On-line Tuning of a Wavelet Neural Proportional-integral Controller for Excitation Control of a Generator

Abstract

The effective control of generators needs intelligent techniques and advanced modeling. In this article, a wavelet neural proportional-integral controller, which combines the capabilities of a proportional-integral and wavelet neural network controller, is used to control a turbo-generator connected to an infinite bus via a transmission line. In the structure of the proposed controller, a wavelet neural network is also used to identify the complex non-linear dynamic of a power system. The conventional automatic voltage regulator is replaced by the proposed controller. To demonstrate the ability of the proposed controller, several simulations are performed. The simulation results show that the wavelet neural proportional-integral controller can return the terminal voltage of generator to the pre-specified level with less oscillation under a wide range of loading conditions and different disturbances. In addition, it is shown that unlike the fixed parameters based conventional automatic voltage regulator, the performance of the proposed controller is satisfactory even when the parameters of a power system are varied.

Keywords:

• on-line tuning
• intelligent controller
• wavelet neural network
• excitation control