Wooden box steel belt transmission high-precision displacement correction method and device
Through high-precision grating scale and wavelet noise reduction technology, combined with adaptive deviation correction control and digital twin optimization, the displacement error problem of wooden box steel belt transmission system under high speed conditions is solved, achieving high-precision displacement correction and system stability improvement.
Patent Information
- Application Number
- CN202411122910.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-15
- Publication Date
- 2024-11-01
- Estimated Expiration
- 2044-08-15
AI Technical Summary
The traditional wooden box steel belt transmission system is prone to displacement offset, vibration and creep under high-speed operation and long-term working conditions, resulting in the accumulation of displacement errors. The existing displacement correction technology is difficult to meet the high-precision requirements, and the system model is uncertain. and parameter drift brings control challenges.
High-precision grating scales are used for installation and calibration, combined with wavelet noise reduction and multi-scale trend analysis, adaptive deviation correction control is carried out, and deviation correction actions are performed through linear motor drive and air float guide rails, and dynamic optimization of calibration curves is used for digital twin simulation and multi-objective optimization. and compensation model.
Improve the basic accuracy of displacement measurement, reduce noise and interference effects, realize rapid response and precise adjustment to complex dynamic environments, enhance the long-term stability and adaptability of the system, build a more comprehensive and accurate compensation model, and improve the The corrective effect is achieved.
Smart Images

Figure CN118877445A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of displacement correction, and in particular to a high-precision displacement correction method and device for a wooden box steel belt transmission. Background Art
[0002] The steel belt transmission system of wooden boxes is widely used in packaging, logistics and other industries, and its displacement accuracy directly affects product quality and production efficiency. However, due to mechanical errors, environmental factors and system nonlinearity, traditional displacement control methods are difficult to meet high-precision requirements. Especially under high-speed operation and long-term working conditions, the steel belt is prone to deviation, vibration and creep, resulting in the accumulation of displacement errors.
[0003] Existing displacement correction technologies mainly rely on simple PID control and fixed parameter compensation, which cannot effectively cope with system changes in complex dynamic environments. At the same time, the traditional displacement measurement method has limited accuracy and is difficult to provide reliable data support for high-precision correction. In addition, the uncertainty of the system model and parameter drift also bring challenges to precise control. Summary of the invention
[0004] The present invention provides a method and device for high-precision displacement correction of a wooden box steel belt transmission, which are used to achieve high-precision displacement correction of a wooden box steel belt transmission system and improve the overall performance and reliability of the system.
[0005] In a first aspect, the present invention provides a high-precision displacement correction method for a wooden box steel belt transmission, the high-precision displacement correction method for a wooden box steel belt transmission comprising: Install and calibrate the high-precision grating ruler, and create the first calibration curve and the first compensation model; According to the first calibration curve and the first compensation model, high dynamic range acquisition is performed on the displacement data of the steel strip to obtain original displacement data; Performing wavelet denoising and multi-scale trend analysis on the original displacement data to obtain denoised displacement data and displacement trend information; Performing adaptive deviation correction control based on the noise reduction displacement data and the displacement trend information, and generating a deviation correction control instruction; The correction control instruction is accurately executed by a linear motor drive and an air-floating guide rail to obtain an actual correction action and a corresponding execution result; According to the actual deviation-correcting action and the execution result, digital twin simulation and multi-objective optimization are performed on the first calibration curve and the first compensation model to obtain a second calibration curve and a second compensation model.
[0006] In a second aspect, the present invention provides a high-precision displacement correction device for a wooden box steel belt transmission, the high-precision displacement correction device for a wooden box steel belt transmission comprising: An installation module, used for installing and calibrating the high-precision grating ruler, and creating a first calibration curve and a first compensation model; An acquisition module, used for acquiring the steel strip displacement data in a high dynamic range according to the first calibration curve and the first compensation model to obtain original displacement data; An analysis module, used for performing wavelet denoising and multi-scale trend analysis on the original displacement data to obtain denoised displacement data and displacement trend information; A control module, configured to perform adaptive deviation correction control based on the noise reduction displacement data and the displacement trend information, and generate a deviation correction control instruction; An execution module, used to drive the linear motor and the air-floating guide rail to accurately execute the deviation correction control instruction, and obtain the actual deviation correction action and the corresponding execution result; An optimization module is used to perform digital twin simulation and multi-objective optimization on the first calibration curve and the first compensation model according to the actual correction action and the execution result to obtain a second calibration curve and a second compensation model.
[0007] The third aspect of the present invention provides a high-precision displacement correction device for a wooden box steel belt transmission, comprising: a memory and at least one processor, wherein the memory stores instructions; the at least one processor calls the instructions in the memory to enable the high-precision displacement correction device for a wooden box steel belt transmission to execute the above-mentioned high-precision displacement correction method for a wooden box steel belt transmission.
[0008] A fourth aspect of the present invention provides a computer-readable storage medium, in which instructions are stored, and when the computer-readable storage medium is run on a computer, the computer executes the above-mentioned high-precision displacement correction method for wooden box steel belt transmission.
[0009] In the technical solution provided by the present invention, the application and precise calibration of high-precision grating rulers improve the basic accuracy of displacement measurement, and high dynamic range data acquisition and multi-level signal processing technology effectively improve the quality of original displacement data and reduce the influence of measurement noise and interference. Wavelet denoising and multi-scale trend analysis methods can effectively separate noise and effective information in displacement data, thereby improving the accuracy of subsequent control. The adaptive correction control strategy combines fuzzy control, PID control and reinforcement learning to achieve rapid response and precise adjustment to complex dynamic environments. The precise actuators of linear motor drive and air-floating guide rails improve the execution accuracy of the correction action and reduce the influence of mechanical errors. Digital twin simulation and multi-objective optimization methods realize dynamic optimization of calibration curves and compensation models, thereby improving the long-term stability and adaptability of the system. Taking into account a variety of influencing factors such as temperature and speed, a more comprehensive and accurate compensation model is constructed, which effectively improves the overall effect of correction. BRIEF DESCRIPTION OF THE DRAWINGS
[0010] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.
[0011] Figure 1 A schematic diagram of the process flow of a high-precision displacement correction method for a wooden box steel belt drive provided in an embodiment of the present application; Figure 2 A schematic block diagram of the structure of a high-precision displacement correction device for a wooden box steel belt drive provided in an embodiment of the present application. DETAILED DESCRIPTION
[0012] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0013] The flowcharts shown in the accompanying drawings are only examples and do not necessarily include all the contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may also be decomposed, combined or partially merged, so the actual execution order may change based on actual conditions.
[0014] It should also be understood that the terms used in this application specification are only for the purpose of describing specific embodiments and are not intended to limit the application. As used in this application specification and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.
[0015] It should be further understood that the term “and / or” used in the specification and appended claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.
[0016] In conjunction with the accompanying drawings, some embodiments of the present application are described in detail below. In the absence of conflict, the following embodiments and features in the embodiments can be combined with each other.
[0017] See also Figure 1 , Figure 1 A schematic diagram of the process flow of the high-precision displacement correction method for the wooden box steel belt transmission provided in the embodiment of the present application is shown in FIG. Figure 1As shown, the high-precision displacement correction method for wooden box steel belt transmission provided in the embodiment of the present application includes steps S100 to S600.
[0018] Step S100, installing and calibrating the high-precision grating ruler, and creating a first calibration curve and a first compensation model; It is understandable that the execution subject of the present invention can be a wooden box steel belt transmission high-precision displacement correction device, or a terminal or a server, which is not limited here. The embodiment of the present invention is described by taking the server as the execution subject as an example.
