Power Quality Improvement in a Micro Grid with ELM based Nonlinear Autoregressive Neural network

Abstract

Various types of renewable energy generation sources (DGs) forms the Micro Grid concept and are absolutely preferred to meet the energy scarcity in present scenario. The renewable energy integration with the conventional grids distorts the signal quality. Improvement of power quality disturbance (PQD) increases efficiency of suppliers and consumers. In this paper the extreme learning based nonlinear Auto regressive neural network with exogenous output(ELM-NARX), has been implemented to distribution static compensator (D-STATCOM) integrated with an AC Micro Grid, to improve the power quality disturbances under various operating conditions effectively. The performance of (ELM-NARX) controller is investigated through various power quality issues like, voltage sag /swell, voltage deviation, unbalancing, communication delay etc and compared with the traditional controller like PI and NARX controller.