Neural Network Based DSTATCOM Controller for Three-phase, Three-wire System

This paper deals with the design, analysis and simulation of neural network based DSTATCOM controller. Conventional controllers being fixed structure provide optimum performance only over a limited range of operating conditions for which they are designed. Alternate controllers based on fuzzy logic and neural networks are more robust and can be designed to operate well under a wide range of operating conditions. This paper highlights the performance of artificial neural network (ANN) based DSTATCOM controller with respect to PI controllers.