Design of Adaptive PID Controller for Coke Oven Cooling Blower Drum System

Steady state of coke oven cooling blower system directly affects process parameters of coking production. According to three different conditions(normal operating conditions, maintenance insulation conditions,abnormal conditions) on site, the nearest neighbor clustering algorithm was employed to train RBF neural network identification, and the simulation models of coke oven cooling blower system and its control system were established, the identified Jacobian information was used for BP neural network to tune PID parameters, and self-tuning PID control under different conditions was implemented. Simulation results show that with the established model and its control system the primary cooler anterior suction of cooling blower system can be controlled within a certain range quickly and effectively, with high control accuracy and good stability, which can ensure the stable operation of different condition, and its adaptability has a certain validity in stabilizing the production process parameters.