Abstract:
High-precision wind power forecasting provides reliable basis for dispatching of power system. This paper proposes a forecasting procedure that applies Long Short-time Me...Show MoreMetadata
Abstract:
High-precision wind power forecasting provides reliable basis for dispatching of power system. This paper proposes a forecasting procedure that applies Long Short-time Memory (LSTM) neural network to forecast ultra-short term wind power. First, Spearman rank correlation coefficient method is utilized to determine the hyper-parameters of LSTM forecasting model. Then, case study using the practical data collected from a wind farm in western China is performed to verify the LSTM forecasting model. Study results indicate that the proposed method has higher forecasting accuracy than traditional artificial neural networks.
Date of Conference: 07-09 September 2019
Date Added to IEEE Xplore: 27 April 2020
ISBN Information: