Application of PSO-OSVRMPC in Coke Oven Cold Drum System

Aiming at the problems of high control complexity, strong nonlinearity, weak anti-interference ability and the inability to use mathematical models to accurately fit the traditional coke oven cold drum system, the strategy of model predictive control (MPC) based on particle swarm optimization (PSO) algorithm and nonlinear online support vector regression (OSVR) was proposed. The predictive control model was established with the identification method of OSVR, and the model was corrected online according to the online learning ability of the identification strategy. The objective function of the system was optimized by PSO algorithm to complete the predictive control of the system, and the control strategy was verified by simulation experiments. The results show that this strategy can shorten the response time of coke oven cold drum control system and improve the operation efficiency of the system. The PSO strategy has strong self-learning ability, can improve the anti-interference ability of the system and enhance the robustness of the system.