An enameled wire intelligent production line collaborative control method and system

Abstract

The application discloses a kind of enameled wire intelligent production line collaborative control method and system, it is related to intelligent manufacturing and industrial automation control technical field, including by obtaining operating parameter and quality index, calculate real-time influence weight and compare with historical weight and generate deviation signal;Response deviation signal suspends original rule, filters key parameter and generates candidate linkage rule;Execute candidate rule one by one and calculate quality improvement degree, keep effective rule configuration priority and enter database;Periodically evaluate effective rule and dynamically adjust weight.The application realizes the adaptive correction of regulation and control logic, avoids blind trial and error, significantly improves the level of production line autonomous decision and operation reliability.

Description

Technical Field

[0001] This invention relates to the field of intelligent manufacturing and industrial automation control technology, and in particular to a collaborative control method and system for an intelligent production line for enameled wire. Background Technology

[0002] In the field of enameled wire manufacturing, advanced production lines generally employ centralized control systems. These systems use a linkage and control mechanism to collaboratively manage processes such as wire feeding, annealing, coating, baking, cooling, and winding. The system operates based on pre-stored combinations of process parameters and adjusts each actuator in conjunction with online monitoring of tension, temperature, and wire speed to ensure product quality stability.

[0003] However, when replacing with a new batch of insulating varnish, even if the macroscopic physicochemical indicators of the new material are qualified, its microscopic process characteristics (such as solvent evaporation rate and resin curing reaction curve) may still have slight differences from the old batch. The main drawback of the existing technology is that the "process parameter-quality response" correlation model on which the control system is based is established based on the old batch of materials and cannot detect changes in the characteristics of the new material. When online detection detects a decrease in withstand voltage breakdown performance, the system still increases the baking oven temperature according to the old model. As a result, not only does it fail to improve the withstand voltage performance, but it also causes new defects such as granular protrusions, reduced flexibility, and cracking on the varnish surface. If the line speed is reduced and the baking time is extended, production efficiency decreases and quality problems still exist. This control logic is out of touch with the actual physicochemical process, making the adjustment actions a blind trial and error, and unable to stably control product quality under the conditions of the new batch of raw materials. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a collaborative control method and system for an intelligent production line for enameled wire. The aim is to solve the problem that when a new batch of insulating varnish is introduced into an intelligent production line for enameled wire, the differences in the microscopic process characteristics of the new varnish compared to the old batch cause the original linkage control logic based on the characteristics of the old material to fail. Consequently, the adjustment actions of the control system based on online quality feedback cannot effectively improve quality issues, and instead introduce new defects such as particle protrusion on the varnish surface and reduced flexibility, even leading to a decrease in production efficiency.

[0005] To achieve the above objectives, the present invention provides a collaborative control method for an intelligent production line of enameled wire, comprising the following steps: S1. Obtain the temperature of each heating zone of the baking oven, the operating speed of the production line, and the paint supply of the coating device as a set of operating parameters during the production of enameled wire; obtain the online detection of the withstand voltage breakdown performance, the paint film flexibility index, and the quantitative value of the paint film surface defects as a set of quality indicators; obtain the rule base, which contains several linkage control rules. S2. Based on the set of operating parameters and the set of quality indicators, calculate the real-time impact weight of each operating parameter on each quality indicator, and compare it with the historical impact weight calculated from the most recent N batches of operating parameter sets and quality indicator sets; when the deviation between any real-time impact weight and the historical impact weight exceeds a preset threshold, generate an association mode deviation signal. S3. In response to the deviation signal of the correlation mode, the original linkage control rules are suspended and linkage control rule exploration is initiated: Based on the real-time influence weight, the operating parameters whose influence weight on the quality indicators is higher than the preset effective threshold are selected as key control parameters. The control direction and control amplitude are determined for each key control parameter, and multiple candidate linkage control rules containing the target set values ​​of each key control parameter are generated. S4. Execute each candidate linkage control rule one by one. After executing each candidate linkage control rule, collect the corresponding new quality index set and calculate the improvement degree of the quality index before and after the execution of the corresponding candidate linkage control rule. S5. Retain candidate linkage control rules whose quality indicator improvement degree is higher than the preset improvement threshold, configure priority weights based on the improvement degree of the candidate linkage control rules, and add the candidate linkage control rules with configured priority weights to the rule base; discard candidate linkage control rules whose quality indicator improvement degree is lower than the preset improvement threshold, and execute the linkage control rule with the highest priority weight in the rule base.

[0006] Furthermore, the acquisition of the temperature of each heating zone of the baking oven, the operating speed of the production line, and the paint supply of the coating device during the enameled wire production process as a set of operating parameters, and the acquisition of online detection of the withstand voltage breakdown performance, paint film flexibility index, and paint film surface defect quantification value as a set of quality indicators, includes: The output signals of the temperature sensors in each heating zone of the baking oven are collected, and after analog-to-digital conversion and temperature correction, the real-time temperature values ​​of each heating zone are generated. The pulse signal from the rotary encoder of the take-up drum is collected, and the pulse count and unit time conversion are used to generate the real-time operating speed of the production line. The output signal of the flow sensor of the painting device is collected, and after calibration and conversion, a real-time paint supply value is generated. The real-time temperature values ​​of each heating zone, the real-time operating speed of the production line, and the real-time paint supply values ​​are combined into a set of operating parameters. The breakdown detection signal of the high-voltage electrode array is collected to generate a quantitative index of the withstand voltage breakdown performance; The signal from the paint film bending detection device is collected to generate a paint film flexibility index. Surface images are acquired using a surface scanning camera, and quantitative values ​​of surface defects in the paint film are generated using a defect recognition algorithm. The quantitative indicators of pressure resistance and breakdown performance, the indicators of paint film flexibility, and the quantitative values ​​of paint film surface defects are combined into a set of quality indicators.

[0007] Furthermore, based on the set of operating parameters and the set of quality indicators, the real-time impact weight of each operating parameter on each quality indicator is calculated and compared with the historical impact weights calculated from the most recent N batches of operating parameter sets and quality indicator sets; when the deviation between any real-time impact weight and the historical impact weight exceeds a preset threshold, a correlation pattern deviation signal is generated, including: The set of operating parameters and the set of quality indicators are synchronized and aligned to obtain the synchronized parameter sequence and indicator sequence. Using each quality indicator in the quality indicator set as the dependent variable and all operating parameters in the operating parameter set as independent variables, the regression coefficients are calculated using the multiple linear regression method. The regression coefficients are used as real-time impact weights to construct a real-time impact weight matrix; Obtain the historical weight matrix calculated from the parameter set and quality index set of the most recent N batches of operation; Calculate the absolute value of the difference between the elements at the same position in the real-time influence weight matrix and the historical influence weight matrix, and divide it by the historical influence weight to obtain the relative deviation; When any relative deviation exceeds a preset deviation threshold, a correlation mode deviation signal is generated.

[0008] Furthermore, the calculation of regression coefficients using a multiple linear regression method with each quality indicator in the quality indicator set as the dependent variable and all operating parameters in the operating parameter set as independent variables includes: Construct an independent variable matrix using the running parameter values ​​at all sampling times in the synchronized parameter sequence. The rows of the independent variable matrix correspond to the sampling times, and the columns of the independent variable matrix correspond to the running parameters. Construct a dependent variable vector using the values ​​of the current quality index at each sampling time in the synchronized index sequence; The regression coefficients are calculated using the least squares method based on the independent variable matrix and the dependent variable vector.

