Industrial robot dynamic monitoring and precision compensation system

Abstract

The application discloses an industrial robot dynamic monitoring and precision compensation system, which comprises a multi-jaw wear state monitoring module, a gripper fine control module and a joint wear limit judgment module. The multi-jaw wear state monitoring module collects the wear state information of the jaws in real time through a micro-deformation monitoring unit and a frequency response analysis unit, and evaluates the wear degree of the jaws. The gripper fine control module calculates the clamping correction parameters according to the wear information and the workpiece clamping data, and adjusts the clamping force and the clamping position. The joint wear limit judgment module monitors the wear state through a single-jaw threshold judgment unit and an overall mean judgment unit, and outputs a warning and provides maintenance suggestions when the threshold is exceeded. The application corrects the wear limit threshold by inputting the jaw and workpiece characteristics, improves the wear monitoring precision and real-time compensation capability, and improves the precision and reliability of the end gripper of the industrial robot.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of industrial robot control, in particular to an industrial robot dynamic monitoring and precision compensation system. BACKGROUND

[0002] In modern industrial production, industrial robots have been widely used in automatic processing, assembly, transportation and other tasks, especially in high-precision and high-repetitive operations, the performance of robots becomes more and more important. The gripper at the end of the robot, as one of its key components, is responsible for accurately controlling the grabbing, positioning and placing of workpieces. In order to ensure efficient and stable work, the performance of the gripper must maintain high accuracy and consistency.

[0003] However, long-term use and repeated operation can cause wear and tear of the gripper, especially in the case of frequent contact between the contact surface of the gripper jaw and the workpiece. The occurrence of wear and tear usually leads to the weakening of the gripper contact force and the deviation of the clamping position, which further affects the accuracy of the gripper, causing defective products or production accidents. The existing technology mainly solves the problem of wear and tear through regular inspection and maintenance, but these methods cannot monitor and accurately compensate the wear and tear state of the gripper in real time, often leading to a decrease in the grabbing accuracy of the workpiece, affecting production efficiency and quality.

[0004] In order to solve this problem, the existing technology has proposed the use of wear sensors and monitoring systems to monitor the working state of the gripper. However, these technologies still have the following shortcomings: the existing wear monitoring method is only limited to rough wear detection, and cannot accurately judge the wear degree and wear position of each gripper jaw, resulting in inaccurate compensation; the existing technology usually relies on regular inspection and manual intervention, causing the equipment to continue working in a worn state, and failing to achieve timely warning and accurate maintenance.

[0005] With the development of industrial automation, the requirements for accuracy and real-time performance of mechanical equipment are becoming higher and higher. Therefore, how to monitor and compensate the wear state of the gripper in real time without affecting production has become a technical problem to be solved. SUMMARY

[0006] The purpose of the present application is to provide an industrial robot dynamic monitoring and precision compensation system, which has the advantages of improving the clamping accuracy and reliability.

[0007] The above technical purpose of the present application is achieved by the following technical scheme:

[0008] The industrial robot dynamic monitoring and precision compensation system comprises a multi-gripper wear state monitoring module, a gripper fine control module and a joint wear limit judgment module,

[0009] The multi-jaw wear state monitoring module is used for acquiring wear state information of multiple jaws in an end gripper of an industrial robot in a repeated clamping operation, and specifically comprises:

[0010] A micro-deformation monitoring unit is arranged at a clamping contact area of each jaw and is used for acquiring micro-deformation information based on a capacitive micro-deformation sensor to obtain a normal compression deformation amount of each jaw in a clamping action, so as to represent a wear trend of a local contact position;

[0011] A frequency response analysis unit is arranged at a structure area of each jaw and is used for acquiring a vibration response frequency of each jaw in a clamping state based on a laser excitation means to obtain frequency response information, so as to represent a change of structure stiffness with use time;

[0012] A multi-jaw wear state fusion unit is used for fusion processing of the micro-deformation information and the frequency response information of each jaw, and outputs wear state information of the corresponding jaw;

[0013] The gripper fine control module is used for calculating clamping correction parameters according to the wear state information of each jaw and clamping information corresponding to a target workpiece, and adjusting a clamping force and a clamping position compensation amount of each jaw according to the clamping correction parameters.

[0014] The joint wear limit judgment module judges whether the end gripper reaches a limit standard according to the wear state information, and outputs early warning information and corresponding maintenance suggestions if it is judged that the limit standard is reached.

[0015] Further setting: the gripper fine control module specifically comprises:

[0016] A target parameter input unit is used for inputting or importing clamping information of a target workpiece, and the clamping information includes a clamping position, surface characteristics and structure parameters;

[0017] A deviation calculation unit is used for calculating a deviation characteristic quantity of each jaw according to the clamping information of the target workpiece and the wear state information corresponding to each jaw;

[0018] A compensation control unit is used for generating corresponding clamping control parameters according to the deviation characteristic quantity, and adjusting a clamping force and a clamping position compensation amount of each jaw according to the clamping control parameters.

