Intelligent control method for packaging machine

Abstract

The invention discloses an intelligent control method for a packaging machine, and belongs to the technical field of intelligent control of lithium battery production equipment, and the method comprises the following steps: S1, deploying a comprehensive detection unit at a key node of a lithium battery pole piece winding packaging machine, generating a pole piece state data packet, and storing the pole piece state data packet in a preset database; s2, when the thickness change trend value of the active coating of the pole piece exceeds a preset fluctuation threshold value, generating a heat conduction compensation grade and a pole piece deviation correction guide roller adjusting instruction set; s3, the thermocompression bonding temperature of a pole piece thermocompression bonding station is adjusted based on the heat conduction compensation grade, and a guide roller system of a pole piece traction area is adjusted according to the pole piece deviation rectification guide roller adjusting instruction set; and S4, calculating a prediction execution deviation value of the heat conduction compensation grade and the guide roller adjustment instruction set in real time, and updating the instruction generation process of the step S2. The electrochemical performance and the safety characteristic of the battery are optimized through the heat conduction compensation level and the guide roller adjusting instruction set.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent control of lithium battery production equipment, and particularly relates to an intelligent control method of a packaging machine. BACKGROUND

[0002] As a core component of modern high-performance energy storage, the manufacturing process of lithium batteries is precise and complex. Among them, the winding of the pole piece and the packaging of the battery are key processes, which directly affect the energy density, safety and consistency of the battery. Winding needs to accurately control the tension, speed and alignment of the pole piece and the separator to prevent defects such as deformation and short circuit; packaging uses specific processes (such as laser welding, heat sealing) to ensure sealing and structural stability according to different forms such as square, cylindrical or soft package, to meet the diversified application requirements.

[0003] In the manufacturing process of lithium batteries, the winding and packaging of the pole piece are key processes. The working principle of the traditional intelligent control method of the packaging machine for lithium battery manufacturing is as follows: first, the tension, position and alignment data of the pole piece and the separator are collected in real time by a high-precision sensor network and transmitted to a central processor; the processor calculates and outputs control instructions to the actuator (such as a servo motor, a deviation correction device) based on pre-set process parameters (such as winding speed, tension threshold) and dynamic adjustment algorithms (such as adaptive PID control, Kalman filtering), to adjust the unwinding / winding speed and tension roller position in real time, ensuring that the pole piece and the separator maintain constant tension and accurate alignment during winding; at the same time, a human-machine interface (HMI) is integrated to monitor the working condition in real time, and a data tracing system is used to record the whole process parameters to form a quality management closed loop, finally realizing high-precision, high-speed and stable production of battery winding.

[0004] The traditional lithium battery winding and packaging machine is difficult to adapt to the dynamic changes of the pole piece tension, temperature and coating thickness during control, resulting in poor winding alignment accuracy, pole piece damage or insufficient hot pressure sealing strength, which directly affects the electrochemical performance and safety characteristics of the battery. The existing technology lacks a closed-loop control mechanism for the material properties of lithium batteries (such as active material coating thickness, heat sensitivity), resulting in production yield fluctuations.

[0005] Therefore, it is necessary to provide an intelligent control method of a packaging machine to solve the above problems. SUMMARY

[0006] The technical problem to be solved by the present application is to overcome the shortcomings of the traditional lithium battery winding and packaging machine in control, which is difficult to adapt to the dynamic changes of the pole piece tension, temperature and coating thickness, resulting in poor winding alignment accuracy, pole piece damage or insufficient hot pressure sealing strength, which directly affects the electrochemical performance and safety characteristics of the battery. The existing technology lacks a closed-loop control mechanism for the material properties of lithium batteries (such as active material coating thickness, heat sensitivity), resulting in production yield fluctuations, and provides an intelligent control method of a packaging machine.

[0007] To solve the above technical problems, one technical solution adopted by the present application is to provide a packaging machine intelligent control method, comprising the following steps:

[0008] S1, deploying a comprehensive detection unit at the key nodes of the lithium battery pole piece winding packaging machine, synchronously collecting the thickness change trend value of the pole piece active coating in the preset pole piece unwinding mechanism, the temperature distribution characteristic map of the preset pole piece hot pressing work station and the tension fluctuation change rate of the preset pole piece traction area, generating a pole piece state data packet, and storing it in a preset database;

[0009] S2, when the thickness change trend value of the pole piece active coating exceeds the preset fluctuation threshold value, generating a heat conduction compensation level and a pole piece deviation correction guide roller adjustment instruction set according to the temperature distribution characteristic map and the tension fluctuation change rate;

[0010] S3, adjusting the hot pressing temperature of the pole piece hot pressing work station based on the heat conduction compensation level, and adjusting the guide roller system of the pole piece traction area according to the pole piece deviation correction guide roller adjustment instruction set;

[0011] S4, real-time calculating the predicted execution deviation value of the heat conduction compensation level and the guide roller adjustment instruction set, when the predicted execution deviation value exceeds the preset model deviation threshold value, dynamically updating the pre-trained prediction model parameters based on the pole piece state data packet, and synchronizing the updated model parameters to the instruction generation process of step S2.

[0012] The present application is further provided: the comprehensive detection unit comprises a non-contact thickness sensor arranged at the pole piece unwinding mechanism, an infrared thermal imaging array arranged at the pole piece hot pressing work station and a strain type tension sensor installed at the pole piece traction area, and the non-contact thickness sensor, the infrared thermal imaging array and the strain type tension sensor are synchronously output to the preset central processor through the preset high-speed data bus.

[0013] The present application is further provided: the generation step of the pole piece state data packet in step S1 is as follows:

[0014] S11, the central processor receives the pole piece coating thickness signal sequence output by the non-contact thickness sensor in real time, and extracts the thickness change trend value of the pole piece active coating through time domain noise reduction processing;

[0015] S12, the central processor synchronously analyzes the infrared radiation data of the infrared thermal imaging array, and generates a temperature distribution characteristic map based on a preset algorithm;

[0016] S13, the central processor dynamically processes the original electric signal of the strain type tension sensor, and converts and outputs a tension fluctuation change rate function curve;

[0017] S14, the central processor compresses the thickness change trend value of the pole piece active coating, the temperature distribution characteristic map and the tension fluctuation change rate function curve by time alignment, generates a pole piece state data packet, and stores it in the database.

