An industrial automation control method and system

Abstract

The application discloses an industrial automation control method and system, relates to the technical field of lithium battery production automation, and comprises a multi-source state acquisition module, a coating parameter-tension dynamic adaptation module, a drying-winding coordination time sequence optimization module and an execution control module. The coating parameter-tension dynamic adaptation module adopts an improved weighted least squares support vector machine algorithm to establish a dynamic matching relationship between the substrate tension and the coating parameter; and the drying-winding coordination time sequence optimization module realizes the dynamic coordination of the drying temperature, the air speed and the winding speed through an improved time sequence coordination algorithm. Through the cooperative work of the two creative modules, the application improves the uniformity of the pole piece coating thickness and the consistency of drying, guarantees the production quality of the lithium battery pole piece, and has strong industry pertinence and practical value.

Description

Technical Field

[0001] This invention relates to the field of industrial automation control, specifically to an industrial automation control method and system. Background Technology

[0002] Electrode coating is a core process in lithium battery production, requiring a coating machine, dryer, and winding machine to work in strict sequence: the coating machine evenly coats the copper / aluminum foil substrate, the dryer removes the solvent from the electrode, and the winding machine smoothly winds up the dried electrode. The uniformity of the electrode coating thickness and the consistency of drying directly determine the key performance characteristics of lithium batteries, such as energy density and cycle life.

[0003] The current automated control of lithium battery electrode coating production lines has specific problems: First, the adaptability of coating parameters is poor. Traditional systems use fixed parameters such as coating roller speed and coating pressure, which cannot cope with the tension fluctuations of the electrode substrate (such as sudden tension changes at the substrate joint) and changes in the solid content of the slurry (such as uneven slurry mixing), which easily leads to excessive electrode thickness deviation (the industry requires a thickness deviation of ≤±2μm, while traditional systems often have a deviation of 3-5μm). Second, the drying-winding coordination sequence is rigid. The temperature and wind speed of the dryer are matched with the winding speed of the winding machine using a preset sequence. When the coating thickness changes dynamically, the fixed sequence cannot guarantee the uniformity of electrode drying, which easily leads to local over-drying (electrode embrittlement) or under-drying (solvent residue). At the same time, the mismatch between the winding speed and the tension of the electrode after drying can also cause electrode wrinkles.

[0004] Existing technologies mostly employ simple proportional-integral-derivative (PID) algorithms to fine-tune parameters of individual devices, failing to achieve dynamic and coordinated control across the entire process of tension-coating-drying-winding. This approach cannot fundamentally solve the specific problems mentioned above, hindering the improvement of lithium battery electrode production quality. Therefore, there is an urgent need for an industrial automation control method and system for lithium battery electrode coating production lines. Summary of the Invention

[0005] The purpose of this invention is to provide an industrial automation control method and system to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: an industrial automation control method, comprising the following steps: Step 1: Collect real-time operating status data of the coating machine, dryer, and winding machine in the lithium battery electrode coating production line, including but not limited to electrode substrate tension, slurry solid content, coating roller speed, drying temperature, and winding speed. Step 2: Using the coating parameter-tension dynamic adaptation module, the real-time status data collected in Step 1 is processed based on the improved weighted least squares support vector machine algorithm to generate coating parameters that adapt to the current tension and slurry status. Step 3: Through the drying-rewinding coordinated timing optimization module, based on the improved timing coordination algorithm, and combined with coating parameters and electrode humidity data, dynamically adjust the drying machine operating parameters and the winding machine coordinated timing. Step 4: Based on the adjusted coating parameters, drying parameters, and coordination timing, perform closed-loop control operation on the coating-drying-rewinding equipment.

[0007] Preferably, the specific implementation steps of step 1 are as follows: Step 1.1 Deploy dedicated data acquisition equipment: Install tension sensors on the substrate transfer path of the coating machine, install online solids content detectors on the slurry supply pipeline, install encoders on the ends of the coating roller shafts, install temperature sensors in the drying chamber, and install encoders on the ends of the winding machine shafts; Step 1.2: Accurately collect target parameters: Simultaneously collect data on electrode substrate tension (N), slurry solid content (%), coating roller speed (r / min), drying temperature (°C), and winding speed (r / min) at a sampling frequency of 20Hz to obtain the collected data; Step 1.3, Data Preprocessing: The Kalman filter algorithm is used to filter noise in the collected data, eliminate vibration interference from the production environment, and output clean real-time status data to the coating parameter-tension dynamic adaptation module.

