The Application Of Fuzzy Neural Network In Circulating Fluidized Bed

Bed temperature and mainstream pressure are important parameter of circulating fluidized bed boiler (CFBB) combustion control system. To maintain them normal is key to the stable and economic operation of CFBB. Based on the dynamic character analysis of CFBB combustion system, the mathematical model describing the dynamic characteristic of bed temperature and mainstream pressure of 440t/h CFBB is proposed in this thesis. A fuzzy neural network control system is designed, and the parameters of the fuzzy neural network controller are decided off-line firstly by genetic algorithm and particle swarm optimization, adjusted on-line by use of BP algorithm. The simulation results are satisfactory in case of control effects. The results indicate that the control quality, the speediness, the robustness and the ability of anti-disturbance of fuzzy neural network control system are all superior to the general PID control system.