[0009] Furthermore, in response to the correlation mode deviation signal, the original linkage control rules are suspended, and linkage control rule exploration is initiated: based on real-time influence weights, operating parameters with influence weights on quality indicators higher than preset effective thresholds are selected as key control parameters, the control direction and control amplitude are determined for each key control parameter, and multiple candidate linkage control rules containing target set values ​​for each key control parameter are generated, including: In response to the correlation mode deviation signal, the original linkage control rules are suspended, and the real-time impact weight matrix is ​​read; For any operating parameter, if the real-time impact weight of at least one associated quality indicator is higher than a preset effective threshold, then the operating parameter is marked as a key control parameter. Obtain the adjustable range of each key control parameter, and set multiple incremental control amplitudes based on the adjustable range; Using the target setpoints of the key control parameters in the original linkage control rules as the baseline, multiple candidate target setpoints are generated for each key control parameter based on the control direction and multiple control amplitudes. The multiple candidate target setpoints of each key control parameter are combined by Cartesian product to generate multiple candidate linkage control rules. Each candidate linkage control rule consists of a set of target setpoints for each key control parameter. The control direction is determined according to the positive or negative sign of the real-time influence weight. When the real-time influence weight is positive, the control direction is to increase the value of the operating parameter; when the real-time influence weight is negative, the control direction is to decrease the value of the operating parameter.

[0010] Furthermore, the generation of multiple candidate linkage control rules also includes processing of the original linkage control rules, as follows: For any operating parameter that is not marked as a key control parameter, obtain the original linkage control rules in the rule base that contain the control logic of that operating parameter; Calculate the average of the real-time impact weights of all quality indicators associated with this operating parameter; If the average value is lower than the preset effective threshold, the original linkage control rule will be marked as an inefficient rule.

[0011] Furthermore, the process involves executing each candidate linkage control rule one by one, collecting the corresponding new quality indicator set after executing each candidate linkage control rule, and calculating the improvement degree of the quality indicators before and after the execution of the corresponding candidate linkage control rule, as follows: Before executing any candidate linkage control rule, the effective linkage control rule with the highest priority weight in the rule base is run, and the set of running parameters and quality indicators are collected. The set of quality indicators collected at this time is used as the benchmark quality indicator set. Execute each candidate linkage control rule one by one. For each candidate linkage control rule: The target setpoints of each operating parameter are parsed from the candidate linkage control rules; Based on the target setpoint, the production line is adjusted, the breakdown detection signal of the high voltage electrode array is re-acquired, and the pressure resistance breakdown performance after execution is generated; the signal of the paint film bending detection device is re-acquired, and the paint film flexibility index after execution is generated; the surface image is re-acquired through the surface scanning camera, and the surface defect quantification value of the paint film after execution is generated by the defect recognition algorithm. The pressure breakdown performance, the paint film flexibility index, and the paint film surface defect quantification value after implementation will be combined into a new set of quality indicators after implementation. Obtain the baseline quality index set, calculate the change of each quality index in the new quality index set after execution relative to the baseline quality index set, and perform a weighted summation of the changes in each quality index to obtain the quality index improvement degree of the candidate linkage control rule. After completing the execution and evaluation of a candidate linkage control rule, the production line is restored to the state defined by the effective linkage control rule before the execution of the candidate linkage control rule, and then the next candidate linkage control rule is executed.

[0012] Furthermore, the process of retaining candidate linkage control rules whose quality indicator improvement degree is higher than a preset improvement threshold, configuring priority weights based on the improvement degree of the candidate linkage control rules, and adding the candidate linkage control rules with configured priority weights to the rule base; and discarding candidate linkage control rules whose quality indicator improvement degree is lower than the preset improvement threshold, and executing the linkage control rule with the highest priority weight in the rule base, includes: Compare the improvement rate of the quality index of each candidate linkage control rule with the preset improvement threshold; When the improvement of the quality indicator is greater than the preset improvement threshold, the candidate linkage control rule is retained, the priority weight is determined based on the improvement of the quality indicator, and the candidate linkage control rule and its priority weight are added to the rule base. When the improvement of the quality indicator is less than or equal to the preset improvement threshold, the candidate linkage control rule is discarded. After the candidate linkage control rules are screened and added to the database, all valid linkage control rules in the rule base are sorted in descending order according to their priority weight, and the linkage control rule with the highest priority weight is selected as the currently effective linkage control rule.

[0013] Furthermore, adding the candidate linkage control rule and its priority weight to the rule base also includes adjusting the priority weights of existing linkage control rules in the rule base, as follows: Based on the current effective linkage control rules, during the operation process, the new quality indicator set after execution is periodically collected, and the improvement degree of quality indicators in this operation is calculated. The improvement rate of the quality index calculated in this study is compared with the current priority weight of the currently effective linkage control rule; If the improvement degree is greater than the current priority weight, the priority weight of the linkage control rule is increased by a preset increment; if the improvement degree is less than the current priority weight, the priority weight of the linkage control rule is decreased by a preset decrement; if the improvement degree is equal to the current priority weight, the priority weight remains unchanged. The adjusted priority weights are written back to the rule base, and when the rules of the linkage control rule base are sorted again, the currently effective linkage control rules are re-determined based on the updated priority weights.

[0014] A collaborative control system for an intelligent production line of enameled wire includes: Parameter acquisition module: used to acquire the temperature of each heating zone of the baking oven, the operating speed of the production line, and the paint supply of the coating device as a set of operating parameters during the production of enameled wire; to acquire the online detection of the withstand voltage breakdown performance, the paint film flexibility index, and the quantitative value of the paint film surface defects as a set of quality indicators; and to acquire the rule base, which contains several linkage control rules. The correlation signal module is used to calculate the real-time influence weight of each operating parameter on each quality indicator based on the set of operating parameters and the set of quality indicators, and compare it with the historical influence weight calculated from the most recent N batches of operating parameter sets and quality indicator sets; when the deviation between any real-time influence weight and the historical influence weight exceeds a preset threshold, a correlation mode deviation signal is generated. New rule generation module: In response to the correlation mode deviation signal, it suspends the original linkage control rules and starts linkage control rule exploration: Based on the real-time influence weight, it selects the operating parameters whose influence weight on the quality index is higher than the preset effective threshold as key control parameters, determines the control direction and control amplitude for each key control parameter, and generates multiple candidate linkage control rules containing the target set values ​​of each key control parameter. Parameter adjustment module: used to execute each candidate linkage control rule one by one, collect the corresponding new quality index set after executing each candidate linkage control rule, and calculate the improvement degree of the quality index before and after the execution of the corresponding candidate linkage control rule; The optimization detection module is used to retain candidate linkage control rules whose quality indicator improvement degree is higher than the preset improvement threshold, configure priority weights based on the improvement degree of the candidate linkage control rules, and add the candidate linkage control rules with priority weights to the rule base; it discards candidate linkage control rules whose quality indicator improvement degree is lower than the preset improvement threshold, and executes the linkage control rule with the highest priority weight in the rule base.

[0015] The present invention provides a collaborative control method and system for an intelligent production line of enameled wire, the beneficial effects of which are mainly reflected in the following aspects: 1. This invention significantly improves the production line's ability to perceive and adapt to changes in the characteristics of raw material batches. By calculating the real-time impact weight of each operating parameter on each quality indicator and comparing it with historical impact weights, it can promptly generate a correlation mode deviation signal when changes in the microscopic characteristics of raw materials cause a shift in the process correlation mode. It keenly "detects" changes in the process response pattern after a new batch of insulating varnish is added, effectively overcoming the lag and misjudgment defects of traditional control systems that rely on fixed process models, and avoiding product quality fluctuations caused by control logic failures.

