A nut servo pressing depth control system and control method

Abstract

The present invention relates to the field of mechanical manufacturing technology, and in particular to a nut servo pressing depth control system and a control method, comprising the following steps: S1, workpiece detection: acquiring characteristic data of the workpiece; S2, parameter setting: setting initial parameters of a servo motor; S3, pressing force path optimization: generating an optimal pressing force path through a pressing force path prediction model; S4, servo pressing control: performing a nut pressing operation according to the optimized pressing force path; S5, pressing depth detection and correction: checking the verticality and position of the nut; S6, learning feedback: collecting the verticality and position of the nut after fine-tuning and correction in real time, and continuously optimizing the pressing force path and the pressing operation; the present invention improves the accuracy and adaptability of the pressing force path prediction, significantly improves the controllability of the pressing process, and meets the demand for high-precision assembly of complex workpieces.

Description

Technical Field

[0001] The invention relates to the technical field of mechanical manufacturing, and in particular to a nut servo pressing depth control system and a control method. Background Art

[0002] In the field of industrial manufacturing, nuts are a common connecting fastener, and their assembly quality is directly related to the reliability and performance of the final product. The traditional nut pressing method usually relies on mechanical force or pneumatic equipment for operation. The accuracy and stability of the pressing process have a vital impact on the assembly quality. With the development of high-precision manufacturing technology, higher requirements are put forward for the position, pressing force and path control during the nut pressing process to ensure the correct positioning and reliable connection of the nut in the complex workpiece.

[0003] There are generally some shortcomings in the existing nut pressing methods. First, the traditional methods lack comprehensive detection and analysis of workpiece characteristic data, resulting in a lack of accuracy in the pressing force path design, which affects the assembly quality. Second, the existing system lacks real-time feedback and dynamic adjustment mechanism during the pressing process, which is prone to problems such as uneven pressing force, path deviation and stress concentration, increasing the risk in the assembly process. In addition, the existing technology fails to effectively utilize historical data for operation optimization, and cannot continuously improve the pressing accuracy and efficiency, limiting its application in high-precision assembly.

[0004] The present invention aims to overcome the deficiencies of the prior art and provide a nut servo pressing depth control system and a control method, which can significantly improve assembly quality and production efficiency and meet the needs of the high-precision manufacturing field. Summary of the invention

[0005] The invention provides a nut servo pressing depth control system and a control method.

[0006] A nut servo pressing depth control method comprises the following steps:

[0007] S1, workpiece detection: multiple sensors are used to detect the surface shape, material properties and initial positioning of the workpiece, and obtain the characteristic data of the workpiece, including geometric shape, hardness distribution and surface roughness information;

[0008] S2, parameter setting: according to the results of workpiece detection, set the initial parameters of the servo motor, including pressing force, pressing speed and target depth;

[0009] S3, press force path optimization: Based on the set initial parameters and combined with the characteristic data of the workpiece, the optimal press force path is generated through the press force path prediction model, including:

[0010] S31, data preprocessing: preprocessing the characteristic data of the workpiece and the initial parameters of the servo motor, including data cleaning, standardization and feature extraction;

[0011] S32, multi-path prediction: inputting the pre-processed characteristic data and initial parameters into the pressing force path prediction model to predict and generate multiple pressing force paths;

[0012] S33, multi-path evaluation and optimization: compare the performance of multiple generated pressure paths and determine the optimal pressure path;

[0013] S4, servo press-in control: adopts a closed-loop control mechanism, drives the press-in force head through a servo motor, and performs the nut press-in operation according to the optimized press-in force path under the dual feedback of the press-in force and displacement;

[0014] S5, press-in depth detection and correction: After the nut is pressed in, the final press-in depth is detected by the displacement sensor, and the verticality and position of the nut are checked. If a deviation from the target value is detected, fine-tuning correction is automatically performed;

[0015] S6, learning feedback: collect the verticality and position of the nut after fine-tuning and correction in real time, continuously optimize the pressing force path and pressing operation, and provide parameter and path suggestions for future pressing operations.

[0016] Optionally, the workpiece detection in S1 includes:

[0017] S11, sensor selection and configuration: multiple sensors are configured to detect the surface shape, material properties and initial positioning of the workpiece. The sensors include laser displacement sensors, ultrasonic sensors and optical profilers;

[0018] S12, geometric shape detection: use a laser displacement sensor to scan the workpiece surface, obtain the workpiece's geometric shape data, and measure the workpiece's surface profile and size through the displacement change of the reflected light beam;

[0019] S13, hardness distribution evaluation: The internal structure and hardness distribution of the workpiece material are measured by an ultrasonic sensor, and the density change of the internal material of the workpiece is evaluated by transmitting and receiving sound wave signals to determine the hardness distribution;

[0020] S14, surface roughness measurement: Use an optical profiler to scan the workpiece surface and measure the surface roughness. The optical profiler generates roughness data of the workpiece surface by analyzing the intensity and distribution of the reflected light on the surface.