[0019] Specifically, the grating ruler is calibrated using a precision level and an angle meter to obtain initial installation error data. The horizontal or angular deviation of the grating ruler during the installation process is identified through the initial installation error data. According to the initial installation error data, the high-precision grating ruler is adjusted to ensure that the grating ruler is in the calibrated installation position. After the installation and adjustment of the grating ruler is completed, a laser interferometer is used to perform multi-point measurement on it to obtain the corresponding data of the grating ruler reading and the interferometer reading, reflecting the relationship between the reading of the grating ruler at different positions and the precise measurement results of the laser interferometer. The corresponding data of the grating ruler reading and the interferometer reading are fitted by the least squares method to obtain the initial calibration curve equation. This equation is a mathematical model between the grating ruler reading and the actual position, which preliminarily reflects the systematic errors that may exist in the grating ruler during the measurement process. Based on the initial calibration curve equation, the grating ruler is subjected to nonlinear error analysis to obtain the distribution function of the nonlinear error, which describes the distribution of the nonlinear error within the measurement range of the grating ruler, and the initial calibration curve equation is corrected accordingly to generate a more accurate first calibration curve. The first calibration curve is expanded by Fourier series to extract the periodic error compensation term in the grating ruler measurement process. The periodic error is usually related to the mechanical characteristics of the grating ruler, and the Fourier series expansion can decompose these errors into several harmonic components, each of which corresponds to a specific physical phenomenon. The extracted harmonic components are used to construct a temperature compensation model. In actual operation, temperature has a significant impact on the measurement accuracy of the grating ruler. Based on the periodic error compensation term and the temperature data obtained from the temperature sensor, a temperature compensation model is constructed to obtain a temperature compensation function. Combining the temperature compensation function and the periodic error compensation term, a first compensation model is constructed. This compensation model is used to correct the reading deviation of the grating ruler at different temperatures and different measurement positions, thereby improving the measurement accuracy. The compensation model can be expressed as C(x, T)=x+Σ[Ai*sin(2πix / P+φi)]+k(T-T0), where C(x, T) is the actual position value after compensation, x is the original reading of the grating ruler, Ai is the amplitude of the i-th harmonic in the Fourier series, P is the basic period of the grating ruler, φi is the phase of the i-th harmonic, k is the temperature compensation coefficient, T is the current ambient temperature, and T0 is the reference temperature. Through this model, the original reading of the grating ruler can be effectively corrected, the influence of periodic errors and temperature changes can be eliminated, and high-precision displacement measurement can be achieved.
[0020] Step S200, performing high dynamic range acquisition on the steel strip displacement data according to the first calibration curve and the first compensation model to obtain original displacement data; Specifically, the grating ruler signal is pre-amplified to obtain an amplified analog signal. The amplified signal has a higher signal-to-noise ratio and can capture displacement changes more effectively. The amplified analog signal is band-pass filtered to eliminate high-frequency and low-frequency noise in the signal and retain only the effective frequency band related to the displacement. Based on the programmable gain amplifier, the filtered signal is dynamically gain adjusted to ensure that the dynamic range of the signal is adapted to the input range of the ADC (analog-to-digital converter) to avoid ADC distortion or inaccurate sampling due to excessively high or low signal amplitude. After the signal is dynamically gain adjusted, the signal adapted to the ADC input range is analog-to-digital converted to convert the analog signal into digital displacement data. According to the first calibration curve, the digital displacement data is nonlinearly corrected to correct the measurement error caused by the grating ruler at different positions or under different conditions to ensure that the data is closer to the actual displacement. Based on the periodic error compensation term in the first compensation model, the corrected displacement data is periodically compensated. Periodic errors are usually related to mechanical structures or periodic vibrations of the system. By compensating these errors, the accuracy and reliability of the measurement are improved. Combined with the data of the temperature sensor and the temperature compensation function in the first compensation model, the compensated displacement data is temperature compensated to reduce or eliminate the influence of temperature on the displacement measurement, so as to obtain more accurate displacement data after temperature compensation. The displacement data after temperature compensation is subjected to moving average filtering to smooth the data and remove accidental high-frequency noise, so as to obtain the displacement data after preliminary filtering. Based on the Kalman filter algorithm, the displacement data after preliminary filtering is subjected to noise suppression. Kalman filtering is an adaptive filtering algorithm applied to dynamic systems, which can provide the best estimate under various noise conditions. After Kalman filtering, the random noise in the displacement data is effectively suppressed, and smoother and more reliable noise-suppressed displacement data is obtained. Data integrity check and outlier detection are performed on the noise-suppressed displacement data to identify and eliminate potential erroneous data, so as to obtain high-quality original displacement data.
[0021] Step S300, performing wavelet denoising and multi-scale trend analysis on the original displacement data to obtain denoised displacement data and displacement trend information; Specifically, the original displacement data is subjected to wavelet decomposition, and the displacement signal is decomposed into multiple wavelet coefficients of different scales, which reflect the characteristic information of the signal at different frequencies. The multi-layer wavelet coefficients are subjected to soft threshold processing to remove the mixed noise components in the signal. By setting a suitable threshold and performing soft threshold processing on the wavelet coefficients, the influence of noise is effectively reduced, and the denoised wavelet coefficients are obtained. The denoised wavelet coefficients are subjected to wavelet reconstruction, and the denoised wavelet coefficients are recombined to restore the displacement data close to the original signal, and the initial denoised displacement data is obtained. The initial denoised displacement data is decomposed to obtain multiple singular value components. Singular value components are different components in the signal, through which the inherent structure and trend of the data are revealed. Multiple singular value components are selected according to the cumulative contribution rate criterion. The cumulative contribution rate criterion helps determine which components contribute the most to the overall trend of the signal. After selecting these target trend components, they are reconstructed to obtain long-term trend data, which reflects the main change direction and long-term trend in the signal. The long-term trend data is the low-frequency component of the signal. At the same time, the initial denoised displacement data is modally decomposed to obtain multiple intrinsic mode functions (IMFs), which reflect the oscillation components of the signal on different time scales. By analyzing their energy ratio and frequency characteristics, the IMF combinations representing short-term fluctuations are screened out. Short-term fluctuations are high-frequency components in the signal, reflecting the rapid changes and local vibrations of the data. The IMF combinations representing short-term fluctuations are reconstructed to generate short-term fluctuation data. By weighted combination of long-term trend data and short-term fluctuation data, denoised displacement data and complete displacement trend information are obtained. De-noised displacement data can provide more accurate displacement measurement results, while displacement trend information reveals the long-term trend and short-term fluctuation characteristics in the signal.
[0022] Step S400, performing adaptive deviation correction control based on the noise reduction displacement data and the displacement trend information, and generating a deviation correction control instruction; Specifically, the noise reduction displacement data and displacement trend information are fuzzified, and the continuous displacement data and trend information are converted into fuzzy input variables. Fuzzy variables can effectively handle the uncertainty and nonlinear characteristics in the system. The fuzzy input variables are inferred using the preset fuzzy rule base, and the fuzzy output variables are generated according to the logical relationship and empirical rules in the rule base, which preliminarily reflect the control adjustment direction and amplitude required by the system. The fuzzy output variables are defuzzified using the center of gravity method, and the fuzzy output is converted into a specific numerical value to obtain the initial PID parameters. The initial PID parameters include proportional, integral and differential coefficients, which are used as preliminary control parameters in the PID controller to perform preliminary PID control calculations on the noise reduction displacement data to obtain the initial control quantity. The initial control quantity is feedforward compensated in combination with the displacement trend information. Feedforward compensation uses the displacement trend information to predict the future state of the system, thereby correcting possible errors in advance and obtaining the compensated control quantity. Reinforcement learning is performed on the compensated control quantity and the current system state. Reinforcement learning continuously adjusts and improves the control action according to the current system state, and gradually optimizes the control strategy through interactive learning with the environment to obtain the optimized control action. According to the optimized control action, the initial PID parameters are adjusted online to obtain adaptive PID parameters. The adaptive PID parameters can adaptively adjust the response characteristics of the controller according to the changes in the real-time working conditions, ensuring that the system can maintain good control effects under different working conditions. Based on the adaptive PID parameters, the noise reduction displacement data is subjected to secondary PID control calculation to obtain the optimized control quantity. The optimized control quantity is saturated and limited to ensure that the control quantity does not exceed the physical limit or safety range of the system in actual execution, and the limited control quantity is obtained. The limited control quantity is finally converted into a correction control instruction, which specifically includes the motor drive current and direction information. The motor drive current determines the strength of the correction action, while the direction information indicates the direction of the correction.
