PID controller parameter estimator using ant colony system

Abstract

This work presents a proportional-integral-derivative (PID) estimator that is a parallel ant colony system (para-ACS) to regulate and obtain the precise parameters of PID controller. The framework consists of a main controller that is a PID controller and an para-ACS in which is offline tuned for the parameters of PID. The structure of para-ACS consists of three ACS sets that are parallel. Each ACS set individually estimates the three parameters of a PID controller. The searching pattern of each ACS set has five layers in which each layer has ten nodes, so as to represent each parameter is a five-bit decimal real number. In addition, the integral of time and absolute error (ITAE) and the tracking error are used in the cost function of each ACS set to update the pheromone of each ACS set. Finally, the effectiveness of the proposed method has been verified by simulations using the inverted pendulum system.