[0008] Preferably, the specific implementation logic of step 2 is as follows: Step 2.1, Feature Parameter Extraction: Features are extracted from the preprocessed data collected in Step 1 to determine the input feature vector. and output control vector Input feature vector ,in The real-time tension (N) of the electrode substrate. Real-time solids content of the slurry (%); Output control vector ,in The target rotational speed (r / min) of the coating roller. The target pressure for coating (MPa); Step 2.2, Improved WLSSVM Model Initialization: Traditional WLSSVM models assign consistent weights to different input features. This invention introduces the influence of slurry solids content on the weights. Optimize the model training process; weights The calculation method (in layman's terms: the greater the impact of changes in solid content on thickness, the higher the weight): In the process of improving the initialization of the WLSSVM model, the influence of slurry solid content on the weights is addressed. The calculation method (i.e., the interpretation is: the greater the impact of solid content changes on electrode thickness, the higher the weight of this parameter, making the model more focused on adapting to solid content fluctuations): ,in The weighting is determined by the solids content of the slurry; it has no unit and a value greater than 1. The greater the deviation of the solids content from the standard value, the higher the weighting. The larger; The solid content influence coefficient is an inherent coefficient of the equipment, calibrated through lithium battery electrode production experiments, with a value range of 1.2-1.5, used to adjust the degree of influence of solid content deviation on the weight. The real-time solid content of the material, expressed in %, is collected in real time by the online solid content detector of the multi-source status acquisition module. It is one of the core parameters affecting the coating thickness. The standard solid content of the material is the process standard value for lithium battery electrode production (the industry standard value is such as 50%), which is determined by the production formula. The absolute deviation between the real-time solid content and the standard solid content reflects the degree of fluctuation in the real-time solid content. Step 2.3, Model Training and Parameter Solving: This involves using historical production data, i.e., data containing different... , The corresponding optimal , The electrode thickness data were divided into training and validation sets, and then used to train the improved WLSSVM model. The optimization objective of the improved WLSSVM model is to minimize the electrode thickness deviation. Minimize, that is, make the electrode thickness corresponding to the predicted coating parameters reach the target thickness. The optimization objective function is: The constraints for optimizing the objective function are: ,in To optimize the objective function value, which is a comprehensive indicator representing the model's prediction error and complexity, the objective is to minimize... ; This is the model weight vector, used to adjust the degree of influence of the input feature vector on the output control vector; Weight vector The squared L2 norm is used to control model complexity and avoid overfitting. This is a regularization parameter with a value of 100, used to balance model fitting accuracy (reducing prediction error) and model complexity (avoiding overfitting). To determine the number of training samples, we selected an effective sample size (e.g., 1000 groups) from historical lithium battery electrode production data. The samples included coating parameters and corresponding electrode thickness data under different tensions and solid contents. For the first The influence weight of the slurry solid content of each training sample is calculated in step 2.2 and is used to increase the weight of samples with fluctuating solid content in model training. For the first The prediction error of each training sample. This reflects the deviation between the model's predicted values ​​and the actual process requirements; For the first The output value of each training sample is the optimal coating parameter (coating roller speed or coating pressure) under the corresponding input features, where the speed is in r / min and the pressure is in MPa. Weight vector Transpose of; For the input feature vector The high-dimensional mapping function is used to map linearly inseparable "tension-solid content" feature data to a high-dimensional space to achieve linear fitting; for The input feature vector of each training sample, i.e. ,in For the first Real-time tension of the substrate for each sample. For the first Real-time solids content of slurry for each sample; The bias term is used to adjust the model's fitting baseline and improve the model's prediction accuracy. The target thickness of the electrode is in μm, which is the standard value for lithium battery electrode production (e.g., 120μm), determined by product design requirements. The optimal model parameters are obtained by solving the objective function using the Lagrange multiplier method. and Complete the training of the improved WLSSVM model; Step 2.4, Output Adaptive Coating Parameters: Real-time Acquisition Input a pre-trained improved WLSSVM model, and the model outputs coating parameters adapted to the current state. The data is then transmitted to the drying-rewinding coordinated timing optimization module and the execution control module.

[0009] Preferably, the specific implementation steps of step 3 are as follows: Step 3.1, Basic Data Acquisition: Receive coating parameters - coating parameters output by the tension dynamic adaptation module. By combining real-time data from the multi-source status acquisition module, key intermediate parameters are calculated: a. Theoretical thickness The calculations are used to predict the electrode thickness based on real-time coating parameters, providing a basis for subsequent drying parameter adjustments. ,in, This represents the theoretical coating thickness of the electrode, expressed in μm. It is a theoretical value calculated based on real-time coating parameters and is used to guide the adaptation of drying-winding parameters. The coating thickness coefficient is 1, and the unit is 1. , is the inherent coefficient of the equipment, which is determined through experiments using lithium battery electrode coating equipment, and characterizes the comprehensive influence coefficient of slurry solid content and coating roller speed on coating thickness; The coating thickness coefficient is 2, and the unit is... , is the inherent coefficient of the equipment, which is determined through equipment experiments and characterizes the attenuation coefficient of coating pressure on coating thickness; This refers to the real-time solids content of the slurry, expressed as a percentage. Positive correlation: the higher the solid content, the greater the coating thickness under the same rotation speed and pressure. The target rotational speed of the coating roller, in units of Output from the coating parameter-tension dynamic adaptation module, and Negative correlation: the faster the rotation speed, the thinner the coating thickness. The target pressure for coating is expressed in units of... Output from the coating parameter-tension dynamic adaptation module, and Negative correlation: the greater the pressure, the thinner the coating thickness. b. Current tension of the winding machine The calculations are used to monitor the tension of the electrode sheet in real time during the winding process, preventing electrode sheet wrinkles caused by tension fluctuations. ,in The real-time tension of the electrode sheet during winding, measured in N, is a key parameter to ensure flat winding of the electrode sheet and must be controlled within the standard tension range. nearby; This is the tension coefficient for the winding speed, in units of... , is the inherent coefficient of the equipment, determined through experiments on the winding machine equipment; , represents the degree of influence of the winding speed on the tension. This is the winding torque tension coefficient, in units of... The coefficient is determined through equipment testing and is the inherent coefficient of the equipment, representing the degree of attenuation of tension by the winding torque. The current speed of the winding machine, in units of The data is collected in real time by the encoder of the multi-source status acquisition module; The output torque of the winding machine, in units of , is the core operating parameter of the winding machine, and is negatively correlated with winding tension; Step 3.2: Construct an improved time-series coordination objective function: based on electrode drying uniformity. and winding tension stability To achieve the dual optimization objectives, a dynamic adjustment factor is introduced. Adjust the priorities of the two objectives (i.e., prioritize ensuring uniform drying when drying is uneven, and prioritize ensuring tension stability when tension fluctuations are large), and the objective function is (minimize the objective function value, i.e., simultaneously ensure uniform drying and tension stability): ,in This function optimizes a dual objective value, is dimensionless, and comprehensively reflects both drying uniformity and tension stability; the objective is to minimize... ; Electrode drying uniformity The calculation reflects the overall drying consistency of the electrode by measuring the humidity deviation at multiple detection points. ,in Electrode drying uniformity, unit: , which is the average deviation of the humidity at each detection point from the standard humidity; The number of humidity detection points at the dryer outlet is not specified; in this embodiment, five points are used, evenly arranged along the width of the electrode to ensure comprehensive detection of the drying condition. The detection point number has a specified value. ; For the first Real-time humidity of the electrode at each detection point, in units of The humidity data is collected in real time by the humidity sensor in the multi-source status acquisition module. Standard dry humidity of the electrode, unit: This is the standard value for the production process of lithium battery electrodes (industry standard value such as 8%RH). Meeting the humidity standard can avoid solvent residue from affecting the performance of lithium batteries. For the first The absolute deviation of humidity at each detection point from the standard humidity, in units of... This reflects the drying deviation at a single detection point; Winding tension stability The calculation uses the sum of squared tension deviations from multiple sampling points to reflect the degree of tension fluctuation during the winding process. ,in For winding tension stability, the unit is... The smaller the value, the smaller the tension fluctuation, and the less likely the electrode is to wrinkle; This represents the number of tension sampling points, which has no unit. In this embodiment, 20 points are used, representing the number of consecutive sampling points within one winding cycle. The sampling point number is used to determine the value. ; For the first The winding tension at each sampling point, in N, is collected in real time by the tension sensor of the multi-source state acquisition module; The standard tension for winding is expressed in N, and is the standard value for the winding process of lithium battery electrode sheets (e.g., 5N) to ensure that the electrode sheets are wound flat and not stretched or damaged. For the first The square of the deviation between the tension at each sampling point and the standard tension, in N², amplifies the effect of tension fluctuations, making the model more focused on suppressing large fluctuations. Dynamic adjustment factor The calculation enables adaptive adjustment of the optimization priorities for drying uniformity and tension stability: ,in This represents the highest historical drying unevenness, measured in units of... Recorded during system operation Maximum value (e.g.) This reflects the maximum possible drying deviation in the production of lithium battery electrodes; The current electrode drying uniformity is expressed in units of... ; This represents the historical maximum tension fluctuation value, in units of Recorded during system operation Maximum value (e.g.) This reflects the maximum tension fluctuation that may occur during the production of lithium battery electrodes; when When the molecular weight increases (with a prominent issue of uneven drying), The rate of decrease has slowed. Increase the optimization priority of improving drying uniformity; when When the denominator increases (with tension fluctuations becoming more prominent), Increase Prioritize reducing and improving tension stability. Step 3.3, Solving the timing adjustment scheme: Based on the improved particle swarm optimization algorithm, the optimal solution of the dual objective function shown by the improved timing coordination objective function is solved, and finally the target parameters (temperature) of the dryer adapted to the current coating state are obtained. Wind speed and the target speed of the winding machine This ensures that the drying-rewinding process is coordinated with the dynamic matching of coating parameters. Step 3.4, Output timing optimization instructions: Output the obtained dryer parameters ( , and winding speed The commands are converted into timing control instructions and transmitted to the execution control module to achieve dynamic coordination of drying-winding and coating parameters.

