Wind abandonment decrement method based on ultra-short-term wind abandonment curve prediction

Abstract

The invention discloses a wind abandonment decrement method based on ultra-short-term wind abandonment curve prediction, and belongs to the active dispatching field of electric engineering. Environment data, wind generator data, load data and self-defined data which are related to wind abandonment quantity are monitored and recorded; and five important parameters, including a weak correlation environment prediction index, a strong correlation environment prediction index, a strong correlation wind generator prediction index, a region load correction total quantity and region load correction difference between peak and valley, are calculated according to the four types of data, and then a wind abandonment curve prediction function is established. By virtue of the curve prediction function,an effect of prediction precision adjustment can be realized by improving data measurement frequency and adjusting the prediction calculation time intervals; and by virtue of accurate prediction on the wind abandonment power of a wind power plant, basis is provided for power grid coordination dispatching, so as to optimize frequency modulation and spinning reserved capacity, perform online unit combination optimization and economic load dispatching, and reduce wind abandonment quantity.

Description

technical field [0001] The invention belongs to the field of electrical engineering active power dispatching, and the invention relates to a curtailment wind reduction method based on ultra-short-term curtailment curve prediction. Background technique [0002] With the increasing scale of wind power access to the grid, the phenomenon of wind abandonment is also becoming more and more obvious, resulting in the waste of wind energy resources and the loss of power system resources, and greatly restricting the further development of wind energy. Therefore, the prediction of wind curtailment curve becomes more and more important. [0003] The existing ultra-short-term wind curtailment prediction method uses the method of subtracting the load curve from the wind power output curve, and uses the similar day method, artificial neural network method, wavelet analysis method, and support vector machine. The similar date method can only conduct quantitative analysis on existing histor...

AI Technical Summary

Problems solved by technology

[0009] Existing technology 3 has a hysteresis phenomenon when the wind power output fluctuates greatly

Images