CAREER: Real-Time Fault Diagnosis and Failure Prognosis of Next Generation Power Electronics Systems

宽带隙（WBG）功率半导体的发展将在广泛的电力电子系统和应用中发挥关键作用。与目前的同类产品相比，WBG器件具有更高的阻断电压、开关频率、功率处理能力和导热性，这可以推动向交通电气化和高度自动化的现代电网的迁移，以及其他应用。在恶劣的工作环境中，这些组件会受到各种机械和电气应力、磨损和振动的影响，从而增加设备故障的可能性，其中失效的组件可能会导致意外中断、严重的安全问题，或者很容易导致数百万美元的维修成本。提出的研究计划将通过提供状态监测工具，扩大对WBG器件退化过程的理解，早期故障诊断和剩余使用寿命预测，帮助实现可靠的WBG电力电子系统的自我监测。建立先进的可靠性框架将带来新的科学工具，加快WBG系统在电网集成、电力运输和国防系统中的部署。PI的长期愿景和职业规划是整合与该计划相关的几个重点，即：WBG功率半导体，电力电子系统和故障诊断/预测理论作为核心学科。在教育方面，目标是开发教育材料，为培训下一代电力电子工程师提供相关领域的综合视图。该项目负责人将继续在他的实验室向高中生提供切实的、以项目为中心的经验。根据拟议计划开发的技术和实际应用将成为激励和鼓励学生从事STEM职业的工具。更具体地说，该建议旨在研究宽带隙半导体器件的早期故障，并创建在线故障诊断/故障预测工具，为未来的电力电子系统建立早期预警系统。由于有限的现场数据和潜在的不确定性，宽带隙半导体的快速和广泛部署引起了对可靠性的重大关注，特别是对于任务和安全关键系统。持续监测这些系统对于防止意外停机和灾难性故障至关重要，这些故障可能导致致命事故或运营中的重大损失。然而，目前的诊断工程工具还不够成熟，无法实时检测或识别故障前兆，导致基于wbg的电力转换系统存在很大的可靠性差距。为了可靠地推动WBG技术向前发展，本提案研究了：(i) WBG器件在各种加速老化条件下的退化过程和用于预警系统的故障前兆；（ii）新颖的在线退化监测工具，利用现成的系统组件提高整体可靠性；（iii）为故障生长模型推导提供了新的分析见解，以动态评估故障严重程度，获得自适应故障阈值，并预测剩余使用寿命；（iv）为自我监测智能能源转换系统实现自主早期故障诊断和健康评估的集成框架；(v)二次控制/调制方案，将故障WBG组件的电气和热应力降至最低，以延长使用寿命。这些工具将为自我监测提供基础，智能能源转换系统可以在早期阶段识别故障前兆，从而帮助防止灾难性故障。

Advancements in wide band-gap (WBG) power semiconductors will play a critical role in a wide spectrum of power electronics systems and applications. Compared to their current counterparts, WBG devices are capable of higher blocking voltage, switching frequencies, power handling capacity and thermal conductivity, which can motivate migration toward electrification of transportation and a highly automated modern grid, among other applications. In harsh operating environments, these components are subjected to various mechanical and electrical stresses, wear out, and vibration that contribute to increased equipment failure potential, where a failed component can cause unexpected interruptions, serious safety issues, or easily lead to millions of dollars in repair costs. The proposed research plan will help achieve self-monitoring reliable WBG power electronics systems by providing condition monitoring tools that broaden understanding of the WBG device degradation process, incipient fault diagnosis and remaining useful lifetime prediction. Establishing an advanced reliability framework will lead to new scientific tools, accelerating deployment of WBG systems in grid integration, electric transportation and defense systems. The long-term vision and career plan of the PI is to integrate several thrusts relevant to this program, namely: WBG power semiconductors, power electronics systems, and fault diagnosis/prognosis theory as core disciplines. On the educational side, the goal is to develop educational materials that provide an integrated view of relevant fields for training the next generation of power electronics engineers. The PI will continue to reach out to high school students with tangible, project centric experience in his laboratory. The technologies and tangible applications developed as a result of the proposed plan will serve as instruments to inspire and encourage students to pursue STEM careers.More specifically, the proposal aims to investigate incipient faults of wide band-gap semiconductor devices and create online fault diagnosis / failure prognosis tools to establish early warning systems for future power electronics systems. The rapid and widespread deployment of wide band-gap semiconductors raises significant reliability concerns, particularly for mission and safety critical systems due to limited field data and potential uncertainties. Continuously monitoring these systems is essential to prevent unexpected shutdowns and catastrophic failures that could result in fatal accidents or significant loss in operations. However, current diagnostic engineering tools are not mature enough to detect or identify failure precursors in real time, leading to a major reliability gap in WBG-based power conversion systems. In order to reliably move WBG technologies forward, this proposal investigates: (i) WBG device degradation processes under various accelerated aging conditions and failure precursors for use in early warning systems; (ii) novel online degradation monitoring tools to improve overall reliability using readily available system components; (iii) new analytical insights for fault growth model derivation to dynamically evaluate the fault severity level, obtain an adaptive fault threshold, and predict the remaining useful lifetime; (iv) an integrated framework enabling autonomous incipient fault diagnosis and health assessment for self-monitoring smart-energy conversion systems; and (v) secondary control/modulation schemes which minimize electrical and thermal stresses on faulty WBG components for lifetime extension. These tools will provide a foundation for self-monitoring, smart-energy conversion systems that can recognize failure precursors at an early stage and thus help prevent catastrophic failures.