Adaptive Adaline Neural PQ Strategy-Based Multipurpose PV Interfaced Shunt Active Power Filter

Abstract

This work introduces an adaptive Adaline neural PQ based approach in order to examine the performance of a multipurpose photovoltaic system interfaced shunt active filter. Besides supplying solar power to the nonlinear load and injecting the surplus into the distribution system, the model has a great potential of reactive power compensation, DC link voltage adjustment and Harmonic current attenuation. So as to improve dynamic performance of the system by an online and a precise identification of harmonic terms, an Adaptive Adaline Neural PQ control strategy is used. In order to explore the effectiveness of these performances, a comparison has been established between the conventional instantaneous PQ strategy and the neural one under an abrupt change of the load and of the insolation. The simulation results show that the used strategy provides better performance in terms of references tracking with aforementioned functionalities. Moreover, the proposed neural scheme leads to a total harmonic distortion, of the distribution system currents, compliant with IEEE-519 standard under the nonlinear load variation.