[This article belongs to Volume - 56, Issue - 01, 2024]
Gongcheng Kexue Yu Jishu/Advanced Engineering Science
Journal ID : AES-30-04-2024-43

Title : ENHANCING ACCURACY AND RELIABILITY IN DENSITY FORECASTING OF WIND POWER GENERATION USING DEEP LEARNING MODEL
1*Deepak Rathod, 2Dr. Lata Gidwani

Abstract :

Planning for renewable energy systems and running the grid depend a lot on how well predictions of wind energy production work. Using normal methods, it's not always easy to record the complicated and unpredictable events that happen when wind energy is produced. This is a new way to make density forecasts more accurate and reliable during actual measurements of wind energy output. It uses deep learning models. Its goal is to make a deep learning system that can handle the spatiotemporal aspects of wind data well. Neural networks that are both recurrent and convolutional are used. The suggested method might be able to make probabilistic density predictions for wind energy generation by finding complex patterns and correlations in data on wind direction and speed. The usefulness of our model is checked by comparing it to baseline methods and using real data from wind power production. The results of our deep learning process show that it is more accurate and reliable than other methods, especially when it comes to finding strange events and uncertainty. There are also sensitivity tests that are done to see how stable the proposed framework is in different situations. The research results show that the use of deep learning models can greatly enhance the precision and dependability of estimates of wind power output. In turn, this could make it easier for green energy sources to be added to the power grid.