【摘要】：變壓器是電力系統(tǒng)中的重要設(shè)備,其安全與穩(wěn)定直接影響著國(guó)民經(jīng)濟(jì)的健康發(fā)展。油中溶解氣體分析(Dissolved Gas Analysis,DGA)是分析變壓器故障類(lèi)別的重要手段。卷積神經(jīng)網(wǎng)絡(luò)是深度學(xué)習(xí)的一種模型,廣泛應(yīng)用于圖像識(shí)別、語(yǔ)音處理等領(lǐng)域,具有非常好的分類(lèi)能力。文章選取了變壓器的五種油中溶解氣體含量作為模型輸入量,在借鑒傳統(tǒng)淺層BP神經(jīng)網(wǎng)絡(luò)油中氣體分析方法的基礎(chǔ)上,針對(duì)BP神經(jīng)網(wǎng)絡(luò)表達(dá)能力不足以及容易過(guò)擬合的缺點(diǎn),將卷積神經(jīng)網(wǎng)絡(luò)應(yīng)用于變壓器故障診斷,并與BP神經(jīng)網(wǎng)絡(luò)的分類(lèi)效果進(jìn)行了對(duì)比,通過(guò)算例研究證明了卷積神經(jīng)網(wǎng)絡(luò)的效果更優(yōu)。文章也對(duì)卷積神經(jīng)網(wǎng)絡(luò)的卷積核個(gè)數(shù)、卷積核大小以及采樣寬度對(duì)分類(lèi)效果的影響進(jìn)行了探討。
[Abstract]:Transformer is an important equipment in power system. Its safety and stability directly affect the healthy development of national economy. Dissolved gas analysis (Dissolved Gas Analysis,DGA) in oil is an important method to analyze transformer fault types. Convolutional neural network is a kind of model of deep learning, which is widely used in image recognition, speech processing and other fields, and has a very good classification ability. In this paper, the dissolved gas content in five kinds of oil of transformer is selected as the input quantity of the model. On the basis of drawing lessons from the traditional gas analysis method of shallow BP neural network oil, the expression ability of BP neural network is insufficient and it is easy to over-fit. The convolutional neural network is applied to transformer fault diagnosis, and compared with the classification effect of BP neural network. The numerical example shows that the convolution neural network is more effective. The effects of the number of convolution kernels, the size of convolution kernels and the sampling width on the classification effect of convolution neural networks are also discussed in this paper.
【作者單位】： 華南理工大學(xué)電力學(xué)院;
【基金】：國(guó)家重點(diǎn)基礎(chǔ)研究發(fā)展計(jì)劃(973計(jì)劃)(2013CB228205) 國(guó)家自然科學(xué)基金資助項(xiàng)目(51477055)
【分類(lèi)號(hào)】：TM407;TP183

【相似文獻(xiàn)】

相關(guān)期刊論文 前10條

1 王雪峰,鄔建華,馮英浚,王建元;運(yùn)用樣本更新的實(shí)時(shí)神經(jīng)網(wǎng)絡(luò)進(jìn)行短期電力負(fù)荷預(yù)測(cè)[J];系統(tǒng)工程理論與實(shí)踐;2003年04期

2 文漢云;;硫化氫燃燒的神經(jīng)網(wǎng)絡(luò)PID控制及其仿真[J];自動(dòng)化與儀器儀表;2006年01期

3 張U，

本文編號(hào)：2414862