[0019] Further setting: the deviation calculation unit comprises:

[0020] A reference parameter extraction subunit is used for extracting a clamping position and a required clamping force of a jaw-workpiece in a non-worn state of the jaw according to the clamping information of the target workpiece, and generating a reference clamping parameter set;

[0021] The attenuation modeling subunit is configured to calculate the clamping force attenuation and the clamping position offset of the clamping jaw under the current wear state based on the wear state information of each clamping jaw, and generate a clamping capability attenuation index;

[0022] The deviation calculation subunit is configured to compare the reference clamping parameter set with the clamping capability attenuation index, and calculate a deviation feature quantity between the clamping performance of each clamping jaw under the current state and the target clamping performance.

[0023] Further, the joint wear limit judgment module specifically comprises:

[0024] The single-clamping-jaw threshold judgment unit is configured to set a corresponding single-clamping-jaw wear limit threshold for each clamping jaw, and output a determination result that the gripper reaches the limit standard when the wear state information of any one clamping jaw exceeds the corresponding single-clamping-jaw wear limit threshold;

[0025] The overall mean value judgment unit is configured to statistically calculate the wear state information of all clamping jaws to obtain an average wear level of the overall gripper, and output a determination result that the gripper reaches the limit standard when the average level exceeds a set group tolerance upper limit;

[0026] The determination result output unit is configured to output early warning information and maintenance suggestions that the wear of the gripper reaches the limit when the wear state information of any one clamping jaw exceeds the single-clamping-jaw wear limit threshold, or the average wear level of the overall gripper exceeds the group tolerance upper limit.

[0027] Further, the multi-clamping-jaw wear state fusion unit specifically comprises:

[0028] The feature extraction subunit is configured to perform multi-scale analysis on the micro-deformation information and the frequency response information of each clamping jaw to extract multi-dimensional features;

[0029] The feature fusion subunit is configured to fuse the multi-dimensional features according to a preset dimension to obtain a fusion feature vector;

[0030] The wear classification subunit is configured to input the fusion feature vector into a trained machine learning model to output a wear grade or a continuous wear score corresponding to each clamping jaw, and simultaneously output corresponding physical wear parameters;

[0031] The wear state information output unit is configured to output the physical wear parameters together with the wear grade or the continuous score as wear state information.

[0032] Further, the multi-scale analysis on the micro-deformation information and the frequency response information of each clamping jaw to extract multi-dimensional features specifically comprises:

[0033] The micro-deformation information is subjected to wavelet transform or short-time Fourier transform to extract micro-deformation features in the signal, the micro-deformation features including maximum deformation, deformation rate of change and time domain energy distribution;

[0034] The frequency response information is subjected to main frequency extraction and harmonic analysis to obtain frequency response features, the frequency response features including main frequency offset and frequency bandwidth change.

[0035] Further, the joint wear limit judgment module further comprises a threshold correction unit, which is configured to generate a threshold correction parameter according to the artificial input jaw characteristics and workpiece characteristics, and correct the single jaw wear limit threshold and the group tolerance upper limit to obtain new single jaw wear limit threshold and group tolerance upper limit according to the threshold correction parameter.

[0036] Further, the generation of the threshold correction parameter according to the artificial input jaw characteristics and workpiece characteristics specifically comprises:

[0037] The wear resistance parameter of the jaw is calculated according to the input jaw characteristics, and a jaw correction factor is generated based on the wear resistance parameter;

[0038] The wear sensitivity of the workpiece is calculated according to the input workpiece characteristics, and a workpiece correction factor is generated based on the wear sensitivity;

[0039] The jaw correction factor and the workpiece correction factor are used to calculate the threshold correction parameter according to a preset weighting value.

[0040] In summary, the present application has the following advantages: through the multi-jaw wear state monitoring module, the system can obtain the wear state information of each jaw in real time, including micro-deformation and frequency response data, accurately evaluate the wear condition, and based on the real-time wear data, the gripper fine control module can immediately calculate the required clamping force correction and position compensation parameters, to ensure that the clamping precision of the jaw is always maintained at the best state. Through continuous monitoring and accurate compensation of the wear state, the system effectively avoids the clamping force attenuation and position deviation caused by wear, thereby improving the precision and working stability of the robot gripper. The system can automatically adjust the working parameters of the gripper, avoiding manual adjustment and error accumulation, and improving the accuracy and reliability of the overall operation. The joint wear limit judgment module can detect the risk of reaching the wear limit in advance through single jaw threshold judgment and overall mean value judgment functions. When the wear degree of a jaw exceeds the set threshold, or the wear level of the overall gripper reaches the group tolerance upper limit, the system will automatically issue a warning, prompt timely maintenance or replacement, provide maintenance suggestions, help reduce sudden failures, reduce maintenance costs, and prolong the service life of the equipment.

[0041] The threshold correction unit allows the adjustment of the wear limit threshold according to the manually inputted jaw characteristics and workpiece characteristics, enabling the system to adaptively adjust according to actual use, providing personalized precision control. The generation process of the jaw correction factor and the workpiece correction factor ensures that the system can flexibly cope with the wear characteristics under different working conditions and operating conditions, further improving the adaptability and precision of the system.

