Wind power prediction apparatus and prediction method

Abstract

A wind power prediction apparatus and a prediction method are disclosed. The prediction apparatus consists of a fan data test device, a historical numerical weather prediction data interface, a data acquisition device, a router, a database, a data processor and a network server. According to the wind power prediction apparatus and method, a non-stable wind speed time sequence is decomposed into components of different scales through empirical mode decomposition, then corresponding models are established for different components respectively to perform prediction, and prediction results of all the components are combined through improved weighted prediction; and compared to a wind power modeling prediction method in the prior art, the prediction accuracy can be effectively improved.

Description

technical field [0001] The invention belongs to power system wind power prediction, in particular to the technical field of power system wind power prediction device and method. Background technique [0002] In recent years, with the rapid economic development, energy consumption has gradually increased, traditional fuels such as fossil fuels have been depleted, and environmental pollution has become increasingly serious. People have a strong desire for new forms of clean energy. In this context, new energy power generation technology has developed rapidly. At present, in the field of new energy, due to the relatively mature wind power technology, many countries take the development of wind power as an important measure to improve the energy structure, protect the ecological environment and reduce environmental pollution, and incorporate it into the national development plan. Therefore, wind power has been paid more and more attention and vigorously developed. [0003] Acc...

AI Technical Summary

Problems solved by technology

At present, artificial intelligence and other causal relationship analysis methods are commonly used to establish prediction models, and power prediction is performed on the established models, mainly including artificial neural networks (ANN), support vector machines (SVM), fuzzy logic (Fuzzy Logic), etc. Statistical prediction requirements A large amount of historical data is used to train the model, and the statistical model has strong generalization ability, but the processing accuracy of abnormal sample points is not enough

Images