[0016] 2. This invention achieves intelligent exploration and self-correction of the linkage control logic, eliminating blind trial and error. Upon identifying a deviation in the correlation pattern, it automatically suspends the failed original linkage control rules and, based on real-time impact weights, filters out key control parameters that significantly affect quality indicators, generating multiple candidate linkage control rules in a targeted manner. This fundamentally solves the problem in existing technologies where ineffective adjustment actions, or even new quality defects such as surface particles and decreased flexibility of the paint film, are caused by the failure of old experience.

[0017] 3. This invention calculates the improvement in quality indicators by executing and evaluating candidate linkage control rules one by one, and only retains and prioritizes the execution of rules that have been verified to be effective in practice. This iterative optimization mechanism of "observation-understanding-adjustment-verification" enables the control rule base to be continuously updated and enriched with the production process. This ensures that the production line can consistently and efficiently output enameled wire products that meet quality requirements when faced with changes in external conditions such as batch replacement of insulating varnish, significantly improving the autonomous decision-making level and operational reliability of the intelligent production line. Attached Figure Description

[0018] Figure 1 This is a flowchart illustrating a collaborative control method for an intelligent production line of enameled wire according to the present invention. Figure 2 This is a functional block diagram of a collaborative control system for an intelligent production line of enameled wire according to the present invention. Detailed Implementation

[0019] 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.

[0020] Please see Figure 1 This invention provides a collaborative control method for an intelligent production line of enameled wire, comprising the following steps: S1. Obtain the temperature of each heating zone of the baking oven, the operating speed of the production line, and the paint supply of the coating device as a set of operating parameters during the production of enameled wire; obtain the online detection of the withstand voltage breakdown performance, the paint film flexibility index, and the quantitative value of the paint film surface defects as a set of quality indicators; obtain the rule base, which contains several linkage control rules. In this embodiment, the acquisition of the temperature of each heating zone of the baking oven, the operating speed of the production line, and the paint supply of the coating device during the enameled wire production process are used as a set of operating parameters. The acquisition of online detection of the withstand voltage breakdown performance, paint film flexibility index, and quantified values ​​of paint film surface defects are used as a set of quality indicators, including: The output signals of the temperature sensors in each heating zone of the baking oven are collected, and after analog-to-digital conversion and temperature correction, the real-time temperature values ​​of each heating zone are generated.

[0021] Specifically, temperature sensors refer to temperature-measuring elements, such as thermocouples or resistance temperature detectors (RTDs), installed in each independent temperature control zone of the baking oven. Their output signals correspond to the actual temperature inside the oven. Analog-to-digital conversion and temperature correction are standard signal processing methods in industrial measurement and control. Real-time temperature values ​​refer to digital quantities that, after processing, reflect the actual oven temperature of each heating zone at the current moment.

[0022] The pulse signal from the rotary encoder of the take-up drum is collected, and the pulse count and unit time conversion are used to generate the real-time operating speed of the production line.

[0023] Specifically, the take-up reel refers to the rotating end component used to wind the finished enameled wire. The rotary encoder is an angular displacement sensor mounted coaxially with the take-up reel shaft, and its output pulse count is proportional to the angular displacement of the shaft. After pulse counting and conversion to unit time, the real-time operating speed of the production line is obtained, that is, the linear speed of the enameled wire moving along the production direction.

[0024] The output signal of the flow sensor of the painting device is collected and converted into a real-time paint supply value after calibration.

[0025] Specifically, the coating unit refers to the process unit that uniformly coats the surface of bare copper wire with liquid insulating varnish. The flow sensor is a measuring device installed in the varnish supply pipeline, whose output signal is proportional to the instantaneous flow rate of the varnish. Calibration conversion refers to converting the original signal into a physically meaningful flow rate value using a pre-established signal-flow relationship. The real-time varnish supply value refers to the instantaneous flow rate of insulating varnish supplied by the coating unit at the current moment.

[0026] The real-time temperature values ​​of each heating zone, the real-time operating speed of the production line, and the real-time paint supply values ​​are combined into a set of operating parameters.

[0027] Specifically, merging refers to combining the above three types of real-time production data into a unified data set after aligning them by timestamps. The operating parameter set refers to a combination of multiple operational variables that characterize the operating status of the production line, collected at the same time or within a short period of time.

[0028] The breakdown detection signal of the high-voltage electrode array is collected to generate a quantitative index of the withstand voltage breakdown performance.

[0029] Specifically, a high-voltage electrode array refers to a group of non-contact discharge electrodes arranged along the path of the enameled wire, used to apply a test voltage to the cured enamel film. The breakdown detection signal refers to the voltage drop or current pulse signal captured by the detection circuit when a weak point exists in the enamel film. Quantitative indicators convert the frequency of breakdown events or the breakdown voltage value into comparable numerical values. Withstand voltage breakdown performance refers to the ability of the enameled wire insulation layer to withstand a specified voltage without breakdown damage.

[0030] The signal from the paint film bending detection device is collected to generate the paint film flexibility index.

[0031] Specifically, a coating film bending test device refers to an online testing unit that applies a specified mechanical bend to enameled wire. The device signal refers to the electrical signal collected during the bending process that reflects the degree of cracking or deformation of the coating film. The flexibility index refers to the quantitative value of the coating film's ability to withstand bending deformation without cracking or separating from the conductor.

[0032] Surface images are acquired using a surface scanning camera, and surface defect quantification values ​​are generated using a defect recognition algorithm.

[0033] Specifically, a surface scanning camera refers to an industrial vision camera that captures high-speed images of the circumferential surface of enameled wire. A surface image refers to a series of digital image frames continuously acquired by the camera, reflecting the appearance of the paint film. A defect recognition algorithm refers to a software program that uses image processing or deep learning techniques to automatically detect abnormal morphologies such as particles and scratches from surface images. A paint film surface defect quantification value refers to a numerical index that integrates information such as the number, area, or density of defects.

[0034] The quantitative indicators of pressure resistance and breakdown performance, the indicators of paint film flexibility, and the quantitative values ​​of paint film surface defects are combined into a set of quality indicators.

[0035] Specifically, merging refers to aligning the three numerical indicators that respectively characterize electrical performance, mechanical performance, and appearance integrity into a unified data set by time or location. A quality indicator set refers to a combination of multiple indicators obtained within the same testing period to comprehensively evaluate the quality level of enameled wire.

[0036] S2. Based on the set of operating parameters and the set of quality indicators, calculate the real-time impact weight of each operating parameter on each quality indicator, and compare it with the historical impact weight calculated from the most recent N batches of operating parameter sets and quality indicator sets; when the deviation between any real-time impact weight and the historical impact weight exceeds a preset threshold, generate an association mode deviation signal. In this embodiment, the real-time impact weight of each operating parameter on each quality indicator is calculated based on the set of operating parameters and the set of quality indicators, and compared with the historical impact weights calculated from the most recent N batches of operating parameter sets and quality indicator sets; when the deviation between any real-time impact weight and the historical impact weight exceeds a preset threshold, a correlation pattern deviation signal is generated, including: The set of operating parameters and the set of quality indicators are synchronized and aligned to obtain the synchronized parameter sequence and indicator sequence.

[0037] Specifically, the operating parameter set refers to the data set composed of operational variables such as temperature of each section of the baking oven, production line operating speed, and paint supply, collected at the same time or within a short period. The quality index set refers to the data set composed of quality indicators such as pressure breakdown performance, paint film flexibility index, and surface defect quantification value obtained within the same testing period. Synchronization alignment refers to matching the values ​​of each sampling moment in the operating parameter set with the corresponding detection results in the quality index set based on the timestamp information of each data acquisition channel, eliminating time-domain misalignment caused by signal transmission delay or sampling period differences. The parameter sequence refers to the queue of operating parameter values ​​arranged in chronological order after alignment. The index sequence refers to the queue of quality index values ​​arranged in chronological order after alignment.

