ROBUST CONTROL OF AN INDUCTIVE LOAD: OPTIMAL PID CONTROLLER TUNING VIA GENETIC ALGORITHM
Ali Abderrazak Tadjeddine1,2, Salih Lachache1, Med Sofiane Bendelhoum1, Mohamed Alami1, Ridha Ilyas Bendjillali1
1LSETER Laboratory, Technology Institute, Nour Bachir University Center, El-Bayadh, Algeria
2SCAMRE Laboratory, Maurice Audin National Polytechnic School of Oran, Algeria
Abstract: The precise control of inductive loads is a persistent challenge in industrial automation, where Proportional-Integral-Derivative (PID) controllers remain ubiquitous due to their simplicity and effectiveness. However, optimal tuning is critical to meet modern performance demands for both speed and stability. This paper presents a comprehensive methodology for designing and validating a high-performance PID controller for a specific RRL inductive load model. The study begins with system identification to establish a baseline with classic tuning methods, namely Ziegler-Nichols (Z-N) and Internal Model Control (IMC). These are then bench marked against optimization-based strategies using both a local optimizer (F_min.search) and a global Genetic Algorithm (GA). The optimization objective was to minimize the Integral of Time-weighted Absolute Error (ITAE), a criterion chosen for its efficacy in reducing settling time and eliminating long-duration errors. The results overwhelmingly identify the Genetic Algorithm as the superior tuning method. The GA-optimized controller achieved a remarkable balance, with a settling time of 20ms and virtually zero overshoot (less than 0.001%). This performance starkly contrasts with the classic methods; the Z-N controller was faster but produced an unacceptably high 38.9% overshoot, while the IMC controller was stable but sluggish with a settling time of 26.5ms. Furthermore, the GA controller demonstrated superior efficiency, reducing the required control effort by a significant 88.7% compared to the energy-intensive Z-N method. Extensive robustness tests confirmed the controller’s practical viability, showing it could maintain stability and precise tracking despite component parameter variations of up to ±50%, external load disturbances, and significant measurement noise. This study conclusively validates that employing a Genetic Algorithm is a highly effective strategy for developing robust, high-performance PID controllers well suited for demanding real-world applications.
Keywords: PID Controller Optimization, Transient regimes, RRL circuits, Genetic Algorithm, Controller Tuning, Inductive Load, Robustness Analysis, ITAE Criterion.