Neural network current controller for a boost rectifier

This paper investigates the use of a continuously on-line trained artificial neural network current controller for a boost rectifier, to determine the viability for its use in variable speed drives, to overcome voltage dips on the mains supply. The paper analyzes simulation results of a continuous on-line trained artificial neural network controller for a variable speed drive system, and shows that this current controller is dependent on the magnitude and frequency of the supply or load currents, and the artificial neural network´s internal variables. The paper indicates the advantages and disadvantages of a neural network current controller, and show that an artificial neural network current controller can be used with a boost rectifier, to overcome mains voltage dips