Self-Evolving Digital Twin-based Online Health Monitoring of Multi-Phase Boost Converters

Component degradation in power electronic converters severely threatens the system’s reliability. These components degrade over long operation time due to electrical, thermal, and mechanical stress. This calls for accurate monitoring of component health considering cost, accuracy, and adaptability. This paper develops and validates a real-time Digital Twin (DT)-based condition monitoring for Multi-Phase Interleaved Boost Converters (IBCs). The DT model employs state-space modeling to twin the real-physical hardware attributes and performance. Subsequently, outputs from both the physical hardware and the DT model undergo comparison to determine the least squared error in a multi-objective optimization setting. Techniques such as particle swarm optimization and the genetic algorithm are employed for assessing the health of the converter’s components. Furthermore, this suggested approach can be adapted for various inductor coupling methods, functioning under both continuous-conduction-mode (CCM) and discontinuous-conduction-mode (DCM). The paper proposes decoupling and hybrid approaches to improve estimation accuracy by 9.4% and reduce embedded computational requirements by 22%, respectively. A 75 kW, 60 kHz SiC IBC hardware prototype is built and tested for concept validation.

Speaker(s): Kushan Choksi

Hauppauge Radisson, 110 Motor Parkway, Hauppauge, New York, United States, 11788