[0018] The application is further provided that the specific content in step S2 is:

[0019] When the thickness change trend value of the pole piece active coating exceeds the preset fluctuation threshold:

[0020] S2a, when the thickness change trend value of the pole piece active coating continuously exceeds the fluctuation threshold, generating a heat conduction compensation level according to the temperature distribution characteristic map, and generating a pole piece deviation correction guide roller adjustment instruction set according to the coupling relationship between the thickness change trend value of the pole piece active coating and the tension fluctuation change rate;

[0021] S2b, if the thickness change trend value of the pole piece active coating has a step mutation and exceeds the fluctuation threshold in a continuous period, a preset pole piece switching mode is activated;

[0022] The generation step of the heat conduction compensation level in step S2a is as follows:

[0023] S2a1, the central processor performs time sequence analysis on the temperature distribution characteristic map, inputs the analysis result into a pre-trained prediction model, and outputs a temperature field evolution thermal map in the future period;

[0024] S2a2, scanning the temperature field evolution thermal map, when detecting that the temperature difference anomaly exceeding the preset safety threshold appears continuously, recording the abnormal period and the temperature difference amplitude, and synchronously capturing the thickness change trend value of the pole piece active coating in the abnormal period;

[0025] S2a3, combining the pre-stored current pole piece material characteristic parameters in the database, and the time domain coupling analysis of the temperature difference amplitude, the thickness change trend value of the pole piece active coating in the abnormal period, generating a dynamic heat field balance evaluation value;

[0026] S2a4, when the dynamic thermal field balance evaluation value is in a preset basic interval, a basic level compensation instruction is generated; if the dynamic thermal field balance evaluation value is in a preset enhanced interval, an enhanced level compensation instruction is activated; when the dynamic thermal field balance evaluation value exceeds a preset emergency interval threshold value, an emergency level compensation instruction is triggered, and a fluctuation threshold value is calculated in real time by the central processor based on an intensity value of the currently generated compensation instruction and a thickness change trend value of the active coating of the pole piece in the database; when the thickness change trend value of the active coating of the pole piece actually detected exceeds the fluctuation threshold value, the enhanced level compensation instruction or the emergency level compensation instruction is activated as the heat conduction compensation level; and when the thickness change trend value of the active coating of the pole piece actually detected is less than or equal to the fluctuation threshold value, the basic level compensation instruction is activated as the heat conduction compensation level.

[0027] The application further provides that the specific step of synchronously capturing the thickness change trend value of the active coating of the pole piece in the abnormal period in step S2a2 is:

[0028] S2a21, extracting a temperature change gradient curve corresponding to the abnormal period from the temperature field evolution thermodynamic diagram;

[0029] S2a22, inputting the temperature change gradient curve into a pre-trained thickness response prediction model to output a thickness response prediction value of the active coating of the pole piece in the abnormal period;

[0030] S2a23, correcting the thickness response prediction value of the active coating of the pole piece in combination with the thickness change trend value of the active coating of the pole piece in the historical same period of the abnormal period stored in the database;

[0031] S2a24, taking the corrected thickness response prediction value of the active coating of the pole piece as the capture result of the thickness change trend value of the active coating of the pole piece in the current abnormal period.

[0032] The application further provides that the generation step of the pole piece deviation correction guide roller adjustment instruction set in step S2a is as follows:

[0033] S2a01, the central processor compares the thickness change trend value of the active coating of the pole piece in the pole piece state data packet with the tension fluctuation change rate, compares a fluctuation pattern curve of the thickness change trend value of the active coating of the pole piece in the historical same period of the abnormal period stored in the database, identifies an abnormal section position and a duration in the thickness change trend value of the active coating of the pole piece;

[0034] S2a02, correcting the offset amount of the abnormal section position based on the pole piece material elastic recovery capability parameter in the database, and extracting an extreme value distribution characteristic of the tension fluctuation change rate in the same time period to generate a guide roller displacement compensation amount and a dynamic response priority coefficient.

[0035] S2a03, the guide roller displacement compensation, the dynamic response priority coefficient and the thickness change trend value of the active coating of the pole piece in the database are introduced into a preset coupling relationship operation model, and a pole piece deviation correction guide roller adjustment instruction set is output, the pole piece deviation correction guide roller adjustment instruction set includes a guide roller displacement amplitude, an angle correction amount and an action time length.

[0036] The application is further provided that: the specific content of the pole piece switching mode in step S2b is:

[0037] S2b1, based on the current thickness change trend value of the active coating of the pole piece and the historical thickness change trend value of the active coating of the pole piece in the database, a fluctuation threshold value is calculated in real time by the central processor, when the current thickness change trend value of the active coating of the pole piece occurs step mutation and exceeds the fluctuation threshold value in the continuous period, the guide roller system of the pole piece traction area is controlled to implement position locking and keep the current tension stable;

[0038] S2b2, based on the current thickness change trend value of the active coating of the pole piece, the matching heat conduction compensation level and the pole piece deviation correction guide roller adjustment instruction set in the database are retrieved and called.

[0039] The application is further provided that: the specific content of adjusting the hot pressing temperature of the pole piece hot pressing station in step S3 is:

[0040] S31, the central processor calls different heat power correction coefficients corresponding to the basic level compensation instruction, the enhanced level compensation instruction and the emergency level compensation instruction in the database according to the heat conduction compensation level;

[0041] S32, the weight of the heat power correction coefficient is adjusted in combination with the pre-stored heat conduction rate of the current pole piece material in the database, and the hot pressing temperature adjustment is implemented in the pole piece hot pressing station according to the adjusted heat power correction coefficient.

[0042] The application is further provided that: the specific content of adjusting the guide roller system of the pole piece traction area according to the pole piece deviation correction guide roller adjustment instruction set in step S3 is as follows: the central processor calculates the real-time displacement amplitude and angle correction amount of the guide roller system based on the guide roller displacement compensation and the dynamic response priority coefficient, generates an adaptive adjustment weight factor of the guide roller action time length in combination with the pre-stored current pole piece material elastic recovery ability parameter and the extreme value distribution characteristics of the tension fluctuation rate in the database, generates a guide roller displacement execution instruction set according to the adaptive adjustment weight factor, and synchronously drives the guide roller system to execute dynamic deviation correction according to the displacement amplitude, angle correction amount and action time length in the guide roller displacement execution instruction set.

[0043] The application is further provided that: the specific content of the step S4 is that the central processor calculates the predicted execution deviation value of the heat conduction compensation level and the guide roller adjustment instruction set in real time, when the predicted execution deviation value exceeds a preset model deviation threshold, based on the thickness change trend value of the pole piece active coating, the temperature distribution feature map and the tension fluctuation change rate function curve in the pole piece state data package, dynamically evaluate the confidence decay gradient of the pre-trained prediction model, and reconstruct the weight distribution and feature extraction layer structure of the pre-trained prediction model through a preset algorithm, then map the updated model parameters to the instruction generation process of the step S2 in real time, and optimize the generation logic of the heat conduction compensation level and the guide roller adjustment instruction set synchronously.