[0010] Preferably, the specific implementation logic for solving the timing adjustment scheme in step 3.3 is as follows: Step 3.31: Determine the optimization variables and constraints: Based on the requirements of lithium battery electrode production process and equipment operating limits, the optimization variable is determined to be the target temperature of the dryer. (°C), Target wind speed and the target speed of the winding machine (r / min); Simultaneously, strict equipment parameter constraints are set: drying temperature constraint: (Too low a temperature will result in solvent residue, while too high a temperature will easily cause the electrode to become brittle); Winding speed constraint: ,in The coating roller target speed output by the coating parameter-tension dynamic adaptation module ensures that the winding speed matches the coating speed, avoiding electrode stretching or accumulation; wind speed constraint: (Too low a wind speed affects drying efficiency, while too high a wind speed can easily cause electrode misalignment). Step 3.32, Improved Particle Swarm Initialization: Each particle in the particle swarm corresponds to a set of candidate solutions for optimization variables, i.e. To avoid the initial candidate solutions exceeding the operating range of the equipment and to improve the convergence efficiency of the algorithm, the initialization phase directly generates a particle population randomly within the range of the constraints (population size is set to 50, iteration number is set to 100). The initial position of each particle satisfies the above temperature, velocity, and wind speed constraints to ensure the engineering feasibility of the initial solution. Step 3.33, Fitness Function Definition: [The text abruptly ends here, likely due to an incomplete sentence or a formatting error.] Dual objective optimization function As the fitness function of particle swarm optimization, i.e. ,in (Drying uniformity) (Tension stability) (Dynamic adjustment factors) are all calculated in step 3.2 above; the smaller the fitness value, the stronger the corresponding particle. The combination can better meet the requirements of uniform drying and stable tension; Step 3.34, Particle Iterative Update: An improved formula for updating particle velocity and position is used to achieve population evolution. Speed ​​update formula: ,in The inertial weight is linearly decreasing, with an initial value of 0.9 and a final value of 0.4 at the end of the iteration, balancing global search and local optimization. 1. The learning factor is 2. , for Random numbers, For the first The optimal position of a single particle in the next iteration. The optimal position for the population; Position update formula: After the update, a constraint check is performed on the particle position. If it exceeds the constraint range of step 3.31, it is corrected to the constraint boundary (e.g., ...). When the calculated value is 125℃, it should be corrected to 120℃. Step 3.35, Optimal Solution Determination and Output: During the iteration process, the optimal fitness value of a single particle is recorded in each round. ) and the population's optimal fitness value ( When the iteration reaches the preset number of iterations (100 times) or the optimal fitness value of the population remains unchanged for 10 consecutive rounds, the iteration stops, and the optimal position of the population is determined. corresponding This is the optimal timing adjustment parameter, ensuring that the combination of parameters can maximize both the uniformity of electrode drying and the stability of winding tension within the safe operating range of the equipment.

[0011] Preferably, an industrial automation control system includes a multi-source status acquisition module for acquiring real-time operating status data of the coating machine, dryer, and winding machine, as well as electrode quality inspection data; a coating parameter-tension dynamic adaptation module for establishing a dynamic matching relationship between electrode substrate tension, slurry solid content, and coating parameters based on an improved weighted least squares support vector machine algorithm, and outputting adapted coating parameters; a drying-winding coordinated timing optimization module for dynamically adjusting dryer operating parameters and winding machine coordinated timing based on an improved timing coordination algorithm, combined with coating parameters and electrode humidity data; and an execution control module for receiving various optimized parameters and timing instructions, and driving the coating-drying-winding equipment to operate collaboratively.