[0021] Optionally, the parameter setting in S2 includes:

[0022] S21, calculation of initial pressing force and pressing speed: Calculate the initial pressing force and pressing speed of the current workpiece based on the workpiece characteristic data, expressed as:

[0023] ;

[0024] ;

[0025] in, is the initial pressing force, is the workpiece surface area, is the average hardness of the workpiece, is the workpiece surface roughness coefficient, is the initial pressing speed, is the material density, is the geometric shape factor, and is the empirical coefficient;

[0026] S22, target depth setting and servo motor adjustment: combined with the target pressing depth of the workpiece , by adjusting the control parameters of the servo motor, set the final initial pressing force , Pressing speed and target depth , the control parameter adjustment of the servo motor is expressed as:

[0027] ;

[0028] in, is the adjustment parameter of the servo motor, is the response factor of the servo motor.

[0029] Optionally, the data preprocessing in S31 includes:

[0030] S311, data cleaning: using mean filtering method to check and remove outliers and noise data in workpiece characteristic data and initial parameters;

[0031] S312, standardization processing: standardizing the characteristic data and initial parameters of the cleaned workpiece;

[0032] S313, feature extraction: extract key features from the standardized workpiece characteristic data and initial parameters through the principal component analysis (PCA) algorithm, including the shape characteristics of the workpiece, the material hardness change rate, and the surface roughness gradient.

[0033] Optionally, the pressing force path prediction model in S32 adopts a convolutional neural network (CNN) model, and the convolutional neural network (CNN) model includes:

[0034] S321, data input and multi-channel feature processing: multi-channel processing is performed on the input data, and each channel corresponds to a characteristic data, which is expressed as:

[0035] ;

[0036] ;

[0037] ;

[0038] in, , , They are the characteristic diagrams of geometry, hardness distribution and surface roughness, , , are the convolution kernels of different channels respectively. , , is the input data of different channels, , , is the bias term, is the activation function;

[0039] S322, feature fusion and deep convolution: The feature maps of multiple channels are fused, and multi-level complex features are extracted through deep convolution layers to capture the relationship between the multi-dimensional characteristics of the workpiece and the indentation force path, which is expressed as:

[0040] ;

[0041] ;

[0042] in, is the feature fusion operation, is the feature map after deep convolution, is the convolution kernel of the depth convolution layer, is the fused feature map, is the bias term;

[0043] S323, multi-scale feature extraction: After the depth convolution, a multi-scale convolution operation is introduced to extract features at different scales through convolution kernels of different sizes (such as 3×3, 5×5, 7×7), which is expressed as:

[0044] ;

[0045] ;

[0046] ;

[0047] ;

[0048] in, , , It is the feature map extracted under different convolution kernel sizes. , , are convolution kernels of different sizes, , , is the corresponding bias term, It is the feature map after multi-scale feature fusion;

[0049] S324, fully connected layer path prediction: Flatten the feature map after multi-scale convolution and input it into the fully connected layer to predict the pressure path. The fully connected layer is responsible for integrating all extracted features and outputting multiple pressure paths, which are expressed as:

[0050] ;

[0051] in, For the predicted indentation force paths, is the weight of the fully connected layer, is the flattened feature map, is the bias term, is the softmax activation function.

[0052] Optionally, the multipath evaluation and optimization in S33 includes:

[0053] S331, path performance index calculation: Perform performance analysis on each indentation force path, calculate its indentation force distribution uniformity, path smoothness and stress concentration, expressed as:

[0054] ;

[0055] ;

[0056] ;

[0057] ;

[0058] in, For the The comprehensive performance index of each path, is the uniformity of the pressure distribution of the path, For path The pressure at each point, is the average pressure force along the path, is the smoothness of the path, For path The location of the point, For path The location of the point, is the stress concentration of the path, For path The contact area of ​​each point, , , is the weight coefficient, is the number of data points on the path;

[0059] S332, path cost function calculation: Comprehensively calculate the performance indicators of each path and calculate its cost function , expressed as:

[0060] ;

[0061] in, For the The cost function of the path, , , is the weight coefficient in the cost function;

[0062] S333, optimal path selection: According to the calculation result of the cost function, the path with the smallest cost function value is selected as the optimal pressing force path, and it is used as the final output to guide the pressing operation of the nut, which is expressed as:

[0063] ;

[0064] in, is the optimal pressing force path, The index of the path.

[0065] Optionally, the servo push control in S4 includes:

[0066] S41, Pressing force feedback control: collect the data of the pressing force sensor in real time, compare it with the target pressing force, and calculate the pressing force error The output torque of the servo motor is adjusted by the proportional-integral-differential PID controller. To reduce the error, it can be expressed as:

[0067] ;

[0068] in, is the pressing force error, is the target pressure, is the actual measured pressing force, , , are the proportional, integral and differential coefficients of the PID controller respectively;

[0069] S42, displacement feedback control: real-time monitoring of displacement sensor data, comparison of actual displacement with target displacement, and calculation of displacement error , adjust the moving speed of the servo motor through the PID controller , expressed as:

[0070] ;

[0071] in, is the displacement error, is the target displacement, is the actual measured displacement, , , are the proportional, integral and differential coefficients of the displacement control PID controller respectively;

[0072] S43, closed-loop control integration: The pressing force and displacement feedback control are integrated into the closed-loop control mechanism. During the entire pressing process, the servo motor dynamically adjusts the torque and position of the pressing head according to the real-time feedback of the pressing force and displacement data, so that the nut is pressed into the predetermined depth along the optimized pressing force path.

