Proportional Integral Derivative (PID) controllers have become commonplace in various industries, finding applications in diverse fields such as industrial automation, robotics, and process control. Despite their widespread use, the challenge of parameter tuning remains a significant bottlenec k in realizing optimal controller performance. While methods like the Ziegler-Nichols method have been traditionally employed for PID parameter tuning, they often require significant expertise and may not always yield satisfactory results. Consequently, trial and error remains a prevalent albeit laborious approach to tuning PID controllers. This paper investigates the utilization of a novel Particle Swarm Optimization (PSO) algorithm as an alternative method for tuning PID controllers. PSO is a metaheuristic optimization technique inspired by the social behavior of birds flocking and fish schooling. It operates by iteratively updating a population of candidate solutions (particles) based on their individual and collective performance, with the aim of finding the optimal solution to a given optimization problem. By integrating PSO with PID controller tuning, this study seeks to overcome the limitations of traditional tuning methods and improve controller performance. The proposed approach involves formulating the PID controller parameters as optimization variables and defining an objective function that quantifies the controller's performance in terms of desired control objectives such as stability, overshoot, and se ttling time. Through a series of simulations, the effectiveness of PSO-based PID tuning is evaluated across different objective functions. The results demonstrate the capability of our algorithm to efficiently search the parameter space and converge to optimal or near-optimal PID settings, thereby enhancing control system performance while reducing the need for extensive manual tuning. Millonas' adaptability principle [2] of swarm intelligence states that the population must be able to change its behavior mode when it is worth the computational price. There has been no particle swarm optimization algorithm that decreases the number of particles as the distance to the global optimum is decreased and that is what makes our approach unique. The number of particles should be proportional to the search space, and this has not been the case historically. In PSO, the computational cost is directly proportional to the number of particles and as particles get closer to the global optimum, less of them are needed to converge on the target.