[0042] By reducing the clamping precision error caused by wear, the damage of the workpiece caused by improper clamping is avoided, the production efficiency is improved, the system can adjust the clamping parameters in real time, the idle time and failure rate of the robot are reduced, and continuous and efficient production is ensured. BRIEF DESCRIPTION OF DRAWINGS

[0043] Figure 1 is the overall structure block diagram of the embodiment. DETAILED DESCRIPTION

[0044] The present application will be further described in detail below in conjunction with the drawings.

[0045] Embodiment:

[0046] As shown in Figure 1 The industrial robot dynamic monitoring and precision compensation system comprises a multi-jaw wear state monitoring module, a gripper fine control module and a joint wear limit judgment module.

[0047] The multi-jaw wear state monitoring module is used to obtain the wear state information of multiple jaws in the end gripper of the industrial robot during repeated clamping operations. The multi-jaw wear state monitoring module specifically comprises:

[0048] A micro-deformation monitoring unit is arranged at the clamping contact area of each jaw and is used to obtain micro-deformation information based on a capacitive micro-deformation sensor to collect the normal compression deformation amount of each jaw during clamping action, so as to represent the wear trend of the local contact position.

[0049] A frequency response analysis unit is arranged at the structure area of each jaw and is used to obtain frequency response information based on laser excitation means to obtain the vibration response frequency of each jaw in the clamped state, so as to represent the change of structural stiffness with use time.

[0050] A multi-jaw wear state fusion unit is used to fuse and process the micro-deformation information and the frequency response information of each jaw, and output the wear state information of the corresponding jaw.

[0051] The gripper fine control module is used to calculate the clamping correction parameters according to the wear state information of each jaw and the clamping information corresponding to the target workpiece, and adjust the clamping force and the clamping position compensation amount of each jaw according to the clamping correction parameters.

[0052] The joint wear limit judgment module judges whether the end gripper reaches the limit standard according to the wear state information, and outputs a warning information and a corresponding maintenance suggestion if it is judged that the limit standard is reached.

[0053] The main function of the system is to ensure the high efficiency and accuracy of the end gripper of the robot in repeated work by monitoring the wear state of the gripper in real time and making accurate clamping corrections according to the characteristics of the workpiece. The system monitors the wear of the gripper in real time through the multi-gripper wear state monitoring module, and obtains the wear state information by using frequency response analysis and micro-deformation monitoring technology. The information is transmitted to the gripper fine control module for adjusting the clamping force and clamping position of the gripper. At the same time, the wear limit judgment module dynamically adjusts the wear threshold by analyzing the wear state information, evaluates the wear state of the gripper, and gives an early warning when the wear approaches the limit. The main goal of the system is to ensure that the gripper can maintain the best working state in the environment of long-time high-frequency operation of industrial robots by accurate wear monitoring and adaptive clamping correction strategy, and to avoid insufficient clamping force or positioning error caused by wear.

[0054] The multi-gripper wear state fusion unit specifically comprises:

[0055] A feature extraction subunit is configured to perform multi-scale analysis on the micro-deformation information and the frequency response information of each gripper to extract multi-dimensional features.

[0056] A feature fusion subunit is configured to fuse the multi-dimensional features according to a preset dimension to obtain a fusion feature vector.

[0057] A wear classification subunit is configured to input the fusion feature vector into a trained machine learning classification model to output a wear grade or a continuous wear score corresponding to each gripper, and simultaneously output corresponding physical wear parameters.

[0058] A wear state information output unit is configured to output the physical wear parameters together with the wear grade or the continuous score as wear state information.

[0059] The multi-scale analysis on the micro-deformation information and the frequency response information of each gripper to extract multi-dimensional features specifically comprises:

[0060] Performing wavelet transform or short-time Fourier transform on the micro-deformation information to extract micro-deformation features, wherein the micro-deformation features include maximum deformation, deformation rate of change, and time domain energy distribution.

[0061] Performing main frequency extraction and harmonic analysis on the frequency response information to obtain frequency response features, wherein the frequency response features include main frequency offset and frequency bandwidth change.

[0062] The multi-jaw wear state monitoring module is mainly responsible for acquiring the wear state information of multiple jaws in the end gripper of the industrial robot. The module includes three main parts: a micro-deformation monitoring unit, a frequency response analysis unit, and a multi-jaw wear state fusion unit.

[0063] The micro-deformation monitoring unit is used to collect the micro-deformation information of the jaws generated during the clamping process to represent the wear trend of the jaws. A capacitive micro-deformation sensor is arranged on the clamping contact area of each jaw. The sensor measures the normal deformation of the jaw under the clamping force and collects the micro-deformation of the jaw in real time. The sensor array is arranged in the clamping contact area but away from the clamping surface in contact with the workpiece to avoid affecting the clamping process. The micro-deformation data collected by the sensor is digitized by an analog signal acquisition system, and the data acquisition module ensures that the micro changes in the clamping action can be processed through high-speed sampling and data caching functions.

