Performance Comparison of Neural Network Based Multi Output SMPS with Improved Power Quality and Voltage Regulation.

Abstract

In this paper a neural network based Switched Mode Power Supply (SMPS) of 175W rating is designed with ±5V and ±12V employing various converter topologies and simulated in MATLAB to  demonstrate improved power quality in terms of low total harmonic distortion of supply current, high power factor, distortion factor and displacement factor. The performance of SMPS is analyzed based on input power quality, output voltage regulation and transient response under varying source and load conditions. Comparison of performances with neural network and conventional controllers highlight the neural network control. The converter with superior performance is adopted for developing hardware.