Predictive control optimization based PID control for temperature in an industrial surfactant reactor

• A novel PID controller optimized by model predictive control (MPC) is proposed.
• The design is applied to temperature in the industrial surfactant reactor.
• Improved closed-loop control performance is achieved in terms of set-point tracking and disturbance rejection.

Due to the character of nonlinearity, uncertainties, time delays and so on in the industrial reactors, the performance of proportional-integral-derivative (PID) control cannot always achieve the desired effect. Model predictive control (MPC) is a useful control strategy in the fact that the process models do not need to be accurately known. However, limited by the cost, hardware and so on, the application of MPC is less convenient than PID. In this paper, the temperature control in an industrial surfactant reactor is studied, where an improved PID controller optimized by extended non-minimal state space model predictive control (ENMSSMPC) framework is employed. The temperature in the surfactant reactor is first modeled as a typical step-response model and then a corresponding improved state space transformation with subsequent MPC design is done. The overall strategy combines the advantages of both PID's simple structure and MPC's good control performance. The proposed method is compared with traditional PID and MPC controllers and results show that it provides improved performance.