[0073] Optionally, the pressing depth detection and correction in S5 includes:

[0074] S51, depth detection: the actual pressing depth of the nut is measured by a displacement sensor and press it and the target into depth Compare and calculate the depth error , expressed as:

[0075] ;

[0076] S52, verticality and position detection: detect the verticality of the nut and location and compared with the target value to calculate the verticality error and position error , expressed as:

[0077] ;

[0078] ;

[0079] in, is the target verticality, is the target location;

[0080] S53, fine-tuning correction: According to the detected depth error, verticality error and position error, the control parameters of the servo motor are adjusted through the fine-tuning algorithm and , to correct the position and angle of the nut, expressed as:

[0081] ;

[0082] ;

[0083] in, is the adjusted servo motor torque, is the adjusted servo motor speed, , , is the corresponding correction factor.

[0084] Optionally, the learning feedback in S6 includes:

[0085] S61, data collection and recording: real-time acquisition of the verticality of the nut after fine-tuning correction and location , and the initial pressing force path and operating parameters are recorded in the database together to form a historical data set of the pressing operation. The operating parameters include pressing force, pressing speed, target depth, displacement control parameters, servo motor torque and servo motor acceleration;

[0086] S62, Path and Operation Optimization: Initial Press Force Path Optimization Using Regression Analysis and operating parameters Optimize, expressed as:

[0087] ;

[0088] ;

[0089] in, is the optimized pressing force path, is the feedback path adjustment amount, For the optimized operating parameters, is the parameter adjustment amount of feedback, is the learning rate;

[0090] S63, parameter and path suggestion: obtained through and Serves as a suggested parameter and path for future push operations.

[0091] A nut servo pressing depth control system, used to implement the above-mentioned nut servo pressing depth control method, includes the following modules:

[0092] Workpiece detection module: detects the surface shape, material properties and initial positioning of the workpiece through multiple sensors, and obtains the characteristic data of the workpiece, including geometric shape, hardness distribution and surface roughness information;

[0093] Parameter setting module: according to the characteristic data of the workpiece, the initial parameters of the servo motor are set, including the pressing force, pressing speed and target depth;

[0094] Press-in force path optimization module: Based on the set initial parameters and combined with the characteristic data of the workpiece, the optimal press-in force path is generated through the press-in force path prediction model;

[0095] Servo press-in control module: adopts a closed-loop control mechanism, drives the press-in force head through a servo motor, and performs the nut press-in operation according to the optimized press-in force path under the dual feedback of the press-in force and displacement;

[0096] Pressing depth detection and correction module: After the nut is pressed in, the displacement sensor detects the final pressing depth and checks the verticality and position of the nut. If a deviation from the target value is detected, fine-tuning and correction are automatically performed;

[0097] Learning feedback module: collects the verticality and position of the nut after fine-tuning and correction in real time, continuously optimizes the pressing force path and pressing operation, and provides parameter and path suggestions for future pressing operations.

[0098] Beneficial effects of the present invention:

[0099] The present invention effectively improves the accuracy and stability of the nut pressing process through comprehensive detection and precise analysis of workpiece characteristic data, combined with parameter setting and pressing force path optimization, and adopts a closed-loop control mechanism to adjust the torque and movement speed of the servo motor in real time to ensure that the nut is accurately pressed along the optimized path. Through multi-level detection and correction steps, not only the uniformity of the pressing force and the smoothness of the path are guaranteed, but also the stress concentration phenomenon is effectively reduced, and the consistency and reliability of nut assembly are significantly improved.

[0100] The present invention, through a convolutional neural network model, uses multi-channel data processing and multi-scale feature extraction, can accurately capture the complex relationship between the geometric shape, material properties and surface roughness of the workpiece, generate multiple possible pressing force paths, and combine multi-path evaluation and optimization to ensure the selection of the optimal pressing force path for the nut pressing operation, effectively improving the accuracy and adaptability of the pressing force path prediction, significantly improving the controllability of the pressing process, and meeting the needs of high-precision assembly of complex workpieces.

[0101] The present invention, by introducing a learning feedback mechanism, collects in real time the verticality and position data of the nut after fine-tuning and correction, and continuously improves the pressing accuracy and path control through regression analysis and optimization of the initial pressing force path and operating parameters. Through the accumulation of historical data and feedback optimization, the accuracy and efficiency of future pressing operations are continuously improved. This mechanism not only effectively reduces errors and deviations in the assembly process, ensures the stability and consistency of the operation, but also greatly improves production efficiency and product quality, and has important industrial application value. BRIEF DESCRIPTION OF THE DRAWINGS

[0102] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only for the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0103] Figure 1 A schematic diagram of a control method flow chart of an embodiment of the present invention;

[0104] Figure 2 Schematic diagram of system function modules according to an embodiment of the present invention. DETAILED DESCRIPTION

[0105] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. At the same time, it is explained here that in order to make the embodiments more detailed, the following embodiments are the best and preferred embodiments, and those skilled in the art may also adopt other alternatives to implement some known technologies; and the accompanying drawings are only for more specific description of the embodiments, and are not intended to specifically limit the present invention.

