RBF Neural Networks Base On Particle Swarm Optimization And Its Application In Control System Of Flatness And Gauge

Strip steel is substantial raw materials in national economy. Flatness quality and gauge precious are two key indices of performance in strip hot mills, and form a complex multivariable system about the flatness and gauge control system considering strong coupling. Thus the research on the automatic flatness control-automatic gauge control (AFC-AGC) system is already at the front line of current study on strip rolling technology.With the development of intelligence control theory and technology, these theories are used in AFC-AGC. The AFC-AGC is a complex system with nonlinear, strong coupling and big time delay, so the general control methods cannot get satisfying results. As a result, modern control technology associated with the intelligence technology has been the development trend of AFC-AGC.The main contributions are as follows:An AFC-AGC integrated control model is set up by means of analyzing the main factors which affect the flatness quality and gauge precious.The particle swarm optimization (PSO) is researched and analyzed. Considering the weakness of local optimism and early convergence, this article is improved particle swarm optimization algorithm (PSO) and it has proved that this method useful.This text describes two types of neural networks, including RBF neural network and BP neural network. By comparing two types of neural networks and selecting RBF neural network. Use the improved PSO algorithm to optimize RBF neural network. By comparing various algorithms optimization results to test and verity this method’s useful.Design a RBF neural network decoupling controller based on PSO algorithm, and decoupling flatness and gauge complex control. By the computer simulation results show that the control scheme of this paper has good decoupling and robustness.In short, the neural network decoupling scheme based on PSO algorithm has simple structure, good capability of decoupling and robustness, thus it can be realized in engineering easily. This scheme expend the application range of neural network and PSO algorithm in industry control and bring forward a new approach to solve the problem of flatness and gauge complex control system. In a word, the studies in this article promote the development of neural network and PSO algorithm.