[0023] Step S500, the linear motor drives and the air-floating guide rails accurately execute the deviation correction control instruction to obtain the actual deviation correction action and the corresponding execution result; Specifically, the correction control instruction is converted into a current signal. A driving current signal of the linear motor is generated through a special current signal conversion device. These driving current signals are the core parameters for controlling the linear motor. Based on the driving current signal, the linear motor is controlled by current closed loop to accurately control the actual output force of the linear motor. The current closed loop control ensures that the linear motor can output stable and accurate force according to the expected instruction. In the process of controlling the driving current, the position of the linear motor is detected in real time based on the Hall sensor to obtain the motor position feedback signal. By analyzing the feedback signal, the position change of the motor can be monitored in real time, and the position closed loop control can be performed based on this. The position closed loop control process ensures that the linear motor can perform the displacement correction operation according to the preset trajectory by continuously correcting the position error of the motor, and obtains an accurate position control result. At the same time, the air pressure of the air floating guide rail is adaptively adjusted to ensure that the air floating guide rail is in the optimal air floating state. The air pressure of the air floating guide rail directly affects the floating performance of the guide rail. Through adaptive adjustment, the smooth operation of the guide rail can be maintained under different load conditions. The floating gap of the air-floating guide rail is sampled at high frequency using a capacitive displacement sensor to obtain floating gap data, which reflects the tiny floating of the guide rail during operation. Based on these data, the guide rail is dynamically balanced. Through dynamic balance control, the vibration and drift in the movement of the guide rail can be effectively reduced, and a stable guide rail movement state can be obtained. In order to achieve the best correction effect, the movement of the linear motor and the air-floating guide rail is coordinated and optimized to form a comprehensive execution strategy. The strategy comprehensively considers the force output, position control and floating characteristics of the linear motor and the air-floating guide rail to ensure that the two work together during operation to achieve the best correction effect. Based on the comprehensive execution strategy, the correction mechanism is precisely driven, the correction operation is performed, and the actual correction action is obtained. After the actual correction action is completed, the correction result is measured by a high-precision grating ruler to obtain an accurate execution result. Through this process, the accuracy and effect of the correction operation can be verified, and the subsequent correction operation can be adjusted and optimized according to the execution result, ensuring that the entire system can maintain high-precision correction control under various working conditions.
[0024] Step S600: According to the actual correction action and the execution result, digital twin simulation and multi-objective optimization are performed on the first calibration curve and the first compensation model to obtain the second calibration curve and the second compensation model.
[0025] Specifically, data preprocessing is performed on the actual correction action and execution results to eliminate noise, normalize data of different scales, and obtain a standardized input data set. Based on the standardized data set, a system virtual simulation environment is constructed, which truly simulates the behavior of the physical system through digital twin technology. In the system virtual simulation environment, parameter sensitivity analysis is performed on the first calibration curve, and the key factors affecting the system performance are identified through analysis. Based on the key influencing factors, the first calibration curve is optimized by polynomial fitting to generate a more accurate second calibration curve. At the same time, the first compensation model is structurally decomposed into multiple sub-models, including a periodic error compensation sub-model and a temperature compensation sub-model. Each sub-model handles different error factors involved in the compensation process. Based on the genetic algorithm, the parameters of multiple sub-models are optimized. The genetic algorithm simulates the process of natural selection, and gradually optimizes the parameters of the sub-model through iterative selection and crossover mutation to obtain the optimized sub-model parameter set. Based on the optimized parameter set, the model is globally optimized to find a global optimal solution to ensure that the compensation model can achieve optimal performance under various working conditions. Based on the global optimal solution, a multi-objective optimization problem is constructed, in which the objective function not only considers the displacement accuracy, but also incorporates the complexity of the model into the optimization objective, thereby obtaining the Pareto optimal solution set. Each solution in the Pareto optimal solution set is a trade-off between the two objectives, reflecting the highest accuracy that can be achieved under a certain complexity. In order to select the optimal solution from the Pareto optimal solution set, the TOPSIS decision method is used to evaluate and select it. The TOPSIS method selects the solution closest to the ideal solution by weighted calculation of multiple evaluation indicators, and obtains the second compensation model. The second compensation model is expressed as: C(x, T, v) = x + Σ[Aisin(2πix / P + φi)] + k (T-T0) + f (v), where C (x, T, v) represents the actual position value after compensation, x is the original reading of the grating ruler, Ai is the amplitude of the i-th harmonic, P is the basic period of the grating ruler, φi is the phase of the i-th harmonic, k is the temperature compensation coefficient, T is the current temperature, T0 is the reference temperature, and f (v) is the speed-related compensation function in the form of f (v) = αv + βv^2 + γsign (v), where α, β, γ are speed compensation coefficients, and sign (v) is the speed direction function. Through this comprehensive model, not only temperature and periodic errors are considered, but also the influence of speed is introduced, so that the compensation model can more comprehensively cope with various complex working conditions and improve the overall accuracy and response speed of the correction system.
[0026] In the embodiment of the present invention, the application and precise calibration of high-precision grating ruler improve the basic accuracy of displacement measurement, and the high dynamic range data acquisition and multi-level signal processing technology effectively improve the quality of the original displacement data and reduce the influence of measurement noise and interference. Wavelet noise reduction and multi-scale trend analysis methods can effectively separate the noise and effective information in the displacement data, and improve the accuracy of subsequent control. The adaptive correction control strategy combines fuzzy control, PID control and reinforcement learning to achieve rapid response and precise adjustment to complex dynamic environments. The precise actuators of linear motor drive and air-floating guide rails improve the execution accuracy of the correction action and reduce the influence of mechanical errors. Digital twin simulation and multi-objective optimization methods realize the dynamic optimization of calibration curves and compensation models, and improve the long-term stability and adaptability of the system. Taking into account a variety of influencing factors such as temperature and speed, a more comprehensive and accurate compensation model is constructed, which effectively improves the overall effect of correction.
[0027] In a specific embodiment, the process of executing step S100 may specifically include the following steps: Calibrate the high-precision grating ruler with a precision level and an inclinometer to obtain initial installation error data, and adjust the high-precision grating ruler according to the initial installation error data to obtain the installation position of the calibrated grating ruler; Based on the laser interferometer, the calibrated grating ruler is measured at multiple points to obtain the corresponding data of the grating ruler reading and the interferometer reading, and the corresponding data of the grating ruler reading and the interferometer reading are fitted by the least square method to obtain the initial calibration curve equation; According to the initial calibration curve equation, a nonlinear error analysis is performed on the grating ruler to obtain a nonlinear error distribution function, and based on the nonlinear error distribution function, the initial calibration curve equation is corrected to obtain a first calibration curve; Performing Fourier series expansion on the first calibration curve to obtain a periodic error compensation term, and constructing a temperature compensation model based on the periodic error compensation term and temperature sensor data to obtain a temperature compensation function; Based on the temperature compensation function and the periodic error compensation term, the first compensation model is constructed, where the first compensation model is: C(x, T)=x+Σ[Ai*sin(2πix / P+φi)]+k(T-T0), where C(x, T) is the compensated position value, x is the original reading of the grating ruler, Ai is the i-th harmonic amplitude, P is the basic period of the grating ruler, φi is the i-th harmonic phase, k is the temperature compensation coefficient, T is the current temperature, and T0 is the reference temperature.
