Lifetime Evaluation of Traction Converter IGBT Based on Operating Conditions and Component Aging

Autor: LI Dong, HE Haopeng, PENG Yangbin, XIN Kaiyue, YAO Xianping, GE Xinlai
Jazyk: čínština
Rok vydání: 2022
Předmět:
Zdroj: Kongzhi Yu Xinxi Jishu, Iss 4, Pp 89-94 (2022)
Druh dokumentu: article
ISSN: 2096-5427
DOI: 10.13889/j.issn.2096-5427.2022.04.015
Popis: IGBT's lifetime is greatly influenced by the operation conditions of train and the current aging degree of IGBT. However, the influences of train operation conditions and current life of an IGBT on the life assessment are normally not considered. This paper takes the IGBT of traction converter of high-speed EMU as the research subject to evaluate its reliability. Simulink is used to build a traction converter circuit, and simulate a complete operating condition of a train by changing load power, so as to extract the information of voltage and current of the converters, and calculate driving signal of IGBT. Considering the IGBT devices with different aging degrees, a Foster thermal network model is established, and the thermal loss of the IGBT device is calculated to obtain the information of device junction temperature by Simulink and PLECS electro-thermal co-simulation. Data of junction temperature is statistically calculated by rain flow counting method. Life prediction model is used to calculate the cumulative loss and the residual lifetime of an IGBT, and finally, the cumulative loss of IGBTs with different aging degrees is obtained based on the operating conditions of one complete process. Two life prediction models of LESIT and Norris-Landzberg are used in this paper and the lifetime of an IGBT is calculated with the results of 21.02 and 27.75 years respectively. As aging degree increases, the cumulative loss of every run will increase linearly. This scheme gives an IGBT lifetime evaluation process based on a complete operating condition. Compared with other IGBT lifetime evaluation methods, the factors are considered more comprehensively, and the prediction results are more accurate.
Databáze: Directory of Open Access Journals