Model reference input for an optimal PID tuning using PSO

Optimization of PID parameters had been a popular issue among academia and industrial players. This is because it is undoubtedly a simple and robust controller for most applications. Recently, many intelligent approaches using optimization techniques had emerged in trying to propose an efficient way of finding the optimal setting for the PID. Among all is using Particle Swarm Optimization (PSO). The PSO was utilized to search for optimum Kp, Ki and Kd that will minimized some objective functions typically IAE and ITSE. These objective functions will manage to find the optimal PID setting in terms of error elimination but yet they cannot guarantee satisfaction in transient response requirements in specific. This research proposed a new approach in PID optimization by introducing a model reference that represents the actual desired response of the controlled variable. This approach will satisfy not only error elimination, but specific transient requirements can be fulfilled perfectly. PSO was applied to find the PID setting by minimizing the error between model reference and process output signal. This paper presents simulation results using MATLAB Simulink to demonstrate the effectiveness of using model reference over step reference input alone. The proposed method was found to be superior in terms of accuracy and consistency in the results over using a step response reference signal alone.