[0012] Compared with the prior art, the beneficial effects of the present invention are: The coating parameter-tension dynamic adaptation module of this invention introduces an improved WLSSVM algorithm that incorporates the influence weight of slurry solid content, which can accurately match the dynamic changes of substrate tension and slurry solid content, resulting in high precision control of electrode thickness deviation and significantly improving electrode thickness uniformity. The drying-winding coordinated timing optimization module of this invention achieves dynamic coordination of parameter timing with dual objectives and dynamic adjustment factors, which improves the drying uniformity of the electrode sheet, reduces the fluctuation of winding tension, and avoids problems such as over-drying, under-drying, and electrode sheet wrinkling. This invention is specifically designed for lithium battery electrode coating industrial production lines. The collected parameters and algorithm models are closely aligned with the physical characteristics of lithium battery slurry coating and electrode drying, making it highly targeted and able to directly improve the energy density and cycle life of lithium batteries. The multi-source status acquisition module and execution control module of this invention adopt mature industrial sensors and PLC equipment, and the development cost is controllable. The algorithms of the two innovative modules can be integrated into the existing production line through software upgrades without large-scale equipment modification, making them easy to promote in the industry. Attached Figure Description

[0013] Figure 1 This is a schematic diagram of the method flow of the present invention; Figure 2This is a schematic diagram illustrating the specific workflow of step 1 of the present invention; Figure 3 This is a schematic diagram illustrating the specific workflow of step 2 of the present invention; Figure 4 This is a schematic diagram illustrating the specific workflow of step 3 of the present invention; Figure 5 This is a schematic diagram of the system structure of the present invention. Detailed Implementation

