-

The aim of this paper is to investigate the methods of representation of a generator in power systems and a new design for a power system stabiliser based on genetic algorithms and artificial neural networks. For this purpose the most important models which are usually used to simulate the generator and power system are introduced. These models are applied in a multimachine system.
A new simple genetic algorithm is applied to a power system stabiliser to find the optimum parameters of a PID controller at different fixed operating points. Then an artificial neural network is proposed to adapt the parameters of the stabiliser to different operating points and different conditions such as a three phase short circuit. This stabiliser is applied to the multimachine systems with and without infinite bus.