DEVELOPMENT OF A HYBRID NEURO-FUZZY-PID CONTROLLER FOR INTELLIGENT SPEED CONTROL OF WOUND-ROTOR INDUCTION MOTORS

Abstract

This research discusses the advancements in intelligent speed controllers for wound-rotor induction motors, specifically highlighting the integration of neuro-fuzzy adaptation with traditional PID regulation. While fixed-gain PID controllers serve as effective baseline regulators, they often face challenges under varying motor parameters and load conditions. Experimental results demonstrate that the Neuro-Fuzzy-PID controller significantly outperforms both the Fuzzy-PID and traditional PID controllers. The Neuro-Fuzzy-PID achieves a rise time of 0.495 seconds, which is an improvement over the Fuzzy-PID’s 0.525 seconds and the PID’s 0.615 seconds. Additionally, the settling time for the Neuro-Fuzzy-PID is 0.705 seconds, better than the Fuzzy-PID’s 0.720 seconds and the PID’s notably longer 5.025 seconds. In terms of performance, the Neuro-Fuzzy-PID exhibits the lowest overshoot at 7.73%, compared to 8.96% for the Fuzzy-PID and a substantial 28.40% for the PID. The steady-state error for the Neuro-Fuzzy-PID is reduced to an impressive 0.074364 rad/s, outperforming the Fuzzy-PID at 0.199703 rad/s and the PID at 4.617537 rad/s. These findings indicate that the Neuro-Fuzzy-PID controller not only provides faster and more accurate speed tracking but also demonstrates superior transient performance. This represents a significant advancement in motor control strategies, highlighting the potential benefits of integrating neuro-fuzzy systems with conventional PID regulation.