[0106] It should be noted that the references to "one embodiment", "embodiment", "exemplary embodiments", "some embodiments" and the like in the specification indicate that the embodiments described may include specific features, structures or characteristics, but not every embodiment may include the specific features, structures or characteristics. In addition, when a specific feature, structure or characteristic is described in conjunction with an embodiment, it should be within the knowledge of a person skilled in the art to implement such feature, structure or characteristic in conjunction with other embodiments (whether or not explicitly described).

[0107] In general, a term can be understood, at least in part, from its use in context. For example, depending, at least in part, on the context, the term "one or more" as used herein can be used to describe any feature, structure, or characteristic in the singular sense, or can be used to describe a combination of features, structures, or characteristics in the plural sense. Additionally, the term "based on" can be understood as not necessarily intended to convey an exclusive set of factors, but can instead, depending, at least in part, on the context, allow for the presence of other factors that are not necessarily explicitly described.

[0108] like Figure 1 As shown, a nut servo pressing depth control method comprises the following steps:

[0109] S1, workpiece detection: multiple sensors are used to detect the surface shape, material properties and initial positioning of the workpiece, and obtain the characteristic data of the workpiece, including geometric shape, hardness distribution and surface roughness information;

[0110] S2, parameter setting: according to the results of workpiece detection, set the initial parameters of the servo motor, including pressing force, pressing speed and target depth;

[0111] S3, press force path optimization: Based on the set initial parameters and combined with the characteristic data of the workpiece, the optimal press force path is generated through the press force path prediction model, including:

[0112] S31, data preprocessing: preprocessing the characteristic data of the workpiece and the initial parameters of the servo motor, including data cleaning, standardization and feature extraction;

[0113] S32, multi-path prediction: inputting the pre-processed characteristic data and initial parameters into the pressing force path prediction model to predict and generate multiple pressing force paths;

[0114] S33, multi-path evaluation and optimization: compare the performance of multiple generated pressure paths and determine the optimal pressure path;

[0115] S4, servo press-in control: adopts a closed-loop control mechanism, drives the press-in force head through a servo motor, and performs the nut press-in operation according to the optimized press-in force path under the dual feedback of the press-in force and displacement;

[0116] S5, press-in depth detection and correction: After the nut is pressed in, the final press-in depth is detected by the displacement sensor, and the verticality and position of the nut are checked. If a deviation from the target value is detected, fine-tuning correction is automatically performed to ensure that the nut is pressed into the predetermined depth completely according to the optimized path;

[0117] S6, learning feedback: collect the verticality and position of the nut after fine-tuning and correction in real time, continuously optimize the pressing force path and pressing operation, and provide parameter and path suggestions for future pressing operations;

[0118] Through the above content, accurate prediction and real-time adjustment of the pressing force path are achieved, ensuring high precision and high stability of the nut pressing process. At the same time, combined with the learning feedback mechanism, future operations are continuously optimized, providing reliable technical support for quality improvement and efficiency optimization in the manufacturing process.

[0119] Artifact detection in S1 includes:

[0120] S11, sensor selection and configuration: multiple sensors are configured to detect the surface shape, material properties and initial positioning of the workpiece. The sensors include laser displacement sensors, ultrasonic sensors and optical profilers;

[0121] S12, geometric shape detection: use a laser displacement sensor to scan the workpiece surface, obtain the workpiece's geometric shape data, and measure the workpiece's surface profile and size through the displacement change of the reflected light beam;

[0122] S13, hardness distribution evaluation: The internal structure and hardness distribution of the workpiece material are measured by an ultrasonic sensor, and the density change of the internal material of the workpiece is evaluated by transmitting and receiving sound wave signals to determine the hardness distribution;

[0123] S14, surface roughness measurement: use an optical profiler to scan the workpiece surface and measure the surface roughness. The optical profiler generates roughness data of the workpiece surface by analyzing the intensity and distribution of the reflected light on the surface;

[0124] Through the above content, high precision and multi-dimensional analysis of the test results are ensured, providing accurate and reliable data support for subsequent optimization of the pressing force path, thereby improving the accuracy and consistency of the overall operation.

[0125] The parameter settings in S2 include:

[0126] S21, calculation of initial pressing force and pressing speed: Calculate the initial pressing force and pressing speed of the current workpiece based on the workpiece characteristic data, expressed as:

[0127] ;

[0128] ;

[0129] in, is the initial pressing force, is the workpiece surface area, is the average hardness of the workpiece, is the workpiece surface roughness coefficient, is the initial pressing speed, is the material density, is the geometric shape factor, and is the empirical coefficient;

[0130] S22, target depth setting and servo motor adjustment: combined with the target pressing depth of the workpiece , by adjusting the control parameters of the servo motor, set the final initial pressing force , Pressing speed and target depth , the control parameter adjustment of the servo motor is expressed as:

[0131] ;

[0132] in, is the adjustment parameter of the servo motor, is the response coefficient of the servo motor;

[0133] Through the above content, not only the accuracy and stability of the pressing process are improved, but also it is ensured that the servo motor can accurately perform the pressing operation, thereby reducing the possible errors in the pressing process and improving the reliability and consistency of the overall operation.

