Control method of tension adjusting equipment on composite copper foil continuous production line

Abstract

The invention provides a control method of tension adjusting equipment on a composite copper foil continuous production line, which comprises the following steps: acquiring a tension set value and a tension actual value of adjacent stations on the composite copper foil continuous production line, and calculating a difference value between the tension set value and the tension actual value to obtain tension deviation data; in the process of adjusting the tension between the adjacent stations, judging whether the copper foil snakes or not, and if yes, feeding back snaking offset data of the copper foil to a feed-forward compensation function and a feedback controller so as to cooperatively adjust the control of the tension; on the basis of tension balance between adjacent stations, operation parameters of all stations of the composite copper foil continuous production line are collected in real time, and a production line operation mathematical description function is established; and optimizing control parameters of production line operation in combination with a preset rule in the preset knowledge base.

Description

Technical Field

[0001] The invention relates to the field of information technology, and in particular to a control method for tension regulating equipment on a composite copper foil continuous production line. Background Art

[0002] In the continuous production line of composite copper foil, when the copper foil passes through multiple stations such as coating, drying, and cooling, there is a tension difference between adjacent stations due to the physical distance between the stations and the elastic-plastic deformation characteristics of the copper foil. If the tension of the copper foil changes suddenly during its movement, it is very easy to cause the problem of copper foil meandering, resulting in a decrease in product quality. Specifically, when the copper foil enters the drying station from the coating station, due to the sudden increase in temperature, the copper foil will produce thermal expansion and contraction effects, causing local stress changes in the copper foil, and then causing a sudden change in tension. Similarly, when the copper foil enters the cooling station from the drying station, the sudden drop in temperature will also cause a similar tension change problem. In addition, the inconsistent tension applied to the copper foil at different stations will also lead to tension differences between stations. For example, the coating station usually needs to apply a large tension to the copper foil to ensure the uniformity of the coating, while the drying station needs to appropriately reduce the tension to prevent the copper foil from deforming. The cooling station has different requirements for tension. These complex process requirements and the physical properties of the copper foil itself make the tension control of the composite copper foil production line face many challenges. How to ensure the stability of the copper foil tension while taking into account the specific needs of each workstation, coordinate the tension balance between adjacent workstations, and avoid copper foil snaking caused by sudden tension changes is a key technical problem that needs to be solved urgently. Summary of the invention

[0003] The present invention provides a control method for a tension regulating device on a composite copper foil continuous production line, which mainly includes:

[0004] Obtaining the tension setting value and the actual tension value of adjacent workstations on the composite copper foil continuous production line, and calculating the difference between the tension setting value and the actual tension value to obtain tension deviation data;

[0005] Combine the tension fluctuation data in the historical production data and the tension deviation data to predict the tension mutation that may occur under the current production conditions, and generate a feedforward compensation signal based on the prediction result to form a preliminary tension gradient correction signal;

[0006] At the same time, combined with the tension deviation data and the real-time operation status data of the copper foil, the tension control parameters are dynamically adjusted to generate a feedback compensation signal, and the preliminary tension gradient correction signal is corrected to form a corrected tension gradient correction signal;

[0007] According to the corrected tension gradient correction signal, it is judged whether the tension between adjacent workstations is balanced. If it is unbalanced, the corrected tension gradient correction signal is sent to the tension adjustment devices of the adjacent workstations respectively, and the tension between the workstations is dynamically adjusted by the adjustment device until the tension between the adjacent workstations reaches a balanced state;

[0008] In the process of adjusting the tension between adjacent workstations, it is determined whether the copper foil is snaking. If snaking occurs, the snaking offset data of the copper foil is fed back to the feedforward compensation function and the feedback controller to coordinately adjust the tension control.

[0009] On the basis of tension balance between adjacent workstations, the operation parameters of each workstation of the composite copper foil continuous production line are collected in real time, and a mathematical description function of the production line operation is established;

[0010] The control parameters of the production line operation are optimized by combining the preset rules in the preset knowledge base.

[0011] The technical solution provided by the embodiment of the present invention may have the following beneficial effects:

[0012] The invention discloses a control method for tension regulating equipment on a continuous production line of composite copper foil. The method obtains tension setting value and actual value and calculates deviation data by setting a tension coordination control unit between adjacent workstations. The deviation data is input into a feedforward and feedback control model, and a tension gradient correction signal is generated by combining historical data and real-time operation status. The tension between workstations is dynamically adjusted according to the correction signal to achieve balance. At the same time, the problem of copper foil snaking is monitored, an alarm is triggered and feedback is given to the control model adjustment strategy. The invention also collects the operating parameters of each workstation in real time, establishes a mathematical model through big data analysis, and adaptively optimizes the control model parameters. The method realizes precise coordinated control of the tension of the composite copper foil production line by combining feedforward prediction and feedback compensation. At the same time, big data analysis and adaptive optimization are introduced to continuously improve the control accuracy and stability. The invention effectively solves the problems of unstable tension control and snaking in the production process of composite copper foil, and significantly improves the consistency of product quality and production efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 The present invention is a flow chart of a method for controlling a tension regulating device on a continuous production line of composite copper foil. DETAILED DESCRIPTION

[0014] In order to further understand the content of the present invention, the present invention is described in detail in conjunction with the accompanying drawings and embodiments. The present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It is understood that the specific embodiments described herein are only used to explain the relevant inventions, rather than to limit the invention. It is also necessary to explain that, for ease of description, only the parts related to the invention are shown in the accompanying drawings.