[0038] The regression coefficients are calculated using a multiple linear regression method, with each quality indicator in the quality indicator set as the dependent variable and all operating parameters in the operating parameter set as independent variables.

[0039] Specifically, the dependent variable refers to the variable being predicted or explained in mathematical modeling; in this case, it is a specific quality indicator among pressure breakdown performance, paint film flexibility index, or surface defect quantification value. The independent variable refers to the variable used to explain changes in the dependent variable; in this case, it is all operating parameters such as the temperature of each section of the baking oven, the production line operating speed, and the paint supply. Multiple linear regression is a statistical modeling technique that establishes a linear relationship between multiple independent variables and a dependent variable. The regression coefficient refers to the multiplier factor obtained by the optimization algorithm before each independent variable; its magnitude and sign reflect the degree and direction of influence of the operating parameter on the current quality indicator, respectively.

[0040] The basic expression for this multiple linear regression model is: ; in, This represents the vector of observed values ​​for the k-th quality indicator. These represent the observation vectors of the 1st to mth operating parameters, respectively. This represents the total number of runtime parameters. The intercept term constant; These are the regression coefficients of the 1st to mth operating parameters on the kth quality index, respectively. This is the random error term.

[0041] The regression coefficients are used as real-time impact weights to form a real-time impact weight matrix.

[0042] Specifically, real-time impact weights refer to numerical values ​​calculated based on data collected in the current batch or most recent time period, reflecting the degree of influence of each operating parameter on each quality indicator. Each real-time impact weight corresponds to a regression coefficient between an operating parameter and a quality indicator. The real-time impact weight matrix is ​​a two-dimensional table organized in a row-column format, where the rows of the matrix correspond to each quality indicator, the columns correspond to each operating parameter, and the elements in the matrix represent the impact weight values ​​between the quality indicator represented by the row and the operating parameter represented by the column. The structure can be represented as: ; in, This indicates the total number of quality indicators. Indicates the total number of runtime parameters; matrix elements Indicates the first The first operating parameter affects the second The real-time impact weight of each quality indicator, i.e., the regression coefficient. .

[0043] Obtain the historical weight matrix calculated from the parameter set and quality index set of the most recent N batches of operation.

[0044] Specifically, a batch refers to a continuous production period between nodes such as changes in insulating varnish raw materials or production task switching. The most recent N batches refer to the N batches that have been completed most recently, where N is a preset positive integer. The historical weight matrix refers to a table of influence weights pre-calculated and stored using the production data of the most recent N batches, following the same multiple linear regression method as the real-time influence weight matrix. Its row and column meanings are consistent with the real-time influence weight matrix, representing the correlation between operating parameters and quality indicators under the raw material characteristics of the old batches.

[0045] Calculate the absolute value of the difference between the elements at the same position in the real-time influence weight matrix and the historical influence weight matrix, and divide it by the historical influence weight to obtain the relative deviation.

[0046] Specifically, elements at the same position refer to elements in the real-time influence weight matrix and the historical weight matrix that have the same row and column indices, corresponding to the same quality indicator and the same operating parameter. The absolute value of the difference is the non-negative result of subtracting the two weight values, used to measure the magnitude of change without considering direction. The historical influence weight refers to the value of the element at that position in the historical weight matrix. The relative deviation is the quotient of the absolute value of the difference divided by the historical influence weight, used to eliminate the influence of differences in the order of magnitude of different weight values ​​on the comparison results.

[0047] The formula for calculating relative deviation is: ; in, This represents the relative deviation of the j-th operating parameter from the i-th quality index; This represents the value of the element in the i-th row and j-th column of the weight matrix that affects real-time effects; This represents the value of the element in the i-th row and j-th column of the historical weight matrix; This indicates the operation of taking the absolute value.

[0048] When any relative deviation exceeds a preset deviation threshold, a correlation mode deviation signal is generated.

[0049] Specifically, the preset deviation threshold refers to the upper limit of the allowable fluctuation range set in advance, and its specific value is determined according to the sensitivity requirements of the enameled wire production process. The correlation mode deviation signal is a marker indicating that the response relationship between the current operating parameters and quality indicators has significantly deviated from the historical baseline state. This step analyzes all relative deviation values ​​one by one. Once any relative deviation exceeds the preset threshold, it is determined that the microscopic process characteristics of the new batch of insulating varnish have caused the original process correlation model to fail, and a correlation mode deviation signal is generated.

[0050] In this embodiment, the step of calculating the regression coefficients using a multiple linear regression method with each quality indicator in the quality indicator set as the dependent variable and all operating parameters in the operating parameter set as independent variables includes: An independent variable matrix is ​​constructed using the operating parameter values ​​at all sampling times in the synchronized parameter sequence. The rows of the independent variable matrix correspond to the sampling times, and the columns of the independent variable matrix correspond to the operating parameters.

[0051] Specifically, the synchronized parameter sequence refers to a time-aligned queue of running parameter values. The independent variable matrix is ​​a two-dimensional array used in regression analysis to store all observed values ​​of the independent variables. In this step, each row of the matrix corresponds to a sampling time point, each column corresponds to a running parameter, and the matrix elements are the measured values ​​of that running parameter at the corresponding time point. This structure converts the time series data into a standard format suitable for matrix operations.

[0052] Independent variable matrix The structure is as follows: ; in, The total number of sampling times. Total number of runtime parameters Indicates the first At the sampling time, the first The values ​​of each running parameter.

[0053] The dependent variable vector is constructed using the values ​​of the current quality index at each sampling time in the synchronized index sequence.

[0054] Specifically, the current quality indicator refers to one of the following three: the quantitative value of pressure breakdown performance, the indicator of paint film flexibility, or the quantitative value of paint film surface defects. The dependent variable vector is a one-dimensional array with a length equal to the total number of sampling times n, and its t-th element is the observed value of the quality indicator at the t-th sampling time.

[0055] The dependent variable vector corresponding to the k-th quality indicator for: ; in, Number the quality indicators. For the first The observed value of a quality index at the nth sampling time; For the first The dependent variable corresponding to each quality indicator.

[0056] The regression coefficients are calculated using the least squares method based on the independent variable matrix and the dependent variable vector.

[0057] Specifically, the least squares method obtains the optimal estimate of the regression coefficients by minimizing the sum of squared residuals.

[0058] Utilize the already constructed The regression coefficients of all operating parameters on the current quality index can be obtained in one step through matrix operations.

[0059] The least squares solution is: ; in, Let K be the regression coefficient vector corresponding to the k-th quality indicator. for transpose, for The inverse matrix.

[0060] S3. In response to the deviation signal of the correlation mode, the original linkage control rules are suspended and linkage control rule exploration is initiated: Based on the real-time influence weight, the operating parameters whose influence weight on the quality indicators is higher than the preset effective threshold are selected as key control parameters. The control direction and control amplitude are determined for each key control parameter, and multiple candidate linkage control rules containing the target set values ​​of each key control parameter are generated. In this embodiment, in response to the correlation mode deviation signal, the original linkage control rules are suspended and linkage control rule exploration is initiated: based on real-time influence weights, operating parameters with influence weights on quality indicators higher than a preset effective threshold are selected as key control parameters. The control direction and amplitude are determined for each key control parameter, and multiple candidate linkage control rules containing target set values ​​for each key control parameter are generated, including: In response to the deviation signal of the associated mode, the original linkage control rules are suspended and the real-time impact weight matrix is ​​read.

