses an error correction system for a force sensor, including a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, an error correction method for a force sensor according to the present invention is implemented.

[0084] The above system also includes other components well known to those skilled in the art, such as a communication bus and a communication interface, and their configuration and functions are known in the art, so they will not be described in detail here.

[0085] In the present invention, the aforementioned memory may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. For example, a computer-readable storage medium may be any suitable magnetic storage medium or magneto-optical storage medium, such as a resistive random access memory, a dynamic random access memory, a static random access memory, an enhanced dynamic random access memory, a high bandwidth memory, a hybrid storage cube, etc., or any other medium that can be used to store the required information and can be accessed by an application, a module, or both. Any such computer storage medium may be part of the device or accessible or connectable to the device.

[0086] Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, various alternatives to the embodiments of the present invention described herein may be employed.

[0087] The above are all preferred embodiments of the present invention, and are not intended to limit the protection scope of the present invention. Therefore, any equivalent changes made based on the structure, shape, and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for error correction of a force sensor, characterized in that: Includes steps: The first stress value of each position on the surface of the structure is collected by using a sensor, and the second stress value of each position on the surface of the structure is fitted by using finite element analysis software; According to the variation law of the first stress value in the preset range of each position, the first fitting value is fitted, and the reliability K1 of the first fitting value is calculated. , S represents a sequence of structural change sharpness of all positions within the preset range of the position, L1 represents a sequence of first stress values ​​of all positions within the preset range of the position, represents the correlation between sequence S and sequence L1, represents the similarity of the variation law of the first stress value of the position and the kth other position, W represents the number of positions on the structure surface, and max() represents the maximum value function; the second fitting value is fitted according to the second stress value within the preset range of each position, and the reliability of the second fitting value is calculated; the first fitting stress value and the second fitting stress value are weighted and summed using the reliability as the weight to obtain the comprehensive fitting value of each position; Error correction is achieved based on the comprehensive fitting value at each position.

2. The error correction method of a force sensor according to claim 1, characterized in that: The method for obtaining the preset range includes: The mean of the first stress value and the second stress value of each position is recorded as the comprehensive stress value, and all the comprehensive stress values ​​in the preset window of each position are fitted with a three-dimensional polynomial, which is recorded as the first analysis model of each position. The Euclidean distance of the coefficient vector of the first analysis model of every two adjacent positions is used as the clustering distance. All positions are clustered to obtain several categories, and the mean of the radius of the circumscribed circle of the positions of all categories is used as the preset radius; a circle with each position as the center and the preset radius as the radius is used as the preset range.

3. The error correction method of a force sensor according to claim 2, characterized in that: The method for obtaining the coefficient vector includes: Get the coefficients of each item in each dimension of the fitting model, arrange the coefficients of all items in ascending order according to the degree of the item, and record the obtained vectors as the local coefficient vectors of each dimension. The vector obtained by splicing the local coefficient vectors of all dimensions of the fitting model together is recorded as the coefficient vector.

4. The error correction method of a force sensor according to claim 2, characterized in that: The method for obtaining the sharpness of the structural change comprises: A preset number of directions are obtained through any position. For any direction, a position within a preset radius on one side of the position is recorded as a first reference position, and a position within a preset radius on the other side is recorded as a second reference position; a polynomial is fitted using the coordinates of all first reference positions, and a vector consisting of the coefficients of the polynomial is recorded as a first coefficient vector; a polynomial is fitted using the coordinates of all second reference positions, and a vector consisting of the coefficients of the polynomial is recorded as a second coefficient vector; the modulus of the difference between the first coefficient vector and the second coefficient vector is taken as the sharpness of the local structural change of the position in the direction; the average of the sharpness of the local structural change of the position in all directions is taken as the sharpness of the structural change of the position.

5. The error correction method of a force sensor according to claim 1, characterized in that: The method for obtaining the reliability of the second fitting value includes: ； Wherein, L2 represents a sequence of second stress values ​​of all positions within the preset range of the position, It indicates the similarity of the variation law of the second stress value of this position with the kth other position, and K2 indicates the reliability of the second fitting value of this position.

6. The error correction method of a force sensor according to claim 3, characterized in that: The method for obtaining the similarity of the change rules includes: The coefficient vector of the fitting model at a position is obtained, and the reciprocal of the Euclidean distance between the coefficient vector of the fitting model at the position and the coefficient vectors of the fitting models at other positions is recorded as the variation pattern similarity.

7. The error correction method of a force sensor according to claim 1, characterized in that: The error correction is realized based on the comprehensive fitting value of each position, including: The comprehensive fitting value at each position is taken as the stress value after error correction.

8. An error correction system for a force sensor, characterized in that: include: A processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, an error correction method for a force sensor according to any one of claims 1 to 7 is implemented.

Images