Study On Multi-level Hybrid Modeling Method And Control Performance Of Hydraulic AGC System

The high-speed development of national economy promotes the development of modern equipment manufacturing industry towards the direction of high efficiency and intelligentization.As an important pillar industry,the steel industry is developing rapidly.The requirements to the quality of cold-rolled strip are higher and higher in all walks of life,and the thickness is an important index to measure the quality of cold-rolled strip.Hydraulic automatic gauge control(AGC)system is the core control system to ensure the accuracy of the strip thickness.The system is characterized by high control precision and fast response speed,which is the typical representation of modern machine electro-hydraulic integrated system.Therefore,hydraulic AGC system was employed as the research object.Combining with real-time data,the hybrid modeling method of complex nonlinear system was investigated,in order to improve the precision of the model and optimize the control performance of the system.Firstly,in this paper,the structure and function of hydraulic AGC system were analyzed in detail.A multi-level hybrid modeling method was proposed to solve the problem of nonlinear modeling of large-scale complex systems,and the working process of hydraulic AGC system was divided into several simple sub-processes;Secondly,from the perspective of machine electro-hydraulic,considering the nonlinear factors of the system,a mechanism model based on state space equation was established;Thirdly,the influence of key parameters on the system were studied,focusing on the Lugre nonlinear friction model.Then,intelligent optimization algorithm was used to identify unknown parameters accurately and validated by experiment.At last,the complete multi-level hybrid model of hydraulic AGC system was set up with MATLAB software.Furthermore,through the optimization of PID control parameters,the system control performance was further improved and the simulation results were verified by the relevant experimental data of 650 rolling mill.In this paper,a new modeling method is proposed for the large-scale complex nonlinear system such as hydraulic AGC system.The multi-level hybrid modeling method combined mechanism model and parameter identification can greatly reduce the difficulty of modeling.The method can also improve the accuracy of the model and make the hybrid model closer to the actual situation.This research provides a theoretical basis for further research on optimizing the control performance of the system.