CN-121973493-A - Self-adaptive control method and system for lithium battery pole piece roller press

CN121973493ACN 121973493 ACN121973493 ACN 121973493ACN-121973493-A

Abstract

A self-adaptive control method and system for a lithium battery pole piece roller press comprises the steps of obtaining real-time data of a controlled variable and real-time data of an online quality index in a rolling production process, wherein the controlled variable comprises at least one of master cylinder pressure, bending cylinder pressure, roller temperature and rolling speed, the online quality index comprises pole piece thickness or surface density, a transfer function matrix model is built based on historical data of the controlled variable and the online quality index and used for describing a dynamic response relation of the controlled variable to the online quality index, real-time adjustment quantity of the controlled variable is calculated through a model predictive control algorithm based on the transfer function matrix model and the online quality index real-time data, a control instruction is generated according to the real-time adjustment quantity and used for controlling the lithium battery pole piece roller press to adjust the controlled variable according to the real-time adjustment quantity. The invention can ensure the consistency of mass production of the lithium battery pole pieces and obviously improve the quality stability of products.

Inventors

CHU DANLEI
JIANG JINGBO

Assignees

厦门奥普拓自控科技有限公司

Dates

Publication Date: 20260505
Application Date: 20260225

Claims (10)

1. The self-adaptive control method for the lithium battery pole piece roller press is characterized by comprising the following steps of: acquiring real-time data of a controlled variable and real-time data of an online quality index in the rolling production process, wherein the controlled variable comprises at least one of main cylinder pressure, bending cylinder pressure, roller temperature and rolling speed, and the online quality index comprises pole piece thickness or surface density; Based on the historical data of the controlled variable and the online quality index, a transfer function matrix model is established, and the transfer function matrix model is used for describing the dynamic response relation of the controlled variable to the online quality index; Calculating the real-time adjustment quantity of the controlled variable through a model predictive control algorithm based on the transfer function matrix model and the real-time data of the online quality index; generating a control instruction according to the real-time adjustment quantity, wherein the control instruction is used for controlling a lithium battery pole piece roller press to adjust the controlled variable according to the real-time adjustment quantity.
2. The adaptive control method for a lithium battery pole piece roll squeezer of claim 1, wherein establishing a transfer function matrix model based on the controlled variable and historical data of an online quality index comprises: Injecting a preset excitation signal into the controlled variable, and acquiring online quality index real-time data, wherein the preset excitation signal is a pseudo-random binary sequence excitation signal; Performing low-pass filtering processing on the preset excitation signal and the corresponding online quality index real-time data thereof, and determining a time delay parameter of a system through a cross-correlation analysis method; based on the time delay parameter, an initial high-order model is obtained through an autoregressive model identification algorithm comprising exogenous input, wherein the exogenous input is a measurable disturbance variable; Reducing the order of the initial high-order model to an initial low-order model with a preset order; And carrying out fitting degree verification on the initial low-order model based on the historical data of the controlled variable and the online quality index, and outputting the initial low-order model as a transfer function matrix model when the verification accuracy meets the preset condition.
3. The adaptive control method for a lithium battery pole piece roll squeezer of claim 2, further comprising a measurable disturbance feedforward step of: Acquiring a set vehicle speed or set roller temperature of a roller press and taking the set vehicle speed or set roller temperature as a measurable disturbance variable; and in the process of calculating the real-time adjustment quantity of the controlled variable through a model predictive control algorithm, carrying out feedforward compensation correction on the real-time adjustment quantity of the controlled variable according to the real-time measurement value of the measurable disturbance variable or the change trend of the measurable disturbance variable.
4. The self-adaptive control method for a lithium battery pole piece roll squeezer according to claim 3, wherein the real-time adjustment quantity of the controlled variable is calculated through a model predictive control algorithm, specifically, the following objective function J is solved: ; Wherein, the A predicted sequence representing the controlled variable; a model uncertainty factor representing the transfer function matrix model; A model uncertainty factor representing a disturbance model, the disturbance source of which comprises the measurable disturbance variable; a value representing the online quality index of step i, predicted onward in a kth solution iteration period; a target value representing the online quality index; q represents the weight diagonal matrix of the online quality index; Representing predicting the value of the controlled variable of step i in the kth solution iteration period; A target value representing the controlled variable; Represents the input prediction step size, and 0< < R represents a weight diagonal matrix of the controlled variable; W represents a weight diagonal matrix of the relaxation variables at the lower boundary of the line quality index; and T represents a weight diagonal matrix of the relaxation variables at the upper boundary of the line quality index.
