CN-122020450-A - Method and system for optimally setting process parameters of three-roller coating machine

CN122020450ACN 122020450 ACN122020450 ACN 122020450ACN-122020450-A

Abstract

The invention discloses a method and a system for optimally setting process parameters of a three-roller coating process, and belongs to the technical field of cold-rolled strip steel treatment lines. According to the method, production data are collected in real time, the film thickness of the coating is predicted by using an artificial neural network model, key process variables are determined through correlation analysis, multi-objective optimization is finally carried out with the aim of minimizing the target film thickness and energy medium consumption, and process parameters are automatically set. The system adopts a closed loop architecture of a data acquisition and execution layer, a control and communication layer and an optimization decision layer, and integrates roller diameter online detection, film thickness prediction, correlation analysis and optimization setting modules. The invention effectively solves the problems of unstable coating quality and high cost caused by traditional manual adjustment, and realizes high-precision, high-stability and low-cost control of the coating thickness.

Inventors

XIA ZHI
WANG YAO
XIONG JUNWEI
ZHU BIAO
CHEN JIE
AI JING

Assignees

中冶南方工程技术有限公司

Dates

Publication Date: 20260512
Application Date: 20251229

Claims (10)

1. The process parameter optimization setting method for the three-roller coating is characterized by comprising the following steps of: Acquiring process parameter data and coating film thickness data of the three-roller coater in real time through an industrial camera, a field sensor and a film thickness online detector; Based on the acquired data, predicting the film thickness of the coating by utilizing an artificial neural network model, and determining a key process variable with the film thickness correlation of the coating being greater than a preset threshold value through correlation analysis; and combining and optimizing the key process variables by taking the minimum film thickness of the target coating and the minimum consumption of the energy medium as optimization targets, generating optimized process parameter instructions, and issuing the optimized process parameter instructions to a three-roller coater for execution through a unit PLC.
2. The method of claim 1, wherein the process parameter data collected in real time comprises strip thickness, strip speed, strip accumulated coating length, strip tension, coating roller diameter, coating liquid composition, concentration and viscosity, speed ratio and gap of metering roller and liquid-taking roller, pressure between liquid-taking roller and coating roller, speed ratio of liquid-taking roller, direction and speed ratio of coating roller, pressure of coating roller and strip steel, wherein the coating film thickness data is obtained by online detection of a film thickness online detector, and the coating roller diameter is obtained by online detection of an industrial camera shooting real-time image of the coating roller and adopting an image processing algorithm.
3. The method of claim 1, further comprising the steps of sample generation and preprocessing of collecting historical process parameter data and corresponding coating film thickness data to form an initial sample set, and removing abnormal data in the initial sample set by at least one of a thresholding method, an angular distance method and a robust regression method to generate a three-roll coating sample set for model training.
4. The method of claim 1, wherein the artificial neural network model is of a three-layer structure and comprises an input layer, an hidden layer and an output layer, wherein the input parameters of the input layer are the process parameter data, the output parameters of the output layer are the predicted coating film thickness, the artificial neural network model adopts a Sigmoid function as a conversion function and adopts a BP algorithm for training, and when training is performed, one part of the three-roller coating sample set is used as training data, and the other part is used as verification data.
5. The method according to claim 1, wherein the correlation analysis calculates the correlation coefficient between each process parameter and the coating film thickness by the following formula: Wherein x is a process parameter independent variable, y is a coating film thickness dependent variable, As the mean value of the independent variable(s), The method comprises the steps of selecting a process parameter with an absolute value of the correlation coefficient larger than a preset threshold value as the key process variable, wherein the average value of the dependent variable is the correlation coefficient R (t) which is positive and negative and positive, and the process parameter with the absolute value of the correlation coefficient larger than the preset threshold value is selected as the key process variable.
6. The method of claim 1, wherein the key process variables include a ratio and gap of metering roll and pick-up roll, a pressure between pick-up roll and applicator roll, a pick-up roll ratio, a direction and ratio of applicator roll, and an applicator roll to strip pressure.
7. The method of claim 1 wherein the optimizing the combination of key process variables specifically comprises calculating predicted coating film thicknesses for a plurality of sets of combinations of key process variables using the trained artificial neural network model, and selecting a combination with a minimum energy medium consumption as an optimal process parameter combination from among combinations where all predicted coating film thicknesses meet a target value.
8. A three roll coating process parameter optimization setting system for implementing the method of any one of claims 1-7, comprising: the data acquisition and execution layer comprises a field sensor, a film thickness on-line detector, an industrial camera and a three-roller coater, wherein the field sensor and the film thickness on-line detector are used for acquiring unit operation parameters and coating film thickness data in real time, the industrial camera is used for acquiring real-time images of the coating roller, and the three-roller coater is used for executing optimized technological parameter instructions; the control and communication layer comprises a unit PLC and an industrial Ethernet, wherein the unit PLC is connected with and controls equipment in the data acquisition and execution layer and is used for collecting various real-time data from the data acquisition and execution layer and issuing control instructions, and the industrial Ethernet provides a communication backbone for a system; The optimizing decision layer is a roll coating model computer which is communicated with the unit PLC and the industrial camera through the industrial Ethernet and is used for receiving the real-time data and the images, running roll coating model software, calculating an optimized technological parameter instruction according to the real-time data, the images and the target film thickness, and then transmitting the optimized technological parameter instruction to the three-roll coater through the unit PLC.
9. The system of claim 8, wherein the roll coating model software running on the roll coating model computer comprises four functional modules that cooperate in sequence: the coating roller diameter online detection module is used for receiving the real-time image shot by the industrial camera and identifying and outputting real-time roller diameter data of the coating roller online through an image processing algorithm; the coating film thickness prediction module receives various real-time process parameters and real-time roller diameter data collected by the unit PLC, calls a pre-trained artificial neural network model and outputs predicted coating film thickness; The correlation analysis module is used for calculating correlation coefficients between each process parameter and the coating film thickness based on the historical data sequence and screening out a key process variable set according to the absolute value of the coefficients; And the roller coating process optimization setting module performs combination optimization in a parameter space of the key process variable set according to the current unit real-time state and the target film thickness, evaluates by using the coating film thickness prediction module, screens out a process parameter combination meeting the target film thickness and having the minimum energy medium consumption, and outputs the process parameter combination as the optimized process parameter instruction.
10. The system of claim 9, wherein the artificial neural network model used by the coating film thickness prediction module is of a three-layer network structure, input parameters received by an input layer of the artificial neural network model are determined according to unit configuration and at least comprise strip steel thickness, strip steel speed, strip steel tension, coating roller diameter, coating liquid characteristic parameters collected by the data collection and execution layer, speed ratio, gap and pressure parameters among a metering roller, a liquid taking roller and a coating roller in the three-roller coater, and an output layer outputs a single predicted coating film thickness value, and the artificial neural network model is trained through historical production data and has nonlinear mapping capability so as to predict the coating film thickness.

