电容器
电容
储能电容器
去耦电容器
电压
电容式探头
噪音(视频)
可变电容器
电子工程
电解电容器
信号(编程语言)
可靠性(半导体)
逆变器
计算机科学
工程类
控制理论(社会学)
电气工程
功率(物理)
物理
电极
量子力学
图像(数学)
人工智能
程序设计语言
控制(管理)
作者
Hao Li,Dawei Xiang,Xu Han,Xiang Zhong,Xingwu Yang
出处
期刊:IEEE Transactions on Power Electronics
[Institute of Electrical and Electronics Engineers]
日期:2019-12-01
卷期号:34 (12): 12200-12211
被引量:34
标识
DOI:10.1109/tpel.2019.2904551
摘要
Monitoring the condition of a dc-link capacitor is important for the reliability-critical voltage source inverter (VSI) systems. With the ever increasing application of a metalized poly propylene film (MPPF) capacitor, a new challenge for dc-link capacitor condition monitoring (CM) is presented due to the very small equivalent series resistance and capacitance end-of-life criterion (typically 2%-5%) of an MPPF capacitor. To improve the accuracy of capacitance estimation, a novel quasi-online CM method is proposed in this paper, which introduces an LC resonance intentionally when system is not in operation. Experimental work is carried out to validate the method and the results show that 1% capacitance estimation accuracy can be achieved in a 3-kW induction machine variable frequency drive system. Because of the signal measurement at large amplitude (e.g., nominal current) and the time-based multi-parameters identification algorithm, the method is accurate and inherently insensitive to the system parameters variation, measurement noise, and effect of temperature drift. In practice, the method can be implemented using the system available sensors with no need of hardware modification. Moreover, since the resonance current is limited and decays shortly, the condition of a dc-link capacitor can be monitored safely and frequently on-site in a VSI system.
科研通智能强力驱动
Strongly Powered by AbleSci AI