5. The adaptive control method for a lithium battery pole-piece roll squeezer according to claim 4, wherein the objective function J satisfies the following constraint: Wherein, the method comprises the steps of, Representing an online quality index lower bound; Representing an online quality index upper bound; Wherein, the method comprises the steps of, Representing the lower bound of the controlled variable; representing the upper bound of the controlled variable; 0 Wherein, the method comprises the steps of, Representation of Is set to be a constant value of the upper bound of (c), Representation of Is a lower constant of (2); ; , Wherein G is the transfer function matrix model; an upper bound for a model uncertainty factor of the transfer function matrix model; an upper bound for a model uncertainty factor of the perturbation model; Wherein Y represents the online quality index, U represents the controlled variable; Representing the measurable disturbance variable.
6. The adaptive control method for a lithium battery pole piece roll squeezer according to any one of claims 1 to 5, further comprising a transfer function matrix model adaptive updating step of: periodically calculating process performance statistics according to the online quality index real-time data, and updating the transfer function matrix model when the process performance statistics exceed a preset threshold; and after the transfer function matrix model is updated, recalculating the real-time adjustment quantity based on the updated transfer function matrix model, and generating a new control instruction.
7. The method of claim 6, wherein the process performance statistics include at least one of a statistical range, a standard deviation, and a process capability index calculated based on the on-line quality index real-time data.
8. The adaptive control method for a lithium battery pole piece roll squeezer of claim 6, wherein updating the transfer function matrix model comprises: Determining updated excitation signal parameters for model identification based on the controlled variable real-time data and the online quality index real-time data, wherein the updated excitation signal parameters comprise amplitude, duration and application direction; If the online quality index real-time data is in the preset specification range, injecting an updating excitation signal into the controlled variable, and acquiring corresponding online quality index real-time data; Based on the updated excitation signals and the corresponding online quality index real-time data, reestablishing a transfer function matrix model as a new model; And comparing the precision of the new model with that of the transfer function matrix model, and updating the transfer function matrix model into the new model if the precision of the new model is higher than that of the transfer function matrix model and the precision of the new model is greater than or equal to a preset model switching threshold value.
9. The adaptive control method for a lithium battery pole piece roll squeezer according to claim 8, wherein reestablishing a transfer function matrix model based on the updated excitation signal and the newly acquired on-line quality index real-time data as a new model comprises: performing low-pass filtering processing on the updated excitation signal and the corresponding online quality index real-time data thereof, and determining an updated time delay parameter of the system through a cross-correlation analysis method; Based on the updated time delay parameters, obtaining an updated higher-order model through an autoregressive model identification algorithm comprising exogenous input; Reducing the order of the updated high-order model to an updated low-order model with a preset order; and carrying out fitting degree verification on the updated low-order model based on the historical data of the controlled variable and the online quality index, and outputting the updated low-order model as a new model when the verification accuracy meets the preset condition.
10. An adaptive control system for a lithium battery pole piece roll squeezer, comprising: The device comprises a data acquisition module, a rolling production module and a rolling control module, wherein the data acquisition module is used for acquiring real-time data of controlled variables and real-time data of online quality indexes in the rolling production process, wherein the controlled variables comprise at least one of main cylinder pressure, bending cylinder pressure, roller temperature and rolling speed, and the online quality indexes comprise pole piece thickness or surface density; the model building module is used for building a transfer function matrix model based on the historical data of the controlled variable and the online quality index, and the transfer function matrix model is used for describing the dynamic response relation of the controlled variable to the online quality index; The control calculation module is used for calculating the real-time adjustment quantity of the controlled variable through a model predictive control algorithm based on the transfer function matrix model and the real-time data of the online quality index; And the control execution module is used for generating a control instruction according to the real-time adjustment quantity, and the control instruction is used for controlling the lithium battery pole piece roller press to adjust the controlled variable according to the real-time adjustment quantity.

