基于相似日和CAPSO-SNN的光伏發(fā)電功率預測
發(fā)布時間：2018-10-29 18:20
【摘要】：針對光伏發(fā)電功率預測精度不高的問題,提出一種基于相似日和云自適應粒子群優(yōu)化(CAPSO)算法優(yōu)化Spiking神經(jīng)網(wǎng)絡(SNN)的發(fā)電功率預測模型�？紤]到季節(jié)類型、天氣類型和氣象等主要影響因素,提出以綜合相似度指標進行相似日選取;以SNN強大的計算能力和其善于處理時間序列問題的特點為基礎,結(jié)合CAPSO算法搜索的隨機性和穩(wěn)定性優(yōu)化SNN的多突觸連接權值,減少對權值的約束,提高算法的收斂精度。根據(jù)某光伏電站的實測功率數(shù)據(jù)對所提模型進行測試和評估,結(jié)果表明,該模型比傳統(tǒng)預測模型具有更高的預測精度和更好的適用性。
[Abstract]:In order to solve the problem of low precision of photovoltaic power prediction, a generation power prediction model based on similar day and cloud adaptive particle swarm optimization (CAPSO) algorithm to optimize Spiking neural network (SNN) is proposed. Considering the main influencing factors such as season type, weather type and meteorology, a comprehensive similarity index is proposed to select similar days. Based on the strong computing power of SNN and its ability to deal with time series problems, combined with the randomness and stability of CAPSO algorithm, the multi-synaptic connection weights of SNN are optimized, which reduces the constraints on weights and improves the convergence accuracy of the algorithm. The proposed model is tested and evaluated according to the measured power data of a photovoltaic power plant. The results show that the proposed model has higher prediction accuracy and better applicability than the traditional model.
【作者單位】： 河海大學能源與電氣學院;ALSTOM
【基金】：國家自然科學基金資助項目(51277052,51507052)~~
【分類號】：TM615
[Abstract]:In order to solve the problem of low precision of photovoltaic power prediction, a generation power prediction model based on similar day and cloud adaptive particle swarm optimization (CAPSO) algorithm to optimize Spiking neural network (SNN) is proposed. Considering the main influencing factors such as season type, weather type and meteorology, a comprehensive similarity index is proposed to select similar days. Based on the strong computing power of SNN and its ability to deal with time series problems, combined with the randomness and stability of CAPSO algorithm, the multi-synaptic connection weights of SNN are optimized, which reduces the constraints on weights and improves the convergence accuracy of the algorithm. The proposed model is tested and evaluated according to the measured power data of a photovoltaic power plant. The results show that the proposed model has higher prediction accuracy and better applicability than the traditional model.
【作者單位】： 河海大學能源與電氣學院;ALSTOM
【基金】：國家自然科學基金資助項目(51277052,51507052)~~
【分類號】：TM615
【相似文獻】
相關期刊論文 前10條
1 劉士榮;李松峰;寧康紅;周嘯波;榮延澤;;基于極端學習機的光伏發(fā)電功率短期預測[J];控制工程;2013年03期
2 ;科技文摘[J];中學物理教學參考;1994年08期
3 王麗婕;廖曉鐘;高陽;高爽;;風電場發(fā)電功率的建模和預測研究綜述[J];電力系統(tǒng)保護與控制;2009年13期
4 田麗;鄧閱;;聯(lián)合條件下風力發(fā)電風速預測[J];安徽工程大學學報;2012年03期
5 盧靜;翟海青;劉純;王曉蓉;;光伏發(fā)電功率預測統(tǒng)計方法研究[J];華東電力;2010年04期
6 徐星;張虹;樂海洪;徐敏;;采用氣象信息的神經(jīng)網(wǎng)絡應用于短期風力發(fā)電功率預測[J];南昌大學學報(工科版);2011年01期
7 阿碧;;身背核電站移民外星[J];發(fā)明與創(chuàng)新(綜合科技);2011年11期
8 許昌;李e，
本文編號：2298442
本文鏈接：http://sikaile.net/kejilunwen/dianlidianqilunwen/2298442.html
最近更新
教材專著