[0015] like Figure 1 In this embodiment, a control method for a tension adjustment device on a composite copper foil continuous production line may specifically include:

[0016] Step S101, obtaining the tension setting value and the tension actual value of adjacent workstations on the composite copper foil continuous production line, and calculating the difference between the tension setting value and the tension actual value to obtain tension deviation data.

[0017] A station sensor array is used to collect tension monitoring data on the composite copper foil transmission line, and the station tension value is obtained by signal processing through a tension data acquisition unit; the station tension value is compared with a preset tension threshold range, and position information is obtained from an optical encoder and converted into a digital quantity; if the station tension value exceeds the preset tension threshold range, the tension deviation value is obtained from the tension data acquisition unit and transmitted to the tension compensation unit.

[0018] Exemplarily, a station sensor array is used to arrange multiple groups of tension monitoring probes between adjacent stations on the composite copper foil transmission line and calculate the real-time tension monitoring data. The collected data is processed by the tension data acquisition unit to obtain the station tension value. An optical encoder is installed at the end of the adjacent station transmission shaft to collect speed information, and the station tension value is compared according to the preset tension threshold range, and the position information is obtained from the optical encoder and converted into a digital quantity. If the station tension value exceeds the preset tension threshold range, the tension deviation value is obtained from the tension data acquisition unit and input into the tension compensation unit, and the compensation coefficient is calculated by the recursive least squares algorithm in the compensation unit. For the compensation coefficient output by the tension compensation unit, a state space equation group including the tension deviation value, the compensation coefficient, and the station speed is established, and a compensation control signal is output. The speed of the transmission shaft drive motor is adjusted according to the compensation control signal, and the adjusted speed signal is fed back to the tension detection unit. The station tension value after speed adjustment is collected by the tension detection unit, and the new tension deviation data is output to the tension compensation unit after comparison with the tension setting reference value. The tension control of the composite copper foil production line involves dynamic coordination among multiple stations. Taking a specific working condition as an example, when the copper foil is transferred from the first station to the second station, the real-time value of the tension is collected through the tension monitoring probe. The probe adopts a strain sensor with a sensitivity of up to 0.1N and a sampling frequency of 200Hz. After signal conditioning, it outputs a 4-20mA standard signal. The tension data acquisition unit converts the analog signal into a digital quantity through a 16-bit AD converter. The station tension setting reference value is set to 50N, and the preset tension fluctuation threshold range is ±5N. The optical encoder is installed at the end of the transmission shaft with a resolution of 2500 lines / revolution. It can accurately detect speed changes. When the tension is detected to exceed the threshold range, compensation control is required. The compensation control adopts the recursive least squares algorithm. By constructing the state vector [tension deviation, speed, compensation coefficient], the state space equation X(k+1)=AX(k)+Bu(k) is established, where A is the state transfer matrix, B is the control matrix, u(k) is the compensation input, and X(k) is the state vector. Assume that at a certain moment, the tension deviation detected is 6N, the speed is 1200rpm, and the initial compensation coefficient is 0.8. Substitute it into the equation to calculate the new compensation coefficient of 0.92. The speed of the drive motor is closed-loop controlled according to the compensation coefficient. The drive motor adopts a permanent magnet synchronous motor with a rated speed of 1500rpm. The speed is adjusted by the frequency converter, and the response time is less than 50ms. After speed regulation, the tension detection unit re-collects the workstation tension value. If there is still a deviation, compensation control continues until the tension returns to the set range. The discrete Kalman filter algorithm is used to solve the state space equation, which can effectively suppress the influence of measurement noise. The system noise covariance matrix Q and the measurement noise covariance matrix R are set as

[0019] Q=diag(0.01,0.01,0.01), R=0.1. The prediction step calculates the estimated value of the state at the next moment through the state equation, and the correction step corrects the estimated value in combination with the measured value to finally obtain the optimal state estimate. During the tension compensation control process, the update cycle of the compensation coefficient is 100ms to avoid system oscillation caused by frequent adjustments. The speed regulation of the drive motor uses a ramp function to achieve a smooth transition, and the slope is set to 200rpm / s to ensure the continuity of the tension change. When the distance between workstations is 2000mm and the copper foil running speed is 60m / min, the coupling effect of the tension of adjacent workstations needs to be eliminated through feedforward compensation.

[0020] Step S102, combining the tension fluctuation data in the historical production data and the tension deviation data, predicting the tension mutation that may occur under the current production conditions, and generating a feedforward compensation signal according to the prediction result to form a preliminary tension gradient correction signal.

[0021] A sliding time window is used to sample the tension deviation data, and a filtered tension deviation sequence is obtained from the sampled data through a median filter; the tension fluctuation law is obtained through long short-term memory network training according to the tension deviation sequence and the tension fluctuation characteristic values ​​in the historical production database; according to the tension fluctuation law, the least squares method is used to fit the tension change curve to obtain the tension mutation warning threshold; a feedforward compensation function is constructed according to the tension mutation warning threshold, the feedforward compensation function is smoothed by an adaptive filter to obtain an initial compensation signal, and the gradient descent method is used to optimize and calculate the initial compensation signal to obtain a tension gradient correction signal.

