【摘要】：隨著半導(dǎo)體技術(shù)的不斷發(fā)展,功率器件的研究日益成熟,其性能也在進(jìn)行不斷的完善,在諸多領(lǐng)域得到了廣泛的應(yīng)用。在非平穩(wěn)工況下,IGBT在運(yùn)行過程中因使用不當(dāng),工作周期長(zhǎng)久而引發(fā)各種故障,加速器件的老化和失效程度,大大降低器件的使用壽命,影響整個(gè)裝置的正常工作。因此,對(duì)IGBT的壽命進(jìn)行評(píng)估顯得十分的迫切。本文主要針對(duì)IGBT在加速老化實(shí)驗(yàn)條件下的失效問題,從IGBT的失效機(jī)理和壽命評(píng)估方法分析出發(fā),在對(duì)IGBT的加速老化實(shí)驗(yàn)數(shù)據(jù)研究后,選取集電極-發(fā)射極關(guān)斷電壓尖峰值作為IGBT壽命評(píng)估的特征參數(shù),利用Saber仿真模型分析該參數(shù)在關(guān)斷過程的變化趨勢(shì),根據(jù)NASA PCoE研究中心提供的IGBT加速老化數(shù)據(jù),對(duì)數(shù)據(jù)進(jìn)行一次指數(shù)平滑和二次指數(shù)平滑處理,采用BP神經(jīng)網(wǎng)絡(luò)、MEA-BP以及NAR動(dòng)態(tài)神經(jīng)網(wǎng)絡(luò)建立IGBT的壽命評(píng)估模型。根據(jù)預(yù)測(cè)結(jié)果采用評(píng)價(jià)指標(biāo)均方誤差MSE,決定系數(shù)R2對(duì)比模型的預(yù)測(cè)準(zhǔn)確性。結(jié)果表明,數(shù)據(jù)經(jīng)過二次指數(shù)平滑處理后的MEA-BP壽命評(píng)估模型的預(yù)測(cè)結(jié)果最好,從而為IGBT壽命評(píng)估的相關(guān)分析提供依據(jù)。本文的分析結(jié)果均以NASA PCoE研究中心提供的IGBT加速老化數(shù)據(jù)為基礎(chǔ),實(shí)驗(yàn)數(shù)據(jù)的分析結(jié)果表明本文所提出的方法能夠運(yùn)用于IGBT的壽命評(píng)估,具有一定的工程應(yīng)用價(jià)值。
[Abstract]:With the continuous development of semiconductor technology, the research of power devices has become more and more mature, and its performance has been continuously improved, which has been widely used in many fields. Under the non-stationary working condition, the malfunction of IGBT is caused by improper use and long working cycle, which accelerates the aging and failure degree of the device, greatly reduces the service life of the device, and affects the normal operation of the whole device. Therefore, it is urgent to evaluate the life of IGBT. In this paper, aiming at the failure problem of IGBT under accelerated aging conditions, the failure mechanism and life evaluation method of IGBT are analyzed, and the experimental data of accelerated aging of IGBT are studied. The peak value of collector-emitter turn-off voltage is selected as the characteristic parameter of IGBT life evaluation. The change trend of this parameter in the turn-off process is analyzed by using Saber simulation model. According to the IGBT accelerated aging data provided by NASA PCoE Research Center, The data are processed by exponential smoothing and quadratic exponential smoothing. The IGBT life evaluation model is established by using BP neural network, MEA-BP and NAR dynamic neural network. According to the prediction results, the prediction accuracy of R ~ 2 model is compared with the evaluation index mean square error (MSE,) determination coefficient R2. The results show that the prediction results of MEA-BP life evaluation model after quadratic exponential smoothing are the best, which provides the basis for the analysis of IGBT life evaluation. The results of this paper are based on the accelerated aging data of IGBT provided by NASA PCoE Research Center. The experimental results show that the proposed method can be applied to the life evaluation of IGBT and has certain engineering application value.
【學(xué)位授予單位】：安徽理工大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類號(hào)】：TN322.8

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 岳新征;李磊民;孫飛;;基于NAR動(dòng)態(tài)神經(jīng)網(wǎng)絡(luò)的石英撓性加速度表參數(shù)建模與預(yù)測(cè)[J];西南科技大學(xué)學(xué)報(bào);2016年01期

2 陳君;張小玲;謝雪松;田蘊(yùn)杰;袁芳;楊友才;;一種IGBT熱阻測(cè)試系統(tǒng)的研制[J];半導(dǎo)體技術(shù);2015年01期

3 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

4 馬立新;吳興鋒;費(fèi)少帥;;基于FFT和神經(jīng)網(wǎng)絡(luò)的APF故障診斷方法[J];機(jī)電工程;2014年11期

5 李碩;王紅;楊士元;;基于柵極信號(hào)的功率器件老化特征分析[J];中國(guó)科學(xué):信息科學(xué);2014年10期

6 劉賓禮;唐勇;羅毅飛;劉德志;王瑞田;汪波;;基于電壓變化率的IGBT結(jié)溫預(yù)測(cè)模型研究[J];物理學(xué)報(bào);2014年17期

7 唐勇;汪波;陳明;劉賓禮;;高溫下的IGBT可靠性與在線評(píng)估[J];電工技術(shù)學(xué)報(bào);2014年06期

8 楊曉冬;王崇林;史麗萍;;H橋逆變器IGBT開路故障診斷方法研究[J];電機(jī)與控制學(xué)報(bào);2014年05期

9 方鑫;周雒維;姚丹;杜雄;孫鵬菊;吳軍科;;IGBT模塊壽命預(yù)測(cè)模型綜述[J];電源學(xué)報(bào);2014年03期

10 徐玲;周洋;張澤峰;陳明祥;劉勝;;IGBT模塊焊料層空洞對(duì)模塊溫度影響的研究[J];中國(guó)電子科學(xué)研究院學(xué)報(bào);2014年02期

，

本文編號(hào)：2330713