Wavelet packet-neural network-based wind/photovoltaic power prediction method

Abstract

The invention relates to the technical field of wind / photovoltaic power generation, and discloses a wavelet packet-neural network-based wind / photovoltaic power prediction method. The method comprises the following steps of: firstly analyzing weather forecast data and history data, and assessing the influences, on the photovoltaic power prediction accuracy, of the weather forecast data and the history data; analyzing the influences, on the wind power prediction accuracy, of factors such as temperature, humidity and atmospheric pressure in the natural environment; carrying out wavelet packet decomposition on the data to obtain different frequency component sequences of a meteorological factor sequence and of a wind / photovoltaic active power sequence to ensure that the representation of change characteristics on various spatial and temporal scales is more accurate, wherein the low-frequency components and the high-frequency components respectively map the interior change rules of predicted objects from two scales of profile and detail; aiming at each sequence component, establishing a plurality of BP networks to carry out prediction; and finally reconstructing each network prediction output to obtain a complete prediction result. The method disclosed by the invention has the effects of improving the ability of mapping wind / photovoltaic power change characteristics and effectively improving the wind / photovoltaic power prediction accuracy.

Description

technical field [0001] The invention relates to the technical field of wind power generation, and relates to a wind power prediction method based on a wavelet packet-neural network. Background technique [0002] With the increasing installed capacity of wind power and photovoltaic power generation, the problems brought about by large-scale renewable energy power generation are also becoming more and more obvious. For example, wind energy and solar power generation rely on meteorological characteristics, showing obvious randomness, and the quality of the generated energy is low, which brings difficulties to power grid dispatching and control. Therefore, a more accurate forecast of the output power of wind power and photovoltaic power generation systems can help to rationally configure the output power of wind and solar power, so that it can achieve optimal regulation to meet the load and related grid-connected requirements. The accuracy of existing prediction methods such as...

AI Technical Summary

Problems solved by technology

[0007] The present invention provides, overcomes the above-mentioned deficiencies in the prior art, and it can effectively solve the problem that the statistical method based on historical data has a simple data structure, is not suitable for medium and long-term power prediction, and the special moment of the physical model-based prediction method The detailed fluctuation of the output power caused by sudden meteorological changes, the problem of insufficient forecasting ability

Images