Generalized predictive control based on multiple model switch for batch reactor process

摘要

This paper, aimed at the dynamic characteristics of a series of intermittent reaction process, using multi-model switching of generalized predictive control algorithm to study the batch reaction quality control problem. First according to the the nonlinear and time-varying characteristics of the batch reaction process, using recursive least squares (RLS) method with forgetting factors to identify temperature object accurate mathematical model parameters, multi-model switch strategy is put forward, so as to guarantee the accuracy and smooth of the multiple model switching; Then using generalized predictive cascade control (GPC-PID) algorithm to realize the reactor temperature fast tracking control. Based on a kind of ideal batch reactor temperature object simulation experiment, and compare the control effect respectively with multiple model PID and adaptive GPC - PID control, and analyzes the simulation results, which show that the proposed advanced control method can well track the given best temperature setting trajectory, and the overshoot and adjustment time is much smaller than the other two kinds of control strategy.