[0044] The beneficial effects of the application are as follows:

[0045] 1. The application synchronously collects the thickness change trend value of the pole piece active coating, the temperature distribution feature map and the tension fluctuation change rate, dynamically generates the heat conduction compensation level and the guide roller adjustment instruction set, accurately adapts to the dynamic changes of tension, temperature and coating thickness in the production process of lithium battery pole pieces, significantly improves the winding alignment accuracy, reduces the damage of pole pieces, enhances the heat sealing strength, and thus optimizes the electrochemical performance and safety characteristics of the battery;

[0046] 2. The application optimizes the heat conduction compensation level and the guide roller adjustment instruction set in real time, forms a continuously iterative intelligent control closed loop, effectively overcomes the influence of lithium battery material characteristic fluctuation, stabilizes the production yield, and improves the consistency and reliability of the manufacturing process. BRIEF DESCRIPTION OF DRAWINGS

[0047] Figure 1 The method flowchart of the application;

[0048] Figure 2 The generation step flowchart of the heat conduction compensation level of the application;

[0049] Figure 3 The specific capture step flowchart of the thickness change trend value of the pole piece active coating of the application. DETAILED DESCRIPTION

[0050] The preferred embodiments of the application will be described in detail below with reference to the accompanying drawings, so that the advantages and features of the application can be more easily understood by those skilled in the art, and the protection scope of the application can be more clearly defined.

[0051] Please refer to Figure 1 - Figure 3 A packaging machine intelligent control method, comprising the following steps:

[0052] S1, deploying a comprehensive detection unit at a key node of a lithium battery pole piece winding packaging machine, synchronously collecting thickness change trend values of pole piece active coating in a preset pole piece unwinding mechanism, temperature distribution characteristic maps of a preset pole piece hot pressing work station, and tension fluctuation change rates of a preset pole piece traction area, generating a pole piece state data packet, and storing the pole piece state data packet in a preset database;

[0053] The comprehensive detection unit comprises a non-contact thickness sensor arranged at the pole piece unwinding mechanism, an infrared thermal imaging array arranged at the pole piece hot pressing work station, and a strain type tension sensor arranged at the pole piece traction area. The non-contact thickness sensor, the infrared thermal imaging array, and the strain type tension sensor are synchronously output to a preset central processing unit through a preset high-speed data bus.

[0054] The high-speed data bus is a special data communication channel connecting the non-contact thickness sensor, the infrared thermal imaging array, and the strain type tension sensor in the comprehensive detection unit. Time-triggered deterministic transmission protocol is used to realize microsecond-level synchronous data output of the three sensors. The central processing unit is an embedded industrial computing platform for executing control logic of steps S1-S4. The central processing unit has a multi-core heterogeneous architecture and can process data streams of the non-contact thickness sensor, the infrared thermal imaging array, and the strain type tension sensor in real time.

[0055] In step S1, the pole piece state data packet is generated as follows:

[0056] S11, the central processing unit receives a pole piece coating thickness signal sequence output by the non-contact thickness sensor in real time, extracts thickness change trend values of the pole piece active coating through time domain noise reduction processing;

[0057] S12, the central processing unit synchronously analyzes infrared radiation data of the infrared thermal imaging array, and generates a temperature distribution characteristic map based on a preset algorithm;

[0058] S13, the central processing unit dynamically processes original electric signals of the strain type tension sensor, and converts and outputs a tension fluctuation change rate function curve;

[0059] S14, the central processing unit performs time scale alignment and compression on the thickness change trend values of the pole piece active coating, the temperature distribution characteristic map, and the tension fluctuation change rate function curve, generates a pole piece state data packet, and stores the pole piece state data packet in the database.

[0060] The central processing unit uniformly calibrates time series data of the thickness change trend values of the pole piece active coating, acquisition time stamps of the temperature distribution characteristic map, and time coordinate axes of the tension fluctuation change rate function curve to a microsecond-level accurate time reference. On the premise of retaining key change points of each data stream, the central processing unit eliminates inherent sampling noise and environmental interference signals of the non-contact thickness sensor, the infrared thermal imaging array, and the strain type tension sensor, and finally integrates the data into a material state data packet.

[0061] S2, when the thickness variation trend value of the pole piece active coating exceeds the preset fluctuation threshold, generating a heat conduction compensation level and a pole piece deviation correction guide roller adjustment instruction set according to the temperature distribution characteristic map and the tension fluctuation change rate;

[0062] The specific content in step S2 is:

[0063] When the thickness variation trend value of the pole piece active coating exceeds the preset fluctuation threshold:

[0064] S2a, when the thickness variation trend value of the pole piece active coating continuously exceeds the fluctuation threshold, generating a heat conduction compensation level according to the temperature distribution characteristic map, and generating a pole piece deviation correction guide roller adjustment instruction set according to the coupling relationship between the thickness variation trend value of the pole piece active coating and the tension fluctuation change rate;

[0065] S2b, if the thickness variation trend value of the pole piece active coating has a step mutation and exceeds the fluctuation threshold in a continuous period, activating a preset pole piece switching mode;

[0066] The generation step of the heat conduction compensation level in step S2a is as follows:

[0067] S2a1, the central processing unit performs time sequence analysis on the temperature distribution characteristic map, inputs the analysis result into a pre-trained prediction model, and outputs a temperature field evolution heat map in the future period;

[0068] The prediction model refers to a spatiotemporal feature mapping network constructed by deep supervision training based on the million-level temperature distribution characteristic map samples and corresponding temperature field evolution data collected by the lithium battery pole piece winding packaging machine during operation, which can predict the conduction path and heat balance evolution trend of the pole piece hot pressing joint temperature field in a specific future period according to the gradient distribution, abnormal heat zone position and temperature change rate characteristics of the current temperature distribution characteristic map.

[0069] S2a2, scanning the temperature field evolution heat map, when detecting that the temperature difference anomaly exceeding the preset safety threshold continuously appears, recording the abnormal period and the temperature difference amplitude, and simultaneously capturing the thickness variation trend value of the pole piece active coating in the abnormal period;

[0070] Preferably, the safety threshold is dynamically generated by the pre-stored pole piece thermal deformation safety critical parameter in the database, and when the temperature difference in the temperature field evolution heat map is greater than the pole piece thermal deformation safety critical value for two or more consecutive sampling periods, it is determined that the temperature difference is abnormal; in a preferred embodiment, for an NCM811 ternary lithium battery pole piece with an energy density of 280 Wh / kg, the pre-stored pole piece thermal deformation safety critical parameter can be set to 5℃, that is, when the temperature difference is greater than 5℃ for two consecutive sampling periods, it is determined that the temperature difference is abnormal.