[0028] Specifically, the high-precision grating ruler is calibrated with a precision level and an inclinometer. By using a precision level and an inclinometer, the grating ruler is initially calibrated to detect the angular error and horizontal error of the grating ruler during installation. These errors may be caused by a slight tilt or distortion when the grating ruler is installed, affecting the measurement accuracy. The level is used to detect the inclination of the grating ruler in the horizontal direction, while the inclinometer is used to measure the angular deviation between the grating ruler and the ideal installation position. Through measurement, the initial installation error data is obtained. The grating ruler is adjusted according to the initial installation error data to achieve the ideal calibration position. The adjustment process may involve fine adjustment of the grating ruler support structure to ensure that the errors of the grating ruler in the horizontal and angular directions are minimized to achieve the accuracy standard that meets the measurement requirements. The calibrated grating ruler is measured at multiple points by a laser interferometer to ensure that it has consistent accuracy throughout the entire measurement range. The laser interferometer is a high-precision measurement tool that provides higher measurement accuracy than the grating ruler. By selecting multiple measuring points within the measuring range of the grating ruler, measuring the grating ruler reading and the interferometer reading, obtaining the corresponding data between the grating ruler reading and the laser interferometer reading, reflecting the actual measurement error of the grating ruler at different positions. The least squares method is used to fit the grating ruler reading and the interferometer reading. The least squares method is a mathematical method that minimizes the square error of the measured data to obtain a best fitting curve, which is the initial calibration curve equation. The equation represents the relationship between the grating ruler reading and the actual position, and describes the error characteristics of the grating ruler in an uncompensated state in mathematical form. The initial calibration curve equation is subjected to nonlinear error analysis to obtain a nonlinear error distribution function. The nonlinear error distribution function describes the nonlinear deviation of the grating ruler at different measuring points. By analyzing these deviations, the initial calibration curve equation is corrected to obtain a more accurate first calibration curve. For example, assuming that the initial calibration curve equation of the grating ruler is y=ax+b, where y is the grating ruler reading, x is the actual position, and a and b are fitting parameters. Through nonlinear error analysis, it is found that the equation has systematic deviations in certain sections, resulting in inaccurate measurement results. According to the nonlinear error distribution function, the equation is corrected, for example, a nonlinear term y=ax+b+c*sin(dx) is added, where c and d are correction parameters obtained through nonlinear error analysis. The corrected equation is the first calibration curve, which is closer to the actual measurement of the grating ruler. In order to compensate for the error caused by the change of ambient temperature during the use of the grating ruler, the first calibration curve is expanded by Fourier series to obtain the periodic error compensation term. The Fourier series expansion can decompose the periodic error in the calibration curve into multiple harmonic components, expressed as Σ[Ai*sin(2πix / P+φi)], where Ai is the amplitude of the i-th harmonic, P is the fundamental period of the grating ruler, and φi is the phase of the i-th harmonic. These harmonic components reflect the errors in the grating ruler due to the mechanical periodic characteristics.Consider the influence of temperature on the measurement accuracy of the grating ruler. Real-time temperature data is obtained through the temperature sensor, and a temperature compensation model is constructed based on these data to obtain a temperature compensation function. The temperature compensation function can be formalized as k(T-T0), where k is the temperature compensation coefficient, T is the current temperature, and T0 is the reference temperature. This function is used to correct the measurement error of the grating ruler at different temperatures. Combining the temperature compensation function and the periodic error compensation term, the first compensation model is constructed. The mathematical expression of the model is C(x, T)=x+Σ[Ai*sin(2πix / P+φi)]+k(T-T0), where C(x, T) represents the actual position value after compensation, and x is the original reading of the grating ruler. The model takes into account the linear and nonlinear errors of the grating ruler, and introduces compensation for periodic errors and temperature errors to improve the measurement accuracy.
[0029] In a specific embodiment, the process of executing step S200 may specifically include the following steps: Pre-amplifying the grating ruler signal to obtain an amplified analog signal, and performing band-pass filtering on the amplified analog signal to obtain a filtered signal; Based on a programmable gain amplifier, the filtered signal is dynamically adjusted for gain to obtain a signal that matches the ADC input range, and the signal that matches the ADC input range is converted into digital form to obtain digital displacement data; According to the first calibration curve, nonlinear error correction is performed on the digitized displacement data to obtain corrected displacement data, and based on the periodic error compensation term in the first compensation model, periodic error compensation is performed on the corrected displacement data to obtain compensated displacement data; According to the temperature sensor data and the temperature compensation function in the first compensation model, temperature compensation is performed on the compensated displacement data to obtain temperature compensated displacement data, and moving average filtering is performed on the temperature compensated displacement data to obtain preliminary filtered displacement data; Based on the Kalman filter algorithm, noise suppression is performed on the displacement data after preliminary filtering to obtain noise-suppressed displacement data, and data integrity checks and outlier checks are performed on the noise-suppressed displacement data to obtain original displacement data.
[0030] Specifically, the grating ruler signal is preamplified, and the weak signal output by the grating ruler is amplified to obtain an analog signal with a higher amplitude, thereby effectively improving the signal-to-noise ratio. The amplified analog signal is subjected to bandpass filtering. The bandpass filter allows signals within a specific frequency range to pass through and filters out frequency components beyond the range. By selecting appropriate bandpass filter parameters, the effective components in the grating ruler signal are retained, while useless noise is removed to obtain a filtered signal. The signal is dynamically gain adjusted based on a programmable gain amplifier (PGA). PGA is an amplifier that can dynamically adjust the gain according to the characteristics of the input signal to ensure that the amplitude of the signal is within the input range of the analog-to-digital converter (ADC) before entering the ADC. If the signal amplitude is too low, signal details may be lost; if the amplitude is too high, signal saturation may occur. Through the dynamic gain adjustment of the PGA, the signal amplitude is optimized to the optimal input range of the ADC to ensure high-precision conversion of the signal. After the dynamic gain adjustment is completed, the signal is sent to the ADC for analog-to-digital conversion. The analog-to-digital conversion process converts the analog signal into a digital signal to form digital displacement data. The digital displacement data is corrected for nonlinear errors according to the first calibration curve. The first calibration curve is usually obtained through error analysis and fitting of the grating ruler, which describes the systematic error of the grating ruler during the measurement process. By applying this calibration curve, the nonlinear error in the digitized displacement data is corrected to obtain corrected data that is closer to the actual displacement. For example, if the calibration curve is in the form of: ; in, is the actual location, is the original reading of the grating ruler, , , is the parameter of the calibration curve. According to this relationship, the digitized data is corrected to eliminate the influence of nonlinear errors. Based on the periodic error compensation term in the first compensation model, the corrected displacement data is corrected. The periodic error compensation term is usually in the form of Fourier series. Through this compensation term, the error caused by periodic factors in the measurement process of the grating ruler is effectively eliminated to obtain the compensated displacement data. According to the data of the temperature sensor and the temperature compensation function in the first compensation model, the compensated displacement data is temperature compensated. The temperature compensation function is usually expressed as a temperature compensation coefficient With temperature difference The product of . Through this function, the displacement data is adjusted to eliminate the deviation caused by temperature changes, and the displacement data after temperature compensation is obtained. In order to smooth the data and reduce the influence of random noise, the displacement data after temperature compensation is subjected to moving average filtering. The moving average filter smoothes short-term fluctuations and emphasizes long-term trends by calculating the average value of a set of data. For example, using a length of The window of smoothing the data is then the value of each data point is the value of its previous and next The average value of the data points is calculated, thereby effectively reducing the high-frequency noise in the measured data and obtaining the displacement data after preliminary filtering. The Kalman filter algorithm is used to suppress the noise of the displacement data after preliminary filtering. Kalman filtering is a recursive filter that suppresses noise by estimating the system state and is suitable for processing noise in dynamic systems. Through Kalman filtering, the true state of the system is estimated to obtain the displacement data after noise suppression. Data integrity check and outlier detection are performed on the displacement data after noise suppression to ensure the accuracy of the data in the entire processing chain, and possible abnormal data points can be identified and eliminated to obtain reliable original displacement data.