[0064] The frequency response analysis unit obtains the vibration response frequency of the jaws through laser excitation and analyzes the change in structural stiffness. A pulsed laser is used as the excitation source to excite the natural vibration mode of the jaws through laser reflection. The laser directly irradiates the specific stress area on the surface of the jaws to excite their structural vibration. The laser reflection signal is received by a photodetector and converted into a frequency response signal. The excitation is selected when the jaws are not performing a clamping task. After the clamping action is completed, the vibration response of the jaws is continuously collected, and the laser reflection signal is recorded by the photodetector. Through the high-speed data acquisition module, the vibration frequency of the jaws is obtained, and the frequency response characteristics are measured.

[0065] The multi-jaw wear state fusion unit is responsible for multi-dimensional feature extraction and fusion processing of signals from the micro-deformation monitoring unit and the frequency response analysis unit to generate the wear state information of each jaw.

[0066] The feature extraction subunit performs multi-scale analysis on the micro-deformation information and the frequency response information through wavelet transform or short-time Fourier transform (STFT) to extract multi-dimensional features. The specific process is as follows:

[0067] Wavelet transform is performed on the collected micro-deformation signal to extract the instantaneous frequency and time domain features of the signal. The wavelet transform formula is:

[0068]

[0069] where ψ* is the mother wavelet, a is the scale factor, b is the translation factor, and x(t) is the signal.

[0070] From the results of wavelet transform, the following key features are extracted: maximum deformation: calculate the peak deformation of the signal; deformation rate of change: calculate the rate of change of deformation, reflecting the wear rate of the jaw contact point; time domain energy distribution: by calculating the energy spectrum density of the signal, the time domain distribution of the deformation signal is obtained.

[0071] Perform fast Fourier transform (FFT) on the frequency response signal, extract the main frequency shift and frequency bandwidth change of the signal from the frequency domain, and the fast Fourier transform formula is:

[0072]

[0073] Where X(f) is the frequency response function, x(t) is the time domain signal, and f is the frequency.

[0074] From the FFT results, the following key features are extracted: main frequency shift, reflecting the attenuation of the jaw stiffness, the frequency shift is related to the degree of jaw wear; frequency bandwidth change, the amount of change of the frequency bandwidth is used to represent the degree of structural loss of the jaw.

[0075] The feature fusion sub-unit fuses the micro-deformation features and frequency response features according to the pre-set dimensions to obtain a fusion feature vector. The specific process is as follows:

[0076] Pre-set dimensions: micro-deformation feature dimensions: including maximum deformation, deformation rate of change, time domain energy distribution, etc. Frequency response feature dimensions: including main frequency shift, frequency bandwidth change, etc.

[0077] Assign different weights to each feature dimension. Common weighting methods include weighted average method, principal component analysis (PCA), etc.

[0078] After fusing the micro-deformation features and the frequency response features according to the pre-set dimensions, a fusion feature vector is obtained, which contains multi-dimensional wear information of the jaw.

[0079] The wear classification sub-unit inputs the fusion feature vector into the trained machine learning model, outputs the corresponding wear grade or continuous wear score of each jaw, and simultaneously outputs the corresponding physical wear parameters.

[0080] The fusion feature vector is obtained by multi-scale feature extraction of the micro-deformation information and frequency response information of the jaw, and the fusion of the two. The resulting feature vector contains multiple dimensions (e.g., deformation rate of change, main frequency shift, frequency bandwidth change, etc.). These feature vectors will be input into a trained machine learning classification model. According to the complexity and non-linear relationship of the feature vectors, commonly used machine learning classification models include: Support Vector Machine (SVM): suitable for handling small sample, high-dimensional classification problems, and can effectively handle non-linear problems; Random Forest (RF): through the construction of multiple decision trees to classify data, which can handle high-dimensional feature classification problems; Neural Network (ANN): especially suitable for handling complex non-linear relationships, which can capture complex relationships between features through multiple levels of hidden layers.

[0081] Support Vector Machine (SVM) is a widely used machine learning method for classification and regression problems, and is particularly suitable for handling non-linear relationships. For the classification problem of wear grade, SVM can separate the data by constructing a hyperplane, achieving effective classification. For continuous wear score, SVM can apply Support Vector Regression (SVR) to realize regression analysis.

[0082] Input feature vector (X): features extracted through multi-scale analysis, including micro-deformation information of the jaw (such as maximum deformation, deformation rate of change, etc.) and frequency response information (such as main frequency shift, frequency bandwidth change, etc.), which will be combined into a multi-dimensional fusion feature vector.

[0083] Training data set: includes the frequency response data and micro-deformation data of each jaw as input features, and the target variable is the wear grade (such as "mild wear", "moderate wear", "severe wear") or continuous wear score (e.g., a score of 0-100).

[0084] For wear grade, SVM maps the feature vector to a high-dimensional space and finds the optimal hyperplane in the high-dimensional space to achieve accurate classification. For continuous wear score, Support Vector Regression (SVR) is used to minimize the regression error to predict the wear score. The training process selects the optimal kernel function (such as Gaussian Radial Basis Kernel (RBF)) and optimizes the parameters of the model through cross-validation.

[0085] Through the trained SVM classification model or SVR regression model, input the fusion feature vector X, and output the following results: wear grade (e.g., "mild wear", "moderate wear", or "severe wear"); continuous wear score (e.g., a wear score of 0-100).

