Various Models for Predicting Wind Energy Production ^†

Abstract

Windmills represent one of the most virtually limitless sources of energy that may be used to generate electricity. It is regarded as a renewable source, but more investigation is indeed required to acquire the scientific knowledge and techniques that guarantee homogeneity in creation, improving the contribution of this origin to the electricity sector. This is because the wind exhibits sudden variants in speed, surface area, and other crucial factors. Comprehensive data collection methods of wind speed and phase are required for the assessment of wind resources in a location. Wind energy is produced when the wind makes contact well with the wind turbine’s rotors. The rotor rotates, converting wind speed into kinetic energy that powers the wind generator’s rotor and produces energy. In addition to assessing the energy production for the coming periods, which is valuable knowledge in the deployment of the producing units and the regulation of the power system, it is crucial to estimate the forecasts of wind activity a minimum of once a day in advance. This study creates a wind speed forecasting model for the ultra-short, short-, medium-, and long-term development of computational techniques. Utilizing wavelet-based prediction, artificial neural network approaches, including autoregressive integrated moving average (ARIMA), are used, as well as other hybrid models.