[0061] Specifically, the correlation mode deviation signal refers to the indicator information that the response relationship between current operating parameters and quality indicators has significantly deviated from the historical benchmark. Suspending existing linkage control rules means suspending the currently executing control strategy based on the characteristics of old batches of insulating varnish. The real-time influence weight matrix is ​​a two-dimensional table calculating the degree of influence of each operating parameter on each quality indicator based on current operating data.

[0062] For any operating parameter, if the real-time impact weight of at least one associated quality indicator is higher than a preset effective threshold, then the operating parameter is marked as a key control parameter.

[0063] Specifically, operating parameters refer to the temperature of each section of the baking oven, the operating speed of the production line, and the real-time paint supply. The real-time influence weight is the regression coefficient of the operating parameter on a specific quality indicator, representing the change in the quality indicator caused by a unit change in the operating parameter. The preset effective threshold is a pre-defined minimum influence limit; only operating parameters with an absolute weight value higher than this threshold are considered to have a significant regulatory effect. Key regulatory parameters refer to operating parameters that still maintain a significant influence on at least one quality indicator under new batches of insulating paint conditions.

[0064] Obtain the adjustable range of each key control parameter, and set multiple incremental control amplitudes based on the adjustable range.

[0065] Specifically, the adjustable range refers to the upper and lower limits within which the key control parameter is allowed to be adjusted under the premise of safe equipment operation. The incremental control amplitude refers to multiple adjustment levels divided into certain steps within the adjustable range.

[0066] Using the target setpoints of the key control parameters in the original linkage control rules as the baseline, multiple candidate target setpoints are generated for each key control parameter based on the control direction and multiple control amplitudes. The multiple candidate target setpoints of each key control parameter are combined by Cartesian product to generate multiple candidate linkage control rules. Each candidate linkage control rule consists of a set of target setpoints for each key control parameter. The control direction is determined according to the positive or negative sign of the real-time influence weight. When the real-time influence weight is positive, the control direction is to increase the value of the operating parameter; when the real-time influence weight is negative, the control direction is to decrease the value of the operating parameter.

[0067] Specifically, the baseline value refers to the target value set for key control parameters in the original linkage control rules. The control direction refers to the direction of change in the operating parameters needed to improve quality indicators. The positive or negative sign of the real-time influence weight indicates the directional relationship between the operating parameters and quality indicators; a positive weight increases the operating parameters to improve quality indicators, and the control direction is increase; a negative weight decreases the operating parameters to improve quality indicators, and the control direction is decrease. Candidate target setpoints refer to a series of proposed operating parameter setpoints calculated starting from the baseline value, according to the control direction and multiple incremental control amplitudes. The Cartesian product combination refers to the complete permutation and pairing of multiple candidate target setpoints for each key control parameter, forming a set covering various parameter combination schemes. A candidate linkage control rule refers to a complete control attempt scheme composed of a set of specific key control parameter target setpoints.

[0068] In this embodiment, the generation of multiple candidate linkage control rules also includes processing of the original linkage control rules, as follows: For any operating parameter that is not marked as a key control parameter, retrieve the original linkage control rules in the rule base that contain the control logic for that operating parameter.

[0069] Specifically, operating parameters not marked as key control parameters refer to those whose real-time impact weight on each quality indicator did not reach the preset effective threshold during the screening process. The rule base refers to the storage unit containing the linkage control rules for each production line. Existing linkage control rules refer to the linkage control rules that existed in the rule base before the generation of the correlation mode deviation signal. This step retrieves all existing linkage control rules related to the adjustment of each operating parameter not included in the key control parameter set from the rule base.

[0070] Calculate the average real-time impact weights of all quality metrics associated with this operating parameter.

[0071] Specifically, all quality indicators associated with this operating parameter refer to all quality indicator weight elements in the column corresponding to this operating parameter in the real-time influence weight matrix. The real-time influence weight refers to the regression coefficient of this operating parameter on a certain quality indicator, calculated based on current operating data. The average value refers to the arithmetic mean obtained by adding the absolute values ​​of the above multiple real-time influence weights and dividing by the total number of quality indicators, used to comprehensively measure the average influence of this operating parameter on the overall quality level under the conditions of a new batch of insulating varnish.

[0072] If the average value is lower than the preset effective threshold, the original linkage control rule will be marked as an inefficient rule.

[0073] Specifically, the preset effective threshold refers to the minimum influence limit set in advance. An average value lower than the preset effective threshold indicates that the overall influence of the operating parameter on various quality indicators under the new batch of insulating varnish is very weak. Inefficient rules refer to the original linkage control rules that have been determined to have significantly weakened or failed under the new material characteristics. This step marks the original linkage control rules associated with operating parameters whose overall average influence weight is lower than the threshold as inefficient.

[0074] S4. Execute each candidate linkage control rule one by one. After executing each candidate linkage control rule, collect the corresponding new quality index set and calculate the improvement degree of the quality index before and after the execution of the corresponding candidate linkage control rule. In this embodiment, each candidate linkage control rule is executed one by one. After executing each candidate linkage control rule, a new set of quality indicators is collected, and the improvement degree of the quality indicators before and after the execution of the corresponding candidate linkage control rule is calculated, as follows: Before executing any candidate linkage control rule, the effective linkage control rule with the highest priority weight in the current rule base is run, and the set of running parameters and quality indicators are collected. The set of quality indicators collected at this time is used as the benchmark quality indicator set.

[0075] Specifically, priority weight refers to the weight value associated with each effective linkage control rule in the rule base. This value reflects the degree of improvement of quality indicators by the rule in historical execution; a higher weight indicates a better control effect of the rule. Effective linkage control rules refer to linkage control rules in the rule base that have not been marked as inefficient. The operating parameter set refers to the combination of operational variables such as the temperature of each section of the baking oven, the operating speed of the production line, and the real-time paint supply collected at the current moment.

[0076] The quality index set refers to the combination of quantitative values ​​of pressure breakdown performance, paint film flexibility, and paint film surface defects collected at the current moment. The baseline quality index set refers to the set of quality index data produced by the production line under the currently effective optimal rule control before any candidate linkage control rule is executed, serving as a reference benchmark for subsequent evaluation of the improvement effect of candidate rules.

[0077] Execute each candidate linkage control rule one by one. For each candidate linkage control rule, extract the target setpoint of each operating parameter from the candidate linkage control rule.

[0078] Specifically, "execution one by one" refers to applying all generated candidate linkage control rules sequentially to the production line for effect verification. A candidate linkage control rule refers to a complete control scheme generated through Cartesian product combinations, containing a set of target setpoints for key control parameters. "Parsing" refers to extracting the specific target values ​​corresponding to each operating parameter from the rule's data structure. The target setpoint refers to the target value that each key control parameter should be adjusted to, as specified by the candidate rule.

[0079] Based on the target setpoint, the production line is adjusted, the breakdown detection signal of the high-voltage electrode array is re-acquired, and the pressure resistance breakdown performance after execution is generated; the signal of the paint film bending detection device is re-acquired, and the paint film flexibility index after execution is generated; the surface image is re-acquired through the surface scanning camera, and the surface defect quantification value of the paint film after execution is generated by the defect recognition algorithm.

[0080] Specifically, adjusting the production line based on target setpoints refers to sending the analyzed target values ​​of each operating parameter to the corresponding actuators, adjusting operational variables such as the temperature of each section of the baking oven, the production line operating speed, or the paint supply to the values ​​specified in the rules. Re-acquisition refers to re-acquiring quality-related signals through various online detection devices after the production line has stabilized under the new parameter conditions. Post-execution withstand voltage breakdown performance refers to the quantified value of the insulation withstand voltage capability of the enameled wire produced by the production line under the new parameter combination. Post-execution paint film flexibility index refers to the quantified value of the paint film's ability to withstand bending deformation under the new parameter combination. Post-execution paint film surface defect quantification value refers to the quantified value of the degree of appearance defects in the paint film under the new parameter combination.

