Motor-Oriented Discrete State Event-Driven Method for Multitime-Scale Simulation of Power Traction Systems
Jiahe Ju,Zhengming Zhao,Bochen Shi,Yicheng Zhu,Zhujun Yu
DOI: https://doi.org/10.1109/TTE.2021.3053027
IF: 6.519
2021-01-01
IEEE Transactions on Transportation Electrification
Abstract:With the development of electric vehicles and high-speed train, the demand for power traction system (PTS) is on the rise. Its stability greatly relies on the performance of power traction converters, whose faults are mainly due to the microsecond- or nanosecond-level switching transient process. To avoid overheat or overvoltage failures, researchers should consider this ultrafast dynamics in power switches during the design phase, preanalyzing transient features such as voltage peak and power loss. Nevertheless, the simultaneous simulation of system dynamics and switching transient, namely multitime-scale simulation, is not easily achievable with current software because of their low efficiency and convergence issue. While the discrete state event-driven (DSED) approach can realize efficient multitime-scale simulation for converters, it is incapable of solving circuits connected with nonlinear elements, e.g., motors. This article demonstrates a motor-oriented interfacing derivative (MID) method for connecting motors with power electronic circuits based on DSED. With this method, a multitime-scale platform capable of simulating large-scale PTSs while providing detailed information such as switching transients is built, which is potentially an accurate and efficient tool for system design. Using this platform, the simulation of modular multilevel converter traction system could be finished 200 times faster than the existing commercial multitime-scale simulation software while being capable of capturing switching transient. The proposed MID method enables an efficient multitime-scale numerical platform for PTSs, which facilitates the overall and iterative design for traction systems, especially those of large scale.