Predictive Control for Coke Oven Blowing Cooler System Based on SVR

Abstract

Coke oven blowing cooler system is one of the important parts in the process of coking production. It is a complex system with the characteristics of nonlinear time-varying, multi-variable, strong random disturbance and so on which cause the system to be difficult to establish the accurate mathematics model. In this paper, a predictive control strategy based on support vector machine regression (SVR) is proposed. The support vector machine regression (SVR) modeling based on the structural risk minimization is used to predict model and the adaptive weight particle swarm optimization (APSO) algorithm is used to optimize SVR parameters. Then using online rolling optimization and feedback correction to forecast and compensate the error in the future. The simulation results show that the control strategy has strong anti-interference and robustness, and ensure the rapid and effective stability of the pre-cooling device pressure in the process.