Model Predictive Control of Continuous Stirred-Tank Reactor

This paper is concerns with the study of model predictive control of continuous stirred-tank reactor (CSTR). Firstly, a nonlinear mechanical model is established for a typical CSTR. Based on mass balance and energy balance, a set of ordinary differential equations are derived. Then, the model is linearized at steady-state point to obtain state space model, which is used in model predictive control (MPC). Finally, MPC strategy is applied, controller parameters and process constrains are discussed.