Controle de tensão e harmônicos por compensador estático de reativos com ajuste de parâmetros via redes neurais artificiais

Abstract

In this work, an artificial neural network-based static var compensator tuning is proposed for voltage and harmonic distortion control. Due to intense harmonic current injection and possible voltage sags produced by industrial facilities such as arc furnaces, an efficient robust control system is needed. Besides, electrical power systems face a new scenario with high penetration of distributed generation and non-linear loads and increased smart grid and microgrid trends. Therefore, the available reactive power sources must be able to provide system control in order to operate the system in a fast, accurate and reliable way. The application of a static var compensator (SVC) with additional filtering function at the controlled node is a possible solution. However, a precise SVC parameters tuning is needed, in order to make the system to work properly. In this work, a control strategy based on artificial neural networks is proposed. The neural networks are trained to recognize steadystate operating patterns and give the SVC adjustment. The proposed technique was implemented in the MatLab® environment. The methodology is validated by simulations in test-systems available in technical literature, using the three-phase harmonic current injection method power flow. Results show the advantages of the proposed methodology.