In automatic control of industrial processes, the determination of controller parameters is very important to prevent dangerous accidents and to increase quality and production. However, it is not easy to calculate controller parameters suitable for real plants, since most real processes have immeasurable model uncertainties.
PID controllers are most frequently applied in many fields where automatic control including process control is necessary. A PID controller has a great variation in control performance depending on how the parameters are determined, and thus many methods have been developed to adjust the parameters appropriately.
Conventional feedback control methods may not be good for time-delay plants because the control action is lagged. A suitable alternative to this kind of plant is the predictive control.
Hong Kwang Hyok, a researcher at the Faculty of Automatics, has investigated an improved internal model control based PID (IMC-PID) controller by combining particle swarm optimization (PSO) together with predictive functional control (PFC) framework.
First, he determined the optimal filter time constant, which is the core element of IMC, using PSO algorithm. Then, by employing the PFC idea to eliminate the effect of delay, he constructed a modified PID control system with PFC features. According to the framework of PFC, he carried out output prediction of the plant with delay and determined the optimal manipulated value by the IMC-PID control strategy.
The control method he investigated proved effective through the control of a first plus dead time (FOPDT) plant.
If further details are needed, please refer to his paper “A novel optimal design of IMC-PID controller incorporating PSO and predictive functional control framework” in “Second International Conference on Electronics, Electrical, and Control System” (EI).