Artificial neuron controller for DC drive

Artificial neural networks (ANN) have become very popular in many control applications due to their high computation rate and ability to handle nonlinear functions. This paper presents a new scheme for the speed control of a separately excited DC motor. The signals corresponding to the motor speed error and change in speed error are used as the inputs. The artificial neural network outputs the appropriate control signals for achieving the desired speed response even under wide variations in motor parameters. The training patterns are generated using fuzzy logic principles, and the effectiveness of the proposed scheme is illustrated using simulation studies