[0031] In a specific embodiment, the process of executing step S300 may specifically include the following steps: Perform wavelet decomposition on the original displacement data to obtain multi-layer wavelet coefficients, and perform soft threshold processing on the multi-layer wavelet coefficients to obtain the denoised wavelet coefficients; Performing wavelet reconstruction on the denoised wavelet coefficients to obtain initial denoised displacement data, and decomposing the initial denoised displacement data to obtain multiple singular value components; According to the cumulative contribution rate criterion, multiple singular value components are selected to obtain the target trend component, and the target trend component is reconstructed to obtain long-term trend data; Perform modal decomposition on the displacement data after preliminary noise reduction to obtain multiple intrinsic mode functions (, and screen multiple intrinsic mode functions according to energy ratio and frequency characteristics to obtain an IMF combination representing short-term fluctuations; The IMF combination representing short-term fluctuations is reconstructed to obtain short-term fluctuation data, and the long-term trend data is weightedly combined with the short-term fluctuation data to obtain the noise-reduced displacement data and displacement trend information.
[0032] Specifically, the original data is subjected to wavelet decomposition. Wavelet decomposition is a signal processing tool that can decompose complex signals into wavelet coefficients of different scales. These coefficients represent the characteristic information of the signal in different frequency ranges. By performing multi-level wavelet decomposition on the displacement data, a series of wavelet coefficients are obtained. The wavelet coefficients of each layer correspond to the performance of the signal at different scales, revealing the multi-scale characteristics of the signal. Soft threshold processing is performed on the multi-level wavelet coefficients. Soft threshold processing is a commonly used denoising method that suppresses those wavelet coefficients with small amplitudes that may represent noise by applying a threshold to the wavelet coefficients. Assume that a wavelet coefficient is , the threshold value is set to , then the processed wavelet coefficients It can be expressed as: ; Through this formula, the noise is reduced and the main features of the signal are retained to obtain the denoised wavelet coefficients. The denoised wavelet coefficients are reconstructed by wavelet, and the decomposed multi-layer wavelet coefficients are recombined into a complete signal, that is, the displacement data after preliminary denoising. In order to extract the main trend information in the signal, the displacement data of the preliminary denoising is decomposed, and the singular value decomposition is used to decompose the signal into several singular value components, each of which represents a specific pattern in the signal. According to the cumulative contribution rate criterion, those singular value components that contribute the most to the signal are selected as the target trend components. Assume that the data matrix Perform SVD decomposition and the decomposition form is: ; in, and is an orthogonal matrix, is a diagonal matrix containing the singular values of the signal. According to the cumulative contribution rate criterion, The component corresponding to the singular value that contributes the most to the signal is selected, and the main trend part of the signal, that is, the long-term trend data, is obtained after reconstruction. After selecting the target trend component, reconstruction is performed to obtain the long-term trend data. The long-term trend data reflects the low-frequency components in the signal, which is the overall trend and long-term change trend of the signal. At the same time, the displacement data that has been initially denoised is modally decomposed, and the commonly used tool is empirical mode decomposition (EMD). EMD decomposes the signal into several intrinsic mode functions (IMFs), each of which represents an inherent vibration mode in the signal. The IMF combination usually contains short-term fluctuation components of the signal, which may be caused by environmental changes or random fluctuations in the system. In order to extract the short-term fluctuation information of the signal, the IMF is screened according to the energy ratio and frequency characteristics, and those representative IMF combinations are selected. The selected IMF combination can accurately reflect the short-term fluctuation characteristics in the signal. These short-term fluctuation data often reflect the rapidly changing part of the signal, which complements the long-term trend data. The long-term trend data and short-term fluctuation data are weightedly combined to balance the long-term trend and short-term fluctuations in the signal, so that the final result can reflect the overall trend of the signal while not ignoring subtle changes, and the final noise-reduced displacement data and displacement trend information are obtained.
[0033] In a specific embodiment, the process of executing step S400 may specifically include the following steps: Perform fuzzification processing on the noise reduction displacement data and displacement trend information to obtain fuzzy input variables, and perform reasoning on the fuzzy input variables according to a preset fuzzy rule base to obtain fuzzy output variables; The fuzzy output variables are defuzzified by the centroid method to obtain the initial PID parameters, and based on the initial PID parameters, the PID control calculation is performed on the noise reduction displacement data to obtain the initial control quantity; According to the displacement trend information, the initial control amount is feedforward compensated to obtain the compensated control amount; Reinforce learning is performed on the compensated control quantity and the current system state to obtain the optimized control action, and the initial PID parameters are adjusted online according to the optimized control action to obtain the adaptive PID parameters; Based on the adaptive PID parameters, secondary PID control calculation is performed on the noise reduction displacement data to obtain the optimized control quantity; The optimized control quantity is saturated and limited to obtain the limited control quantity, and the limited control quantity is converted into a deviation correction control instruction, which includes the motor drive current and direction information.
[0034] Specifically, the noise reduction displacement data and displacement trend information are fuzzified, the input data is converted into fuzzy input variables, and the uncertainty and nonlinear characteristics of the input data are handled by using fuzzy sets, such as "high", "low" or "medium". Assume that the input noise reduction displacement data is , after fuzzification processing, the fuzzy input variable can be obtained . After getting the fuzzy input variable Finally, the fuzzy input variables are inferred using the preset fuzzy rule base to generate fuzzy output variables. The fuzzy rule base consists of a series of "if...then..." rules that reflect the expert experience or preset logic in the control system. By substituting the fuzzy input variables into these rules, the fuzzy output variables are obtained. , represents the fuzzy expression of the control action that the system should take. Convert it into an actual executable control signal and perform defuzzification. For example, the centroid method is used to calculate the centroid of the fuzzy output variable to determine an accurate control signal value. Assume that the membership function corresponding to the fuzzy output variable is , then the precise control signal calculated by the center of gravity method is It can be expressed as: ; The initial PID controller parameters are obtained through this formula. The PID controller consists of three parts: proportional (P), integral (I) and differential (D). These parameters determine the control performance of the system. Perform PID control calculations and use the initial PID parameters to obtain the initial control quantity , directly reflects the control action that the system should perform according to the current displacement state. Combined with the displacement trend information, the initial control amount is feedforward compensated. Feedforward compensation is to adjust the control amount in advance according to the expected changes of the system to reduce lag and error. Through feedforward compensation, the initial control amount is corrected to obtain the compensated control amount The compensated control quantity can better adapt to the dynamic changes of the system and improve the control accuracy. Reinforcement learning is a feedback-based adaptive optimization method that continuously adjusts the control strategy through interaction with the environment to obtain the best control effect. In this process, the system will perform tentative actions based on the current state and the compensated control amount, and adjust the control strategy by evaluating the results, and finally obtain the optimized control action. .
