發(fā)布時(shí)間：2018-10-16 18:25

【摘要】：作為電力電子設(shè)備中重要的大功率開(kāi)關(guān)器件,IGBT(Insulated Gate Bipolar Transistor,絕緣柵雙極晶體管)已經(jīng)被應(yīng)用到眾多領(lǐng)域,發(fā)展前景十分廣闊。然而長(zhǎng)時(shí)間工作在高強(qiáng)度環(huán)境中的IGBT模塊會(huì)逐漸老化,甚至失效。因此能夠?qū)GBT模塊進(jìn)行實(shí)時(shí)故障預(yù)測(cè),不但可以減少日常維護(hù)成本,更能避免因器件失效而造成重大事故�；诖藛�(wèn)題,本文結(jié)合機(jī)器學(xué)習(xí)相關(guān)算法對(duì)IGBT故障預(yù)測(cè)技術(shù)進(jìn)行研究。本文主要工作如下:首先,對(duì)IGBT模塊的工作特性、失效原因以及退化參數(shù)進(jìn)行研究,選取關(guān)斷狀態(tài)的瞬態(tài)尖峰電壓作為觀測(cè)參數(shù),對(duì)NASAPCoE研究中心提供的老化數(shù)據(jù)進(jìn)行處理,得到預(yù)測(cè)實(shí)驗(yàn)所需要的尖峰電壓退化數(shù)據(jù)。其次,深入研究機(jī)器學(xué)習(xí)領(lǐng)域中的回歸算法和神經(jīng)網(wǎng)絡(luò)算法,利用Google開(kāi)源平臺(tái)TensorFlow搭建算法模型,對(duì)關(guān)斷尖峰電壓退化數(shù)據(jù)進(jìn)行訓(xùn)練和預(yù)測(cè)。結(jié)果表明:使用RMSProp和批標(biāo)準(zhǔn)化優(yōu)化的長(zhǎng)短期記憶循環(huán)神經(jīng)網(wǎng)絡(luò)具有較高的預(yù)測(cè)準(zhǔn)確度和訓(xùn)練速度,可以對(duì)IGBT退化數(shù)據(jù)進(jìn)行預(yù)測(cè)。最后,為了提高設(shè)備維護(hù)效率,設(shè)計(jì)一套采用B/S架構(gòu)的IGBT實(shí)時(shí)故障預(yù)測(cè)軟件系統(tǒng)。歷史數(shù)據(jù)的預(yù)測(cè)使用隨機(jī)梯度下降法對(duì)訓(xùn)練過(guò)程進(jìn)行加速;實(shí)時(shí)數(shù)據(jù)的預(yù)測(cè)采用滾動(dòng)預(yù)測(cè)方式實(shí)現(xiàn),有效避免因長(zhǎng)時(shí)間預(yù)測(cè)帶來(lái)的誤差積累問(wèn)題。該系統(tǒng)實(shí)現(xiàn)了基于循環(huán)神經(jīng)網(wǎng)絡(luò)算法的IGBT故障預(yù)測(cè)功能,具有實(shí)際意義。
[Abstract]:As an important high-power switch device, IGBT (Insulated Gate Bipolar Transistor, insulated gate bipolar transistor (IGBT), it has been applied in many fields and has a broad prospect. However, IGBT modules that work long hours in high-intensity environments may gradually age, or even fail. Therefore, the real-time fault prediction of IGBT module can not only reduce the cost of daily maintenance, but also avoid serious accidents caused by device failure. Based on this problem, this paper studies the IGBT fault prediction technology combined with machine learning algorithm. The main work of this paper is as follows: firstly, the working characteristics, failure reasons and degradation parameters of IGBT module are studied, and the transient peak voltage of turn-off state is selected as the observation parameter to process the aging data provided by the NASAPCoE research center. The peak voltage degradation data needed to predict the experiment are obtained. Secondly, the regression algorithm and neural network algorithm in the field of machine learning are deeply studied, and the algorithm model is built by using Google open source platform TensorFlow to train and predict the voltage degradation data of off peak. The results show that the long and short term memory cycle neural networks optimized by RMSProp and batch standardization have high prediction accuracy and training speed, and can be used to predict IGBT degradation data. Finally, in order to improve the efficiency of equipment maintenance, a IGBT real-time fault prediction software system based on B / S architecture is designed. The prediction of historical data is accelerated by stochastic gradient descent method, and the prediction of real-time data is realized by means of rolling prediction, which effectively avoids the problem of error accumulation caused by long-term prediction. The system realizes the function of IGBT fault prediction based on cyclic neural network algorithm, which has practical significance.
【學(xué)位授予單位】：北京交通大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類(lèi)號(hào)】：TN322.8

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 李亞萍;周雒維;孫鵬菊;;IGBT功率模塊加速老化方法綜述[J];電源學(xué)報(bào);2016年06期

2 謝宏宇;侯艷;李康;;基于正則化回歸的組學(xué)數(shù)據(jù)變量篩選方法[J];中國(guó)衛(wèi)生統(tǒng)計(jì);2016年04期

3 楊會(huì)玲;;高速列車(chē)牽引變流器直流電壓脈動(dòng)的研究[J];電子設(shè)計(jì)工程;2016年15期

4 楊經(jīng)偉;王念春;;IGBT模型優(yōu)化及其在Buck變換器中的功耗分析[J];電源技術(shù);2015年11期

5 杜雄;李高顯;劉洪紀(jì);孫鵬菊;周雒維;;風(fēng)速概率分布對(duì)風(fēng)電變流器中功率器件壽命的影響[J];電工技術(shù)學(xué)報(bào);2015年15期

6 趙翠榮;;基于支持向量機(jī)的汽車(chē)故障預(yù)測(cè)貝葉斯網(wǎng)絡(luò)推理系統(tǒng)研究[J];巢湖學(xué)院學(xué)報(bào);2015年03期

7 毛婭婕;周雒維;杜雄;孫鵬菊;;IGBT加速老化實(shí)驗(yàn)研究[J];電源技術(shù);2014年12期

8 劉賓禮;劉德志;唐勇;陳明;;基于加速壽命試驗(yàn)的IGBT模塊壽命預(yù)測(cè)和失效分析[J];江蘇大學(xué)學(xué)報(bào)(自然科學(xué)版);2013年05期

9 韓東;楊震;許葆華;;基于數(shù)據(jù)驅(qū)動(dòng)的故障預(yù)測(cè)模型框架研究[J];計(jì)算機(jī)工程與設(shè)計(jì);2013年03期

10 范庚;馬登武;張繼軍;;一種基于小樣本數(shù)據(jù)的裝備故障預(yù)測(cè)方法[J];彈箭與制導(dǎo)學(xué)報(bào);2012年04期

相關(guān)碩士學(xué)位論文 前5條

1 邴欣;機(jī)器學(xué)習(xí)在推薦系統(tǒng)中的應(yīng)用[D];山東大學(xué);2016年

2 朱炯炯;IGBT模塊故障預(yù)測(cè)技術(shù)[D];電子科技大學(xué);2014年

3 李鵬;高壓大功率IGBT測(cè)試平臺(tái)的研制及相關(guān)問(wèn)題研究[D];浙江大學(xué);2012年

4 陳永淑;IGBT的可靠性模型研究[D];重慶大學(xué);2010年

5 鄧鵬飛;高速公路供電站風(fēng)力及光伏發(fā)電逆變控制器的研究[D];長(zhǎng)沙理工大學(xué);2008年

，

本文編號(hào)：2275246