Publication

Abstract : A Separately Excited DC Motor (SEDM) is most employed machine, especially in the industrial sector. Tractions, conveyors, heavy planners, actuators are some of the most commonly known machines which are dependent on the SEDM for their stable and efficient operation. This paper proposes an application of the Artificial Neural Networks (ANNs), one of the most accurate and efficient techniques for non-linear systems, to achieve a precise trajectory control of the speed for a real-time system. In this paper, a comparison of distinct controllers such as Proportional Integral (PI) and Fuzzy Logic Controller (FLC), ANNs by analyzing the system attributes like peak overshoot time, steady-state time for varying load conditions to determine the most efficient speed controller for SEDM is implemented. The neural control scheme comprises two parts: the neural identifier and the neural controller which are used to regulate the motor speed and trigger the control signal respectively. Known for its self-adapting, learning ability, and super-fast computing features of ANN, the NARMA L-2 controller is well-suited as a speed controller for DC motors.