[0081] The post-implementation pressure breakdown performance, post-implementation paint film flexibility index, and post-implementation paint film surface defect quantification value will be merged into a new set of post-implementation quality indicators.

[0082] Specifically, merging refers to aligning the three post-execution quality indicators, which respectively reflect electrical performance, mechanical performance, and appearance integrity, by timestamp and combining them into a unified data set. The new post-execution quality indicator set refers to the complete quality data combination produced by the production line under the action of candidate linkage control rules.

[0083] Obtain the baseline quality index set, calculate the change of each quality index in the new quality index set after execution relative to the baseline quality index set, and perform a weighted summation of the changes in each quality index to obtain the quality index improvement degree of the candidate linkage control rule.

[0084] Specifically, the change in each quality indicator refers to the difference between the new quality indicator value after implementation and the baseline quality indicator value. For positive indicators such as pressure resistance and paint film flexibility, a positive change indicates quality improvement; for negative indicators such as the quantification value of paint film surface defects, a negative change indicates quality improvement. Weighted summation refers to pre-assigning weight coefficients to each quality indicator based on its importance in production, multiplying the change in each indicator by its corresponding weight, and then summing them up. The degree of quality indicator improvement refers to a single numerical value that comprehensively reflects the degree of improvement in overall product quality brought about by the candidate linkage control rule; the larger this value, the better the rule's control effect.

[0085] After completing the execution and evaluation of a candidate linkage control rule, the production line is restored to the state defined by the effective linkage control rule before the execution of the candidate linkage control rule, and then the next candidate linkage control rule is executed.

[0086] Specifically, "restoration" refers to adjusting the operating parameters of the production line from the target setpoints of the current candidate rules back to the setpoints specified by the effective linkage control rules in operation when the benchmark quality index set was collected. The state defined by the effective linkage control rules before executing the candidate linkage control rule refers to the combination of operating parameters and control logic of the production line at the time the benchmark quality index set was collected.

[0087] S5. Retain candidate linkage control rules whose quality indicator improvement degree is higher than the preset improvement threshold, configure priority weights based on the improvement degree of the candidate linkage control rules, and add the candidate linkage control rules with configured priority weights to the rule base; discard candidate linkage control rules whose quality indicator improvement degree is lower than the preset improvement threshold, and execute the linkage control rule with the highest priority weight in the rule base.

[0088] In this embodiment, the process of retaining candidate linkage control rules whose quality indicator improvement degree is higher than a preset improvement threshold, configuring priority weights based on the improvement degree of the candidate linkage control rules, and adding the candidate linkage control rules with configured priority weights to the rule base; and discarding candidate linkage control rules whose quality indicator improvement degree is lower than the preset improvement threshold, and executing the linkage control rule with the highest priority weight in the rule base, includes: The improvement rate of the quality index of each candidate linkage control rule is compared with the preset improvement threshold.

[0089] Specifically, the quality indicator improvement degree refers to a single numerical value obtained through weighted summation, comprehensively reflecting the degree of improvement in overall product quality brought about by the candidate linkage control rules. The preset improvement threshold refers to a pre-set minimum improvement margin for determining whether a candidate linkage control rule has practical application value. Comparison refers to determining the relationship between the improvement degree value corresponding to each candidate rule and the preset improvement threshold.

[0090] When the improvement of the quality indicator is greater than the preset improvement threshold, the candidate linkage control rule is retained, the priority weight is determined based on the improvement of the quality indicator, and the candidate linkage control rule and its priority weight are added to the rule base.

[0091] Specifically, "retention" means that the candidate linkage control rule is deemed effective and has a significant positive effect on product quality under the new batch of insulating varnish conditions, and is retained. "Priority weight" refers to the weight value assigned to the rule based on its degree of improvement, used to measure the rule's priority among many effective rules. Rules with higher improvement degrees have larger priority weights, indicating better control effects and should be executed first. "Add to rule base" means that the specific control logic of the candidate linkage control rule, along with its priority weight, is stored in the rule base, making it a formal linkage control rule available for production line use.

[0092] When the improvement of the quality indicator is less than or equal to the preset improvement threshold, the candidate linkage control rule is discarded.

[0093] Specifically, "less than or equal to the preset improvement threshold" means that the improvement degree of the candidate linkage control rule has not reached the minimum application standard, indicating that the rule has a weak, ineffective, or even potentially negative effect on product quality improvement. "Discard" means removing the candidate linkage control rule from the candidate set and not retaining or adding it to the database. This step uses threshold filtering to eliminate rules with poor verification effects, preventing low-quality rules from consuming rule base storage resources or interfering with subsequent rule selection.

[0094] After the candidate linkage control rules are screened and added to the database, all valid linkage control rules in the rule base are sorted in descending order according to their priority weight, and the linkage control rule with the highest priority weight is selected as the currently effective linkage control rule.

[0095] Specifically, all valid linkage control rules refer to all linkage control rules in the rule base that have not been marked as inefficient and are in a usable state after the latest round of screening, including newly added candidate rules and existing rules that remain valid after evaluation. The linkage control rule with the highest priority weight refers to the rule ranked first, representing the control strategy that contributes most to quality improvement in the current rule base. Currently effective linkage control rules refer to the linkage control logic actually executed and followed by the production line during subsequent operation.

[0096] In this embodiment, adding the candidate linkage control rule and its priority weight to the rule base also includes adjusting the priority weights of existing linkage control rules in the rule base, as follows: Based on the currently effective linkage control rules, new quality indicator sets are periodically collected after execution to calculate the improvement degree of quality indicators in this operation.

[0097] Specifically, the currently effective linkage control rule refers to the rule with the highest priority weight selected after the latest rule sorting. Periodic data collection refers to acquiring quality data of the production line under the current rule control at preset time intervals. This time interval is determined comprehensively based on the production line's operating cycle time and quality inspection response time, for example, once every hour of production or once every standard length of spool is completed. The new quality index set after execution refers to the combination of the latest acquired quantitative values ​​of pressure breakdown performance, paint film flexibility index, and paint film surface defect quantitative values ​​during the continuous operation of the currently effective rule. The quality index improvement degree of this operation refers to the comprehensive improvement value calculated after weighted comparison of the new quality index set and the benchmark quality index set.

[0098] The improvement rate of the quality index obtained in this calculation is compared with the current priority weight of the currently effective linkage control rule.

[0099] Specifically, the quality indicator improvement obtained in this calculation refers to the latest comprehensive improvement value generated by periodic data collection and calculation. The current priority weight refers to the weight value currently stored in the rule base for this rule. Comparison refers to comparing the latest measured improvement with the weight value recorded in the rule base to determine whether the rule's performance in actual long-term operation is consistent with the level assessed when it was entered into the rule base.

[0100] If the improvement is greater than the current priority weight, the priority weight of the linkage control rule is increased by a preset increment; if the improvement is less than the current priority weight, the priority weight of the linkage control rule is decreased by a preset decrement; if the improvement is equal to the current priority weight, the priority weight remains unchanged.

