Prophesying Velocity of Wind Using Machine Learning Algorithms

Prophesying Velocity of Wind Using Machine Learning Algorithms

연구 분야: Infrastructure

학회: International Conference on Computers, Management & Mathematical Sciences

One of the fastest-growing renewable energy sources worldwide is wind energy, which is used for wind farms to operate efficiently and be maintained. Precise wind speed forecasts are essential. To address this need, we aim to develop a machine learning model for forecasting wind speed. The proposed model includes meteorological inputs such as temperature, humidity, amount of precipitation, wind direction, and historical wind speed data. The proposed research paper uses different machine learning algorithms, including linear regression, KNN regression, and artificial neural networks, in order to develop a viable wind speed forecast model. To determine whether the suggested tactic is effective, the proposed model will be contrasted with current wind speed forecasting methods. This study’s long-term goal is to develop a reliable wind speed prediction model that can be applied to actual conditions to enhance the performance of wind energy projects. Different regression techniques are utilized in this study to reduce error rates, and machine learning models like long short-term memory (LSTM) and convolutional neural networks (CNN)-LSTM are used to produce the best results. Both CNN-LSTM and LSTM attained values of 6.01 and 6.21 mse, respectively. Predicting the wind speed is done by using a database taken from kaggle.com.