Hardware Implementation of a Real-Time Neural Network Controller Set for Reactive Power Compensation Systems

Abstract

This paper introduces the use of hardware implementation of a real time neural network controller set for reactive power compensation (RPC) systems with synchronous motor. In this study, measurement of parameters required in systems such as current, phase differences, frequency and power are measured by means of a PIC 18F452 microcontroller with high accuracy and then controlled via artificial neural networks;. The performance test based on obtained data using a computer codes written in Visual Basic.Net are implemented. Different ANN controller structures are verified by simulating them on a computer. It is evaluated that the set developed can be easily adapted in real time applications.