[0071] The specific steps of synchronously capturing the thickness change trend value of the active coating of the pole piece in the abnormal period in step S2a2 are as follows:

[0072] S2a21, extracting a temperature change gradient curve corresponding to the abnormal period from the temperature field evolution thermodynamic map;

[0073] S2a22, inputting the temperature change gradient curve into the pre-trained thickness response prediction model to output a thickness response prediction value of the active coating of the pole piece in the abnormal period;

[0074] Preferably, the thickness response prediction model refers to a special analysis engine constructed based on a dataset of temperature gradient change data and thickness change trend values of the active coating of the pole piece synchronously collected in the historical lithium battery pole piece winding packaging machine production line through supervised machine learning, which is an intelligent mapping system for predicting the real-time thickness change trend value of the current active coating of the pole piece under thermal stress according to the temperature change gradient curve in the temperature field evolution thermodynamic map;

[0075] S2a23, combining the thickness change trend value of the active coating of the pole piece in the historical same period (referring to a historical running period of the lithium battery pole piece winding packaging machine with the following similarities in any dimension: pole piece dimension: historical production data of the same pole piece supplier / the same batch number; time dimension: data window of the same equipment running period within the past 24 hours; working condition dimension: historical running record under the same environmental temperature and humidity conditions) of the abnormal period to correct the thickness response prediction value of the active coating of the pole piece;

[0076] The correction method of the thickness response prediction value of the active coating of the pole piece: the central processing unit calculates the statistical characteristic deviation amount of the thickness response prediction value of the current active coating of the pole piece from the historical same period data according to the dataset of the thickness change trend value of the active coating of the pole piece in the historical same period of the abnormal period; the original thickness response prediction value of the active coating of the pole piece is maintained when the thickness response prediction value of the current active coating of the pole piece is within the ±10% interval of the historical thickness change fluctuation range; if the thickness response prediction value of the current active coating of the pole piece exceeds the interval, a weighted fusion algorithm is used to correct the thickness response prediction value of the active coating of the pole piece based on the mean and variance of the historical same period data, and the specific operation is as follows: correction value = original thickness response prediction value of the active coating of the pole piece × variance weight of historical data + historical mean × (1-variance weight), wherein the variance weight is dynamically calculated by the standard deviation multiple of the thickness response prediction value of the active coating of the pole piece deviating from the historical mean;

[0077] S2a24, taking the corrected thickness response prediction value of the active coating of the pole piece as the capture result of the thickness change trend value of the active coating of the pole piece in the current abnormal period;

[0078] S2a3, combine the pre-stored current electrode material characteristic parameters in the database, and the temperature difference amplitude in the abnormal period, the thickness change trend value of the electrode active coating for time domain coupling analysis, and generate a dynamic thermal field balance evaluation value;

[0079] The pre-stored current electrode material characteristic parameters in the database include the type of electrode active material (such as ternary material NCM811, lithium iron phosphate LFP or lithium cobaltate LCO) and its basic physical characteristics (true density, tap density), the pore structure parameters of the coating (porosity, average pore size and pore size distribution, tortuosity), thermal characteristics (thermal conductivity rate, thermal deformation safety critical parameter, specific heat capacity), mechanical characteristics (elastic recovery capability parameter, yield strength, adhesion of coating and current collector), electrochemical characteristics (lithium ion diffusion coefficient, electronic conductivity), component ratio (weight percentage of active material, conductive agent, binder) and geometric characteristics (coating thickness, area density, compaction density);

[0080] The generation steps of the dynamic thermal field balance evaluation value are as follows:

[0081] S2a3-1, the central processor extracts the temperature difference amplitude time sequence in the abnormal period, the thickness change trend value time sequence of the electrode active coating and the current electrode material characteristic parameters (including thermal conductivity rate, thermal deformation safety critical parameter, specific heat capacity, elastic recovery capability parameter, yield strength, adhesion of coating and current collector, porosity, average pore size and pore size distribution, tortuosity), respectively calculates the time sequence characteristic quantity (including mean, variance, gradient change rate) of each parameter;

[0082] S2a3-2, time domain alignment is performed on the temperature difference amplitude time sequence characteristic quantity and the thickness change trend value time sequence characteristic quantity of the electrode active coating, and an initial thermal-force coupling factor is generated by using a weighted fusion algorithm, wherein the weighting coefficients are dynamically adjusted according to the thermal characteristics (thermal conductivity rate, specific heat capacity) and the mechanical characteristics (elastic recovery capability parameter, yield strength) in the current electrode material characteristic parameters;

[0083] S2a3-3, based on the initial thermal-force coupling factor and the coating structure parameters (porosity, average pore size and pore size distribution, tortuosity) in the current electrode material characteristic parameters, multi-dimensional regression analysis is performed, the adhesion parameter is introduced as a constraint condition, and an uncorrected thermal field balance index is calculated;

[0084] S2a3-4, call the dynamic thermal field balance evaluation value data set of the same period in the database, take the type of current electrode active material (ternary material NCM811 / lithium iron phosphate LFP / lithium cobaltate LCO) and geometric characteristics (coating thickness, area density, compaction density) as the screening conditions, extract the statistical reference range, correct the uncorrected thermal field balance index by standardization, and finally output the dynamic thermal field balance evaluation value;

[0085] S2a4, when the dynamic thermal field balance evaluation value is in the preset basic interval, a basic level compensation instruction is generated; if the dynamic thermal field balance evaluation value is in the preset enhanced interval, an enhanced level compensation instruction is activated; when the dynamic thermal field balance evaluation value exceeds the preset emergency interval threshold value, an emergency level compensation instruction is triggered, and a fluctuation threshold value is calculated in real time by the central processor based on the intensity value of the currently generated compensation instruction and the thickness change trend value of the historical active coating of the pole piece in the database; when the actual detected thickness change trend value of the active coating of the pole piece exceeds the fluctuation threshold value, the enhanced level compensation instruction or the emergency level compensation instruction is immediately activated as the heat conduction compensation level, and when the actual detected thickness change trend value of the active coating of the pole piece is less than or equal to the fluctuation threshold value, the basic level compensation instruction is activated as the heat conduction compensation level.

[0086] The basic level compensation instruction is triggered when the dynamic thermal field balance evaluation value is in the basic interval of 0-40. The instruction performs minimum intervention adjustment for slight thermal field fluctuation. The central processor calls the preset basic thermal power correction coefficient (the reference value is 0.5) in the database to implement micro compensation for the temperature of the pole piece hot pressing station (usually the adjustment range is ≤±5°C), while maintaining the existing parameters of the guide roller system unchanged. The core target is to suppress the initial temperature drift and calculate the dynamic fluctuation threshold value (formula: historical standard deviation mean × 1.15) through linear interpolation of the historical standard deviation mean and the compensation intensity value, to ensure that the system prevents over-adjustment while maintaining production continuity.

[0087] The enhanced level compensation instruction is activated when the dynamic thermal field balance evaluation value is in the enhanced interval of 41-80. The instruction performs medium-intensity intervention for persistent thermal field abnormalities. The central processor increases the thermal power correction coefficient to 1.2 level to significantly adjust the temperature of the hot pressing station (the adjustment range is usually ±5-15°C). At the same time, the guide roller displacement compensation instruction (such as guide roller fine adjustment ±0.3-0.8mm) is generated in combination with the tension fluctuation change rate, and the fluctuation threshold formula (historical standard deviation mean × 1.36) is used to dynamically expand the tolerance range to cope with the cumulative effect of the coupling deviation of the pole piece coating thickness and temperature, and to avoid coating peeling or deformation caused by local thermal stress concentration.

