In many industrial applications, parallel connection of power semiconductor switches is widely used to achieve higher power levels and fault-tolerant operation. But there is a current imbalance between parallel connected switches in practical applications. In this research, the impact of unbalanced currents on the junction temperature and the reliability of parallel connected switches are quantified and analyzed. Markov based reliability model is developed for quantitative assessment of reliability for parallel switches under different current sharing. The Mean Time to Failure (MTTF) is calculated for different conditions to be used as reliability metric for comparison of various conditions. A sensitivity analysis is also presented to show the effect of parameter variation on the junction temperatures of devices and reliability of the system. At last, a downscaled setup has been built up and tested. Experimental results verify the theoretical analysis and confirm procedure. Also, junction temperatures of power semiconductor devices are measured using thermal camera.

在许多工业应用中，功率半导体开关的并联被广泛用于实现更高的功率水平和容错操作。但在实际应用中，并联开关之间存在电流不平衡。在这项研究中，不平衡电流对结温和并联开关可靠性的影响进行了量化和分析。基于马尔可夫的可靠性模型被开发用于定量评估并联开关在不同电流共享下的可靠性。平均无故障时间 (MTTF) 是针对不同条件计算的，用作比较各种条件的可靠性指标。还提供了灵敏度分析，以显示参数变化对器件结温和系统可靠性的影响。最后，已经建立并测试了缩小规模的设置。实验结果验证了理论分析并确认了程序。此外，使用热像仪测量功率半导体器件的结温。