[0101] Specifically, the preset increment refers to the pre-set step size for increasing the weight, and the preset decrement refers to the pre-set step size for decreasing the weight. The specific values ​​of these increments and decrements are predetermined based on the sensitivity requirements of the control rule priority in the enameled wire production process, the fluctuation range of quality indicator improvement, and the magnitude of the priority weight values. They are typically taken as 5% to 10% of the priority weight value range. For example, if the priority weight value range is between zero and one, the preset increment can be set to 0.05, and the preset decrement can be set to 0.05. The increment and decrement can use the same value to maintain symmetrical adjustment, or they can be set to different values ​​depending on the different emphases of the process on positive incentives and negative penalties.

[0102] The adjusted priority weights are written back to the rule base, and when the rules of the linkage control rule base are sorted again, the currently effective linkage control rules are re-determined based on the updated priority weights.

[0103] Specifically, writing back to the rule base means storing the adjusted new weight values ​​in the storage location of the corresponding rule in the rule base. The next rule sorting refers to re-ranking all valid linkage control rules in the rule base in descending order of priority when an event triggering rule reordering occurs. Re-determining the currently effective linkage control rule means selecting the new highest priority rule as the control strategy for subsequent production line execution based on the updated weight ranking results.

[0104] Please see Figure 2 This invention provides a collaborative control system for an intelligent production line of enameled wire, comprising: Parameter acquisition module: used to acquire the temperature of each heating zone of the baking oven, the operating speed of the production line, and the paint supply of the coating device as a set of operating parameters during the production of enameled wire; to acquire the online detection of the withstand voltage breakdown performance, the paint film flexibility index, and the quantitative value of the paint film surface defects as a set of quality indicators; and to acquire the rule base, which contains several linkage control rules. The correlation signal module is used to calculate the real-time influence weight of each operating parameter on each quality indicator based on the set of operating parameters and the set of quality indicators, and compare it with the historical influence weight calculated from the most recent N batches of operating parameter sets and quality indicator sets; when the deviation between any real-time influence weight and the historical influence weight exceeds a preset threshold, a correlation mode deviation signal is generated. New rule generation module: In response to the correlation mode deviation signal, it suspends the original linkage control rules and starts linkage control rule exploration: Based on the real-time influence weight, it selects the operating parameters whose influence weight on the quality index is higher than the preset effective threshold as key control parameters, determines the control direction and control amplitude for each key control parameter, and generates multiple candidate linkage control rules containing the target set values ​​of each key control parameter. Parameter adjustment module: used to execute each candidate linkage control rule one by one, collect the corresponding new quality index set after executing each candidate linkage control rule, and calculate the improvement degree of the quality index before and after the execution of the corresponding candidate linkage control rule; The optimization detection module is used to retain candidate linkage control rules whose quality indicator improvement degree is higher than the preset improvement threshold, configure priority weights based on the improvement degree of the candidate linkage control rules, and add the candidate linkage control rules with priority weights to the rule base; it discards candidate linkage control rules whose quality indicator improvement degree is lower than the preset improvement threshold, and executes the linkage control rule with the highest priority weight in the rule base.

[0105] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution.

[0106] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0107] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A collaborative control method for an intelligent production line of enameled wire, characterized in that, Includes the following steps: S1. Obtain the temperature of each heating zone of the baking oven, the operating speed of the production line, and the paint supply of the coating device as a set of operating parameters during the production of enameled wire; obtain the online detection of the withstand voltage breakdown performance, the paint film flexibility index, and the quantitative value of the paint film surface defects as a set of quality indicators; obtain the rule base, which contains several linkage control rules. S2. Based on the set of operating parameters and the set of quality indicators, calculate the real-time impact weight of each operating parameter on each quality indicator, and compare it with the historical impact weight calculated from the most recent N batches of operating parameter sets and quality indicator sets; when the deviation between any real-time impact weight and the historical impact weight exceeds a preset threshold, generate an association mode deviation signal. S3. In response to the deviation signal of the correlation mode, the original linkage control rules are suspended and linkage control rule exploration is initiated: Based on the real-time influence weight, the operating parameters whose influence weight on the quality indicators is higher than the preset effective threshold are selected as key control parameters. The control direction and control amplitude are determined for each key control parameter, and multiple candidate linkage control rules containing the target set values ​​of each key control parameter are generated. S4. Execute each candidate linkage control rule one by one. After executing each candidate linkage control rule, collect the corresponding new quality index set and calculate the improvement degree of the quality index before and after the execution of the corresponding candidate linkage control rule. S5. Retain candidate linkage control rules whose quality indicator improvement degree is higher than the preset improvement threshold, configure priority weights based on the improvement degree of the candidate linkage control rules, and add the candidate linkage control rules with configured priority weights to the rule base. Candidate linkage control rules with quality indicator improvement rates below the preset improvement threshold are discarded, and the linkage control rule with the highest priority weight in the rule base is executed.

2. The collaborative control method for an intelligent production line of enameled wire according to claim 1, characterized in that, The process of acquiring operating parameters, including the temperature of each heating zone in the baking oven, the operating speed of the production line, and the paint supply of the coating device during the enameled wire production process, and acquiring online detection parameters such as the withstand voltage breakdown performance, paint film flexibility index, and quantified values ​​of paint film surface defects, are used as a set of quality indicators, including: The output signals of the temperature sensors in each heating zone of the baking oven are collected, and after analog-to-digital conversion and temperature correction, the real-time temperature values ​​of each heating zone are generated. The pulse signal from the rotary encoder of the take-up drum is collected, and the pulse count and unit time conversion are used to generate the real-time operating speed of the production line. The output signal of the flow sensor of the painting device is collected, and after calibration and conversion, a real-time paint supply value is generated. The real-time temperature values ​​of each heating zone, the real-time operating speed of the production line, and the real-time paint supply values ​​are combined into a set of operating parameters. The breakdown detection signal of the high-voltage electrode array is collected to generate a quantitative index of the withstand voltage breakdown performance; The signal from the paint film bending detection device is collected to generate a paint film flexibility index. Surface images are acquired using a surface scanning camera, and quantitative values ​​of surface defects in the paint film are generated using a defect recognition algorithm. The quantitative indicators of pressure resistance and breakdown performance, the indicators of paint film flexibility, and the quantitative values ​​of paint film surface defects are combined into a set of quality indicators.

3. The collaborative control method for an intelligent production line of enameled wire according to claim 1, characterized in that, The real-time impact weight of each operating parameter on each quality indicator is calculated based on the set of operating parameters and the set of quality indicators, and compared with the historical impact weights calculated from the most recent N batches of operating parameter sets and quality indicator sets. When the deviation between any real-time influence weight and the historical influence weight exceeds a preset threshold, a correlation pattern deviation signal is generated, including: The set of operating parameters and the set of quality indicators are synchronized and aligned to obtain the synchronized parameter sequence and indicator sequence. Using each quality indicator in the quality indicator set as the dependent variable and all operating parameters in the operating parameter set as independent variables, the regression coefficients are calculated using the multiple linear regression method. The regression coefficients are used as real-time impact weights to construct a real-time impact weight matrix; Obtain the historical weight matrix calculated from the parameter set and quality index set of the most recent N batches of operation; Calculate the absolute value of the difference between the elements at the same position in the real-time influence weight matrix and the historical influence weight matrix, and divide it by the historical influence weight to obtain the relative deviation; When any relative deviation exceeds a preset deviation threshold, a correlation mode deviation signal is generated.

4. The collaborative control method for an intelligent production line of enameled wire according to claim 3, characterized in that, The method of calculating regression coefficients using a multiple linear regression approach, with each quality indicator in the quality indicator set as the dependent variable and all operating parameters in the operating parameter set as independent variables, includes: Construct an independent variable matrix using the running parameter values ​​at all sampling times in the synchronized parameter sequence. The rows of the independent variable matrix correspond to the sampling times, and the columns of the independent variable matrix correspond to the running parameters. Construct a dependent variable vector using the values ​​of the current quality index at each sampling time in the synchronized index sequence; The regression coefficients are calculated using the least squares method based on the independent variable matrix and the dependent variable vector.