[0022] Exemplarily, a sliding time window is used to sample the tension deviation data sequence, and the sampled data is subjected to median filtering by a data preprocessing unit to remove outliers, and the filtered tension deviation sequence is output from the data preprocessing unit. According to the tension deviation sequence in the data preprocessing unit, the tension fluctuation characteristic value is read from the historical production database, and a long short-term memory network including the time step and the number of hidden layer nodes is constructed by a time series prediction unit. According to the tension fluctuation law obtained by training the long short-term memory network, the least squares method is used to fit the tension change curve, and the tension mutation warning threshold is output from the curve fitting unit. A feedforward compensation function is constructed according to the tension mutation warning threshold, and the compensation function is input into the feedforward compensation unit, and the compensation amount sequence is calculated by the feedforward compensation unit. An adaptive filter is used to smooth the compensation amount sequence, and an initial compensation signal including the compensation signal amplitude is output. The initial compensation signal is optimized and calculated by the gradient descent method, and a tension gradient correction signal is output from the feedforward compensation unit. The tension control of the composite copper foil production line involves multiple key links. Taking the tension fluctuation control in the production process as an example, when the production line is running, a sampling period of 200ms is used to collect tension deviation data, and a sliding time window with a length of 1000 data points is set. The data is preprocessed by the median filter algorithm, and the filter window takes 5 data points to remove sudden interference in the measurement process. The extraction of tension fluctuation feature values ​​adopts the time series analysis method. The tension fluctuation data of the past 30 days is read from the historical database. Each data record contains three fields: timestamp, tension value, and production speed. Time series prediction is performed through the long short-term memory network. The network structure contains 2 LSTM layers, each with 64 hidden layer nodes, the time step is set to 20, and the input feature dimension is 3. The cubic spline interpolation method is used to fit the tension change curve. The polynomial coefficients are solved between adjacent nodes by the least squares method. The tension mutation warning threshold is set to 3 times the standard deviation, that is, when the tension fluctuation amplitude is detected to exceed 3σ, the warning signal is triggered. The feedforward compensation function u(t) adopts the form of PI controller, u(t) = Kpe(t) + Ki∫e(t)dt, where Kp is the proportional coefficient, the value is 0.8, Ki is the integral coefficient, the value is 0.05, and e(t) is the error signal, which represents the difference between the expected output and the actual output. The smoothing of the compensation signal adopts the adaptive Kalman filter algorithm, the state equation is x(k+1) = x(k) + w(k), and the observation equation is y(k) = x(k) + v(k), where w(k) and v(k) are process noise and observation noise respectively, and the covariance matrix Q = 0.01, R = 0.1. The gain matrix K of the filter is dynamically adjusted with the change of the measured value to ensure the smoothness of the filtering result. During the gradient descent optimization process, the learning rate α was set to 0.01, the number of iterations was 100, and the loss function adopted the mean square error form of L = Σ(y-y')2 / n, where y is the target tension value, y' is the actual tension value, and n is the number of data points.Through iterative calculation, the loss function converges to the minimum value, and finally outputs the optimized tension gradient correction signal. This multi-level linkage control scheme shows good dynamic characteristics in actual production. When the production line speed increases from 50m / min to 80m / min, the tension fluctuation amplitude is controlled within the range of ±2N, which improves the stability by 50% compared with the traditional PID control. The control process forms a closed-loop feedback mechanism, and realizes precise control of tension fluctuations through a combination of feedforward compensation and adaptive optimization.

[0023] Step S103, combining the tension deviation data and the real-time operation status data of the copper foil, dynamically adjusting the tension control parameters, generating a feedback compensation signal, and correcting the preliminary tension gradient correction signal to form a corrected tension gradient correction signal.

[0024] According to the real-time operation status data of the copper foil, the recursive least squares algorithm is used to perform online identification of the copper foil operation parameters, and the online identification results are used to construct a feedback compensator; a tension compensation signal is output through the feedback compensator, and the tension compensation signal is used to dynamically correct the tension gradient correction value; a neural network is used to adaptively adjust the dynamically corrected compensation value, and optimized correction parameters are obtained from the neural network, and the correction parameters are used to adjust the tension control parameters.

[0025] Exemplarily, a sensor array is used to monitor the operation status of the copper foil production line in real time. Real-time data including operation speed and tension fluctuation are collected through multiple sets of sensors, and a complete set of operation status data sets are obtained from the sensor array. For the collected operation status data set, the recursive least squares algorithm is used to identify the copper foil operation parameters online, and a feedback compensator including proportional terms and integral terms is constructed through the identification results, and a compensation signal is output from the feedback compensator. The tension gradient correction value is dynamically corrected according to the compensation signal, and the corrected compensation value is adaptively adjusted using a neural network, and the optimized correction parameters are output from the neural network. The tension control parameters are dynamically adjusted using the optimized correction parameters, and a control function including feedback compensation is established, and the corrected tension gradient correction signal is output through the feedback compensator. The feedback control process of the copper foil production line involves real-time monitoring and dynamic adjustment of multiple key parameters. Taking a certain station on the production line as an example, a strain tension sensor, a photoelectric encoder and a displacement sensor are arranged to form a sensor array, the sampling frequency is set to 100Hz, the data resolution is 16 bits, and the operation parameters are collected in real time. The range of the tension sensor is 0-500N, and the sensitivity is 0.2mV / N; the resolution of the photoelectric encoder is 2500 lines / turn, which is used for speed detection; the range of the displacement sensor is 0-50mm, which is used for position feedback. The online identification of the operating parameters adopts the recursive least squares method to construct the state equation x(k+1)=Ax(k)+Bu(k), and the observation equation y(k)=Cx(k)+Du(k), where A is the state transfer matrix, B is the control matrix, C is the output rectangle, and D is the direct transfer matrix. The state vector x(k) contains two components, speed and tension, and the control input u(k) is the torque command of the drive motor. When the speed is detected to change suddenly from 50m / min to 80m / min, the tension fluctuates transiently. The speed response time constant is calculated by the recursive algorithm to be 0.5s, and the tension response time constant is 0.3s. The feedback compensator adopts PI control structure, and the proportional coefficient Kp is determined to be 0.85 by Ziegler-Nichols tuning method, and the integral time Ti is 2s. Compensator u(t) = Kp[e(t)+1 / Ti∫e(t)dt], where e(t) is the tension deviation. For the tension setting value of 50N, when the measured tension is 45N, the proportional term output is calculated to be 4.25N, the integral term output is 3.75N, and the total compensation amount is 8N. The adaptive neural network adopts a three-layer structure, with 2 nodes in the input layer corresponding to the speed and tension deviation, 8 nodes in the hidden layer, and 1 node in the output layer to output the compensation correction amount. The activation function of the network uses the hyperbolic tangent function, and the learning rate is set to 0.01. When the input speed is 80m / min and the tension deviation is -5N, the compensation correction coefficient of 1.2 is output after neural network mapping to correct the original compensation amount.In the closed-loop feedback control, the compensation signal and the gradient correction signal are combined by weighted summation, with weight coefficients of 0.7 and 0.3 respectively. Under actual operating conditions, when the production line starts to accelerate, the feedback compensation loop can quickly respond to the tension fluctuation caused by speed changes. The tension fluctuation amplitude after compensation is reduced from ±8N to ±2N, and the dynamic response time is shortened by 40%.

