Intelligent Modeling and Optimization of Carbon Anode Baking Temperature

Abstract

Coal anode baking temperature system is a MIMO control system which is nonlinear and has cross-coupling and time-delay. The thermal model and flue model before are usually created based on hydrodynamics and heat transfer theory which take long to do the calculation or have ill conditions sometimes. The PID control systems based on the two mechanism models do not yield satisfactory results. In order to control the baking temperature accurately, an intelligent control model based on GA-NN with the aim of improving control precision of baking temperature of carbon anode is established by gathering the real data from anode baking furnace used for initial predictive models. A new GA-NN predictive control system is thus realized for the control of anode baking temperature of which the simulation and real control results showed that the system is efficient and effective with better control precision and robust properties than tradition PID control systems.