Distributed Generation with Photovoltaic Power Prediction in Remote Microgrid Application

Abstract

Distributed generation with Solar photovoltaic power integration is gaining wide acceptance and popularity into grid electricity networks and remote microgrids are better positioned to benefit from this integration. The cost of fossil fuel consumption has a huge negative impact on our environment, most especially with greenhouse gas emission of CO2. Most of the remote microgrids for electrical power supply are largely powered by diesel fuel and thus increases the amount of environmental pollution around us today. It is possible to reduce drastically the use of fossil fuel by remote microgrids, if solar photovoltaic (SPV) power is considered as an alternative power source to supplement microgrid power generation. It is a known knowledge that SPV is unpredictable due to the variability in its power generation caused by intermittent solar irradiance. It becomes inevitable that SPV power be predicted to allow for efficient energy management in microgrid application. The accurate prediction of SPV electrical power generation, will assist in the accurate estimation of SPV electrical power supply required to support microgrid load and provides for an optimal scheduling of the limited SPV power. By extension, this will reduce the operational cost of diesel fuel consumption in remote microgrids and allow for cleaner energy supply to our environment. In this research paper, the focus lies in the use of artificial neural network as a technique to predict SPV power supply for remote microgrid power management.