[0014] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0015] Example 1 Please see Figure 1-4 The present invention provides a technical solution: an industrial automation control method, which includes the following steps: Step 1: Collect real-time operating status data of the coating machine, dryer, and winding machine in the lithium battery electrode coating production line, including but not limited to electrode substrate tension, slurry solid content, coating roller speed, drying temperature, and winding speed; the specific implementation steps are as follows: Step 1.1 Deploy dedicated data acquisition equipment: Install tension sensors on the substrate transfer path of the coating machine, install online solids content detectors on the slurry supply pipeline, install encoders on the ends of the coating roller shafts, install temperature sensors in the drying chamber, and install encoders on the ends of the winding machine shafts; Step 1.2: Accurately collect target parameters: Simultaneously collect data on electrode substrate tension (N), slurry solid content (%), coating roller speed (r / min), drying temperature (°C), and winding speed (r / min) at a sampling frequency of 20Hz to obtain the collected data; Step 1.3, Data Preprocessing: A Kalman filter algorithm is used to filter noise from the collected data, eliminating interference from vibrations in the production environment, and outputting clean, real-time status data to the coating parameter-tension dynamic adaptation module. Step 2: Using the coating parameter-tension dynamic adaptation module, the real-time status data collected in Step 1 is processed based on an improved weighted least squares support vector machine algorithm to generate coating parameters adapted to the current tension and slurry state; the specific implementation logic is as follows: Step 2.1, Feature Parameter Extraction: Features are extracted from the preprocessed data collected in Step 1 to determine the input feature vector. and output control vector Input feature vector ,in The real-time tension (N) of the electrode substrate. Real-time solids content of the slurry (%); Output control vector ,in The target rotational speed (r / min) of the coating roller. The target pressure for coating (MPa); Step 2.2, Improved WLSSVM Model Initialization: Traditional WLSSVM models assign consistent weights to different input features. This invention introduces the influence of slurry solids content on the weights. Optimize the model training process; weights The calculation method (in layman's terms: the greater the impact of changes in solid content on thickness, the higher the weight): In the process of improving the initialization of the WLSSVM model, the influence of slurry solid content on the weights is addressed. The calculation method (i.e., the interpretation is: the greater the impact of solid content changes on electrode thickness, the higher the weight of this parameter, making the model more focused on adapting to solid content fluctuations): ,in The weighting is determined by the solids content of the slurry; it has no unit and a value greater than 1. The greater the deviation of the solids content from the standard value, the higher the weighting. The larger; The solid content influence coefficient is an inherent coefficient of the equipment, calibrated through lithium battery electrode production experiments, with a value range of 1.2-1.5. In this embodiment, it is set to 1.3. It has no unit and is used to adjust the degree of influence of solid content deviation on the weight. The real-time solid content of the material, expressed in %, is collected in real time by the online solid content detector of the multi-source status acquisition module. It is one of the core parameters affecting the coating thickness. The standard solid content of the material is expressed as a percentage, which is the process standard value for lithium battery electrode production (the industry standard value is such as 50%), and is determined by the production formula. The absolute deviation between the real-time solid content and the standard solid content is expressed as a percentage, reflecting the degree of fluctuation in the real-time solid content. Step 2.3, Model Training and Parameter Solving: This involves using historical production data, i.e., data containing different... , The corresponding optimal , The electrode thickness data were divided into training and validation sets, and then used to train the improved WLSSVM model. The optimization objective of the improved WLSSVM model is to minimize the electrode thickness deviation. Minimize, that is, make the electrode thickness corresponding to the predicted coating parameters reach the target thickness. The optimization objective function is: The constraints for optimizing the objective function are: ,in To optimize the objective function value, which is a comprehensive indicator representing the model's prediction error and complexity, the objective is to minimize... ; This is the model weight vector, used to adjust the degree of influence of the input feature vector on the output control vector; Weight vector The squared L2 norm is used to control model complexity and avoid overfitting. This is a regularization parameter with a value of 100, used to balance model fitting accuracy (reducing prediction error) and model complexity (avoiding overfitting). To determine the number of training samples, we selected an effective sample size (e.g., 1000 groups) from historical lithium battery electrode production data. The samples included coating parameters and corresponding electrode thickness data under different tensions and solid contents. For the first The influence weight of the slurry solid content of each training sample is calculated in step 2.2 and is used to increase the weight of samples with fluctuating solid content in model training. For the first The prediction error of each training sample. This reflects the deviation between the model's predicted values ​​and the actual process requirements; For the first The output value of each training sample is the optimal coating parameter (coating roller speed or coating pressure) under the corresponding input features, where the speed is in r / min and the pressure is in MPa. Weight vector Transpose of; For the input feature vector The high-dimensional mapping function is used to map linearly inseparable "tension-solid content" feature data to a high-dimensional space to achieve linear fitting; for The input feature vector of each training sample, i.e. ,in For the first Real-time tension of the substrate for each sample. For the first Real-time solids content of slurry for each sample; The bias term is used to adjust the model's fitting baseline and improve the model's prediction accuracy. The target thickness of the electrode is in μm, which is the standard value for lithium battery electrode production (e.g., 120μm), determined by product design requirements. The optimal model parameters are obtained by solving the objective function using the Lagrange multiplier method. and Complete the training of the improved WLSSVM model; Step 2.4, Output Adaptive Coating Parameters: Real-time Acquisition Input a pre-trained improved WLSSVM model, and the model outputs coating parameters adapted to the current state. The data is then transmitted to the drying-rewinding coordinated timing optimization module and the execution control module. Step 3: Through the drying-rewinding coordinated timing optimization module, based on the improved timing coordination algorithm and combined with coating parameters and electrode humidity data, dynamically adjust the dryer operating parameters and the winding machine coordinated timing; the specific implementation steps are as follows: Step 3.1, Basic Data Acquisition: Receive coating parameters - coating parameters output by the tension dynamic adaptation module. By combining real-time data from the multi-source status acquisition module, key intermediate parameters are calculated: a. Theoretical thickness The calculations are used to predict the electrode thickness based on real-time coating parameters, providing a basis for subsequent drying parameter adjustments. ,in, This represents the theoretical coating thickness of the electrode, expressed in μm. It is a theoretical value calculated based on real-time coating parameters and is used to guide the adaptation of drying-winding parameters. The coating thickness coefficient is 1, and the unit is 1. , is the inherent coefficient of the equipment, which is calibrated through experiments on lithium battery electrode coating equipment. In this embodiment, it is taken as 0.8, which characterizes the comprehensive influence coefficient of slurry solid content and coating roller speed on coating thickness; The coating thickness coefficient is 2, and the unit is... , is the inherent coefficient of the equipment, which is calibrated through equipment experiments. In this embodiment, it is taken as 0.05, which characterizes the attenuation coefficient of coating pressure on coating thickness; For real-time solids content of slurry, and A positive correlation exists; the higher the solid content, the greater the coating thickness under the same rotation speed and pressure. The target rotational speed of the coating roller, in units of Output from the coating parameter-tension dynamic adaptation module, and Negative correlation: the faster the rotation speed, the thinner the coating thickness. The target pressure for coating is expressed in units of... Output from the coating parameter-tension dynamic adaptation module, and Negative correlation: the greater the pressure, the thinner the coating thickness. b. Current tension of the winding machine The calculations are used to monitor the tension of the electrode sheet in real time during the winding process, preventing electrode sheet wrinkles caused by tension fluctuations. ,in The real-time tension of the electrode sheet during winding, measured in N, is a key parameter to ensure flat winding of the electrode sheet and must be controlled within the standard tension range. nearby; This is the tension coefficient for the winding speed, in units of... , is the inherent coefficient of the equipment, which is calibrated through experiments on the winding machine equipment. In this embodiment, it is taken as 0.02, which represents the degree of influence of the winding speed on the tension; This is the winding torque tension coefficient, in units of... , is the inherent coefficient of the equipment, which is calibrated through equipment experiments. In this embodiment, it is taken as 0.01, representing the degree of attenuation of tension by the winding torque; The current speed of the winding machine, in units of The data is collected in real time by the encoder of the multi-source status acquisition module; The output torque of the winding machine, in units of , is the core operating parameter of the winding machine, and is negatively correlated with winding tension; Step 3.2: Construct an improved time-series coordination objective function: based on electrode drying uniformity. and winding tension stability To achieve the dual optimization objectives, a dynamic adjustment factor is introduced. Adjust the priorities of the two objectives (i.e., prioritize ensuring uniform drying when drying is uneven, and prioritize ensuring tension stability when tension fluctuations are large), and the objective function is (minimize the objective function value, i.e., simultaneously ensure uniform drying and tension stability): ,in This function optimizes a dual objective value, is dimensionless, and comprehensively reflects both drying uniformity and tension stability; the objective is to minimize... ; Electrode drying uniformity The calculation reflects the overall drying consistency of the electrode by measuring the humidity deviation at multiple detection points. ,in Electrode drying uniformity, unit: , which is the average deviation of the humidity at each detection point from the standard humidity; This refers to the number of humidity detection points at the dryer outlet, with no unit. In this embodiment, 5 points are used, evenly arranged along the width of the electrode to ensure comprehensive detection of the drying condition. The detection point number has a specified value. ; For the first Real-time humidity of the electrode at each detection point, in units of The humidity data is collected in real time by the humidity sensor in the multi-source status acquisition module; Standard dry humidity of the electrode, unit: This is the standard value for the production process of lithium battery electrodes (industry standard value such as 8%RH). Meeting the humidity standard can avoid solvent residue from affecting the performance of lithium batteries. For the first The absolute deviation of humidity at each detection point from the standard humidity, in units of... This reflects the drying deviation at a single detection point; Winding tension stability The calculation uses the sum of squared tension deviations from multiple sampling points to reflect the degree of tension fluctuation during the winding process. ,in For winding tension stability, the unit is... The smaller the value, the smaller the tension fluctuation, and the less likely the electrode is to wrinkle; This represents the number of tension sampling points, which has no unit. In this embodiment, 20 points are used, representing the number of consecutive sampling points within one winding cycle. The sampling point number is used to determine the value. ; For the first The winding tension at each sampling point, in N, is collected in real time by the tension sensor of the multi-source state acquisition module; The standard tension for winding is expressed in N, and is the standard value for the winding process of lithium battery electrode sheets (e.g., 5N) to ensure that the electrode sheets are wound flat and not stretched or damaged. For the first The square of the deviation between the tension at each sampling point and the standard tension, in N², amplifies the effect of tension fluctuations, making the model more focused on suppressing large fluctuations. Dynamic adjustment factor The calculation enables adaptive adjustment of the optimization priorities for drying uniformity and tension stability: ,in This represents the highest historical drying unevenness, measured in units of... Recorded during system operation Maximum value (e.g.) This reflects the maximum possible drying deviation in the production of lithium battery electrodes; The current electrode drying uniformity is expressed in units of... ; This represents the historical maximum tension fluctuation value, in units of Recorded during system operation Maximum value (e.g.) This reflects the maximum tension fluctuation that may occur during the production of lithium battery electrodes; when When the molecular weight increases (with a prominent issue of uneven drying), The rate of decrease has slowed. Increase the optimization priority of improving drying uniformity; when When the denominator increases (with tension fluctuations becoming more prominent), Increase Prioritize reducing and improving tension stability. Step 3.3, Solving the timing adjustment scheme: Based on the improved particle swarm optimization algorithm, the optimal solution of the dual objective function shown by the improved timing coordination objective function is solved, and finally the target parameters (temperature) of the dryer adapted to the current coating state are obtained. Wind speed and the target speed of the winding machine This ensures coordinated drying-rewinding and dynamic matching of coating parameters; the specific implementation logic is as follows: Step 3.31: Determine the optimization variables and constraints: Based on the requirements of lithium battery electrode production process and equipment operating limits, the optimization variable is determined to be the target temperature of the dryer. (°C), Target wind speed and the target speed of the winding machine ( At the same time, strict equipment parameter constraints are set: drying temperature constraint: (Too low a temperature will result in solvent residue, while too high a temperature will easily cause the electrode to become brittle); Winding speed constraint: ,in The coating roller target speed output by the coating parameter-tension dynamic adaptation module ensures that the winding speed matches the coating speed, avoiding electrode stretching or accumulation; wind speed constraint: (Too low a wind speed affects drying efficiency, while too high a wind speed can easily cause electrode misalignment). Step 3.32, Improved Particle Swarm Initialization: Each particle in the particle swarm corresponds to a set of candidate solutions for optimization variables, i.e. To avoid the initial candidate solutions exceeding the operating range of the equipment and to improve the convergence efficiency of the algorithm, the initialization phase directly generates a particle population randomly within the range of the constraints (population size is set to 50, iteration number is set to 100). The initial position of each particle satisfies the above temperature, velocity, and wind speed constraints to ensure the engineering feasibility of the initial solution. Step 3.33, Fitness Function Definition: [The text abruptly ends here, likely due to an incomplete sentence or a Dual objective optimization function As the fitness function of particle swarm optimization, i.e. ,in (Drying uniformity) (Tension stability) (Dynamic adjustment factors) are all calculated in step 3.2 above; the smaller the fitness value, the stronger the corresponding particle. The combination can better meet the requirements of uniform drying and stable tension; Step 3.34, Particle Iterative Update: An improved formula for updating particle velocity and position is used to achieve population evolution. Speed ​​update formula: ,in The inertial weight is linearly decreasing, with an initial value of 0.9 and a final value of 0.4 at the end of the iteration, balancing global search and local optimization. 1. The learning factor is 2. , for Random numbers, For the first The optimal position of a single particle in the next iteration. The optimal position for the population; Position update formula: After the update, a constraint check is performed on the particle position. If it exceeds the constraint range of step 3.31, it is corrected to the constraint boundary (e.g., ...). When the calculated value is 125℃, it should be corrected to 120℃. Step 3.35, Optimal Solution Determination and Output: During the iteration process, the optimal fitness value of a single particle is recorded in each round. ) and the population's optimal fitness value ( When the iteration reaches the preset number of iterations (100 times) or the optimal fitness value of the population remains unchanged for 10 consecutive rounds, the iteration stops, and the optimal position of the population is determined. corresponding This is the optimal timing adjustment parameter, ensuring that the combination of parameters can maximize both the uniformity of electrode drying and the stability of winding tension within the safe operating range of the equipment.