[0088] Emergency level compensation instruction: the emergency level compensation instruction is triggered immediately when the dynamic thermal field balance evaluation value exceeds the emergency interval threshold value of 80, which takes the highest intensity intervention for serious thermal field out of control, the central processor sets the thermal power correction coefficient to level 2.0 and starts the extreme temperature compensation (adjustment range ≥ ± 15℃), synchronously forces the guide roller system to perform large position correction (displacement amount ≥ ± 1.0mm) and reduces the winding line speed to reduce the heat transfer delay, while greatly relaxing the safety tolerance through the fluctuation threshold formula (mean of historical standard deviation × 1.6), preferentially inhibiting the thermal runaway chain reaction, preventing the melting of the pole piece coating or the collapse of the structure, and if necessary, linking the material switching mode to interrupt the current production batch;

[0089] Basic interval: the dynamic thermal field balance evaluation value is 0-40; enhanced interval: the dynamic thermal field balance evaluation value is 41-80; emergency interval threshold value: the dynamic thermal field balance evaluation value is greater than 80;

[0090] Fluctuation threshold real-time calculation logic: the central processor extracts the compensation instruction intensity value (basic level = 0.5 / enhanced level = 1.2 / emergency level = 2.0) currently activated, combines the mean of the standard deviation of the thickness change trend value of the active coating of the pole piece batch in the historical database, and generates a dynamic fluctuation threshold value through a linear interpolation algorithm: fluctuation threshold = mean of historical standard deviation × (1 + compensation intensity value × 0.3), when the actual material thickness change trend value exceeds the threshold value, the corresponding compensation instruction is activated.

[0091] Wherein, the generation step of the pole piece deviation correction guide roller adjustment instruction set in step S2a is as follows:

[0092] S2a01, the central processor calls the thickness change trend value of the active coating of the pole piece and the tension fluctuation change rate in the pole piece state data packet, compares the fluctuation pattern curve of the thickness change trend value of the active coating of the pole piece stored in the database with the same period in the abnormal period history, identifies the abnormal section position and duration of the thickness change trend value of the active coating of the pole piece;

[0093] S2a02, based on the database, the elastic recovery ability parameter of the pole piece material corrects the offset of the abnormal section position, and extracts the extreme value distribution characteristics of the tension fluctuation change rate in the same time period, generates the guide roller displacement compensation amount and the dynamic response priority coefficient;

[0094] The specific generation method of the guide roller displacement compensation amount is as follows: the central processing unit calculates a position compensation coefficient based on the identified position offset of the thickness abnormal section and in combination with the elastic recovery capability parameter of the pole piece material in the database (the parameter quantifies the deformation recovery degree of the pole piece after stress relief), the position compensation coefficient is compensation coefficient = 1.0-elastic recovery capability / 100, wherein the elastic recovery capability is a percentage value, for example, if the recovery capability is 80%, the compensation coefficient is 0.2; at the same time, the extreme value distribution characteristics (including the tension peak value, the valley value and the change rate) of the tension fluctuation change rate in the same time period are extracted, and the tension peak value change gradient (that is, the maximum change rate of the tension per unit time) is calculated; the position offset is multiplied by the compensation coefficient to obtain a basic displacement value, and then the absolute value of the tension peak value change gradient is superimposed with the product of a preset tension-displacement conversion coefficient (the conversion coefficient is usually set based on the yield strength and the coating adhesion of the pole piece material), to generate the final guide roller displacement compensation amount, which takes into account the comprehensive influence of the permanent deformation of the pole piece and the instantaneous tension impact on the winding path;

[0095] The specific generation method of the dynamic response priority coefficient is as follows: a dynamic response urgency index is calculated based on the duration (unit: seconds) of the thickness abnormal section and the extreme value occurrence frequency (unit: times / second) of the tension fluctuation change rate, the dynamic response urgency index = duration x extreme value frequency, the higher the index, the more persistent and severe the abnormality; at the same time, the dynamic response urgency index is multiplied by the material dynamic response factor (response factor = 1.5-elastic recovery capability / 100+thermal expansion coefficient x 10, wherein the thermal expansion coefficient is 10 -6 / ℃) stored in the database in combination with the elastic recovery capability parameter and the thermal expansion coefficient (the coefficient reflects the dimensional stability of the material under temperature change), to generate the dynamic response priority coefficient, which is finally used to dispatch system resources, to preferentially process abnormal events corresponding to high coefficients, to ensure that the control response matches the severity of the fault;

[0096] S2a03, the guide roller displacement compensation amount, the dynamic response priority coefficient and the natural decay law of the thickness change trend value of the active coating of the pole piece in the database are introduced into a preset coupling relationship operation model, and a pole piece deviation guide roller adjustment instruction set is output, the pole piece deviation guide roller adjustment instruction set includes a guide roller displacement amplitude, an angle correction amount and an action time, wherein the technical definition of the coupling relationship operation model is that the guide roller displacement compensation amount, the dynamic response priority coefficient and the natural decay law of the thickness change trend value are taken as input vectors, the optimal adjustment parameters of the guide roller system are solved in real time through a multi-physical field coupling equation (thermal-force-deformation), and an intelligent decision system outputting three-dimensional control instructions of the guide roller displacement amplitude, the angle correction amount of the guide roller and the action time is output.

[0097] In step S2b, the specific content of the pole piece switching mode is as follows:

[0098] S2b1, based on the thickness change trend value of the current active coating of the pole piece and the thickness change trend value of the historical active coating of the pole piece in the database, the central processing unit calculates the fluctuation threshold value in real time, when the thickness change trend value of the current active coating of the pole piece has a step mutation and exceeds the fluctuation threshold value in the continuous period, the guide roller system of the pole piece traction area is controlled to implement position locking and keep the current tension stable;

[0099] Wherein, the fluctuation threshold value = the standard deviation mean value of the thickness change trend value of the historical active coating of the pole piece × (1 + the step mutation amplitude of the thickness change trend value of the current active coating of the pole piece (the absolute value of the difference between the mutation value of the thickness change trend value of the current active coating of the pole piece and the mean value before mutation) / the maximum step amplitude of the history) ;

[0100] The continuous period refers to the complete time span from the starting point to the ending point of the step mutation of the thickness change trend value of the active coating of the pole piece, and the determination standard is that the thickness change trend value of the active coating of the pole piece exceeds the fluctuation threshold value for 3 consecutive sampling periods (100 ms per period) ;

[0101] When the central processing unit confirms that the new heat conduction compensation level and the pole piece deviation correction guide roller adjustment instruction set have been completely transmitted to the guide roller system of the traction area of the lithium battery pole piece winding and packaging machine through the high-speed data bus, and the real-time monitoring of the tension fluctuation change rate keeps in the preset stable interval (fluctuation value ≤ reference tension ± 5%) for 3 consecutive sampling periods, the guide roller position locking is immediately released and switched to the new parameter execution mode;

[0102] S2b2, based on the thickness change trend value of the current active coating of the pole piece, the heat conduction compensation level and the pole piece deviation correction guide roller adjustment instruction set matched in the database are retrieved and called.