[0026] Step S104, judging whether the tension between adjacent workstations is balanced according to the corrected tension gradient correction signal, if unbalanced, sending the corrected tension gradient correction signal to the tension adjustment devices of the adjacent workstations respectively, and dynamically adjusting the tension between the workstations by adjusting the devices until the tension between the adjacent workstations reaches a balanced state.

[0027] A sensor array is used to collect real-time tension values ​​of adjacent workstations; based on the real-time tension values, the tension difference between the workstations is calculated to obtain tension deviation data; a correction signal is obtained through a tension gradient correction unit, and the correction signal is generated when the tension deviation data exceeds a preset tension balance threshold; in response to the correction signal, a workstation speed regulation instruction is output from a signal distributor, and the workstation speed regulation instruction is calculated by the signal distributor according to the speed regulation ratio of the transmission shaft of each workstation; a feedback controller is used to process the workstation speed regulation instruction, and a speed regulation control signal is output from the feedback controller, and the speed regulation control signal is obtained by the feedback controller calculating the speed regulation compensation amount according to the workstation operation parameters.

[0028] Exemplarily, a sensor array is used to collect the real-time tension values ​​of adjacent workstations, and the tension difference between the workstations is calculated by a tension comparison unit, and the tension deviation data between the workstations is output from the tension comparison unit. A numerical comparison is performed based on the tension deviation data and the preset tension balance threshold. If the tension deviation value exceeds the balance threshold range, a correction signal is obtained from the tension gradient correction unit. Based on the correction signal, the speed regulation ratio of the transmission shaft of each workstation is calculated by a signal distributor, and the workstation speed regulation instruction is output from the signal distributor. A feedback controller is used to process the workstation speed regulation instruction, and the speed compensation amount is calculated according to the workstation operation parameters, and a speed control signal is output from the feedback controller. The speed of the transmission shaft of the adjacent workstation is adjusted by the speed control signal, and the speed feedback signal is collected from the transmission shaft encoder. The workstation tension value is calculated based on the speed feedback signal. If the tension deviation value still exceeds the balance threshold range, the third step is returned to perform until the tension deviation value is within the balance threshold range. The application of tension balance control in copper foil production lines involves the coordination of multiple links. Taking two adjacent workstations as an example, a strain gauge tension sensor is used for real-time monitoring, with a sampling frequency of 200Hz, a signal resolution of 16 bits, and a measurement range of 0-500N. The sensor outputs a 4-20mA standard signal, which is input into the tension comparison unit after signal conditioning. The tension balance control between workstations adopts a two-level structure. The first level is the tension difference judgment, and the preset balance threshold is ±3N. When the tension value of workstation 1 is detected to be 52N and the tension value of workstation 2 is 47N, the calculated tension deviation value is 5N, which exceeds the balance threshold range and triggers the correction signal output. The amplitude of the correction signal is obtained through PID operation, with a proportional coefficient Kp=0.8, an integral time Ti=2s, and a differential time Td=0.1s. The signal distribution adopts weighted distribution method. Based on the characteristics of the workstation, the distribution matrix is ​​established as [0.6, 0.4; 0.4, 0.6], which means that 60% of the speed control command of workstation 1 acts on the workstation itself and 40% acts on the adjacent workstation, while the opposite is true for workstation 2. For a tension deviation of 5N, the speed control command of workstation 1 is -1.8rpm after distribution, and the speed control command of workstation 2 is 1.2rpm. The feedback controller adopts a model predictive control structure, with a prediction time domain of 10 sampling cycles and a control time domain of 3 sampling cycles. The state equation is x(k+1)=0.95x(k)+0.2u(k), where x is the tension value and u is the speed control command. Substitute the initial state of workstation 1 52N, predict the tension change trend in the future time domain, and calculate the optimal speed control sequence through rolling optimization. The speed regulation of the transmission shaft is realized by a frequency converter, with a rated frequency of 50Hz corresponding to a speed of 1500rpm and a speed regulation resolution of 0.01Hz. The encoder feedback signal is 2500 lines / rev, and provides 10,000 pulses / rev position feedback after 4 times the frequency. In closed-loop control, the speed signal is filtered through a first-order low-pass filter with a cutoff frequency of 5Hz to filter out high-frequency interference components.During the iterative process of tension balance control, the workstation tension value is collected again for comparison after each adjustment. When the tension deviation is detected to be 2.8N, it is within the balance threshold range and the adjustment process is completed. Compared with the traditional single-workstation control method, multi-workstation collaborative control can more effectively suppress the transmission and amplification of tension fluctuations.

