This study investigates the application of Artificial Neural Networks (ANNs) for the speed control of a DC motor. It examines key aspects of ANN design, including learning mechanisms, training procedures, and model development. To evaluate performance, both an ANN-based controller and a conventional Proportional-Integral-Derivative (PID) controller were designed and simulated. A comparative analysis of the simulation results revealed that the ANN controller offered superior stability and accuracy. Specifically, the ANN controller achieved a significantly lower speed control error of 0.242%, compared to the PID controller’s deviation of 1.435%. These findings highlight the effectiveness of the ANN model in enhancing control precision and demonstrate its advantages over the traditional PID approach. The ANN model was trained using the Levenberg–Marquardt algorithm, which ensured efficient convergence and robust performance. Simulations were carried out in MATLAB/Simulink, enabling a detailed comparison of the controllers' dynamic responses. This work underscores the potential of ANN-based intelligent control strategies in industrial automation and motor drive applications.
Comments
No have any comment !