Digital-Twin-Based Multi-Parameter Identification of Three-Phase 3L-ANPC Inverters Using Stepwise Decoupled Optimization

In increasingly electrified aerospace systems, high power density and high reliability are the core requirements of power converters. With its topological advantages, the three-level active neutral point clamp (3L-ANPC) inverter has become the preferred solution for on-board power supply systems. However, the device is exposed to extreme environments and operating conditions over extended periods. These harsh conditions will accelerate the aging of components and lead to parameter drift. The traditional parameter monitoring method has limitations in identifying multiple parameters, and it is difficult to meet the high-precision requirements of aviation-level applications. In order to solve the above problems, this paper proposes a multi-parameter monitoring strategy for the 3L-ANPC inverter based on digital twin (DT). This study adopts the fourth-order Runge-Kutta method to solve the mathematical model of the physical inverter, and on this basis, a high-fidelity digital twin model is constructed. In order to solve the problem of strong coupling between parameters, this paper deeply analyzes the multi-parameter coupling mechanism of inverters and introduces a gradual decoupling particle swarm optimization (SD-PSO) algorithm. Through phased optimization, the algorithm effectively realizes parameter decoupling, thus ensuring the high accuracy of parameter estimation. Simulation and experimental verification show that the proposed method can accurately identify the filter inductance, parasitic resistance, and power device conduction resistance, and the parameter estimation error is always controlled within 5%.