【摘要】：針對光伏發(fā)電系統(tǒng)在不同天氣狀況下發(fā)電功率預測精度不高的問題,在分析傳統(tǒng)方法的基礎上,提出一種無跡卡爾曼濾波神經(jīng)網(wǎng)絡光伏發(fā)電預測方法。該方法利用無跡卡爾曼濾波實時更新神經(jīng)網(wǎng)絡模型的權重,以直流電壓和電流作為系統(tǒng)的輸入,以有功功率和無功功率作為系統(tǒng)的輸出,分別建立兩個獨立的雙輸入單輸出功率預測模型。實驗結果表明:所提出的方法對有功功率和無功功率的預測精度分別為97.3%和94.2%,并且對天氣具有良好的魯棒性。
[Abstract]:In order to solve the problem that the prediction accuracy of photovoltaic power generation system under different weather conditions is not high, based on the analysis of the traditional method, an unscented Kalman filter neural network photovoltaic power generation prediction method is proposed. In this method, the unscented Kalman filter is used to update the weight of the neural network model in real time, DC voltage and current are used as the inputs of the system, and the active power and reactive power are taken as the outputs of the system. Two independent double input and single output power prediction models are established respectively. The experimental results show that the prediction accuracy of the proposed method for active power and reactive power is 97.3% and 94.2% respectively, and it is robust to weather.
【作者單位】： 國網(wǎng)青海省電力公司電力科學研究院(青海省光伏發(fā)電并網(wǎng)技術重點實驗室);重慶大學輸配電裝備及系統(tǒng)安全與新技術國家重點實驗室;
【基金】：青海省光伏發(fā)電并網(wǎng)技術重點實驗室項目(2014-Z-Y34A)~~
【分類號】：TM615;TP183

【相似文獻】

相關期刊論文 前1條

1 趙洪山;田甜;;基于自適應無跡卡爾曼濾波的電力系統(tǒng)動態(tài)狀態(tài)估計[J];電網(wǎng)技術;2014年01期

相關碩士學位論文 前1條

1 謝興;基于優(yōu)化無跡卡爾曼濾波的電網(wǎng)動態(tài)諧波檢測[D];深圳大學;2015年

，

本文編號：2508174