Description

Method and system for optimally setting process parameters of three-roller coating machine Technical Field The invention relates to the technical field of cold-rolled strip steel treatment lines, in particular to a method and a system for intelligently and optimally setting technological parameters of a three-roller coater for surface coating treatment of cold-rolled strip steel. Background The roll coating is adopted to process the surface of the steel plate, so that the method is an effective method for changing the chemical components, the tissue structure and the stress state of the surface of the steel product, can obtain the required surface performance, and is a high-efficiency and high-quality method. The thickness and stability of the coating determine the corrosion resistance, insulation and other properties of the product. At present, the strip steel coating process parameters of the silicon steel, color coating and galvanization units are mainly adjusted by manual experience, so that unstable coating quality and resource waste are easily caused, and in order to ensure the product quality, the coating thickness control generally adopts positive tolerance, so that the production cost is increased. Disclosure of Invention In view of the technical defects and technical drawbacks existing in the prior art, the embodiment of the invention provides a method and a system for optimizing and setting process parameters of a three-roller coating process, which overcome or at least partially solve the problems, and the method and the system have the following specific scheme; As a first aspect of the present invention, there is provided a three-roll coating process parameter optimization setting method, comprising the steps of: Acquiring process parameter data and coating film thickness data of the three-roller coater in real time through an industrial camera, a field sensor and a film thickness online detector; Based on the acquired data, predicting the film thickness of the coating by utilizing an artificial neural network model, and determining a key process variable with the film thickness correlation of the coating being greater than a preset threshold value through correlation analysis; and combining and optimizing the key process variables by taking the minimum film thickness of the target coating and the minimum consumption of the energy medium as optimization targets, generating optimized process parameter instructions, and issuing the optimized process parameter instructions to a three-roller coater for execution through a unit PLC. In some embodiments, the process parameter data collected in real time comprises strip thickness, strip speed, strip accumulated coating length, strip tension, coating roller diameter, coating liquid composition, concentration and viscosity, speed ratio and gap of a metering roller and a liquid taking roller, pressure between the liquid taking roller and the coating roller, speed ratio of the liquid taking roller, direction and speed ratio of the coating roller (namely, rotation direction of the coating roller and speed ratio of the coating roller and other rollers), pressure of the coating roller and strip steel, wherein the coating film thickness data is obtained through online detection of a film thickness online detector, and the coating roller diameter is obtained through online detection of an industrial camera shooting real-time image of the coating roller and adopting an image processing algorithm. In some embodiments, before the artificial neural network model is utilized to predict the coating film thickness, the method further comprises a sample generation and preprocessing step of collecting historical process parameter data and corresponding coating film thickness data to form an initial sample set, and removing abnormal data in the initial sample set by utilizing at least one of a threshold method, an angle distance method and a robust regression method to generate a three-roller coating sample set for model training. In some embodiments, the artificial neural network model is of a three-layer structure and comprises an input layer, an implicit layer and an output layer, wherein the input parameters of the input layer are the process parameter data, the output parameters of the output layer are the predicted coating film thickness, the artificial neural network model adopts a Sigmoid function as a conversion function and adopts a BP algorithm for training, and when training is carried out, one part of the three-roller coating sample set is used as training data, the other part is used as verification data until model errors are controlled within an allowable range. In some embodiments, the correlation analysis calculates the correlation coefficient between each process parameter and the coating film thickness by the following formula: Wherein x is a process parameter independent variable, y is a coating film thickness dependent variable, As the mean value