[0086] Neural networks (ANN) are particularly suitable for handling complex non-linear relationships, capable of capturing non-linear relationships between features through multiple layers of neurons. For the physical wear parameters (amount of stiffness decay and amount of contact area change), we will use a neural network regression model to predict these continuous variables.

[0087] Input feature vector (X): The input feature vector for the neural network is the same as the SVM model, which is the fusion feature vector of the micro-deformation information and the frequency response information.

[0088] Training dataset: includes the frequency response data, micro-deformation data (input features) of each gripper, and the target variables are the amount of stiffness decay (△K) and the amount of contact area change (△A).

[0089] Structure: The neural network contains an input layer, one or more hidden layers (using ReLU activation function), and an output layer (outputting the amount of stiffness decay and the amount of contact area change).

[0090] Regression model: The output of the neural network is two continuous values: the amount of stiffness decay (△K) and the amount of contact area change (△A).

[0091] Training process: Use mean square error (MSE) as the loss function, optimize the network weights to minimize the prediction error. During the training process, the weights are updated through the backpropagation algorithm (Backpropagation), and the parameters are optimized using optimizers such as Adam.

[0092] Through the trained neural network regression model, given the input fusion feature vector X, the network outputs: the amount of stiffness decay (△K), indicating the degree of stiffness decay of the gripper; the amount of contact area change (△A), indicating the change of the contact area between the gripper and the workpiece.

[0093] Through the above two models, the wear classification sub-unit can simultaneously output: wear grade or continuous wear score (using SVM); amount of stiffness decay and amount of contact area change, i.e. physical wear parameters (using ANN). This combination of multiple models can simultaneously perform classification and regression tasks, thereby providing more comprehensive wear state evaluation.

[0094] In this embodiment, the multi-gripper wear state fusion unit effectively fuses micro-deformation information and frequency response information through multi-scale analysis and machine learning methods, generating wear state information of the gripper. This method extracts features through techniques such as wavelet transform, short-time Fourier transform and fast Fourier transform, and uses machine learning models for wear classification and calculation of physical wear parameters. Finally, the system outputs the wear grade or continuous wear score, and provides accurate physical wear parameters for the system.

[0095] The gripper fine control module specifically includes:

[0096] a target parameter input unit configured to input or import clamping information of a target workpiece, the clamping information including clamping positions, surface characteristics, and structural parameters;

[0097] a deviation calculation unit configured to calculate deviation characteristic quantities of the clamping jaws based on the clamping information of the target workpiece and the wear state information of the clamping jaws;

[0098] a compensation control unit configured to generate corresponding clamping control parameters based on the deviation characteristic quantities, and adjust the clamping force and the clamping position compensation amount of each clamping jaw based on the clamping control parameters.

[0099] The deviation calculation unit comprises:

[0100] a reference parameter extraction subunit configured to extract the contact positions of the clamping jaws and the required clamping force in the non-worn state of the clamping jaws based on the clamping information of the target workpiece, and generate a reference clamping parameter set;

[0101] a decay modeling subunit configured to calculate the clamping force decay amount and the clamping position offset amount of the clamping jaws in the current worn state based on the wear state information of the clamping jaws, and generate a clamping capability decay index;

[0102] a deviation calculation subunit configured to compare the reference clamping parameter set and the clamping capability decay index, and calculate the deviation characteristic quantities between the clamping performance of each clamping jaw in the current state and the target clamping performance.

[0103] The gripper fine control module calculates the clamping correction parameters based on the clamping information of the target workpiece and the wear state information of the clamping jaws, and accurately adjusts the clamping force and the clamping position of each clamping jaw.

[0104] The function of the target parameter input unit is to receive the clamping information of the target workpiece, including the clamping position, surface characteristics, and structural parameters. These information are the basis for clamping correction, the clamping position represents the clamping area of the target workpiece, usually obtained through visual sensor or position sensor, for more complex workpieces, the clamping position can be further adjusted through force sensor or torque sensor. The surface characteristics of the workpiece include surface roughness, hardness, friction coefficient, etc., these characteristics affect the contact quality between the clamping jaw and the workpiece, and affect the wear degree of the clamping jaw. It can be obtained through surface sensors such as laser scanning, friction force sensors, etc. The structural parameters of the workpiece include shape size, mass distribution, etc., which affect the clamping mode of the clamping jaw, usually obtained through three-dimensional scanning or mechanical analysis model.

[0105] The deviation calculation unit calculates the deviation characteristic quantities of the clamping jaws based on the clamping information of the target workpiece and the wear state information of the clamping jaws.

