The Gauge Control Based On Radial Basis Function Neural Network Sliding Mode Variable Structure

Hot rolling mill strip steel has high production efficiency,good economical value,so its one of the most important steel products of steel industry,its quality and production technology represent one country's level of steel industry.After perennial development,based on the traditional control theory of the production lines are difficult to satisfy the development of science and technology and the market demand for higher product quality requirement,some difficulties difficult to solve,so the new control theories and methods come out to make up for the inadequacy of the traditional control methods,satisfy the needs of the market.To solve above problems,here puts forward an algorithm based on improved particle swarm optimization optimize RBF neural network of sliding mode variable structure control algorithm.First,in the way dynamic changing the inertia weight to optimize particle swarm optimization,make the particles hard to achieve local optimal value in the optimization process and enhance the search efficiency.Through the experimental simulation proves the validity of this improved method.Secondly, using the improved particle swarm optimization to optimize the structure and parameters of RBF neural network and create a more powerful network model.Through the simulation experiment verifies the network has better fitting performance.The last,based on the equivalent sliding mode variable structure control method,IPSO-RBF network as one part of the control law to identify the nonlinear problems.The method can effectively reduce the chattering and eliminating interference by the simulation.Eventually formed a method based on the IPSO-RBF equivalent control of discrete sliding mode control.Here mainly studies the strip steel thickness control system of an important looper system.Looper system's control level will directly affect the merits of the thickness control.Using the IPSORBF equivalent control of discrete sliding mode control for the decoupling control of looper height and tension.The experiment result shows this method can remove coupling relationship.After decoupling of the tension and height control are not interference each other to improve the control precision of the system.So as to improve the rolling of strip steel thickness control precision and rolling out more high quality products.