PID Tuning based on Classical and Meta-heuristic Algorithms: A Performance Comparison

Abstract

This paper presents a tuning process for PID controllers using meta-heuristic algorithms of artificial intelligence and mathematical optimization, specifically the genetic algorithm (GA), the imperialist competitive algorithm (ICA) and the active-set algorithm. The methodologies are used to calculate the optimal PID controllers for a three-tank level system. This type of system is commonly used in the chemical industry. The optimal controllers are tuned in closed-loop, while minimizing objective functions composed of performance indexes such as Integral Square Error (ISE), Integral Absolute Error (IAE), Integral Time Square Error (ITSE) and Integral Time Absolute Error (ITAE). The results indicate that, for this process, the meta-heuristic algorithms outperform others by minimizing cost function, overshoot and settling time.