[0029] Step S105, in the process of adjusting the tension between adjacent workstations, it is determined whether the copper foil has snaking. If snaking occurs, the copper foil snaking offset data is fed back to the feedforward compensation function and the feedback controller to coordinately adjust the tension control.

[0030] A photoelectric sensor array is used to obtain the edge position information of the copper foil, and the snaking offset data of the center line of the copper foil relative to the baseline is calculated by the position detection unit; based on the snaking offset data, characteristic parameters are extracted by a deep neural network, and snaking correction parameters are obtained from a historical database as feedforward compensation; for the snaking offset data, a dynamic compensation signal is calculated by the feedback controller as feedback compensation; for the feedforward compensation and the feedback compensation, a fusion unit performs weighted calculation to obtain a workstation tension compensation instruction, and the tension control device adjusts the tension distribution of adjacent workstations.

[0031] Exemplarily, a photoelectric sensor array is used to detect the edge position of the copper foil in real time, and the offset distance between the center line of the copper foil and the reference line is calculated by the position detection unit, and the snaking offset data is output from the position detection unit. The snaking offset data is compared with the preset snaking warning threshold value. If the snaking offset exceeds the warning threshold range, the snaking alarm device is triggered and an alarm signal is output. The snaking offset data is feature extracted by a deep neural network, and the snaking correction parameters under similar working conditions are obtained from the historical database, and the feedforward compensation is output. The real-time snaking offset data is processed by a feedback controller, and the feedback compensation is calculated according to the offset change trend, and the dynamic compensation signal is output from the feedback controller. The feedforward compensation and the feedback compensation are weightedly fused, and the comprehensive compensation parameters are calculated by the fusion unit, and the workstation tension compensation instruction is output. The workstation tension compensation instruction is used to adjust the tension distribution of adjacent workstations, and the adjusted snaking offset is collected by the position detection unit. It is judged whether the adjusted snaking offset still exceeds the warning threshold. If it exceeds, it returns to perform the feedforward compensation operation until the snaking offset is within the warning threshold range. The snaking control of the copper foil production line involves a closed-loop adjustment process with multiple levels of linkage. Taking the actual working condition as an example, the photoelectric sensors are arranged on both sides of the production line. The detection range of the sensors is ±50mm, the resolution is 0.1mm, and the sampling frequency is 100Hz. The edge position of the copper foil is detected by 5 sets of sensor arrays, and the center line of the copper foil is fitted by the least squares method to calculate the offset distance from the baseline. The snaking warning threshold is set in a hierarchical manner. When the offset of the copper foil exceeds ±10mm, the first-level warning is triggered and the alarm signal is output; when the offset exceeds ±20mm, the second-level warning is triggered, and the deep neural network is started for feature extraction. The deep neural network adopts a 5-layer structure, including 3 convolutional layers and 2 fully connected layers. The input features include parameters such as offset sequence, running speed and tension distribution. The historical database stores the operation data of the past 6 months, and each record contains fields such as timestamp, offset, speed, tension, etc. Similar working conditions are retrieved through the feature matching algorithm, and the correction parameters are extracted as feedforward compensation. The feedforward compensation adopts proportional-differential control, with proportional coefficient Kp=0.8 and differential coefficient Kd=0.1. For an offset of 20mm, the calculated feedforward compensation is 16N. The feedback controller adopts a fuzzy control structure, with the input variables being the offset e and the offset change rate ec, and the output variable being the compensation u. The fuzzy rules are in the form of a two-dimensional lookup table, containing 49 rules. When the offset is detected to be 15mm and the offset rate is 2mm / s, the fuzzy output obtained by the lookup table is 0.6, and the feedback compensation 12N is obtained after defuzzification. The weighted fusion of the compensation adopts the adaptive weight method, and the weight coefficient is dynamically adjusted through the Kalman filter algorithm. The state equation w(k+1)=w(k)+v(k), where w(k) is the weight vector and v(k) is the process noise.The measurement equation is y(k)=Hw(k)+e(k), where H is the observation matrix and e(k) is the measurement noise. When the feedforward compensation effect is good, its weight coefficient is increased to 0.7, and the feedback compensation weight is reduced to 0.3. The station tension adjustment adopts inverter control, and the closed-loop speed regulation is realized through the PI controller. The control cycle is 50ms, the proportional coefficient Ki=0.5, and the integral time Ti=1s. In actual operation, when the copper foil snaking offset is gradually reduced from 20mm to 5mm, the dynamic adjustment process of the station tension shows good tracking characteristics and stability.

[0032] Step S106, based on the tension balance between adjacent workstations, the operating parameters of each workstation of the composite copper foil continuous production line are collected in real time to establish a mathematical description function of the production line operation.

