Fuzzy-neural PID Decoupling Compensation Control in Circulating Fluidized Bed Boiler Combustion System

An adaptive neural decoupling compensation method and an on-line learning mechanism for the neuron weights were presented based on the principle of neural control and decoupling control. By combining fuzzy control and neuron adaptive PID control, a decoupling control method that is independent of the accurate mathematical models of the controlled plant was proposed. The method was applied to the decoupling control of the combustion system of a circulating fluidized bed boiler (CFBB) where the three controlled variables (bed temperature, main steam pressure, and flue gas oxygen content) are strongly coupled. Simulation results show that the method can achieve good decoupling effects, and can overcome the large time delay and the nonlinearity that exist in the CFBB.