ANN BASED HVDC CONTROLLERS FOR POWER OSCILLATION DAMPING IN MULTI MACHINE POWER SYSTEM USING EXTREME LEARNING MACHINE ALGORITHM

Abstract

This paper discusses the artificial intelligence-based controllers for HVDC in multi machine AC-DC power system. In the last decade Researchers have applied AI based HVDC controllers to improve the stability limit in large inter connected power system. Artificial neural network-based controllers are popular among other controllers falls in this category. The performance of these controllers majorly depends on the method of training the neural network. Even though back propagation algorithm finds to be suitable for many applications, it also suffers from the draw backs like slow convergence, local minima and longtime taking in convergence.These disadvantages are overcome using Extreme learning machine algorithm trained neural networks. The effectiveness of the proposed training algorithm is tested on IEEE 9-Bus test power system using MATLAB/SIMULINK under different fault conditions.