[0106] The reference parameter extraction subunit is used to extract the contact position and the required clamping force of the clamping jaw-workpiece in the non-worn state of the clamping jaw according to the clamping information of the target workpiece, and generate a reference clamping parameter set. The extraction of the contact position: in the non-worn state of the clamping jaw, the contact position is determined by the size, shape of the target workpiece and the contact area of the clamping jaw, the accurate position of the clamping jaw and the workpiece is determined by geometric modeling or mechanical analysis, the coordinates in the coordinate system established with the workpiece as the standard are generated, and the coordinates are taken as the contact position. The calculation of the clamping force: the calculation of the clamping force can adopt the contact mechanics model, especially the Hertz contact model, which is a classic theory for describing the contact force between the surfaces of two objects. When the clamping jaw is in contact with the workpiece, the clamping force is not only related to the deformation amount of the clamping jaw, but also related to the geometric shape and material properties of the contact area. The clamping force calculation formula (Hertz contact theory) is as follows:

[0107]

[0108] Wherein: F is the clamping force, R is the equivalent radius of the contact area, δ is the contact depth (deformation amount), that is, the compression displacement between the clamping jaw and the surface of the workpiece, which can be obtained by experiment, E* is the equivalent Young's modulus of the contact body, and the calculation formula is as follows:

[0109]

[0110] Wherein, E1 and E2 are the Young's moduli of the clamping jaw and the workpiece respectively.

[0111] The reference clamping parameter set contains the contact position and the clamping force, and serves as the ideal clamping parameter of the clamping jaw and the workpiece in the non-worn state.

[0112] The attenuation modeling subunit calculates the clamping force attenuation and the clamping position offset of the clamping jaw in the current worn state according to the stiffness attenuation and the contact area change in the worn state information. The contact area change, stiffness attenuation and clamping force attenuation are combined to obtain a comprehensive nonlinear clamping force attenuation model. The clamping force attenuation is affected by the contact area change and the stiffness attenuation, and we obtain:

[0113]

[0114] Wherein: △A is the change amount of the contact area, △K is the stiffness attenuation, A initial is the contact area in the non-worn state, K initial is the stiffness in the non-worn state, α K and α A are the weight coefficients of the stiffness attenuation and the contact area change, γ A is the nonlinear index of the contact area change to the clamping force attenuation, γ Kis the nonlinear index of the contact area change on the clamping force decay, F base is the reference clamping force in the unworn state.

[0115] The change of the contact area will cause the shift of the clamping position, and the relationship is nonlinear. The influence of the contact area change on the clamping position shift follows the power law relationship:

[0116]

[0117] where: ΔPAis the clamping position shift caused by the contact area change, β A is the influence coefficient of the contact area change on the clamping position, μ A is the nonlinear index of the contact area change on the clamping position shift.

[0118] The stiffness decay will also affect the clamping position. The wear causes the clamping jaw to become softer, and the contact surface position shifts. We can calculate the clamping position shift through the stiffness decay. The relationship follows the power law relationship:

[0119]

[0120] ΔP K is the clamping position shift caused by the stiffness decay, β K is the influence coefficient of the stiffness decay on the clamping position, μ K is the nonlinear index of the stiffness decay on the clamping position shift.

[0121] Combining the influences of the contact area change and the stiffness decay, we get the final clamping position shift:

[0122] ΔP = λ1·ΔP A + λ2·ΔP K

[0123] where: λ 1 and λ2 are the weight coefficients, representing the relative influences of the contact area change and the stiffness decay on the clamping position shift.

[0124] The clamping force decay and the clamping position shift are taken as the clamping ability decay indicators.

[0125] The deviation calculation subunit compares the reference clamping parameter set and the clamping capability attenuation index, that is, the reference clamping parameter set represents the clamping force required by each clamping jaw in the non-worn state and the clamping coordinates on the workpiece, and the clamping capability attenuation index of each clamping jaw represents the amount of clamping force attenuation and the amount of clamping position offset of each clamping jaw under the same operating parameters in the current worn state. By comparing the data of the two, the amount of increase of the clamping force required for each clamping jaw and the adjustment amount of the position of each clamping jaw relative to the workpiece to achieve the clamping force and the clamping position in the non-worn state, that is, the deviation characteristic quantity, can be obtained.

[0126] The compensation control unit converts the deviation characteristic quantity into a clamping control parameter output to the control system of the industrial robot, so that the control system adjusts the control parameter of each clamping jaw and compensates the clamping force and the clamping position to achieve the clamping effect in the non-worn state, thereby improving the accuracy of the clamping of the industrial robot.

[0127] The joint wear limit judgment module specifically includes:

[0128] The single-clamping-jaw threshold judgment unit is configured to set a corresponding single-clamping-jaw wear limit threshold for each clamping jaw, and output a determination result that the gripper reaches the limit standard when the wear state information of any one clamping jaw exceeds the corresponding single-clamping-jaw wear limit threshold. The wear limit threshold of each clamping jaw is related to factors such as the service life, material characteristics, and working load of the clamping jaw. When the system is initially configured, the system sets the wear limit threshold for each clamping jaw according to the characteristics of the clamping jaw and the characteristics of the workpiece. When the system receives the wear state information from the multi-clamping-jaw wear state monitoring module, the single-clamping-jaw threshold judgment unit compares the wear state information of each clamping jaw with the set wear limit threshold. If the wear state information of any one clamping jaw exceeds the threshold, it indicates that the clamping jaw has reached the limit standard.

[0129] The overall mean judgment unit is configured to statistically calculate the wear state information of all clamping jaws to obtain the average wear level of the overall gripper, and output a determination result that the gripper reaches the limit standard when the average level exceeds the set group tolerance upper limit. The function of the overall mean judgment unit is to statistically calculate the wear state information of all clamping jaws to obtain the overall wear level of the gripper, and judge whether the average level exceeds the group tolerance upper limit, so as to judge whether the gripper reaches the wear limit.