[0035] In a specific embodiment, the process of executing step S500 may specifically include the following steps: The correction control command is converted into a current signal to obtain a driving current signal of the linear motor, and the linear motor is controlled in a current closed loop according to the driving current signal to obtain the actual output force; Based on the Hall sensor, the position of the linear motor is detected in real time to obtain the motor position feedback signal. According to the motor position feedback signal, the linear motor is controlled in a closed loop to obtain accurate position control results. The air pressure of the air-floating guide rail is adaptively adjusted to obtain the optimal air-floating state, and the floating gap of the air-floating guide rail is sampled at high frequency based on a capacitive displacement sensor to obtain the floating gap data; According to the floating gap data, the air-floating guide rail is dynamically balanced to obtain a stable guide rail motion state, and the motion of the linear motor and the air-floating guide rail is collaboratively optimized to obtain a comprehensive execution strategy; According to the comprehensive execution strategy, the correction mechanism is precisely driven to obtain the actual correction action, and based on the high-precision grating ruler, the displacement of the actual correction action is measured to obtain the execution result.
[0036] Specifically, the correction control instruction is converted into a current signal to obtain the driving current signal of the linear motor. The correction control instruction is usually given in the form of a voltage signal or a digital signal, but the linear motor needs a current signal to drive. These instructions are converted into a driving current signal through a current amplifier or controller. This signal is directly used to control the action of the linear motor. Assume that the input correction control instruction is , the driving current signal obtained after conversion is ,but is a The linear motor is controlled in a closed loop according to the drive current signal to ensure that the force output by the motor is consistent with the desired control command. By measuring the current output of the motor and comparing it with the set drive current signal, the output is adjusted in time so that the actual output force of the motor reaches the expected effect. Assuming the target output force is , the actual output force of the motor The formula can be Calculate, where is the force constant of the motor. Through closed-loop control, real-time adjustment ,make sure near , achieving precise force control. At the same time, the position of the linear motor is precisely controlled. Through real-time position detection based on the Hall sensor, the position change of the motor is measured and a position feedback signal is generated. The feedback signal is compared with the preset target position, and the motor drive signal is adjusted through position closed-loop control so that the motor position can accurately follow the target trajectory. Assume that the target position is , the actual location is , the purpose of position closed-loop control is to reduce the position error To minimize. At the same time, in order to ensure the stable operation of the linear motor with high precision, the state of the air-floating guide rail is optimized and controlled. The air-floating guide rail reduces friction through air pressure suspension to ensure the smooth movement of the linear motor. The air pressure of the air-floating guide rail is adaptively adjusted to obtain the optimal air-floating state so that the floating gap of the guide rail is kept within the ideal range. The output pressure of the air source is adjusted through the control system to ensure the stability of the floating gap. The accurate measurement of the floating gap depends on the capacitive displacement sensor. The capacitive displacement sensor can sample the floating gap data of the guide rail at a high frequency. And provide real-time feedback. Through the floating gap data, the current state of the air-floating guide is judged, and dynamic balance control is performed to ensure that the guide remains stable and without deviation during the entire movement process. The motion of the linear motor and the air-floating guide is synergistically optimized. By comprehensively considering the force output, position control and floating state of the motor, a comprehensive execution strategy is generated. The comprehensive execution strategy ensures that the linear motor and the air-floating guide cooperate with each other during operation to achieve the optimal operating state. For example, if the floating gap increases, the system may adjust the output force of the motor or the air pressure of the guide to compensate for this change and ensure the overall stability of the system. Based on the comprehensive execution strategy, the correction mechanism is precisely driven to perform the actual correction action. This process requires not only controlling the force output and position of the motor, but also adjusting the state of the air-floating guide in real time to ensure the smoothness and accuracy of the correction process. The actual correction action after execution is a physical response to the original correction instruction, reflecting the actual adjustment effect of the system. In order to verify the accuracy of the actual correction action, the displacement after correction is measured based on a high-precision grating ruler. High-precision grating rulers can provide sub-micron displacement measurement accuracy. The displacement data fed back by the grating ruler can be used to measure the execution results of the measurement system. The execution results are compared with the preset targets to determine the accuracy and effect of the correction action. If there is a deviation, the control strategy is further adjusted according to the feedback information to optimize the subsequent correction operation.
[0037] In a specific embodiment, the process of executing step S600 may specifically include the following steps: Perform data preprocessing on the actual deviation correction actions and execution results to obtain a standardized input data set, and build a system virtual simulation environment based on the standardized input data set; Based on the system virtual simulation environment, a parameter sensitivity analysis is performed on the first calibration curve to obtain key influencing factors, and based on the key influencing factors, a polynomial fitting optimization is performed on the first calibration curve to obtain a second calibration curve; Performing structural decomposition on the first compensation model to obtain a plurality of sub-models, including a period error compensation sub-model and a temperature compensation sub-model; Based on the genetic algorithm, multiple sub-models are optimized to obtain the optimized sub-model parameter set, and global optimization is performed based on the optimized sub-model parameter set to obtain the global optimal solution; Based on the global optimal solution, a multi-objective optimization problem is constructed to obtain the Pareto optimal solution set, where the objective function includes displacement accuracy and model complexity; Based on the TOPSIS decision-making method, the Pareto optimal solution set is evaluated and selected to obtain the second compensation model, where the second compensation model is: C(x, T, v)=x+Σ[Ai*sin(2πix / P+φi)]+k(T-T0)+f(v), where C(x, T, v) is the compensated position value, x is the original reading of the grating ruler, Ai is the i-th harmonic amplitude, P is the basic period of the grating ruler, φi is the i-th harmonic phase, k is the temperature compensation coefficient, T is the current temperature, T0 is the reference temperature, f(v) is the speed-related compensation function, v is the current speed, and f(v)=α*v+β*v 2 +γ*sign(v), where α, β, and γ are velocity compensation coefficients, and sign(v) is the velocity direction function.
[0038] Specifically, data preprocessing is performed on the actual correction actions and execution results to obtain a standardized input data set. In data preprocessing, the data is normalized or standardized to eliminate the influence between different dimensions so that each variable can be compared on the same scale. Based on the standardized input data set, a system virtual simulation environment is constructed to test and optimize the control strategy in a simulation environment. The virtual environment predicts the response of the system under different conditions by simulating physical behavior in reality. For example, in this environment, the impact of temperature changes and speed changes on system performance is simulated to provide a reliable testing platform for subsequent parameter optimization. In the virtual simulation environment, a parameter sensitivity analysis is performed on the first calibration curve to identify those parameters that have the greatest impact on the calibration curve, that is, the key influencing factors. These factors may include temperature, pressure, speed, etc. By analyzing the impact of changes in these factors on the calibration accuracy of the system, the parameters that need to be optimized are determined. Assume that the calibration curve is expressed in the form of ,in , , For the parameter to be optimized, through sensitivity analysis, determine which parameter is most sensitive to the change of the calibration curve. Based on the key influencing factors, the first calibration curve is optimized by polynomial fitting to obtain a more accurate second calibration curve. By introducing higher-order terms, polynomial fitting can better describe the nonlinear behavior of the system. The optimized second calibration curve may be as follows: ; in , , , The first compensation model is structurally decomposed. The compensation model usually contains multiple sub-models for handling different types of errors such as periodic error and temperature error. The process of structural decomposition decomposes the first compensation model into several sub-models, such as a periodic error compensation sub-model and a temperature compensation sub-model. The periodic error compensation sub-model may be based on Fourier series, and the temperature compensation sub-model may be based on temperature sensitivity analysis. In order to optimize the parameters of the sub-models, a genetic algorithm is used for parameter optimization. Genetic algorithm is an optimization method based on natural selection, which gradually optimizes the parameter set of the model by simulating the biological evolution process. The genetic algorithm randomly generates a set of parameter values in the initial population, and then gradually evolves the optimal parameter combination through operations such as crossover, mutation and selection. The optimized sub-model parameter set can effectively reduce the system error, thereby improving the accuracy of the compensation model. After obtaining the optimized sub-model parameter set, global optimization is performed to find the global optimal solution of the entire system. The goal of global optimization is to find a parameter combination that performs well under all working conditions. The global optimal solution ensures that the system can achieve the optimal compensation effect under various conditions by comprehensively considering the performance of all sub-models. Based on the global optimal solution, a multi-objective optimization problem is constructed. In this problem, the objective function includes two main objectives: displacement accuracy and model complexity. Displacement accuracy requires that the model can provide accurate displacement compensation under various conditions, while model complexity requires that the model be as simple as possible to reduce the amount of calculation and system burden. By optimizing these two objectives, the Pareto optimal solution set is obtained, which contains a set of solutions that perform best under different trade-offs. In order to select the optimal solution from the Pareto optimal solution set, the TOPSIS decision method is adopted. TOPSIS is a ranking method based on ideal solutions, which ranks each solution by calculating the distance from the ideal solution and the negative ideal solution. The final selected solution is the solution that is closest to the ideal solution, and this solution will be used as the final form of the second compensation model. The final second compensation model can be expressed as: ; in, is the position value after compensation, is the original reading of the grating ruler, For the The amplitude of the subharmonics, is the basic period of the grating ruler, For the The phase of the subharmonics, is the temperature compensation coefficient, is the current temperature, is the reference temperature, is the speed-related compensation function, expressed as sign ,in is the speed compensation coefficient, is a function of velocity direction.
