Compare and Evaluate One Hour Ahead Solar Power Forecasting with ANN Tuned Using Genetic Algorithm and Hybrid Genetics-Based Particle Swarm Optimization

Abstract

AbstractClimate change is one of the most pressing issues in recent times. The main reason behind this is the emission of greenhouse gases. Burning of fossil fuels in thermal power plants for electricity generation contributes to this majorly. This can be reduced by adopting renewable energy resources like wind, solar etc. Solar PV power generation systems are being increasingly adopted due to technological advancements and cost reduction. Solar power generation is not reliable due to variability in the irradiance and temperature values throughout the day. If this variability can be forecasted this will help in proper load scheduling and make the system more reliable. The major focus of the study is to reduce the time for forecasting and to minimize the error for short-time duration i.e., One Hour ahead forecasting. In this paper solar PV power forecasting is done using two methods Artificial Neural Network (ANN) tuned using Genetic algorithm (GA), Artificial Neural Network tuned using Hybrid Genetics-based Particle Swarm Optimization (PSO) respectively and the results are compared. The ANN is simulated and the tuning of ANN as well as verification of results is done using MATLAB.KeywordsSolar PV power forecastingOne hour ahead forecastingGenetics-based particle swarm optimizationGenetic algorithmArtificial neural network