[0130] The determination result output unit is configured to output warning information and maintenance suggestions that the wear of the gripper reaches the limit when the wear state information of any one clamping jaw exceeds the single-clamping-jaw wear limit threshold or the average wear level of the overall gripper exceeds the group tolerance upper limit, to guide subsequent maintenance or replacement operations.

[0131] The joint wear limit judgment module further comprises a threshold correction unit, configured to generate a threshold correction parameter according to the manually input jaw characteristics and workpiece characteristics, and correct the single-jaw wear limit threshold and the group tolerance upper limit to obtain new single-jaw wear limit threshold and group tolerance upper limit according to the threshold correction parameter.

[0132] The threshold correction parameter is generated according to the manually input jaw characteristics and workpiece characteristics, and specifically comprises:

[0133] The wear resistance parameter of the jaw is calculated according to the input jaw characteristics, and a jaw correction factor is generated based on the wear resistance parameter;

[0134] The wear sensitivity of the workpiece is calculated according to the input workpiece characteristics, and a workpiece correction factor is generated based on the wear sensitivity;

[0135] The jaw correction factor and the workpiece correction factor are calculated to generate the threshold correction parameter according to a preset weighting value.

[0136] The wear resistance parameter can be calculated by factors such as the hardness (H) of the material, the friction coefficient (μ) and the working temperature (T), etc. Assuming that the wear resistance parameter M claw The relationship between these factors is:

[0137] M claw = k1·H·μ·e -α·T

[0138] Wherein H is the material hardness of the jaw (unit: HV, Vickers hardness), μ is the friction coefficient when the jaw surface contacts the workpiece, T is the working temperature (unit: ℃), which affects the thermal expansion and fatigue of the material, k1 is an empirical coefficient determined based on experimental data, and α is the temperature influence factor on the wear resistance of the material. The jaw correction factor is calculated by comparing the wear resistance parameter of the jaw with the preset reference wear resistance parameter.

[0139] The wear sensitivity of the workpiece can be calculated by the following formula, which takes into account the influence of workpiece hardness and surface roughness:

[0140]

[0141] Wherein H workpiece is the hardness of the workpiece (unit: HV), R workpiece is the roughness of the workpiece surface (unit: μm), and k2 is an empirical coefficient determined based on experimental data. The workpiece correction factor is calculated by comparing the wear sensitivity with the preset reference wear sensitivity.

[0142] These correction measures ensure high accuracy of the system in actual application, and can dynamically adjust the wear judgment standard according to the actual use.

[0143] In summary, the present application has the following advantages: through the multi-jaw wear state monitoring module, the system can obtain real-time wear state information of each jaw, including micro-deformation and frequency response data, and accurately assess the wear condition. Based on the real-time wear data, the gripper fine control module can immediately calculate the required clamping force correction and position compensation parameters, ensuring that the clamping accuracy of the jaws is always maintained at the optimal state. Through continuous monitoring and accurate compensation of the wear state, the system effectively avoids the clamping force attenuation and position deviation caused by wear, thereby improving the accuracy and stability of the robot gripper. The system can automatically adjust the working parameters of the gripper, avoiding manual adjustment and error accumulation, and improving the overall operation accuracy and reliability. The joint wear limit judgment module can detect the risk of reaching the wear limit in advance through single-jaw threshold judgment and overall mean value judgment functions. When the wear degree of a jaw exceeds the set threshold, or the overall wear level of the gripper reaches the group tolerance upper limit, the system will automatically issue a warning, prompt timely maintenance or replacement, provide maintenance suggestions, help reduce sudden failures, reduce maintenance costs, and prolong the service life of the equipment.

[0144] The threshold correction unit allows adjustment of the wear limit threshold based on manually inputted jaw characteristics and workpiece characteristics, enabling the system to adaptively adjust according to actual usage, provide personalized precision control, and ensure that the system can flexibly cope with wear characteristics under different working conditions and operating conditions, further improving the adaptability and accuracy of the system.

[0145] By reducing the clamping accuracy errors caused by wear, the system avoids workpiece damage caused by improper clamping, improves production efficiency, and can adjust clamping parameters in real time, reducing the idle time and failure rate of the robot, ensuring continuous and efficient production.

[0146] The above-described embodiments do not constitute a limitation on the scope of protection of the technical solutions. Any modifications, equivalent replacements, and improvements made within the spirit and principles of the above-described embodiments should be included within the scope of protection of the technical solutions.

