發(fā)布時(shí)間：2018-12-08 14:37

【摘要】：母線(xiàn)負(fù)荷分析與預(yù)測(cè)對(duì)電力系統(tǒng)的安全穩(wěn)定具有重要意義。目前我國(guó)采集到的母線(xiàn)負(fù)荷數(shù)據(jù)中含有較多且類(lèi)型不同的壞數(shù)據(jù),給母線(xiàn)負(fù)荷的分析的準(zhǔn)確性與預(yù)測(cè)的精確性帶來(lái)較大影響。文中提出了一種基于低秩矩陣分解的母線(xiàn)壞數(shù)據(jù)辨識(shí)與修復(fù)方法。從母線(xiàn)數(shù)據(jù)本身出發(fā),首先分析了母線(xiàn)數(shù)據(jù)的低秩特性,研究不同類(lèi)型壞數(shù)據(jù)產(chǎn)生的原因;然后建立了一種基于低秩矩陣分解的母線(xiàn)壞數(shù)據(jù)辨識(shí)與修復(fù)的模型,并給出了基于閾值迭代法(iterative thresholding,IT)的模型求解方法;最后,利用廣東省母線(xiàn)負(fù)荷實(shí)際算例進(jìn)行了分析,并利用修復(fù)前后的母線(xiàn)數(shù)據(jù)進(jìn)行虛擬預(yù)測(cè)對(duì)比,結(jié)果實(shí)現(xiàn)了壞數(shù)據(jù)的有效恢復(fù)和預(yù)測(cè)精度的提高。
[Abstract]:Busbar load analysis and forecasting is of great significance to the safety and stability of power system. At present, the busbar load data collected in our country contain more bad data of different types, which has a great impact on the accuracy of bus load analysis and prediction accuracy. In this paper, a method for identifying and repairing bad busbar data based on low rank matrix decomposition is proposed. Starting from the bus data itself, this paper first analyzes the low rank characteristic of the bus data, and studies the causes of different types of bad data. Then a model of bad busbar data identification and repair based on low rank matrix decomposition is established, and the method of solving the model based on threshold iteration (iterative thresholding,IT) is given. Finally, the actual calculation example of bus load in Guangdong Province is used to analyze, and the virtual prediction comparison is made by using the bus data before and after repair. The result realizes the effective recovery of bad data and the improvement of prediction accuracy.
【作者單位】： 電力系統(tǒng)及發(fā)電設(shè)備控制和仿真國(guó)家重點(diǎn)實(shí)驗(yàn)室(清華大學(xué)電機(jī)系);
【基金】：國(guó)家重點(diǎn)研發(fā)計(jì)劃(2016YFB0900101) 國(guó)家杰出青年基金項(xiàng)目(51325702)~~
【分類(lèi)號(hào)】：TM714;TM715

【相似文獻(xiàn)】

相關(guān)期刊論文 前1條

1 馮平;具有分解形式的高維非線(xiàn)性電路唯一穩(wěn)態(tài)研究[J];安徽工業(yè)大學(xué)學(xué)報(bào);2001年02期

，

本文編號(hào)：2368523