基于键合图的级联逆变器系统故障诊断技术.

This paper aims to improve the fault detection adaptive threshold of Single-Phase Cascade Five-Level Inverter (SPFLCI) systems to make the system fault diagnosis more accurate. According to the complex electromechanical structure and various fault modes of the SPFLCI system, we proposed a fault diagnosis technique of cascade inverter system based on bonding diagram. Firstly, established a cascade inverter bond graph model through the bond graph theory and the ideas of controlled knots. Then, the upper and lower bounds of the range of uncertainty values of each parameter were calculated by minimizing the gap between residuals and interval thresholds in a normal state. And generated the optimized interval thresholds using the new ranges of these uncertainties. Finally, the optimized interval type (OIT) threshold is used to detect whether any fault has occurred. The simulation results showed that the OIT threshold can detect more minor faults relative to the absolute type. (AT) thresholds and the optimized absolute type (OAT) threshold. [ABSTRACT FROM AUTHOR]