A new model predictive control strategy for a continuous stirred tank reactor

Abstract

A continuous stirred tank reactor (CSTR) is frequently used reactor in processing engineering. This reactor is normally operated at steady state and is assumed to be perfectly mixed. Within CSTR, the temperature, concentration, or reaction rate do not have time or position dependences. On the other hand, model predictive control (MPC) is a control strategy that can handle state and control multivariable at same time. In this work, we will use MPC to find the control strategy of an eight-state CSTR. In the MPC problem formulation, a nonlinear optimal control problem arises using a final-time performance index, the CSTR model as well as box constraints on the control variables. This optimal control problem is solved by a fast-direct method. For the CSTR states steadiness, additional auxiliary constraints are used. These auxiliary constraints are manipulated by linear parameters. Both auxiliary constraints and linear parameters are inserted to optimal control problem formulation. That is, the solution of the optimal control problem includes these optimal parameters to ensure the CSTR state steadiness. Simulation result using IPOPT under C/C++ environment is presented to show the effectiveness of the proposed control strategy.