Educational Experiments in Renewable Energy Analysis, Forecasting, and Management in Hybrid Power System

In this paper, analysis, forecasting and management of the power generated by a renewable energy farm including both solar energy and wind energy in a hybrid power system will be demonstrated. This renewable energy farm is connected to the utility grid. In order to properly cooperate and balance power between the load and the distributed energy sources, a method of building an accurate power forecasting and management model based on the analysis of the existing data of the load, solar irradiance and wind speed by using neural network is given. Considering realistic factors, a stochastic model of local load, available solar energy, and wind energy is proposed. The detail of how to achieve the optimal size of the renewable energy farm based on the analysis of the forecasting model and the cost function by using genetic algorithm is discussed. Based on predicting the load and power generated by the optimal renewable energy farm, the method to design a fuzzy logic power management controller to adjust the charging/discharging ratio of the energy storage to keep the system voltage and frequency stable is given. The model is built with Simulink and several other toolboxes from Mathworks Corp., such as neural network toolbox, optimization toolbox, and fuzzy logic toolbox.