[0134] The data preprocessing in S31 includes:

[0135] S311, data cleaning: Use the mean filter method to check and remove outliers and noise data in the workpiece characteristic data and initial parameters to ensure the accuracy and consistency of the input data, expressed as:

[0136] ;

[0137] in, is the data point after cleaning, is the original data point, is the size of the sliding window;

[0138] S312, standardization processing: standardize the characteristic data and initial parameters of the cleaned workpiece to ensure that the data are compared and calculated on the same scale, expressed as:

[0139] ;

[0140] in, is the standardized data, For the cleaned data, is the mean of the data set, is the standard deviation of the data set;

[0141] S313, feature extraction: The principal component analysis (PCA) algorithm is used to extract key features from the standardized workpiece characteristic data and initial parameters, including the shape characteristics of the workpiece, the material hardness change rate, and the surface roughness gradient, which are expressed as:

[0142] ;

[0143] in, is the extracted feature vector, is the eigenvector matrix, is the standardized data matrix;

[0144] Through the above content, the accuracy and consistency of the workpiece characteristic data and the initial parameters of the servo motor can be improved, providing a more accurate data basis for subsequent operations, thereby greatly improving the accuracy and stability of the entire pressing process.

[0145] The pressure path prediction model in S32 uses a convolutional neural network (CNN) model, which includes:

[0146] S321, data input and multi-channel feature processing: In view of the diversity of workpiece characteristic data (such as geometric shape, hardness distribution and surface roughness), multi-channel processing is performed on the input data. Each channel corresponds to a characteristic data. The data of each channel is extracted through an independent convolution layer, and then the features are fused in the subsequent layer, which is expressed as:

[0147] ;

[0148] ;

[0149] ;

[0150] in, , , They are the characteristic diagrams of geometry, hardness distribution and surface roughness, , , are the convolution kernels of different channels respectively. , , is the input data of different channels, , , is the bias term, is the activation function;

[0151] S322, feature fusion and deep convolution: The feature maps of multiple channels are fused, and multi-level complex features are extracted through deep convolution layers to capture the relationship between the multi-dimensional characteristics of the workpiece and the indentation force path, which is expressed as:

[0152] ;

[0153] ;

[0154] in, is the feature fusion operation, is the feature map after deep convolution, is the convolution kernel of the depth convolution layer, is the fused feature map, is the bias term;

[0155] S323, multi-scale feature extraction: After deep convolution, a multi-scale convolution operation is introduced to extract features at different scales through convolution kernels of different sizes (such as 3×3, 5×5, 7×7), so as to better adapt to the complex shapes and material properties of different workpieces and enhance the model's prediction ability for different indentation force paths, expressed as:

[0156] ;

[0157] ;

[0158] ;

[0159] ;

[0160] in, , , It is the feature map extracted under different convolution kernel sizes. , , are convolution kernels of different sizes, , , is the corresponding bias term, It is the feature map after multi-scale feature fusion;

[0161] S324, fully connected layer path prediction: Flatten the feature map after multi-scale convolution and input it into the fully connected layer to predict the pressure path. The fully connected layer is responsible for integrating all extracted features and outputting multiple pressure paths, which are expressed as:

[0162] ;

[0163] in, For the predicted indentation force paths, is the weight of the fully connected layer, is the flattened feature map, is the bias term, is the softmax activation function;

[0164] Through the above content, the complex relationship between the workpiece geometry, material properties and surface roughness can be accurately captured. The model generates multiple possible pressing force paths, which provides diversity and flexibility for subsequent evaluation and optimization, effectively improves the accuracy and adaptability of the pressing force path prediction, and better meets the needs of precise control of the pressing process.

[0165] Multipath evaluation and optimization in S33 include:

[0166] S331, path performance index calculation: Perform performance analysis on each indentation force path, calculate its indentation force distribution uniformity, path smoothness and stress concentration, expressed as:

[0167] ;

[0168] ;

[0169] ;

[0170] ;

[0171] in, For the The comprehensive performance index of each path, is the uniformity of the pressure distribution of the path, For path The pressure at each point, is the average pressure force along the path, is the smoothness of the path, For path The location of the point, For path The location of the point, is the stress concentration of the path, For path The contact area of ​​each point, , , is the weight coefficient, is the number of data points on the path;

[0172] S332, path cost function calculation: Comprehensively calculate the performance indicators of each path and calculate its cost function , expressed as:

[0173] ;

[0174] in, For the The cost function of the path, , , is the weight coefficient in the cost function;

[0175] S333, optimal path selection: According to the calculation result of the cost function, the path with the smallest cost function value is selected as the optimal pressing force path, and it is used as the final output to guide the pressing operation of the nut, which is expressed as:

[0176] ;

[0177] in, is the optimal pressing force path, is the index of the path;

[0178] Through the above content, the scientificity and accuracy of path selection are ensured, and key factors such as the uniformity of pressing force distribution, path smoothness and stress concentration can be effectively considered, thereby significantly improving the stability and accuracy of the pressing process. By comprehensively considering various performance indicators, it is ensured that the final selected path shows the best performance in actual operation, minimizing errors and risks in the processing process.