[0033] The output signals of the tension sensor, the rotation speed sensor and the temperature sensor are obtained, and the original parameter data are output from the data acquisition unit; according to the original parameter data, a data preprocessing unit is used to perform normalization processing, and a standardized parameter data set is obtained from the preprocessing unit; for the standardized parameter data set, the tension size, copper foil speed and temperature parameters are subjected to data dimension reduction by a principal component analysis method, and a parameter correlation coefficient matrix is ​​obtained from a correlation analysis unit; a neural network is used to perform nonlinear mapping on the working condition characteristic parameters, and a state prediction value is obtained from a prediction unit, and the state prediction value is output from the prediction unit; a mathematical description function of the production line operation is established according to the state prediction value.

[0034] Exemplarily, a sensor array is used to collect the operating parameters of each station of the production line, and the output signals of the tension sensor, the speed sensor, and the temperature sensor are obtained respectively through a multi-channel data acquisition unit, and the original parameter data is output from the data acquisition unit. The original parameter data is normalized and outliers are eliminated by a data preprocessing unit, and a standardized parameter data set is generated from the preprocessing unit. According to the standardized parameter data set, the principal component analysis method is used to reduce the data dimension of the tension size, copper foil speed, and temperature parameters, and the parameter correlation coefficient matrix is ​​obtained from the correlation analysis unit. For the parameter correlation coefficient matrix, the characteristic decomposition method is used to extract the main influencing factors, and the working condition characteristic parameters are output from the feature extraction unit. A neural network is used to perform nonlinear mapping on the working condition characteristic parameters, and a state predictor containing hidden layer nodes is constructed, and the state prediction value is output from the prediction unit. A multivariate state space equation group is established according to the state prediction value, and the equation parameters are identified online by the recursive least squares method. The identified equation parameters are uploaded to the central control unit by a data communication unit, and the mathematical description function of the production line operation is formed by the control unit storage. The parameter modeling process of the copper foil production line involves data processing and analysis in multiple links. Multiple sensor arrays are arranged on the production line, including tension sensors, speed sensors and temperature sensors. The tension sensor adopts a strain gauge structure with a range of 0-500N and a sensitivity of 0.2mV / N; the speed sensor adopts a photoelectric encoder with a resolution of 2500 lines / turn; the temperature sensor adopts PT100 with a measurement range of 0-200℃ and an accuracy of 0.1℃. The sampling frequency is set to 100Hz and the data resolution is 16 bits. The preprocessing of the original data adopts a standardized method. For the measured value x, the standard score is calculated by the formula z=(x-μ) / σ, where μ is the mean and σ is the standard deviation. The 3σ criterion is used to determine the outlier. When the standard score of a certain measured value exceeds ±3, it is determined to be an outlier. Taking the data of a certain workstation as an example, the tension mean is 50N and the standard deviation is 2N. When the measured value is greater than 56N, it will be eliminated. In the process of principal component analysis, the correlation coefficient matrix R between the parameters is first calculated. Taking three parameters as an example, the correlation coefficient matrix R = [1.0, 0.8, 0.3; 0.8, 1.0, 0.4; 0.3, 0.4, 1.0] is obtained, in which the correlation coefficient between tension and speed is 0.8, showing a strong correlation. Eigenvalue decomposition obtains eigenvalues ​​λ1 = 2.1, λ2 = 0.6, λ3 = 0.3, and the corresponding eigenvectors are

[0035] v1=[0.707,0.707,0.000]、v2=[-0.408,0.408,0.816]、v3=[0.577,-0.577,0.577]The contribution rate of the first principal component reaches 70%, reflecting that tension-speed coupling is the main influencing factor. The neural network adopts a three-layer structure, the 3 nodes of the input layer correspond to the standardized parameters, the 8 nodes of the hidden layer use the hyperbolic tangent activation function, and the 3 nodes of the output layer correspond to the state prediction value. The training adopts the Levenberg-Marquardt algorithm with a learning rate of 0.01. The training is stopped when the error of the validation set does not decrease for 5 consecutive iterations. The state space equation x(k+1) adopts a discrete form, x(k+1)=Ax(k)+Bu(k), where the state vector x(k) contains three components of tension, speed and temperature, the control input u(k) is the torque command of the drive motor, and B is the control matrix, which describes how the control input u(k) affects the state variables of the system. The forgetting factor of the recursive least squares method is 0.95, and the initial covariance matrix P0 = 100I. After 1000 iterations, the state transfer matrix A = [0.95, 0.04, 0.01; 0.03, 0.96, 0.01; 0.01, 0.01, 0.98] is obtained, which reflects the dynamic coupling relationship between parameters. The data communication adopts the OPC UA protocol. The data frame contains three fields: timestamp, parameter value and quality stamp. The communication cycle is 50ms and is uploaded to the central control unit via Ethernet.

[0036] Step S107, optimizing the control parameters of the production line operation in combination with the preset rules in the preset knowledge base.

[0037] The preset rules in the preset knowledge base are parsed by the rule matcher to extract the parameter setting rules; the parameter setting rules are quantified by the fuzzy reasoner, and the parameter adjustment direction is obtained by membership calculation; according to the parameter adjustment direction, the action space is constructed by the deep reinforcement learning method, the tension deviation sequence and the control parameter combination are set as the state space, and the optimized control parameter group is output from the reinforcement learning unit.