[0039] See also Figure 2 , Figure 2 The schematic block diagram of the structure of the high-precision displacement correction device 200 for the wooden box steel belt drive provided in the embodiment of the present application is as follows: Figure 2 As shown, the wooden box steel belt driven high-precision displacement correction device 200 comprises: An installation module 210 is used to install and calibrate the high-precision grating ruler and create a first calibration curve and a first compensation model; The acquisition module 220 is used to perform high dynamic range acquisition on the steel strip displacement data according to the first calibration curve and the first compensation model to obtain original displacement data; An analysis module 230 is used to perform wavelet denoising and multi-scale trend analysis on the original displacement data to obtain denoised displacement data and displacement trend information; A control module 240, configured to perform adaptive deviation correction control based on the noise reduction displacement data and the displacement trend information, and generate a deviation correction control instruction; The execution module 250 is used to drive the linear motor and the air-floating guide rail to accurately execute the deviation correction control instruction, and obtain the actual deviation correction action and the corresponding execution result; The optimization module 260 is used to perform digital twin simulation and multi-objective optimization on the first calibration curve and the first compensation model according to the actual correction action and the execution result, so as to obtain the second calibration curve and the second compensation model.
[0040] Through the synergy of the above components, the application and precise calibration of high-precision grating rulers improve the basic accuracy of displacement measurement. High dynamic range data acquisition and multi-level signal processing technology effectively improve the quality of original displacement data and reduce the impact of measurement noise and interference. Wavelet noise reduction and multi-scale trend analysis methods can effectively separate noise and effective information in displacement data, and improve the accuracy of subsequent control. The adaptive correction control strategy combines fuzzy control, PID control and reinforcement learning to achieve rapid response and precise adjustment to complex dynamic environments. The precise actuators of linear motor drive and air-floating guide rails improve the execution accuracy of correction actions and reduce the impact of mechanical errors. Digital twin simulation and multi-objective optimization methods realize dynamic optimization of calibration curves and compensation models, and improve the long-term stability and adaptability of the system. Taking into account a variety of influencing factors such as temperature and speed, a more comprehensive and accurate compensation model is constructed, which effectively improves the overall effect of correction.
[0041] The present application also provides a high-precision displacement correction device for a wooden box steel belt transmission, the high-precision displacement correction device for a wooden box steel belt transmission comprises a memory and a processor, the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the processor executes the steps of the high-precision displacement correction method for a wooden box steel belt transmission in the above-mentioned embodiments.
[0042] The present application also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. Instructions are stored in the computer-readable storage medium. When the instructions are executed on a computer, the computer executes the steps of the high-precision displacement correction method for wooden box steel belt transmission.
[0043] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems, systems and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.
[0044] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc., various media that can store program codes.
[0045] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.
Claims
1. A high-precision displacement correction method for wooden box steel belt transmission, characterized in that: include: Install and calibrate the high-precision grating ruler, and create the first calibration curve and the first compensation model; According to the first calibration curve and the first compensation model, high dynamic range acquisition is performed on the displacement data of the steel strip to obtain original displacement data; Performing wavelet denoising and multi-scale trend analysis on the original displacement data to obtain denoised displacement data and displacement trend information; Performing adaptive deviation correction control based on the noise reduction displacement data and the displacement trend information, and generating a deviation correction control instruction; The correction control instruction is accurately executed by a linear motor drive and an air-floating guide rail to obtain an actual correction action and a corresponding execution result; According to the actual deviation-correcting action and the execution result, digital twin simulation and multi-objective optimization are performed on the first calibration curve and the first compensation model to obtain a second calibration curve and a second compensation model.
2. The high-precision displacement correction method for wooden box steel belt transmission according to claim 1 is characterized in that: The step of installing and calibrating the high-precision grating ruler and creating a first calibration curve and a first compensation model includes: Calibrate the high-precision grating ruler with a precision level and an inclinometer to obtain initial installation error data, and adjust the high-precision grating ruler according to the initial installation error data to obtain a calibrated grating ruler installation position; Based on the laser interferometer, multi-point measurement is performed on the calibrated grating ruler to obtain corresponding data of the grating ruler reading and the interferometer reading, and the corresponding data of the grating ruler reading and the interferometer reading are fitted by the least square method to obtain the initial calibration curve equation; According to the initial calibration curve equation, a nonlinear error analysis is performed on the grating ruler to obtain a nonlinear error distribution function, and based on the nonlinear error distribution function, the initial calibration curve equation is corrected to obtain a first calibration curve; Performing Fourier series expansion on the first calibration curve to obtain a periodic error compensation term, and constructing a temperature compensation model based on the periodic error compensation term and temperature sensor data to obtain a temperature compensation function; Based on the temperature compensation function and the periodic error compensation term, a first compensation model is constructed, wherein the first compensation model is: C(x, T)=x+Σ[Ai*sin(2πix / P+φi)]+k(T-T0), wherein C(x, T) is the compensated position value, x is the original reading of the grating scale, Ai is the i-th harmonic amplitude, P is the basic period of the grating scale, φi is the i-th harmonic phase, k is the temperature compensation coefficient, T is the current temperature, and T0 is the reference temperature.
3. The high-precision displacement correction method for wooden box steel belt transmission according to claim 1 is characterized in that: The step of performing high dynamic range acquisition on the steel strip displacement data according to the first calibration curve and the first compensation model to obtain original displacement data includes: Pre-amplifying the grating ruler signal to obtain an amplified analog signal, and performing band-pass filtering on the amplified analog signal to obtain a filtered signal; Based on a programmable gain amplifier, dynamically adjust the gain of the filtered signal to obtain a signal adapted to the ADC input range, and perform analog-to-digital conversion on the signal adapted to the ADC input range to obtain digital displacement data; According to the first calibration curve, the digitized displacement data is subjected to nonlinear error correction to obtain corrected displacement data, and based on the periodic error compensation term in the first compensation model, the corrected displacement data is subjected to periodic error compensation to obtain compensated displacement data; According to the temperature sensor data and the temperature compensation function in the first compensation model, temperature compensation is performed on the compensated displacement data to obtain temperature compensated displacement data, and moving average filtering is performed on the temperature compensated displacement data to obtain preliminary filtered displacement data; Based on the Kalman filter algorithm, noise suppression is performed on the displacement data after the preliminary filtering to obtain the noise-suppressed displacement data, and data integrity check and abnormal value check are performed on the noise-suppressed displacement data to obtain the original displacement data.