[0179] The servo push control in S4 includes:

[0180] S41, Pressing force feedback control: collect the data of the pressing force sensor in real time, compare it with the target pressing force, and calculate the pressing force error The output torque of the servo motor is adjusted by the proportional-integral-differential PID controller. To reduce the error, it can be expressed as:

[0181] ;

[0182] in, is the pressing force error, is the target pressure, is the actual measured pressing force, , , are the proportional, integral and differential coefficients of the PID controller respectively;

[0183] S42, displacement feedback control: real-time monitoring of displacement sensor data, comparison of actual displacement with target displacement, and calculation of displacement error , adjust the moving speed of the servo motor through the PID controller , ensuring that the nut moves precisely along the optimized press-in force path, expressed as:

[0184] ;

[0185] in, is the displacement error, is the target displacement, is the actual measured displacement, , , are the proportional, integral and differential coefficients of the displacement control PID controller respectively;

[0186] S43, closed-loop control integration: Integrate the pressing force and displacement feedback control into the closed-loop control mechanism. During the entire pressing process, the servo motor dynamically adjusts the torque and position of the pressing head according to the real-time feedback of the pressing force and displacement data, so that the nut is pressed into the predetermined depth along the optimized pressing force path;

[0187] Through the above content, the torque and movement speed of the servo motor can be accurately adjusted to ensure that the nut is accurately pressed in along the optimized path. This not only ensures the stability of the pressing force and the accuracy of the displacement, but also can dynamically respond to any changes or errors in the operation, thereby significantly improving the reliability, consistency and final assembly quality of the nut pressing process.

[0188] The indentation depth detection and correction in S5 includes:

[0189] S51, depth detection: the actual pressing depth of the nut is measured by a displacement sensor and press it and the target into depth Compare and calculate the depth error , expressed as:

[0190] ;

[0191] S52, verticality and position detection: detect the verticality of the nut and location and compared with the target value to calculate the verticality error and position error , expressed as:

[0192] ;

[0193] ;

[0194] in, is the target verticality, is the target location;

[0195] S53, fine-tuning correction: According to the detected depth error, verticality error and position error, the control parameters of the servo motor are adjusted through the fine-tuning algorithm and , to correct the position and angle of the nut, expressed as:

[0196] ;

[0197] ;

[0198] in, is the adjusted servo motor torque, is the adjusted servo motor speed, , , is the corresponding correction factor;

[0199] Through the above content, the inclination and displacement deviation of the nut during the press-in process can be accurately detected, and any deviation can be identified and corrected in time to ensure that the nut finally reaches the ideal press-in position and posture, thereby improving the accuracy and quality of assembly and reducing potential assembly failures and the need for subsequent adjustments.

[0200] Learning feedback in S6 includes:

[0201] S61, data collection and recording: real-time acquisition of the verticality of the nut after fine-tuning correction and location , and the initial pressing force path and operating parameters are recorded in the database together to form a historical data set of the pressing operation. The operating parameters include pressing force, pressing speed, target depth, displacement control parameters, servo motor torque and servo motor acceleration;

[0202] S62, Path and Operation Optimization: Initial Press Force Path Optimization Using Regression Analysis and operating parameters Optimize, expressed as:

[0203] ;

[0204] ;

[0205] in, is the optimized pressing force path, is the feedback path adjustment amount, For the optimized operating parameters, is the parameter adjustment amount of feedback, is the learning rate, which determines the impact of feedback data on path and parameter optimization;

[0206] S63, parameter and path suggestion: obtained through and As the recommended parameters and paths for future press-in operations, it is used to guide subsequent operations to continuously improve the accuracy and efficiency of the press-in process;

[0207] Through the above content, it is possible to continuously improve the pressing accuracy and path control, effectively reduce errors and deviations, ensure the stability and consistency of the pressing process, and thus significantly improve production efficiency and product quality.

[0208] like Figure 2 As shown, a nut servo pressing depth control system is used to implement the above-mentioned nut servo pressing depth control method, including the following modules:

[0209] Workpiece detection module: detects the surface shape, material properties and initial positioning of the workpiece through multiple sensors, and obtains the characteristic data of the workpiece, including geometric shape, hardness distribution and surface roughness information;

[0210] Parameter setting module: according to the characteristic data of the workpiece, the initial parameters of the servo motor are set, including the pressing force, pressing speed and target depth;

[0211] Press-in force path optimization module: Based on the set initial parameters and combined with the characteristic data of the workpiece, the optimal press-in force path is generated through the press-in force path prediction model;

[0212] Servo press-in control module: adopts a closed-loop control mechanism, drives the press-in force head through a servo motor, and performs the nut press-in operation according to the optimized press-in force path under the dual feedback of the press-in force and displacement;

[0213] Pressing depth detection and correction module: After the nut is pressed in, the displacement sensor detects the final pressing depth and checks the verticality and position of the nut. If a deviation from the target value is detected, fine-tuning and correction are automatically performed;

[0214] Learning feedback module: collects the verticality and position of the nut after fine-tuning and correction in real time, continuously optimizes the pressing force path and pressing operation, and provides parameter and path suggestions for future pressing operations.

[0215] The present invention covers any substitution, modification, equivalent method and scheme made on the essence and scope of the present invention. In order to make the public have a thorough understanding of the present invention, specific details are described in detail in the following preferred embodiments of the present invention, but those skilled in the art can fully understand the present invention without the description of these details. In addition, in order to avoid unnecessary confusion about the essence of the present invention, well-known methods, processes, procedures, components and circuits are not described in detail.

