Wind power plant short-term wind speed prediction method based on space-time correlation

Abstract

The invention discloses a wind power plant short-term wind speed prediction method based on spatial-temporal correlation, and the method comprises the steps: firstly obtaining the spatial-temporal data of the wind speed of a target station based on VMD and a hybrid deep learning model CNN-Attention-LSTM, carrying out the data cleaning, and carrying out the VMD decomposition, and obtaining a stable IMF (intrinsic mode function) component; further, aiming at each IMF component, applying a CNN model of a bottom layer to extract airspace features of the wind speed; extracting wind speed time domain features by applying an item layer LSTM model, obtaining a prediction result of each component, and fusing a channel attention mechanism at the same time; and finally combining to obtain a final predicted wind speed. According to the method, the spatial-temporal correlation of the wind speed is fully utilized, the VMD and CNN-Attention-LSTM networks are combined, the unstable characteristic of the original wind speed is improved, the wind speed prediction precision is effectively improved, the dispatching performance of a power grid comprising a wind power plant can be optimized, and reliable and economical operation of a power system is guaranteed.

Description

technical field [0001] The invention relates to a wind speed prediction method, in particular to a short-term wind speed prediction method of a wind farm based on time-space correlation. Background technique [0002] Compared with traditional energy sources such as coal, oil, and natural gas, wind energy has the characteristics of low pollution, low cost, and sustainability. It has become the mainstream of new energy and is widely used around the world. According to the latest report released by the Global Wind Energy Council (GWEC), the global installed capacity of wind power will reach 743GW in 2020, and the new installed capacity will increase by 53% year-on-year. However, for the power system including wind farms, the randomness, volatility and intermittency of wind farms pose a major challenge to the operation control of the entire power system. [0003] At present, wind speed prediction techniques can be divided into three categories: physical models, statistical mode...

AI Technical Summary

Problems solved by technology

At present, the correlation between this part of the airspace data and the time domain data commonly used for forecasting has not been effectively utilized, resulting in a significant reduction in the accuracy of short-term wind speed forecasting in some scenarios

Images