[0103] In Example 1, taking the production process of NCM811 ternary lithium battery pole piece winding with an energy density of 280 Wh / kg as an example, the central processor monitors the pole piece active coating thickness change trend value in real time through the non-contact thickness sensor, which continuously exceeds the fluctuation threshold value (for example, 2.5 μm) calculated based on the historical standard deviation mean value, while the infrared thermal imaging array detects that the heat pressing station appears temperature difference abnormality (triggering the safety threshold value) with a temperature difference greater than 5℃ for two consecutive sampling periods. The system immediately executes the S2a process: first, generate the temperature field evolution thermal map based on the temperature distribution feature map through the pre-trained spatiotemporal feature mapping network, and synchronously capture the thickness response prediction value (such as a predicted thickness increase of 3.2 μm) in the abnormal period; then, conduct time domain coupling analysis combined with the current pole piece material characteristic parameters (including thermal conductivity rate 2.5 W / mk, thermal deformation safety critical parameter 5℃, elastic recovery ability 80%, porosity 35%, etc.), to generate a dynamic thermal field balance evaluation value of 65 (in the 41-80 enhancement interval), thereby activating the enhanced level compensation instruction - the central processor increases the thermal power correction coefficient to level 1.2, making the heat pressing temperature increase by 12℃ (within the range of the reference value ±10℃), while generating the guide roller displacement compensation amount (calculated as +0.5 mm) according to the extreme value distribution characteristics of the tension fluctuation change rate (detecting a peak change gradient of 15 N / ms); then, in the S3 stage, adjust and scan the quality feedback characteristic value, and find that the pole piece heat pressing peeling force peak value is only 0.9 N / mm (lower than the NCM811 material preset standard 1.0 N / mm), thereby triggering the S4 optimization process: further increase the heat conduction compensation level to the emergency level (thermal power coefficient 2.0) in the temperature difference abnormality area after analyzing the heat sealing strength-temperature correlation matrix, while correcting the guide roller displacement compensation amount to +0.7 mm according to the deformation accuracy-tension correlation vector, and finally stabilizing the peeling force peak value to 1.2 N / mm through closed-loop feedback, controlling the winding alignment deviation amount within 0.3 mm (satisfying the square battery tolerance threshold value ≤0.5 mm), realizing high-precision control of the cell winding.

[0104] The beneficial effects of this embodiment are as follows: by monitoring the pole piece active coating thickness change trend value in real time and triggering the enhanced level compensation instruction, the heat pressing temperature is dynamically increased to suppress the temperature difference abnormality, while the guide roller displacement compensation amount is generated based on the tension fluctuation change rate to optimize the winding path; when the heat pressing peeling force peak value is insufficient, the heat conduction compensation level is further increased to the emergency level through the closed-loop optimization process and the guide roller displacement amount is corrected, finally realizing the coordinated improvement of the pole piece heat pressing sealing strength and the winding alignment accuracy, effectively overcoming the coating damage and deformation deviation caused by the coupling of temperature field fluctuation and mechanical stress in the production process of lithium battery pole piece, and significantly enhancing the battery pole piece packaging quality and electrochemical performance stability.

[0105] S3, adjust the hot-pressing temperature of the electrode sheet hot-pressing station based on the heat conduction compensation level, and adjust the guide roller system of the electrode sheet traction area according to the electrode sheet deviation correction guide roller adjustment instruction set;

[0106] In step S3, the specific content of adjusting the hot-pressing temperature of the electrode sheet hot-pressing station is as follows:

[0107] S31, the central processor calls different heat power correction coefficients of the corresponding basic level compensation instruction, enhanced level compensation instruction and emergency level compensation instruction in the database according to the heat conduction compensation level;

[0108] The different heat power correction coefficients refer to the dynamic power amplification multiple of the heat conduction compensation level instruction (basic level / enhanced level / emergency level) in the heat sealing temperature adjustment. The technical essence is that the basic level compensation instruction corresponds to 1.0-1.05 times heat power fine tuning coefficient (temperature compensation ≤5%); the enhanced level compensation instruction corresponds to 1.1-1.2 times power gain coefficient (temperature compensation 10%-20%); and the emergency level compensation instruction corresponds to 1.3-1.5 times power multiplication coefficient (temperature compensation 30%-50%). Each coefficient value is dynamically calculated and generated from the current electrode sheet thermal deformation safety critical value and real-time temperature deviation stored in the database.

[0109] S32, adjust the weight of the heat power correction coefficient in combination with the pre-stored heat conduction rate of the current electrode sheet material in the database, and implement hot-pressing temperature adjustment in the electrode sheet hot-pressing station according to the adjusted heat power correction coefficient.

[0110] The adjustment method of the weight of the heat power correction coefficient is as follows: the central processor extracts the heat conduction rate value of the current packaging electrode sheet in the database, calculates the ratio of the rate value to the reference heat conduction rate of the same material as the weight factor; and multiplies the heat power correction coefficient obtained in step S31 by the weight factor to generate the final compensation coefficient. When the heat conduction rate value is higher than the reference value, the weight factor is taken from the interval of 0.8-1.0 (the higher the rate, the lower the weight), and when the heat conduction rate value is lower than the reference value, the weight factor is taken from the interval of 1.0-1.2 (the lower the rate, the higher the weight).

[0111] In step S3, the specific content of adjusting the guide roller system of the electrode sheet traction area according to the electrode sheet deviation correction guide roller adjustment instruction set is as follows: the central processor calculates the real-time displacement amplitude and angle correction amount of the guide roller system based on the guide roller displacement compensation amount and the dynamic response priority coefficient, generates an adaptive adjustment weight factor of the guide roller action time length in combination with the pre-stored current electrode sheet material elastic recovery ability parameter and the extreme value distribution characteristics of the tension fluctuation rate in the database, generates a guide roller displacement execution instruction set according to the adaptive adjustment weight factor, and synchronously drives the guide roller system to execute dynamic deviation correction according to the displacement amplitude, angle correction amount and action time length in the guide roller displacement execution instruction set.