[0216] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A nut servo pressing depth control method, characterized in that: The following steps are involved: S1, workpiece detection: multiple sensors are used to detect the surface shape, material properties and initial positioning of the workpiece, and obtain the characteristic data of the workpiece, including geometric shape, hardness distribution and surface roughness information; S2, parameter setting: according to the results of workpiece detection, set the initial parameters of the servo motor, including pressing force, pressing speed and target depth; S3, press force path optimization: Based on the set initial parameters and combined with the characteristic data of the workpiece, the optimal press force path is generated through the press force path prediction model, including: S31, data preprocessing: preprocessing the characteristic data of the workpiece and the initial parameters of the servo motor, including data cleaning, standardization and feature extraction; S32, multi-path prediction: inputting the pre-processed characteristic data and initial parameters into the pressing force path prediction model to predict and generate multiple pressing force paths; S33, multi-path evaluation and optimization: compare the performance of multiple generated pressure paths and determine the optimal pressure path; S4, servo press-in control: adopts a closed-loop control mechanism, drives the press-in force head through a servo motor, and performs the nut press-in operation according to the optimized press-in force path under the dual feedback of the press-in force and displacement; S5, press-in depth detection and correction: After the nut is pressed in, the final press-in depth is detected by the displacement sensor, and the verticality and position of the nut are checked. If a deviation from the target value is detected, fine-tuning correction is automatically performed; S6, learning feedback: collect the verticality and position of the nut after fine-tuning and correction in real time, continuously optimize the pressing force path and pressing operation, and provide parameter and path suggestions for future pressing operations.

2. A nut servo pressing depth control method according to claim 1, characterized in that: The workpiece detection in S1 includes: S11, sensor selection and configuration: multiple sensors are configured to detect the surface shape, material properties and initial positioning of the workpiece. The sensors include laser displacement sensors, ultrasonic sensors and optical profilers; S12, geometric shape detection: use a laser displacement sensor to scan the workpiece surface, obtain the workpiece's geometric shape data, and measure the workpiece's surface profile and size through the displacement change of the reflected light beam; S13, hardness distribution evaluation: The internal structure and hardness distribution of the workpiece material are measured by an ultrasonic sensor, and the density change of the internal material of the workpiece is evaluated by transmitting and receiving sound wave signals to determine the hardness distribution; S14, surface roughness measurement: Use an optical profiler to scan the workpiece surface and measure the surface roughness. The optical profiler generates roughness data of the workpiece surface by analyzing the intensity and distribution of the reflected light on the surface.

3. A nut servo pressing depth control method according to claim 2, characterized in that: The parameter settings in S2 include: S21, calculation of initial pressing force and pressing speed: Calculate the initial pressing force and pressing speed of the current workpiece based on the workpiece characteristic data, expressed as: ； ； in, is the initial pressing force, is the workpiece surface area, is the average hardness of the workpiece, is the workpiece surface roughness coefficient, is the initial pressing speed, is the material density, is the geometric shape factor, and is the empirical coefficient; S22, target depth setting and servo motor adjustment: combined with the target pressing depth of the workpiece , by adjusting the control parameters of the servo motor, set the final initial pressing force , Pressing speed and target depth , the control parameter adjustment of the servo motor is expressed as: ； in, is the adjustment parameter of the servo motor, is the response factor of the servo motor.

4. A nut servo pressing depth control method according to claim 1, characterized in that: The data preprocessing in S31 includes: S311, data cleaning: using mean filtering method to check and remove outliers and noise data in workpiece characteristic data and initial parameters; S312, standardization processing: standardizing the characteristic data and initial parameters of the cleaned workpiece; S313, feature extraction: extract key features from the standardized workpiece characteristic data and initial parameters through the principal component analysis algorithm, including the shape characteristics of the workpiece, the material hardness change rate, and the surface roughness gradient.

5. A nut servo pressing depth control method according to claim 4, characterized in that: The pressing force path prediction model in S32 adopts a convolutional neural network model, and the convolutional neural network model includes: S321, data input and multi-channel feature processing: multi-channel processing is performed on the input data, and each channel corresponds to a characteristic data, which is expressed as: ； ； ； in, , , They are the characteristic diagrams of geometry, hardness distribution and surface roughness, , , are the convolution kernels of different channels respectively. , , is the input data of different channels, , , is the bias term, is the activation function; S322, feature fusion and deep convolution: The feature maps of multiple channels are fused, and multi-level complex features are extracted through deep convolution layers to capture the relationship between the multi-dimensional characteristics of the workpiece and the indentation force path, which is expressed as: ； ； in, is the feature fusion operation, is the feature map after deep convolution, is the convolution kernel of the depth convolution layer, is the fused feature map, is the bias term; S323, multi-scale feature extraction: After deep convolution, a multi-scale convolution operation is introduced to extract features at different scales through convolution kernels of different sizes, which can be expressed as: ； ； ； ； in, , , It is the feature map extracted under different convolution kernel sizes. , , are convolution kernels of different sizes, , , is the corresponding bias term, It is the feature map after multi-scale feature fusion; S324, fully connected layer path prediction: Flatten the feature map after multi-scale convolution and input it into the fully connected layer to predict the pressure path. The fully connected layer is responsible for integrating all extracted features and outputting multiple pressure paths, which are expressed as: ； in, For the predicted indentation force paths, is the weight of the fully connected layer, is the flattened feature map, is the bias term, is the softmax activation function.

