Ultra-Short-Term Wind Power Forecasting Based on Deep Belief Network

Ultra-short-term wind power forecasting is one of the research hotspots of wind power generation and an important part of power system. Aiming at the problems of complex models and general accuracy of existing power forecasting algorithms, deep belief network (DBN) has be proposed for ultra-short-term wind power forecasting. In this paper, historical data are used as input to train the DBN model, through pre-training and reverse fine-tuning process, and finally output the power forecasting value. This method not only solves the problem that traditional forecasting methods cannot dig the potential information of data in depth, but also improves the accuracy of forecasting, and can effectively solve the problem that neural networks and other methods are easy to fall into local optimum. Finally, through the data modeling and Simulation of a wind farm, the results show that using DBN model can improve the forecasting accuracy, the feasibility of this method and high application value.