本文選題：用電負(fù)荷特征　切入點(diǎn)：支持向量回歸機(jī)　出處：《計(jì)算機(jī)工程》2017年09期

【摘要】：針對(duì)用電負(fù)荷的周期性特點(diǎn),將用電負(fù)荷特征學(xué)習(xí)建模為小時(shí)、天數(shù)、負(fù)荷數(shù)3個(gè)維度的回歸問題,提出一種基于支持向量回歸機(jī)的三維回歸模型。將支持向量機(jī)的核函數(shù)設(shè)計(jì)為多個(gè)核函數(shù)的線性組合分別進(jìn)行參數(shù)訓(xùn)練,并給出多路徑逐步逼近的參數(shù)訓(xùn)練算法。仿真結(jié)果表明,與三層神經(jīng)網(wǎng)絡(luò)、最小二乘非線性擬合模型相比,該模型具有較好的用電負(fù)荷特征學(xué)習(xí)與預(yù)測(cè)能力。
[Abstract]:According to the periodicity of electric load, the paper models the characteristic learning of power load as the regression problem of three dimensions: hour, day and load number. A three dimensional regression model based on support vector regression machine is proposed. The kernel function of support vector machine is designed as a linear combination of several kernel functions to train parameters separately. The simulation results show that compared with the three-layer neural network and the least square nonlinear fitting model, this model has a better ability to learn and predict the characteristics of power load.
【作者單位】： 長(zhǎng)沙理工大學(xué)計(jì)算機(jī)與通信工程學(xué)院;
【基金】：國(guó)家自然科學(xué)基金(61303043) 湖南省自然科學(xué)基金(13JJ4052) 湖南省教育廳科學(xué)研究項(xiàng)目(13C1023)
【分類號(hào)】：O212.1;TM743

【相似文獻(xiàn)】

相關(guān)期刊論文 前1條

1 施俊良;;計(jì)算用電負(fù)荷的C—K_L法[J];電氣工程應(yīng)用;1986年02期

，

本文編號(hào)：1674284