Wind power interval prediction method based on kernel density estimation and implementation system thereof

Abstract

The invention discloses a wind power interval prediction method based on kernel density estimation and an implementation system thereof, and the method comprises the steps: firstly carrying out the deterministic point prediction of wind power based on a support vector machine modeling method of continuous time period clustering, and obtaining a comparison graph of a wind power prediction value andan actual value; secondly, establishing a probability density function for the prediction error in each power partition by adopting kernel density estimation and optimal window width selection, and performing comparative analysis with a probability density curve and an error frequency histogram obtained by a recursion method and a sliding window width method; and finally, calculating a wind powerprediction interval meeting a certain confidence probability in combination with a deterministic point prediction result, and performing comparative analysis with a recursion method and a sliding window width method by taking an interval coverage rate and an interval average width as evaluation indexes. According to the scheme, the random change rule of the wind power error is reflected more accurately, the obtained interval prediction effect is better, the precision is higher, and a more accurate wind turbine generator output interval range is provided for making an economic dispatching planfor a power system.

Description

technical field [0001] The invention relates to the technical field of wind power interval prediction, in particular to a wind power interval prediction method based on kernel density estimation and an implementation system thereof. Background technique [0002] In recent years, grid-connected wind power has developed rapidly. However, due to the high randomness and volatility of wind power and the low accuracy of wind power prediction, large-scale grid-connected wind farms will pose severe challenges to the operation and scheduling of power systems. In order to meet the requirements of power grid planning, the wind power forecasting system not only needs to provide certain point forecast values, but also should make a reasonable assessment of the risks contained in the forecast values. Therefore, on the basis of deterministic prediction, it is very necessary to calculate the confidence interval under a given confidence level. It is very important for the balance and economi...

AI Technical Summary

Problems solved by technology

[0007] The main problems of the methods mentioned above are as follows: First, most of them use a single probability distribution model to fit the probability distribution of wind power prediction error, and the effect is not very good; When using the probability density curve, the selection of the window width is not good enough to accurately reflect the random variation law of wind power prediction error

Images