[0038] Exemplarily, an optimization calculation unit is used to analyze the mathematical description function of the production line operation, the control parameters are binary-encoded by a genetic algorithm, a fitness function is constructed based on the tension stability index, and the optimization interval of the control parameters is output from the optimization calculation unit. According to the control parameter optimization interval, the empirical rules in the preset knowledge base are analyzed by a rule matcher, and expert experience data including parameter setting rules are extracted from the rule base. For the expert experience data, a fuzzy reasoner is used to quantify the parameter setting rules, and the parameter adjustment direction and adjustment step length are obtained by membership calculation. According to the parameter adjustment direction, deep reinforcement learning is used to construct the action space, the tension deviation sequence and the control parameter combination are set as the state space, and the optimized control parameter group is output from the reinforcement learning unit. The optimized control parameters are operated online by a proportional integral controller, and the tension output deviation is calculated by a closed-loop control method. For the tension output deviation, an evaluation index function including overshoot and adjustment time is established, and the parameter evaluation result is output from the evaluation calculation unit. The parameter evaluation result is compared with the preset optimization target. If the evaluation index does not reach the optimization target, it returns to execute the action space construction operation until the evaluation index meets the preset optimization target. The parameter optimization of the composite copper foil production line involves iterative calculations in multiple links. Taking the control parameter optimization as an example, the parameters are represented by 16-bit binary coding. The value range of the proportional coefficient Kp is

[0039] [0.1,2.0], the value range of the integral time Ti is [0.5s,5.0s]. The fitness function f adopts a weighted form, f = 0.6 / Mp + 0.4 / ts, where Mp is the overshoot and ts is the adjustment time. The rule base stores the parameter setting rules under typical working conditions, which are described in the form of If-Then. For example, when the tension deviation is positive and the deviation change rate is positive, the corresponding rule is Ife>0andde / dt>0, ThenΔKp=-0.1. Fuzzy reasoning adopts the Mamdani method, the input variable adopts a triangular membership function, the [-1,1] interval is divided into 5 equal parts, and the output adopts a single-valued membership function. The state space of deep reinforcement learning contains 8 dimensions, including the tension deviation sequence [e(k), e(k-1), e(k-2)], control parameters [Kp, Ti] and working condition parameters [v, T]. The action space is a two-dimensional continuous space, corresponding to the parameter increment [ΔKp, ΔTi]. The reward function r is designed as r = -|e|-0.1|de / dt|-0.01|u|, where e is the tension deviation and u is the control output. The controller adopts a cascade PI structure, with the inner loop for speed control and the outer loop for tension control. The control cycle of the speed loop is 10ms, and the control cycle of the tension loop is 50ms. When it is detected that the tension setting value jumps from 50N to 60N, the pre-optimization parameters Kp = 0.8 and Ti = 2.0s are used for control, and the overshoot Mp = 15% and the adjustment time ts = 3.5s are obtained. After 100 generations of genetic optimization iterations, the population size is set to 50, the crossover probability is 0.8, and the mutation probability is 0.05, and the optimized parameters Kp = 1.2 and Ti = 1.5s are obtained. Deep reinforcement learning adopts the Actor-Critic structure. Both the Actor network and the Critic network adopt a three-layer structure with 32 hidden nodes. After 10,000 steps of training, the parameters converged to Kp = 1.25 and Ti = 1.4s. The evaluation index function J = w1Mp + w2ts + w3*σe, where σe is the standard deviation of the steady-state error, and the weight coefficients [w1, w2, w3] = [0.4, 0.4, 0.2]. The optimized parameters were used for step response testing, and the overshoot was reduced to 8%, the adjustment time was shortened to 2.2s, and the standard deviation of the steady-state error was reduced from 0.8N to 0.3N. The optimization process formed a closed-loop optimization link of parameter encoding, rule matching, reinforcement learning and evaluation feedback.

[0040] The above describes the specific embodiments of the present invention. It should be understood that the present invention is not limited to the above specific embodiments, and those skilled in the art can make various changes or modifications within the scope of the claims, which does not affect the essence of the present invention. In the absence of conflict, the embodiments of the present application and the features in the embodiments can be combined with each other arbitrarily.

Claims

1. A control method for a tension adjustment device on a composite copper foil continuous production line, characterized in that: The method comprises: Obtaining the tension setting value and the actual tension value of adjacent workstations on the composite copper foil continuous production line, and calculating the difference between the tension setting value and the actual tension value to obtain tension deviation data; Combine the tension fluctuation data in the historical production data and the tension deviation data to predict the tension mutation that may occur under the current production conditions, and generate a feedforward compensation signal based on the prediction result to form a preliminary tension gradient correction signal; At the same time, combined with the tension deviation data and the real-time operation status data of the copper foil, the tension control parameters are dynamically adjusted to generate a feedback compensation signal, and the preliminary tension gradient correction signal is corrected to form a corrected tension gradient correction signal; According to the corrected tension gradient correction signal, it is judged whether the tension between adjacent workstations is balanced. If it is unbalanced, the corrected tension gradient correction signal is sent to the tension adjustment devices of the adjacent workstations respectively, and the tension between the workstations is dynamically adjusted by the adjustment device until the tension between the adjacent workstations reaches a balanced state; In the process of adjusting the tension between adjacent workstations, it is determined whether the copper foil is snaking. If snaking occurs, the copper foil snaking offset data is fed back to the feedforward compensation function and the feedback controller to coordinately adjust the tension control.

2. The method according to claim 1, characterized in that The method of obtaining the tension setting value and the actual tension value of adjacent workstations on the composite copper foil continuous production line, and calculating the difference between the tension setting value and the actual tension value to obtain tension deviation data includes: A station sensor array is used to collect tension monitoring data on the composite copper foil transmission line, and the station tension value is obtained by signal processing through a tension data acquisition unit; According to the comparison between the workstation tension value and a preset tension threshold range, position information is obtained from the optical encoder and converted into a digital value; If the workstation tension value exceeds a preset tension threshold range, a tension deviation value is acquired from the tension data acquisition unit and sent to a tension compensation unit.