[0112] The calculation formula of the real-time displacement amplitude and the angle correction amount of the guide roller system is as follows:

[0113]

[0114] A = D + R * E * F d A is the real-time displacement amplitude; D c is the guide roller displacement compensation offset; R p is the dynamic response priority coefficient; R max is the normalized reference of the dynamic response priority coefficient (taking the historical maximum value or the theoretical limit value, dimensionless); E is the extreme value distribution characteristic of the tension fluctuation rate (i.e. the first derivative of the tension fluctuation rate function curve); E r is the elastic recovery capability parameter of the sheet material;

[0115] The calculation formula of the angle correction amount is as follows:

[0116]

[0117] θ = L * (A / R) * E base L is the reference action arm length of the guide roller system (determined by the mechanical structure); the remaining parameters are defined in the calculation formula of the real-time displacement amplitude and the angle correction amount of the guide roller system;

[0118] The calculation formula of the adaptive adjustment weight factor is as follows:

[0119]

[0120] σ = W * (σ max / σ ref) max σ is the historical maximum tension fluctuation; W is the adaptive adjustment weight factor; the remaining parameters are defined in the calculation formula of the real-time displacement amplitude and the angle correction amount of the guide roller system;

[0121] The specific steps of generating the guide roller displacement execution instruction set according to the adaptive adjustment weight factor are as follows: the central processing unit dynamically calculates the guide roller action time based on the adaptive adjustment weight factor, synchronously fuses the guide roller displacement compensation amount, the dynamic response priority coefficient and the current elastic recovery capability parameter of the sheet material, and generates the synergistic optimization value of the guide roller displacement amplitude and the angle correction amount in real time; combined with the extreme value distribution characteristic of the tension fluctuation rate and the pre-stored thermal expansion coefficient of the sheet material in the database, the guide roller displacement execution instruction set is output through a multi-physical field coupling operation model, and finally the guide roller system is driven to execute high-precision dynamic deviation correction according to the displacement amplitude, the angle correction amount and the action time in the instruction set, so as to realize real-time adaptive regulation and control of the winding path.

[0122] S4, real-time calculation of the heat conduction compensation level and the predicted execution deviation value of the guide roller adjustment instruction set, when the predicted execution deviation value exceeds the preset model deviation threshold, dynamically updating the pre-trained prediction model parameters based on the pole piece state data packet, and synchronizing the updated model parameters to the instruction generation process of step S2.

[0123] The pre-trained prediction model is trained according to a machine learning and neural network learning model;

[0124] The calculation method of the predicted execution deviation value is that the central processing unit dynamically calculates the difference value of the heat conduction compensation level and the expected execution effect of the guide roller adjustment instruction set, and quantitatively generates a predicted execution deviation value by comparing the real-time data of the pole piece hot pressing station temperature distribution feature map and the tension fluctuation rate of the traction area. When the deviation value exceeds the preset model deviation threshold, the prediction model parameter updating process based on the pole piece state data packet is triggered, and finally the optimized model parameters are synchronized to the instruction generation module in real time to form a closed-loop adaptive control link.

[0125] Embodiment 1, taking NCM811 ternary pole piece winding as an example:

[0126] 1. Hot pressing temperature adjustment: detecting temperature deviation triggers enhanced level compensation (hot power correction coefficient 1.1), combining pole piece heat conduction rate and reference ratio 0.9 to generate weight factor 0.9, and finally compensation coefficient = 1.1 x 0.9 = 0.99, hot pressing temperature is accurately raised to target value;

[0127] 2. Guide roller displacement control: guide roller displacement compensation amount 10 and dynamic response priority coefficient 8 are converted by normalizing reference 10, superimposed with tension fluctuation rate extreme value 3 and elastic recovery ability parameter 2 to generate real-time displacement amplitude 14;

[0128] 3. Weight factor regulation: based on displacement compensation amount 10, tension change rate 3, elastic recovery parameter 2 and historical maximum tension fluctuation 5, calculate the adaptive adjustment weight factor and generate the instruction set to drive the guide roller to execute deviation correction according to displacement amplitude 14, angle correction amount 0.12 radian and action time 80 ms.

[0129] This embodiment realizes accurate temperature compensation by dynamically weighting the hot power correction coefficient and the material heat conduction rate, and synchronously generates high-precision displacement and angle instructions by multi-parameter coupling calculation of guide roller displacement, tension characteristics and material elasticity, finally realizes hot pressing temperature stability deviation ≤±1.5℃, winding alignment control within 0.15mm, significantly improves the consistency and safety performance of lithium battery pole piece packaging.

[0130] The above merely illustrates the embodiments of the present application, and does not limit the patent scope of the present application, and any equivalent structure or equivalent process transformation, or direct or indirect application in other related technical fields, which are made by using the content of the present application specification and drawings, are also included in the patent protection scope of the present application.

Claims

1. A method of intelligent control of a packaging machine, characterized in that: The method comprises the following steps: S1, deploying a comprehensive detection unit at a key node of a lithium battery pole piece winding packaging machine, synchronously collecting thickness change trend values of pole piece active coating in a preset pole piece unwinding mechanism, temperature distribution characteristic maps of a preset pole piece hot pressing work station, and tension fluctuation change rates of a preset pole piece traction area, generating a pole piece state data package, and storing the pole piece state data package in a preset database; S2, when the thickness change trend values of the pole piece active coating exceed a preset fluctuation threshold, generating a heat conduction compensation level and a pole piece deviation rectification guide roller adjustment instruction set according to the temperature distribution characteristic maps and the tension fluctuation change rates; S3, adjusting the hot pressing temperature of the pole piece hot pressing work station based on the heat conduction compensation level, and adjusting the guide roller system of the pole piece traction area according to the pole piece deviation rectification guide roller adjustment instruction set; S4, real-time calculating a predicted execution deviation value of the heat conduction compensation level and the guide roller adjustment instruction set, when the predicted execution deviation value exceeds a preset model deviation threshold, dynamically updating the pre-trained prediction model parameters based on the pole piece state data package, and synchronously updating the updated model parameters to the instruction generation process of step S2.

2. The intelligent control method of a packaging machine according to claim 1, characterized in that: The comprehensive detection unit comprises a non-contact thickness sensor arranged at the pole piece unwinding mechanism, an infrared thermal imaging array arranged at the pole piece hot pressing work station, and a strain type tension sensor arranged at the pole piece traction area, and the non-contact thickness sensor, the infrared thermal imaging array, and the strain type tension sensor are synchronously output to a preset central processing unit through a preset high-speed data bus.

3. The intelligent control method of a packaging machine according to claim 2, characterized in that: The generation step of the pole piece state data package in step S1 is as follows: S11, the central processing unit receives a pole piece coating thickness signal sequence output by the non-contact thickness sensor in real time, extracts the thickness change trend values of the pole piece active coating through time domain noise reduction processing; S12, the central processing unit synchronously analyzes infrared radiation data of the infrared thermal imaging array, and generates a temperature distribution characteristic map based on a preset algorithm; S13, the central processing unit dynamically processes the original electric signal of the strain type tension sensor, and converts and outputs a tension fluctuation change rate function curve; S14, the central processing unit performs time scale alignment compression on the thickness change trend values of the pole piece active coating, the temperature distribution characteristic map, and the tension fluctuation change rate function curve, generates a pole piece state data package, and stores the pole piece state data package in the database.

