Switched Reluctance Motor Direct Torque Control with Adaptive Neural Network

In this paper, the authors present a control strategy that utilizes fuzzy PID and adaptive neural networks for speed loop regulation, where their output serves as the reference input for direct torque control to achieve precise speed control of switched reluctance motors. The key advantage of this approach lies in its sensorless speed control capability, as the fuzzy PID controller and adaptive neural network can self-adjust based on real-time motor operating conditions. From an implementation perspective, the fuzzy PID component typically involves rule-based membership functions for error and error rate processing, while the neural network employs backpropagation algorithms for online weight adaptation. The paper also provides comparative analysis with conventional control methods and discusses potential future enhancements, such as optimizing network architecture for improved control precision and dynamic response. Overall, this research offers a practical framework for engineers to achieve advanced speed control of switched reluctance motors and establishes a foundation for further investigations in adaptive motor control systems.