3. The method according to claim 1, characterized in that The method combines the tension fluctuation data and the tension deviation data in the historical production data to predict the tension mutation that may occur under the current production conditions, and generates a feedforward compensation signal according to the prediction result to form a preliminary tension gradient correction signal, including: The tension deviation data is sampled using a sliding time window, and a filtered tension deviation sequence is obtained from the sampled data through a median filter; According to the tension deviation sequence and the tension fluctuation characteristic values ​​in the historical production database, the tension fluctuation law is obtained through long short-term memory network training; According to the tension fluctuation law, the tension change curve is fitted by the least square method to obtain the tension mutation warning threshold; A feedforward compensation function is constructed according to the tension mutation warning threshold, the feedforward compensation function is smoothed by an adaptive filter to obtain an initial compensation signal, and the initial compensation signal is optimized and calculated by a gradient descent method to obtain a tension gradient correction signal.

4. The method according to claim 1, characterized in that: The method combines the tension deviation data and the real-time operation status data of the copper foil, dynamically adjusts the tension control parameters, generates a feedback compensation signal, corrects the preliminary tension gradient correction signal, and forms a corrected tension gradient correction signal, including: Using a recursive least squares algorithm to perform online identification of copper foil operating parameters according to the real-time operating status data of the copper foil, the online identification results are used to construct a feedback compensator; Outputting a tension compensation signal through the feedback compensator, wherein the tension compensation signal is used to dynamically correct the tension gradient correction value; A neural network is used to adaptively adjust the compensation value after dynamic correction, and an optimized correction parameter is obtained from the neural network. The correction parameter is used to adjust the tension control parameter.

5. The method according to claim 1, characterized in that The method comprises: determining whether the tension between adjacent workstations is balanced according to the corrected tension gradient correction signal; if unbalanced, sending the corrected tension gradient correction signal to the tension adjustment devices of the adjacent workstations respectively, and dynamically adjusting the tension between the workstations by adjusting the devices until the tension between the adjacent workstations reaches a balanced state, including: A sensor array is used to collect the real-time tension values ​​of adjacent workstations; Calculate the tension difference between workstations according to the real-time tension value to obtain tension deviation data; Acquiring a correction signal through a tension gradient correction unit, wherein the correction signal is generated when the tension deviation data exceeds a preset tension balance threshold; In response to the correction signal, a workstation speed regulation instruction is output from the signal distributor, and the workstation speed regulation instruction is calculated by the signal distributor according to the speed regulation ratio of the transmission shaft of each workstation; A feedback controller is used to process the workstation speed regulation instruction, and a speed regulation control signal is output from the feedback controller. The speed regulation control signal is obtained by the feedback controller calculating a speed regulation compensation amount according to the workstation operation parameters.

6. The method according to claim 1, characterized in that In the process of adjusting the tension between adjacent workstations, it is determined whether the copper foil is snaking. If snaking occurs, the copper foil snaking offset data is fed back to the feedforward compensation function and the feedback controller to coordinately adjust the tension control, including: The photoelectric sensor array is used to obtain the edge position information of the copper foil, and the meandering offset data of the center line of the copper foil relative to the reference line is calculated by the position detection unit; According to the meandering offset data, characteristic parameters are extracted through a deep neural network, and meandering correction parameters are obtained from a historical database as feedforward compensation; With respect to the meandering offset data, a dynamic compensation signal is calculated by the feedback controller as a feedback compensation amount; A fusion unit performs weighted calculation on the feedforward compensation amount and the feedback compensation amount to obtain a workstation tension compensation instruction, and a tension control device adjusts the tension distribution of adjacent workstations.

7. The method according to claim 1, characterized in that The method further comprises: based on the tension balance between adjacent workstations, real-time collection of operating parameters of each workstation of the composite copper foil continuous production line, and establishment of a mathematical description function of the production line operation.

8. The method according to claim 7, characterized in that On the basis of the tension balance between adjacent workstations, the operation parameters of each workstation of the composite copper foil continuous production line are collected in real time, and a mathematical description function of the production line operation is established, including: Obtain output signals of the tension sensor, the rotation speed sensor and the temperature sensor, and output original parameter data from the data acquisition unit; According to the original parameter data, a data preprocessing unit is used to perform normalization processing, and a standardized parameter data set is obtained from the preprocessing unit; For the standardized parameter data set, data dimension reduction is performed on the tension magnitude, copper foil speed and temperature parameters by using a principal component analysis method, and a parameter correlation coefficient matrix is ​​obtained from a correlation analysis unit; A neural network is used to perform nonlinear mapping on the operating condition characteristic parameters, a state prediction value is obtained from a prediction unit, and the state prediction value is output from the prediction unit; The mathematical description function of production line operation is established according to the state prediction value.

9. The method according to claim 1, characterized in that: The method further includes: optimizing control parameters of the production line operation in combination with preset rules in a preset knowledge base.

10. The method according to claim 9, characterized in that The control parameters of the production line operation are optimized by combining the preset rules in the preset knowledge base, including: The preset rules in the preset knowledge base are parsed by the rule matcher to extract the parameter setting rules; The fuzzy inference device is used to quantify the parameter setting rules, and the parameter adjustment direction is obtained through membership calculation. According to the parameter adjustment direction, the deep reinforcement learning method is used to construct the action space, the tension deviation sequence and the control parameter combination are set as the state space, and the optimized control parameter group is output from the reinforcement learning unit.

Images