Thickness Control Of Hot Dip Galvanizing Coating For Strip Steel Based On Data Driven And Neural Network

Hot-dip galvanizing of steel strip is an economical and effective anticorrosion method.In this paper,the coating thickness control of continuous hot-dip galvanizing of strip steel is studied.On the basis of studying the basic theory,mechanism model and control principle of hot-dip galvanizing,the coating thickness control system is developed by using data driven and neural network technology,and the industrial application test is carried out.The target of improving coating thickness accuracy,improving corrosion resistance and reducing zinc raw material consumption is realized.The main research contents are as follows:(1)Formation mechanism and equipment technology of coating.The formation process,structure and properties of the coating were studied,and the influencing factors of the coating thickness were analyzed,and the process optimization scheme was proposed.Based on the fluid dynamics analysis,the change law of physical parameters of air knife impingement jet,the flow state of liquid zinc on the surface of steel strip and the influencing factors of coating thickness are analyzed,and the mechanism model of coating thickness is established.The defects and deficiencies of key equipment for hot-dip galvanizing of steel strip are analyzed,and the improvement scheme of equipment is proposed.(2)Quality control of coating thickness.Based on the production process data,the setting range of process parameters for hot-dip galvanizing of steel strip is analyzed by statistical analysis method,and the correlation between coating thickness and influencing factors is studied.The statistical process control was carried out for the hot-dip galvanizing process of steel strip,and the control precision of coating thickness was improved by optimizing the process setting and operating procedures.(3)Process setting and coating thickness prediction model.According to the statistical analysis results of industrial production data and combined with manual operation experience,the process setting value table of air knife distance and air knife height is formulated.BP neural network is used to approximate the mapping relationship between air knife pressure and coating thickness under steady-state conditions.The neural network is trained by industrial production data under steady-state conditions,and the air knife pressure setting model is established.NARX dynamic neural network is used to establish the on-line prediction model of coating thickness,and the time series production process data are used to train the neural network model for on-line prediction.(4)Intelligent control algorithm of coating thickness.The air knife pressure setting model is used to calculate the air knife pressure setting value under different working conditions.The iterative learning control algorithm is used to track and learn the repeated tasks to realize the accurate setting and continuous optimization of the air knife pressure.Neural network multi-step predictive control is used to further improve the control performance of coating thickness.(5)Industrial control system of continuous hot-dip galvanizing coating thickness for steel strip.Based on the above research results,the control system of continuous hot dip galvanizing coating thickness of steel strip was developed by using soft PLC technology,and the industrial application test was carried out to produce high-quality hot-dip galvanized steel strip for automobile and household appliances.This project is a major technology research and development project of a large iron and steel enterprise,aiming to improve the coating thickness accuracy of continuous hot-dip galvanized steel strip products,especially upmarket automobile plate products,enhance the core competitiveness of the products,and reduce the consumption of zinc raw materials.In this paper,mechanism analysis,statistical analysis,data-driven and practical experience are combined to develop the thickness control system of continuous hot-dip galvanizing of steel strip.It has been successfully applied to industrial production,which has theoretical and practical significance.