Prediction of Wind Speed Distribution Using Artificial Neural Network: The Case of Saudi Arabia

摘要

Besides geometric and aerodynamic wind turbine parameters, wind speed and nonlinear fluctuations represent the main components in the prediction of the aerodynamic loads and performance of wind turbines. Determining wind speed characteristics is crucial in the computation of the power generated, loads and stress on rotor blades, and the fatigue of structural components. In this paper, we propose the Artificial Neural Networks (ANNs) method to forecast the daily wind speed at some locations in the Kingdom of Saudi Arabia using multiple local meteorological measurement data provided by K.A.CARE. The available database used for ANN prediction is divided into training and validation sets. The attributes for the training includes among others the time of the day, the year, the latitude and longitude, air temperature, wind direction, humidity, and pressure. The suggested model is a feed-forward neural network model with the administered learning technique using the back-propagation algorithm. The sigmoid function has been adopted as an activation function at the second layer and linear activation in the output layer. Based on the different tests conducted, the best values of the correlation coefficientRand the Root Mean Square ErrorRMSEwere obtained with a learning phase of 60% of the training set 40% of testing. By increasing the number of neurons in the hidden layer, the best structure was obtained for 10 neurons in the hidden layers matching a minimum of MSE and the highest value of R. The model has been implemented using WEKA software where numerical validation with data from meteorological stations has confirmed that the proposed model shows good agreement. The significance of the study relies on its ability to predict the daily wind speed, select the best site for wind turbine installation, ensure a secure and reliable electrical power output, and help the operators in a wind farm to manage efficiently the generated power. Another key element of this study is the outreach and dissemination of wind energy technologies within Effat community to address the challenges facing Saudi Arabia’s in the energy sector.