4. The high-precision displacement correction method for wooden box steel belt transmission according to claim 1 is characterized in that: The performing of wavelet denoising and multi-scale trend analysis on the original displacement data to obtain denoised displacement data and displacement trend information includes: Performing wavelet decomposition on the original displacement data to obtain multi-layer wavelet coefficients, and performing soft threshold processing on the multi-layer wavelet coefficients to obtain denoised wavelet coefficients; Performing wavelet reconstruction on the denoised wavelet coefficients to obtain initial denoised displacement data, and decomposing the initial denoised displacement data to obtain a plurality of singular value components; According to the cumulative contribution rate criterion, the plurality of singular value components are selected to obtain a target trend component, and the target trend component is reconstructed to obtain long-term trend data; Performing modal decomposition on the displacement data after preliminary noise reduction to obtain a plurality of intrinsic mode functions (IMFs), and screening the plurality of IMFs according to energy proportion and frequency characteristics to obtain an IMF combination representing short-term fluctuations; The IMF combination representing the short-term fluctuation is reconstructed to obtain short-term fluctuation data, and the long-term trend data is weightedly combined with the short-term fluctuation data to obtain noise-reduced displacement data and displacement trend information.
5. The high-precision displacement correction method for wooden box steel belt transmission according to claim 1 is characterized in that: The adaptive deviation correction control is performed based on the noise reduction displacement data and the displacement trend information to generate a deviation correction control instruction, including: Performing fuzzification processing on the noise reduction displacement data and the displacement trend information to obtain fuzzy input variables, and performing reasoning on the fuzzy input variables according to a preset fuzzy rule base to obtain fuzzy output variables; Defuzzifying the fuzzy output variable using the centroid method to obtain initial PID parameters, and performing PID control calculation on the noise reduction displacement data based on the initial PID parameters to obtain an initial control amount; According to the displacement trend information, feedforward compensation is performed on the initial control amount to obtain a compensated control amount; Performing reinforcement learning on the compensated control amount and the current system state to obtain an optimized control action, and adjusting the initial PID parameters online according to the optimized control action to obtain adaptive PID parameters; Based on the adaptive PID parameters, a secondary PID control calculation is performed on the noise reduction displacement data to obtain an optimized control amount; The optimized control amount is saturated and limited to obtain a limited control amount, and the limited control amount is converted into a deviation correction control instruction, wherein the deviation correction control instruction includes motor drive current and direction information.
6. The high-precision displacement correction method for wooden box steel belt transmission according to claim 5 is characterized in that: The linear motor drive and the air-floating guide rail are used to accurately execute the deviation correction control instruction to obtain the actual deviation correction action and the corresponding execution result, including: Convert the deviation correction control instruction into a current signal to obtain a driving current signal of the linear motor, and perform current closed-loop control on the linear motor according to the driving current signal to obtain an actual output force; Based on the Hall sensor, the position of the linear motor is detected in real time to obtain a motor position feedback signal, and according to the motor position feedback signal, the linear motor is controlled in a closed loop position to obtain an accurate position control result; Adaptively adjust the air pressure of the air-floating guide rail to obtain an optimal air-floating state, and perform high-frequency sampling on the floating gap of the air-floating guide rail based on a capacitive displacement sensor to obtain floating gap data; According to the floating gap data, the air-floating guide rail is dynamically balanced to obtain a stable guide rail motion state, and the motion of the linear motor and the air-floating guide rail is collaboratively optimized to obtain a comprehensive execution strategy; According to the comprehensive execution strategy, the correction mechanism is precisely driven to obtain the actual correction action, and based on the high-precision grating ruler, the displacement of the actual correction action is measured to obtain the execution result.
7. The high-precision displacement correction method for wooden box steel belt transmission according to claim 1 is characterized in that: The step of performing digital twin simulation and multi-objective optimization on the first calibration curve and the first compensation model according to the actual deviation correction action and the execution result to obtain a second calibration curve and a second compensation model includes: Performing data preprocessing on the actual deviation correction action and the execution result to obtain a standardized input data set, and constructing a system virtual simulation environment based on the standardized input data set; Based on the system virtual simulation environment, performing parameter sensitivity analysis on the first calibration curve to obtain key influencing factors, and based on the key influencing factors, performing polynomial fitting optimization on the first calibration curve to obtain a second calibration curve; Performing structural decomposition on the first compensation model to obtain a plurality of sub-models, including a period error compensation sub-model and a temperature compensation sub-model; Based on a genetic algorithm, the parameters of the multiple sub-models are optimized to obtain an optimized sub-model parameter set, and global optimization is performed based on the optimized sub-model parameter set to obtain a global optimal solution; Based on the global optimal solution, a multi-objective optimization problem is constructed to obtain a Pareto optimal solution set, wherein the objective function includes displacement accuracy and model complexity; Based on the TOPSIS decision-making method, the Pareto optimal solution set is evaluated and selected to obtain a second compensation model, wherein the second compensation model is: C(x, T, v)=x+Σ[Ai*sin(2πix / P+φi)]+k(T-T0)+f(v), wherein C(x, T, v) is the position value after compensation, x is the original reading of the grating ruler, Ai is the i-th harmonic amplitude, P is the basic period of the grating ruler, φi is the i-th harmonic phase, k is the temperature compensation coefficient, T is the current temperature, T0 is the reference temperature, f(v) is the speed-related compensation function, v is the current speed, and f(v)=α*v+β*v 2 +γ*sign(v), where α, β, and γ are velocity compensation coefficients, and sign(v) is the velocity direction function.
8. A high-precision displacement correction device for wooden box steel belt transmission, characterized in that: The method for high-precision displacement correction of a wooden box steel belt drive according to any one of claims 1 to 7 comprises: An installation module, used for installing and calibrating the high-precision grating ruler, and creating a first calibration curve and a first compensation model; An acquisition module, used for acquiring the steel strip displacement data in a high dynamic range according to the first calibration curve and the first compensation model to obtain original displacement data; An analysis module, used for performing wavelet denoising and multi-scale trend analysis on the original displacement data to obtain denoised displacement data and displacement trend information; A control module, configured to perform adaptive deviation correction control based on the noise reduction displacement data and the displacement trend information, and generate a deviation correction control instruction; An execution module, used to drive the linear motor and the air-floating guide rail to accurately execute the deviation correction control instruction, and obtain the actual deviation correction action and the corresponding execution result; An optimization module is used to perform digital twin simulation and multi-objective optimization on the first calibration curve and the first compensation model according to the actual correction action and the execution result to obtain a second calibration curve and a second compensation model.
9. A high-precision displacement correction device for wooden box steel belt transmission, characterized in that: The high-precision displacement correction device for wooden box steel belt transmission comprises: a memory and at least one processor, wherein instructions are stored in the memory; The at least one processor calls the instructions in the memory to enable the wooden box steel belt transmission high-precision displacement correction device to execute the wooden box steel belt transmission high-precision displacement correction method as described in any one of claims 1-7.
10. A computer-readable storage medium having instructions stored thereon, characterized in that: When the instructions are executed by the processor, the high-precision displacement correction method for wooden box steel belt transmission as described in any one of claims 1-7 is implemented.
Citation Information
Patent Citations
Grating ruler signal error compensation process
CN101162139A
Superhigh precision servo driving system based on PID online calibration machine tool
CN105824290A
Shafting jacking data acquiring and processing device
CN106940169A
Method and device for testing comprehensive performance of numerical control system
CN109656196A
Grating ruler dynamic error detection system and detection method
CN111765844A
Cited By
Stay wire type displacement sensor error analysis method and system
CN119128817A
Intelligent material conveying deviation remote monitoring system based on Internet of Things
CN120207895A
Intelligent Material Conveyor Deviation Remote Monitoring System Based on Internet of Things
CN120207895B
Linear displacement sensor calibration analysis system
CN120385283A
A linear displacement sensor calibration and analysis system
CN120385283B