4. The intelligent control method of a packaging machine according to claim 3, characterized in that: The specific content in step S2 is as follows: When the thickness change trend values of the pole piece active coating exceed the preset fluctuation threshold: S2a, when the thickness change trend values of the pole piece active coating continuously exceed the fluctuation threshold, a heat conduction compensation level is generated according to the temperature distribution characteristic map, and a pole piece deviation rectification guide roller adjustment instruction set is generated according to the coupling relationship between the thickness change trend values of the pole piece active coating and the tension fluctuation change rate; S2b, if the thickness change trend values of the pole piece active coating have a step mutation and exceed the fluctuation threshold in a continuous period, a preset pole piece switching mode is activated; The generation step of the heat conduction compensation level in step S2a is as follows: S2a1, the central processor analyzes the temperature distribution feature map in time sequence, inputs the analysis result into a pre-trained prediction model, and outputs a temperature field evolution heat map in a future period; S2a2, the temperature field evolution heat map is scanned, when a temperature difference abnormality exceeding a preset safety threshold continuously is detected, an abnormal period and a temperature difference amplitude are recorded, and a thickness change trend value of the active coating of the pole piece in the abnormal period is synchronously captured; S2a3, time domain coupling analysis is performed on the current pole piece material characteristic parameter pre-stored in the database, and the temperature difference amplitude and the thickness change trend value of the active coating of the pole piece in the abnormal period, a dynamic thermal field balance evaluation value is generated; S2a4, when the dynamic thermal field balance evaluation value is in a preset basic interval, a basic level compensation instruction is generated, if the dynamic thermal field balance evaluation value is in a preset enhanced interval, an enhanced level compensation instruction is activated, when the dynamic thermal field balance evaluation value exceeds a preset emergency interval threshold, an emergency level compensation instruction is triggered, and a fluctuation threshold is calculated in real time by the central processor based on the intensity value of the currently generated compensation instruction and the thickness change trend value of the active coating of the pole piece in the database; When the actually detected thickness change trend value of the active coating of the pole piece exceeds the fluctuation threshold, the enhanced level compensation instruction or the emergency level compensation instruction is activated for thermal conduction compensation level, and when the actually detected thickness change trend value of the active coating of the pole piece is less than or equal to the fluctuation threshold, the basic level compensation instruction is activated for thermal conduction compensation level.

5. The intelligent control method of a packaging machine according to claim 4, characterized in that: The specific steps of synchronously capturing the thickness change trend value of the active coating of the pole piece in the abnormal period in the step S2a2 are as follows: S2a21, a temperature change gradient curve corresponding to the abnormal period is extracted from the temperature field evolution heat map; S2a22, the temperature change gradient curve is input into a pre-trained thickness response prediction model, and a thickness response prediction value of the active coating of the pole piece in the abnormal period is output; S2a23, the thickness response prediction value of the active coating of the pole piece is corrected in combination with the thickness change trend value of the active coating of the pole piece in the same period as the abnormal period stored in the database; S2a24, the corrected thickness response prediction value of the active coating of the pole piece is taken as the capture result of the thickness change trend value of the active coating of the pole piece in the current abnormal period.

6. The intelligent control method of a packaging machine according to claim 5, characterized in that: The generation steps of the pole piece deviation correction guide roller adjustment instruction set in the step S2a are as follows: S2a01, the central processor calls the thickness change trend value of the active coating of the pole piece in the pole piece state data packet and the tension fluctuation change rate, compares the fluctuation form curve of the thickness change trend value of the active coating of the pole piece in the same period as the abnormal period stored in the database, identifies the abnormal section position and duration in the thickness change trend value of the active coating of the pole piece; S2a02, the offset amount of the abnormal section position is corrected based on the pole piece material elastic recovery ability parameter in the database, and the extreme value distribution characteristics of the tension fluctuation change rate in the same period are extracted, a guide roller displacement compensation amount and a dynamic response priority coefficient are generated; S2a03, the guide roller displacement compensation, the dynamic response priority coefficient and the thickness change trend value of the active coating of the pole piece in the database are introduced into a preset coupling relationship operation model, and a pole piece deviation correction guide roller adjustment instruction set is output, the pole piece deviation correction guide roller adjustment instruction set includes a guide roller displacement amplitude, an angle correction amount and an action time length.

7. The intelligent control method of a packaging machine according to claim 6, characterized in that: The specific content of the pole piece switching mode in step S2b is: S2b1, based on the thickness change trend value of the current active coating of the pole piece and the historical thickness change trend value of the active coating of the pole piece in the database, the central processing unit calculates a fluctuation threshold value in real time, when the current thickness change trend value of the active coating of the pole piece has a step mutation and exceeds the fluctuation threshold value in a continuous period, the guide roller system of the pole piece traction area is controlled to implement position locking and keep the current tension stable; S2b2, based on the thickness change trend value of the current active coating of the pole piece, the database is searched and the matching heat conduction compensation level and pole piece deviation correction guide roller adjustment instruction set are called.

8. The intelligent control method of a packaging machine according to claim 7, characterized in that: The specific content of adjusting the hot pressing temperature of the pole piece hot pressing station in step S3 is: S31, the central processing unit calls different heat power correction coefficients corresponding to the basic level compensation instruction, the enhanced level compensation instruction and the emergency level compensation instruction in the database according to the heat conduction compensation level; S32, the weight of the heat power correction coefficient is adjusted in combination with the pre-stored heat conduction rate of the current pole piece material in the database, and the hot pressing temperature is adjusted according to the adjusted heat power correction coefficient in the pole piece hot pressing station.

9. The method of claim 8, wherein: The specific content of adjusting the guide roller system of the pole piece traction area according to the pole piece deviation correction guide roller adjustment instruction set in step S3 is as follows: the central processing unit calculates the real-time displacement amplitude and angle correction amount of the guide roller system based on the guide roller displacement compensation and the dynamic response priority coefficient, generates an adaptive adjustment weight factor of the guide roller action time length in combination with the pre-stored current pole piece material elastic recovery ability parameter and the extreme value distribution characteristic of the tension fluctuation change rate in the database, generates a guide roller displacement execution instruction set according to the adaptive adjustment weight factor, and synchronously drives the guide roller system to execute dynamic deviation correction according to the displacement amplitude, angle correction amount and action time length in the guide roller displacement execution instruction set.

10. The method of claim 9, wherein: The specific content of step S4 is that the central processing unit calculates the predicted execution deviation value of the heat conduction compensation level and the guide roller adjustment instruction set in real time, when the predicted execution deviation value exceeds a preset model deviation threshold value, the confidence decay gradient of the pre-trained prediction model is dynamically evaluated based on the thickness change trend value of the active coating of the pole piece, the temperature distribution characteristic map and the tension fluctuation change rate function curve in the pole piece state data packet, and the weight distribution and feature extraction layer structure of the pre-trained prediction model are reconstructed through a preset algorithm, and then the updated model parameters are mapped to the instruction generation process of step S2 in real time, and the generation logic of the heat conduction compensation level and the guide roller adjustment instruction set is optimized synchronously.

Images