Improvement of multi-machine power system stability with variable series capacitor using on-line learning neural network

Abstract

This paper presents an adaptive control technique for the variable series capacitor using a recurrent neural network (RNN). Since, the parameters of the controller are determined by Genetic Algorithm (GA), which is one of the optimization algorithms, they are optimum only for that operating point and it is not possible to obtain good control performance against variations in the operating and fault point. The adaptive controller proposed in this paper consists of an optimum controller using GA and an RNN. As the RNN was on-line training, robust control performance can be achieved for various operating conditions. The effectiveness of this control method is demonstrated by considering simulation of a multi-machine power system.