5. The collaborative control method for an intelligent production line of enameled wire according to claim 4, characterized in that, In response to the correlation mode deviation signal, the original linkage control rules are suspended, and linkage control rule exploration is initiated: based on real-time influence weights, operating parameters with influence weights on quality indicators higher than preset effective thresholds are selected as key control parameters. The control direction and amplitude are determined for each key control parameter, and multiple candidate linkage control rules containing target set values ​​for each key control parameter are generated, including: In response to the correlation mode deviation signal, the original linkage control rules are suspended, and the real-time impact weight matrix is ​​read; For any operating parameter, if the real-time impact weight of at least one associated quality indicator is higher than a preset effective threshold, then the operating parameter is marked as a key control parameter. Obtain the adjustable range of each key control parameter, and set multiple incremental control amplitudes based on the adjustable range; Using the target setpoints of the key control parameters in the original linkage control rules as the baseline, multiple candidate target setpoints are generated for each key control parameter based on the control direction and multiple control amplitudes. The multiple candidate target setpoints of each key control parameter are combined by Cartesian product to generate multiple candidate linkage control rules. Each candidate linkage control rule consists of a set of target setpoints for each key control parameter. The control direction is determined according to the positive or negative sign of the real-time influence weight. When the real-time influence weight is positive, the control direction is to increase the value of the operating parameter; when the real-time influence weight is negative, the control direction is to decrease the value of the operating parameter.

6. The collaborative control method for an intelligent production line of enameled wire according to claim 5, characterized in that, The process of generating multiple candidate linkage control rules also includes processing the existing linkage control rules, as follows: For any operating parameter that is not marked as a key control parameter, obtain the original linkage control rules in the rule base that contain the control logic of that operating parameter; Calculate the average of the real-time impact weights of all quality indicators associated with this operating parameter; If the average value is lower than the preset effective threshold, the original linkage control rule will be marked as an inefficient rule.

7. The collaborative control method for an intelligent production line of enameled wire according to claim 6, characterized in that, The process involves executing each candidate linkage control rule one by one, collecting the corresponding new quality indicator set after executing each candidate linkage control rule, and calculating the improvement degree of the quality indicators before and after the execution of the corresponding candidate linkage control rule, as follows: Before executing any candidate linkage control rule, the effective linkage control rule with the highest priority weight in the rule base is run, and the set of running parameters and quality indicators are collected. The set of quality indicators collected at this time is used as the benchmark quality indicator set. Execute each candidate linkage control rule one by one. For each candidate linkage control rule: The target setpoints of each operating parameter are parsed from the candidate linkage control rules; The production line is adjusted based on the target setpoint, and the breakdown detection signal of the high voltage electrode array is re-acquired to generate the withstand voltage breakdown performance after execution. The signal from the paint film bending detection device is re-acquired, and the paint film flexibility index after execution is generated. The surface image is re-acquired by a surface scanning camera, and the defect identification algorithm generates the quantified value of the surface defect of the paint film after execution. The pressure breakdown performance, the paint film flexibility index, and the paint film surface defect quantification value after implementation will be combined into a new set of quality indicators after implementation. Obtain the baseline quality index set, calculate the change of each quality index in the new quality index set after execution relative to the baseline quality index set, and perform a weighted summation of the changes in each quality index to obtain the quality index improvement degree of the candidate linkage control rule. After completing the execution and evaluation of a candidate linkage control rule, the production line is restored to the state defined by the effective linkage control rule before the execution of the candidate linkage control rule, and then the next candidate linkage control rule is executed.

8. The collaborative control method for an intelligent production line of enameled wire according to claim 1, characterized in that, The candidate linkage control rules with a quality index improvement degree higher than the preset improvement threshold are retained, and priority weights are configured based on the improvement degree of the candidate linkage control rules. The candidate linkage control rules with the configured priority weights are then added to the rule base. Candidate linkage control rules whose quality indicator improvement is lower than the preset improvement threshold are discarded, and the linkage control rules with the highest priority weight in the rule base are executed, including: Compare the improvement rate of the quality index of each candidate linkage control rule with the preset improvement threshold; When the improvement of the quality indicator is greater than the preset improvement threshold, the candidate linkage control rule is retained, the priority weight is determined based on the improvement of the quality indicator, and the candidate linkage control rule and its priority weight are added to the rule base. When the improvement of the quality indicator is less than or equal to the preset improvement threshold, the candidate linkage control rule is discarded. After the candidate linkage control rules are screened and added to the database, all valid linkage control rules in the rule base are sorted in descending order according to their priority weight, and the linkage control rule with the highest priority weight is selected as the currently effective linkage control rule.

9. The collaborative control method for an intelligent production line of enameled wire according to claim 8, characterized in that, Adding the candidate linkage control rule and its priority weight to the rule base also includes adjusting the priority weight of existing linkage control rules in the rule base, as follows: Based on the current effective linkage control rules, during the operation process, the new quality indicator set after execution is periodically collected, and the improvement degree of quality indicators in this operation is calculated. The improvement rate of the quality index calculated in this study is compared with the current priority weight of the currently effective linkage control rule; If the improvement degree is greater than the current priority weight, the priority weight of the linkage control rule is increased by a preset increment; if the improvement degree is less than the current priority weight, the priority weight of the linkage control rule is decreased by a preset decrement; if the improvement degree is equal to the current priority weight, the priority weight remains unchanged. The adjusted priority weights are written back to the rule base, and when the rules of the linkage control rule base are sorted again, the currently effective linkage control rules are re-determined based on the updated priority weights.

10. A collaborative control system for an intelligent enameled wire production line, used in conjunction with the collaborative control method for an intelligent enameled wire production line according to any one of claims 1 to 9, characterized in that, include: Parameter acquisition module: used to acquire the temperature of each heating zone of the baking oven, the operating speed of the production line, and the paint supply of the coating device as a set of operating parameters during the production of enameled wire; to acquire the online detection of the withstand voltage breakdown performance, the paint film flexibility index, and the quantitative value of the paint film surface defects as a set of quality indicators; and to acquire the rule base, which contains several linkage control rules. The correlation signal module is used to calculate the real-time influence weight of each operating parameter on each quality indicator based on the set of operating parameters and the set of quality indicators, and compare it with the historical influence weight calculated from the most recent N batches of operating parameter sets and quality indicator sets; when the deviation between any real-time influence weight and the historical influence weight exceeds a preset threshold, a correlation mode deviation signal is generated. New rule generation module: In response to the correlation mode deviation signal, it suspends the original linkage control rules and starts linkage control rule exploration: Based on the real-time influence weight, it selects the operating parameters whose influence weight on the quality index is higher than the preset effective threshold as key control parameters, determines the control direction and control amplitude for each key control parameter, and generates multiple candidate linkage control rules containing the target set values ​​of each key control parameter. Parameter adjustment module: used to execute each candidate linkage control rule one by one, collect the corresponding new quality index set after executing each candidate linkage control rule, and calculate the improvement degree of the quality index before and after the execution of the corresponding candidate linkage control rule; The optimization detection module is used to retain candidate linkage control rules whose quality indicator improvement degree is higher than the preset improvement threshold, configure priority weights based on the improvement degree of the candidate linkage control rules, and add the candidate linkage control rules with configured priority weights to the rule base. Candidate linkage control rules with quality indicator improvement rates below the preset improvement threshold are discarded, and the linkage control rule with the highest priority weight in the rule base is executed.

Images