[0016] Step 3.4, Output timing optimization instructions: Output the obtained dryer parameters ( , and winding speed The commands are converted into timing control instructions and transmitted to the execution control module to achieve dynamic coordination of drying-winding and coating parameters.

[0017] Step 4: Based on the adjusted coating parameters, drying parameters, and coordination timing, perform closed-loop control operations on the coating-drying-rewinding equipment. As the core of the closed-loop control process of this invention, it receives output commands from the coating parameter-tension dynamic adaptation module and the drying-rewinding coordination timing optimization module. Simultaneously, it links with a multi-source status acquisition module to achieve status feedback, ensuring precise coordinated operation of all equipment. Its core configuration is an industrial-grade PLC (e.g., Siemens S7-1500), possessing efficient data processing and command-driven capabilities. It achieves high-speed data transmission via the Profinet bus protocol, accurately receiving coating parameters such as coating roller speed and coating pressure, drying parameters such as drying temperature and wind speed, as well as winding speed and timing commands. These parameters are converted into drive signals recognizable by the equipment, controlling the coating machine servo motor, dryer heating element / fan, and winding machine servo motor to operate collaboratively according to the optimized timing. Meanwhile, during equipment operation, this module collects real-time operating status data of each device (such as actual speed and temperature) and feeds it back to the multi-source status acquisition module for preprocessing. This provides real-time data support for the two preceding creative modules to re-optimize parameters, ultimately forming a closed-loop control process of "acquisition-optimization-execution-feedback".

[0018] Example 2 Please see Figure 5 An industrial automation control system includes a multi-source status acquisition module for acquiring real-time operating status data of the coating machine, dryer, and winding machine, as well as electrode quality inspection data; a coating parameter-tension dynamic adaptation module for establishing a dynamic matching relationship between electrode substrate tension, slurry solid content, and coating parameters based on an improved weighted least squares support vector machine algorithm, and outputting adapted coating parameters; a drying-winding coordinated timing optimization module for dynamically adjusting dryer operating parameters and winding machine coordinated timing based on an improved timing coordination algorithm and combining coating parameters and electrode humidity data; and an execution control module for receiving various optimized parameters and timing instructions and driving the coating-drying-winding equipment to operate collaboratively.

[0019] This invention discloses an industrial automation control method and system, relating to the field of lithium battery production automation technology. It aims to solve specific problems in lithium battery electrode coating production lines, such as uneven electrode thickness and unstable drying efficiency caused by fluctuations in electrode substrate tension and changes in slurry solid content during the multi-equipment collaborative process of coating, drying, and winding, which consequently affect lithium battery performance. This invention includes a multi-source state acquisition module, a coating parameter-tension dynamic adaptation module, a drying-winding collaborative timing optimization module, and an execution control module. The coating parameter-tension dynamic adaptation module employs an improved weighted least squares support vector machine algorithm to establish a dynamic matching relationship between substrate tension and coating parameters. The drying-winding collaborative timing optimization module uses an improved timing coordination algorithm to achieve dynamic coordination of drying temperature, air velocity, and winding speed. Through the collaborative work of these two innovative modules, this invention improves the uniformity of electrode coating thickness and drying consistency, ensuring the quality of lithium battery electrode production, and has strong industry relevance and practical value.