6. A nut servo pressing depth control method according to claim 5, characterized in that: The multipath evaluation and optimization in S33 includes: S331, path performance index calculation: Perform performance analysis on each indentation force path, calculate its indentation force distribution uniformity, path smoothness and stress concentration, expressed as: ； ； ； ； in, For the The comprehensive performance index of each path, is the uniformity of the pressure distribution of the path, For path The pressure at each point, is the average pressure force along the path, is the smoothness of the path, For path The location of the point, For path The location of the point, is the stress concentration of the path, For path The contact area of ​​each point, , , is the weight coefficient, is the number of data points on the path; S332, path cost function calculation: Comprehensively calculate the performance indicators of each path and calculate its cost function , expressed as: ； in, For the The cost function of the path, , , is the weight coefficient in the cost function; S333, optimal path selection: According to the calculation result of the cost function, the path with the smallest cost function value is selected as the optimal pressing force path, and it is used as the final output to guide the pressing operation of the nut, which is expressed as: ； in, is the optimal pressing force path, The index of the path.

7. A nut servo pressing depth control method according to claim 6, characterized in that: The servo push control in S4 includes: S41, Pressing force feedback control: collect the data of the pressing force sensor in real time, compare it with the target pressing force, and calculate the pressing force error The output torque of the servo motor is adjusted by the proportional-integral-differential PID controller. To reduce the error, it can be expressed as: ； in, is the pressing force error, is the target pressure, is the actual measured pressing force, , , are the proportional, integral and differential coefficients of the PID controller respectively; S42, displacement feedback control: real-time monitoring of displacement sensor data, comparison of actual displacement with target displacement, and calculation of displacement error , adjust the moving speed of the servo motor through the PID controller , expressed as: ； in, is the displacement error, is the target displacement, is the actual measured displacement, , , are the proportional, integral and differential coefficients of the displacement control PID controller respectively; S43, closed-loop control integration: The pressing force and displacement feedback control are integrated into the closed-loop control mechanism. During the entire pressing process, the servo motor dynamically adjusts the torque and position of the pressing head according to the real-time feedback of the pressing force and displacement data, so that the nut is pressed into the predetermined depth along the optimized pressing force path.

8. A nut servo pressing depth control method according to claim 7, characterized in that: The indentation depth detection and correction in S5 includes: S51, depth detection: the actual pressing depth of the nut is measured by a displacement sensor and press it and the target into depth Compare and calculate the depth error , expressed as: ； S52, verticality and position detection: detect the verticality of the nut and location and compared with the target value to calculate the verticality error and position error , expressed as: ； ； in, is the target verticality, is the target location; S53, fine-tuning correction: According to the detected depth error, verticality error and position error, the control parameters of the servo motor are adjusted through the fine-tuning algorithm and , to correct the position and angle of the nut, expressed as: ； ； in, is the adjusted servo motor torque, is the adjusted servo motor speed, , , is the corresponding correction factor.

9. A nut servo pressing depth control method according to claim 1, characterized in that: The learning feedback in S6 includes: S61, data collection and recording: real-time acquisition of the verticality of the nut after fine-tuning correction and location , and the initial pressing force path and operating parameters are recorded in the database together to form a historical data set of the pressing operation. The operating parameters include pressing force, pressing speed, target depth, displacement control parameters, servo motor torque and servo motor acceleration; S62, Path and Operation Optimization: Initial Press Force Path Optimization Using Regression Analysis and operating parameters Optimize, expressed as: ； ； in, is the optimized pressing force path, is the feedback path adjustment amount, For the optimized operating parameters, is the parameter adjustment amount of feedback, is the learning rate; S63, parameter and path suggestion: obtained through and Serves as a suggested parameter and path for future push operations.

10. A nut servo pressing depth control system, used to implement a nut servo pressing depth control method according to any one of claims 1 to 9, characterized in that: Includes the following modules: Workpiece detection module: detects the surface shape, material properties and initial positioning of the workpiece through multiple sensors, and obtains the characteristic data of the workpiece, including geometric shape, hardness distribution and surface roughness information; Parameter setting module: according to the characteristic data of the workpiece, the initial parameters of the servo motor are set, including the pressing force, pressing speed and target depth; Press-in force path optimization module: Based on the set initial parameters and combined with the characteristic data of the workpiece, the optimal press-in force path is generated through the press-in force path prediction model; Servo press-in control module: adopts a closed-loop control mechanism, drives the press-in force head through a servo motor, and performs the nut press-in operation according to the optimized press-in force path under the dual feedback of the press-in force and displacement; Pressing depth detection and correction module: After the nut is pressed in, the displacement sensor detects the final pressing depth and checks the verticality and position of the nut. If a deviation from the target value is detected, fine-tuning and correction are automatically performed; Learning feedback module: collects the verticality and position of the nut after fine-tuning and correction in real time, continuously optimizes the pressing force path and pressing operation, and provides parameter and path suggestions for future pressing operations.

Images