Description

Self-adaptive control method and system for lithium battery pole piece roller press Technical Field The invention relates to the technical field of lithium battery manufacturing, in particular to a self-adaptive control method and a self-adaptive control system for a lithium battery pole piece roller press. Background The rolling of the lithium battery pole piece is a key process for manufacturing the battery core, and the aim is to roll the coated pole piece to a preset thickness and surface density by accurately regulating parameters such as pressure, temperature and the like of the roller so as to ensure the consistency of electrochemical performance of the battery. The quality index of the rolled pole piece, particularly the uniformity and stability of thickness and surface density, is directly related to the capacity, internal resistance and safety performance of the battery. At present, the control of the roller press in the industrial field mainly adopts the following modes: and (3) manual experience control, namely manually adjusting parameters such as main cylinder pressure, bending cylinder pressure and the like by combining self experience according to an offline sampling detection result (such as measuring thickness by a micrometer and calculating surface density by weighing). This approach suffers from problems of delayed response, low adjustment accuracy, and high personal experience dependence, and it is difficult to ensure mass production consistency. Single variable closed loop control-some advanced devices are equipped with closed loop controllers based on a single quality index (e.g., thickness). For example, the master cylinder pressure is adjusted according to the deviation by measuring the thickness by an on-line thickness gauge and comparing with a set value using a PID algorithm (proportional-integral-derivative control algorithm). But the rolling process is a complex process that is typically multivariable, strongly coupled and time-lapse. For example, adjusting the master cylinder pressure affects the thickness while potentially changing the lateral thickness profile of the pole piece (coupled with the bending cylinder pressure), and adjusting the roll temperature affects the plastic deformation characteristics of the material, which in turn affects the thickness and density. The existing univariate control method cannot process the dynamic coupling relation among multiple variables, so that fluctuation of other quality indexes is easy to be caused when one parameter is regulated, and cooperative optimization control of thickness and areal density is difficult to realize. The preset parameter control based on the fixed model is that few systems try to adopt feedforward or simple models for control, but the internal model is usually a fixed linear model obtained based on specific working conditions and specific materials through offline identification. However, in actual production, the incoming material characteristics (such as coating weight, slurry formulation), ambient temperature, equipment conditions (such as roll surface wear), and production cadence (speed change) are all frequently changed. These variations can lead to drift in process dynamics, significantly degrading or even disabling the performance of the fixed model-based controller, and failing to achieve long-term stable and accurate control. In summary, the prior art cannot achieve adaptive control of multivariate cooperation. Disclosure of Invention In order to solve the technical problems, the invention provides a self-adaptive control method for a lithium battery pole piece roller press, which comprises the following steps: acquiring real-time data of a controlled variable and real-time data of an online quality index in the rolling production process, wherein the controlled variable comprises at least one of main cylinder pressure, bending cylinder pressure, roller temperature and rolling speed, and the online quality index comprises pole piece thickness or surface density; Based on the historical data of the controlled variable and the online quality index, a transfer function matrix model is established, and the transfer function matrix model is used for describing the dynamic response relation of the controlled variable to the online quality index; Calculating the real-time adjustment quantity of the controlled variable through a model predictive control algorithm based on the transfer function matrix model and the real-time data of the online quality index; generating a control instruction according to the real-time adjustment quantity, wherein the control instruction is used for controlling a lithium battery pole piece roller press to adjust the controlled variable according to the real-time adjustment quantity. Optionally, establishing a transfer function matrix model based on the controlled variable and the historical data of the online quality index includes: Injecting a preset excitation signal into the