[0020] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An industrial automation control method, characterized in that, Includes the following steps: Step 1: Collect real-time operating status data of the coating machine, dryer, and winding machine in the lithium battery electrode coating production line, including but not limited to electrode substrate tension, slurry solid content, coating roller speed, drying temperature, and winding speed. Step 2: Using the coating parameter-tension dynamic adaptation module, the real-time status data collected in Step 1 is processed based on the improved weighted least squares support vector machine algorithm to generate coating parameters that adapt to the current tension and slurry status. Step 3: Through the drying-rewinding coordinated timing optimization module, based on the improved timing coordination algorithm, and combined with coating parameters and electrode humidity data, dynamically adjust the drying machine operating parameters and the winding machine coordinated timing. Step 4: Based on the adjusted coating parameters, drying parameters, and coordination timing, perform closed-loop control operation on the coating-drying-rewinding equipment.

2. The industrial automation control method according to claim 1, characterized in that: The specific implementation steps of step 1 are as follows: Step 1.1 Deploy dedicated data acquisition equipment: Install tension sensors on the substrate transfer path of the coating machine, install online solids content detectors on the slurry supply pipeline, install encoders on the ends of the coating roller shafts, install temperature sensors in the drying chamber, and install encoders on the ends of the winding machine shafts; Step 1.2, Accurately collect target parameters: Simultaneously collect electrode substrate tension, slurry solid content, coating roller speed, drying temperature, and winding speed at a sampling frequency of 20Hz to obtain the collected data; Step 1.3, Data Preprocessing: The Kalman filter algorithm is used to filter noise in the collected data, eliminate vibration interference from the production environment, and output clean real-time status data to the coating parameter-tension dynamic adaptation module.

3. The industrial automation control method according to claim 1, characterized in that: The specific implementation logic of step 2 is as follows: Step 2.1, Feature Parameter Extraction: Extract features from the preprocessed data collected in Step 1 to determine the input feature vector. and output control vector Input feature vector ,in For the real-time tension of the electrode substrate, Real-time solids content of the slurry; Output control vector ,in The target rotational speed of the coating roller. The target pressure for coating; Step 2.2, Improved WLSSVM model initialization: Introduce the influence weights of slurry solids content. Optimize the model training process; weights Calculation method: In the process of improving the initialization of the WLSSVM model, the influence of slurry solid content on the weights is addressed. Calculation method: ,in The weighting is determined by the solid content of the slurry; the value is greater than 1, indicating that the solid content deviates more from the standard value. The larger; The solid content influence coefficient is calibrated through lithium battery electrode production experiments, with a value range of 1.2-1.

5. In this embodiment, it is set to 1.

3. It has no unit and is used to adjust the degree of influence of solid content deviation on the weight. The real-time solid content of the material is collected in real time by the online solid content detector of the multi-source status acquisition module; The standard solid content of the material is a process standard value for the production of lithium battery electrode sheets, determined by the production formula. The absolute deviation between the real-time solid content and the standard solid content reflects the degree of fluctuation in the real-time solid content. Step 2.3, Model Training and Parameter Solving: This involves using historical production data, i.e., data containing different... , The corresponding optimal , The electrode thickness data are divided into training set and validation set, and then substituted into the improved WLSSVM model for training. The optimization objective of the improved WLSSVM model is to minimize the electrode thickness deviation. Minimize, that is, make the electrode thickness corresponding to the predicted coating parameters reach the target thickness. The optimization objective function is: The constraints for optimizing the objective function are: ,in To optimize the objective function value, which is a comprehensive indicator representing the model's prediction error and complexity, the objective is to minimize... ; This is the model weight vector, used to adjust the degree of influence of the input feature vector on the output control vector; Weight vector The squared L2 norm is used to control model complexity and avoid overfitting. This is a regularization parameter, with a value of 100, used to balance model fitting accuracy and model complexity. To determine the number of training samples, we selected the effective sample size from the historical data of lithium battery electrode production. The samples included coating parameters and corresponding electrode thickness data under different tensions and solid contents. For the first The influence weight of the slurry solid content of each training sample is calculated in step 2.2; For the first The prediction error of each training sample. ; For the first The output value of each training sample, which corresponds to the optimal coating parameters under the input features; Weight vector transpose; For the input feature vector The high-dimensional mapping function is used to map linearly inseparable "tension-solid content" feature data to a high-dimensional space to achieve linear fitting; for The input feature vector of each training sample, i.e. ,in For the first Real-time tension of the substrate for each sample. For the first Real-time solids content of slurry for each sample; The bias term is used to adjust the model's fitting baseline and improve the model's prediction accuracy. The target thickness is the standard value for lithium battery electrode production (e.g., 120μm), determined by product design requirements. The optimal model parameters are obtained by solving the objective function using the Lagrange multiplier method. and Complete the training of the improved WLSSVM model; Step 2.4, Output Adaptive Coating Parameters: Real-time Acquisition Input a pre-trained improved WLSSVM model, and the model outputs coating parameters adapted to the current state. The data is then transmitted to the drying-rewinding coordinated timing optimization module and the execution control module.

