Small Scale-Wind Power Dispatchable Energy Source Modeling

Document Type


Publication Date


Publication Title

International Journal of Scientific Engineering and Applied Science




Due to the importance of Wind energy as an intermittent renewable resource in Micro-Grids applications; this paper is proposed. So this research proposal seeks to model and analyze components of a wind turbine generator (WTG) system to store energy and supply loads with the stored energy. Focus is placed on the storage of energy into a lead acid battery and using the battery with the inverter as a dispatchable energy source. The storage device and inverter acts as a steam power plant generator. The small-scale system consists of wind turbine, wind generator, loads, dc-dc converter, ac-dc inverter, controller and battery. We use the desired power value delivered to each load to determine characteristics of the wind turbine system. Some characteristics are: wind speed, power, and charging / discharging characteristics for the battery are presented. We build the proposed real system to present a system with its components in details on a small scale. Such model’ components are tested together with other distributed system models in order to evaluate and predict the overall system performance. The proposed research presents to show an operational wind power system, for a small-scale micro-grid application. The experimental test-bed is implemented to supply the Neural Network model with its real training data. Using the Artificial Neural Network (ANN), with feed forward back-propagation technique to introduce discharging ANN model with Time as input and Voltage, Ampere-hours and Power (Watt) as outputs. ANN network consists of two layers one hidden with log-sigmoid function (has two neurons) and the second with pure-line function (has three neurons). This is done to make benefits from the ability of neural network for interpolation between points and also curves. ANN model with Back - Propagation (BP) technique is created with suitable numbers of layers and neurons. The model is checked and verified by comparing actual and predicted ANN values, with good error value and excellent regression factor to imply accuracy.