An online thermal monitoring method for converter system based on the extraction of DC bus leakage current

Qinghao Zhang,Geye Lu,Tianze Meng,Pinjia Zhang
DOI: https://doi.org/10.17775/CSEEJPES.2022.03560
IF: 6.014
2023-01-01
CSEE Journal of Power and Energy Systems
Abstract:The majority of converter failures are caused by overheating and temperature fluctuation. Therefore, online thermal monitoring methods for converters are necessary to enhance the reliable operation capacity. Conventional thermal monitoring methods mainly focus on a single semiconductor device. The monitoring cost is high, and measuring circuits are difficult for integration. Moreover, converter structure must be adjusted in industrial applications for these methods. To solve these problems, a novel online thermal monitoring method for IGBT-based converter system is proposed. The aim is to obtain an overall temperature evaluation of the converter. First, the leakage current in DC bus (I <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">leakDC</inf> ) is selected as an indicator for converter thermal monitoring. Theoretical analysis and offline experiments prove that I <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">leakDC</inf> is quasi-exponential to temperature. Second, temperature-related parameters (T <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">CONU</inf> and T <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">CONL</inf> ) are defined corresponding to I <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">leakDC</inf> respectively to characterize the thermal state of converters. Third, an extraction method for I <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">leakDC</inf> is proposed based on the state-delay converter control method. Using the online extracted I <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">leakDC</inf> , a novel thermal monitoring strategy is developed. Online experiments validate the effectiveness of the proposed strategy in different operating condition. The proposed method gains advantages in low cost, easy integration, feasible and convenient implementation for system-level thermal management.
What problem does this paper attempt to address?