Integrated Load and Renewable Energy Forecasting in Micro-grid Power Supply

Abstract

Integrated Load and Renewable Energy Forecasting for Microgrid Power Supply For power planning and operation activities, accurate forecasting of demand is very important in sustaining the load demand in the electrical power system. There has been increased use of renewable energy generation which involves nature that is beyond human control. Unlike other sources of electricity, like hydro generators there is uncertainty in estimating power from renewable sources. Therefore, proper and accurate techniques for forecasting renewable energy and load demand are of paramount importance. Several forecasting techniques have been researched on in the past and are classified into; physical, statistical and Artificial Intelligence techniques. How input variables relate to output variables has been solved using these techniques, ranging from linear to complex non- parametric relationships. This research involves short term forecasting for integrated load and renewable energy (solar and wind) using Artificial Neural Network and Particle Swamp Optimization Techniques. The proposed models use weather variables and historical data for load, solar power and wind power forecasting. Particle Swamp Optimization is used to train the algorithm and update the weights and bias of the neural network.