Claims

1. A system for dynamic monitoring and precision compensation of an industrial robot, characterized in that Comprise: Multi-jaw wear state monitoring module, gripper fine control module and joint wear limit judgment module, The multi-jaw wear state monitoring module is used to obtain the wear state information of multiple jaws in the end gripper of the industrial robot in the repeated clamping operation process, and the multi-jaw wear state monitoring module specifically comprises: Micro-deformation monitoring unit, which is arranged in each jaw clamping contact area and is used to obtain micro-deformation information based on the normal compression deformation amount of each jaw in the clamping action by a capacitive micro-deformation sensor to represent the wear trend of the local contact part; Frequency response analysis unit, which is arranged in each jaw structure area and is used to obtain frequency response information based on laser excitation means to obtain the vibration response frequency of each jaw in the clamping state to represent the change of structural stiffness with use time; Multi-jaw wear state fusion unit, which is used to fuse the micro-deformation information and frequency response information of each jaw, and output the wear state information of the corresponding jaw; The gripper fine control module is used to calculate the clamping correction parameters according to the wear state information of each jaw and the clamping information corresponding to the target workpiece, and adjust the clamping force and clamping position compensation amount of each jaw according to the clamping correction parameters; The joint wear limit judgment module judges whether the end gripper reaches the limit standard according to the wear state information, and outputs the warning information and the corresponding maintenance suggestion if it is judged that the limit standard is reached; The gripper fine control module specifically comprises: Target parameter input unit, the target parameter input unit is used to input or import the clamping information of the target workpiece, and the clamping information includes the clamping position, surface characteristics and structural parameters; Deviation calculation unit, the deviation calculation unit is used to calculate the deviation characteristic quantity of each jaw according to the clamping information of the target workpiece and the wear state information corresponding to each jaw; Compensation control unit, the compensation control unit is used to generate the corresponding clamping control parameters according to the deviation characteristic quantity, and adjust the clamping force and clamping position compensation amount of each jaw according to the clamping control parameters; The deviation calculation unit comprises: Reference parameter extraction subunit, which is used to extract the contact position of the jaw-workpiece and the required clamping force under the non-worn state of the jaw according to the clamping information of the target workpiece, and generate a reference clamping parameter set; Decay modeling subunit, which is used to calculate the clamping force decay amount and clamping position offset amount of the jaw under the current wear state based on the wear state information of each jaw, and generate a clamping capacity decay index; Deviation calculation subunit, which is used to compare the reference clamping parameter set with the clamping capacity decay index, and calculate the deviation characteristic quantity between the clamping performance of each jaw under the current state and the target clamping performance.

2. The industrial robot dynamic monitoring and accuracy compensation system of claim 1, wherein, The joint wear limit judgment module specifically comprises: Single-jaw threshold judgment unit, which is used to set the corresponding single-jaw wear limit threshold for each jaw, and outputs the determination result that the gripper reaches the limit standard when the wear state information of any one jaw exceeds the corresponding single-jaw wear limit threshold. The overall average judgment unit is configured to statistically calculate the wear state information of all the clamping jaws, obtain an average wear level of the overall gripper, and output a determination result that the gripper reaches a limit criterion when the average wear level exceeds a set group tolerance upper limit. The determination result output unit is configured to output early warning information and maintenance suggestions that the gripper wear reaches a limit when any clamping jaw wear state information exceeds a single clamping jaw wear limit threshold or the average wear level of the overall gripper exceeds the group tolerance upper limit.

3. The industrial robot dynamic monitoring and accuracy compensation system of claim 1, wherein, The multi-clamping jaw wear state fusion unit specifically includes: The feature extraction subunit is configured to perform multi-scale analysis on the micro-deformation information and the frequency response information of each clamping jaw to extract multi-dimensional features. The feature fusion subunit is configured to fuse the multi-dimensional features according to a preset dimension to obtain a fusion feature vector. The wear classification subunit is configured to input the fusion feature vector into a trained machine learning model to output a wear grade or a continuous wear score corresponding to each clamping jaw, and simultaneously output corresponding physical wear parameters. The wear state information output unit is configured to output the physical wear parameters together with the wear grade or the continuous score as wear state information.

4. The industrial robot dynamic monitoring and accuracy compensation system of claim 3, wherein, The multi-scale analysis on the micro-deformation information and the frequency response information of each clamping jaw specifically includes: Performing wavelet transform or short-time Fourier transform on the micro-deformation information to extract micro-deformation features, including maximum deformation, deformation change rate, and time domain energy distribution. Performing main frequency extraction and harmonic analysis on the frequency response information to obtain frequency response features, including main frequency offset and frequency bandwidth change.

5. The industrial robot dynamic monitoring and accuracy compensation system of claim 2, wherein, The joint wear limit judgment module further includes a threshold correction unit configured to generate threshold correction parameters according to manually input clamping jaw characteristics and workpiece characteristics, and correct the single clamping jaw wear limit threshold and the group tolerance upper limit according to the threshold correction parameters to obtain new single clamping jaw wear limit threshold and group tolerance upper limit.

6. The industrial robot dynamic monitoring and accuracy compensation system of claim 5, wherein, The generation of threshold correction parameters according to manually input clamping jaw characteristics and workpiece characteristics specifically includes: Calculating a wear resistance parameter of the clamping jaw according to the input clamping jaw characteristics, and generating a clamping jaw correction factor based on the wear resistance parameter; Calculating a wear sensitivity of the workpiece according to the input workpiece characteristics, and generating a workpiece correction factor based on the wear sensitivity; The clamping jaw correction factor and the workpiece correction factor are used to calculate the threshold correction parameters according to a preset weighting value.

Images