4. The industrial automation control method according to claim 1, characterized in that: The specific implementation steps of step 3 are as follows: Step 3.1, Basic Data Acquisition: Receive coating parameters - coating parameters output by the tension dynamic adaptation module. By combining real-time data from the multi-source status acquisition module, key intermediate parameters are calculated: a. Theoretical thickness The calculations are used to predict the electrode thickness based on real-time coating parameters, providing a basis for subsequent drying parameter adjustments. ,in, The theoretical thickness of the electrode coating is a theoretical value calculated based on real-time coating parameters, used to guide the adaptation of drying-winding parameters; The coating thickness coefficient is 1, which is determined experimentally using a lithium battery electrode coating equipment. In this embodiment, it is set to 0.8, which characterizes the combined influence coefficient of slurry solid content and coating roller speed on coating thickness. The coating thickness coefficient is 2, which is determined by equipment experiments to characterize the attenuation coefficient of coating pressure on coating thickness. For the real-time solids content of the slurry, and A positive correlation exists; the higher the solid content, the greater the coating thickness under the same rotation speed and pressure. The target rotational speed of the coating roller is output by the coating parameter-tension dynamic adaptation module, and... Negative correlation: the faster the rotation speed, the thinner the coating thickness. The target coating pressure is output by the coating parameter-tension dynamic adaptation module, and... Negative correlation: the greater the pressure, the thinner the coating thickness. b. Current tension of the winding machine The calculations are used to monitor the tension of the electrode sheet in real time during the winding process, preventing electrode sheet wrinkles caused by tension fluctuations. ,in This refers to the real-time tension of the electrode sheet during winding; The tension coefficient for winding speed is determined experimentally using winding machine equipment; The winding torque tension coefficient is determined through equipment testing and represents the degree to which the winding torque reduces the tension. The current speed of the winding machine is collected in real time by the encoder of the multi-source status acquisition module; This provides the output torque for the winding machine. Step 3.2: Construct an improved time-series coordination objective function: based on electrode drying uniformity. and winding tension stability To achieve the dual optimization objectives, a dynamic adjustment factor is introduced. Adjust the priorities of the two objectives; the objective function is: ,in The function value is optimized to achieve dual objectives, comprehensively reflecting both drying uniformity and tension stability. The objective is to minimize... ; Electrode drying uniformity The calculation reflects the overall drying consistency of the electrode by measuring the humidity deviation at multiple detection points. ,in Electrode drying uniformity is the average deviation of humidity at each test point from the standard humidity. The number of humidity detection points at the dryer outlet is evenly distributed along the width of the electrode to ensure comprehensive detection of the drying condition. The detection point number is used to determine the value. ; For the first The real-time humidity of the electrode at each detection point is collected in real time by the humidity sensor of the multi-source status acquisition module; This refers to the standard drying humidity of the electrode sheet, which is the process standard value for the production of lithium battery electrode sheets. For the first The absolute deviation of the humidity at each detection point from the standard humidity reflects the drying deviation of a single detection point; Winding tension stability The calculation uses the sum of squared tension deviations from multiple sampling points to reflect the degree of tension fluctuation during the winding process. ,in For winding tension stability; This represents the number of tension sampling points. The sampling point number is used to determine the value. ; For the first The winding tension at each sampling point is collected in real time by the tension sensor of the multi-source state acquisition module; This refers to the standard tension for winding, which is the standard value for the winding process of lithium battery electrode sheets. For the first The squared deviation of the tension at each sampling point from the standard tension; Dynamic adjustment factor The calculation enables adaptive adjustment of the optimization priorities for drying uniformity and tension stability: ,in This represents the highest historical drying unevenness, measured in units of... , ; This represents the current electrode drying uniformity. This represents the historical maximum tension fluctuation value; when When the size increases, the molecules The rate of decrease has slowed. Increase the optimization priority of improving drying uniformity; when When the denominator increases, Increase Prioritize reducing and improving tension stability. Step 3.3, Solving the Timing Adjustment Scheme: Based on the improved particle swarm optimization algorithm, the optimal solution of the dual-objective optimization function shown by the improved timing coordination objective function is obtained, ultimately yielding the target parameters of the dryer and the target speed of the winding machine adapted to the current coating state. This ensures that the drying-rewinding process is coordinated with the dynamic matching of coating parameters. Step 3.4: Output timing optimization instructions: Output the obtained dryer parameters and winding machine speed. The commands are converted into timing control instructions and transmitted to the execution control module to achieve dynamic coordination of drying-winding and coating parameters.

5. The industrial automation control method according to claim 4, characterized in that: The specific implementation logic for solving the timing adjustment scheme in step 3.3 is as follows: Step 3.31: Determine the optimization variables and constraints: Based on the requirements of lithium battery electrode production process and equipment operating limits, the optimization variable is determined to be the target temperature of the dryer. Target wind speed and the target speed of the winding machine At the same time, strict equipment parameter constraints are set: drying temperature constraint: Roll speed constraint: ,in The coating roller target speed output by the coating parameter-tension dynamic adaptation module ensures that the winding speed matches the coating speed, avoiding electrode stretching or accumulation; wind speed constraint: ; Step 3.32, Improved Particle Swarm Initialization: Each particle in the particle swarm corresponds to a set of candidate solutions for optimization variables, i.e. During the initialization phase, a particle population is randomly generated within the constraints. The initial position of each particle satisfies the above-mentioned temperature, velocity, and wind speed constraints, ensuring the engineering feasibility of the initial solution. Step 3.33, Fitness Function Definition: [The text abruptly ends here, likely due to an incomplete sentence or a Dual objective optimization function As the fitness function of particle swarm optimization, i.e. ,in , , All values ​​were obtained through the calculations in step 3.2 above; the smaller the fitness value, the stronger the fitness of the particle. The combination can better meet the requirements of uniform drying and stable tension; Step 3.34, Particle Iterative Update: An improved formula for updating particle velocity and position is used to achieve population evolution. Speed ​​update formula: ,in For inertial weights, 1. The learning factor is 2. , for Random numbers, For the first The optimal position of a single particle in the next iteration. The optimal position for the population; Position update formula: After the update, the particle position is checked for constraints. If it exceeds the constraint range of step 3.31, it is corrected to the constraint boundary. Step 3.35, Optimal Solution Determination and Output: During the iteration process, the optimal fitness value of a single particle and the optimal fitness value of the population are recorded in each round. When the preset number of iterations is reached or the optimal fitness value of the population remains unchanged for 10 consecutive rounds, the iteration stops, and the optimal position of the population is determined. corresponding This is the optimal timing adjustment parameter, ensuring that the combination of parameters can maximize both the uniformity of electrode drying and the stability of winding tension within the safe operating range of the equipment.

6. An industrial automation control system, applied to the industrial automation control method according to any one of claims 1-5, characterized in that: The multi-source status acquisition module is used to collect real-time operating status data of the coating machine, dryer, and winding machine, as well as electrode quality inspection data; The coating parameter-tension dynamic adaptation module is used to establish a dynamic matching relationship between electrode substrate tension, slurry solid content and coating parameters based on an improved weighted least squares support vector machine algorithm, and output the adapted coating parameters; the drying-winding coordinated timing optimization module is used to dynamically adjust the dryer operating parameters and winding machine coordinated timing based on an improved timing coordination algorithm and coating parameters and electrode humidity data; the execution control module is used to receive various optimization parameters and timing instructions and drive the coating-drying-winding equipment to operate in a coordinated manner.

Images