GOALI: Sensorless condition monitoring and prognostics for automotive alternators

本研究的目的是开发一种基于物理、模型和数据的诊断和预测算法，用于广泛应用于汽车的爪极同步交流发电机。该方法仅使用电压和电流传感器来满足这一目标。这是通过创建交流发电机的详细电磁和机械模型来完成的。接下来，通过将输出电流的特征与故障模式联系起来，制定诊断规则。然后，基于试验台数据、启发式进展模型以及与已开发的故障机器模型的比较，开发了预测技术来预测RUL和故障严重程度。本研究的智力价值包括改进了对交流发电机故障机制的理解，以及它们在输出电流特征中的表现，无传感器数据诊断的新算法，以及包括启发式，基于模型和实验衍生信息的预测新技术。这些概念将在实际的硬件测试平台上进行验证，并使用制造商的设计数据。更广泛的影响研究的好处包括及时和准确地检测交流发电机问题，这将减少滞留驾驶者的发生率，并大大减少误诊，从而导致保修成本节省。此外，该项目的成果将提高汽车可靠性，影响整个经济，并降低维修成本和保修成本，直接影响消费者的价格。这相当于每年数百万美元。这项技术应用于美国汽车将提高我国的竞争力。最后，预测工作将使所有类型和应用的电机受益。这项研究也将有助于新的研究生课程内容，以及对本科生和研究生的培训。在与该机构的机会学者计划现有合作的基础上，将努力从代表性不足的群体中招收学生。

The objective of this research is to develop a combination of physics-, model-, and data-based diagnostic and prognostic algorithms for clawpole synchronous alternators that are widely used in automotive applications. The approach is to meet this objective using only voltage and current sensors. This is accomplished by creating detailed electromagnetic and mechanical models of alternators. Next, diagnostic rules are developed by relating the features of the output current to failure modes. Then, prognostic techniques are developed to predict RUL and fault severity based on test bench data, heuristic progression models, and a comparison to developed models for faulty machines.Intellectual MeritThe intellectual merit of this research includes an improved understanding of alternator failure mechanisms, and their manifestation in features of the output current, new algorithms for sensorless data-based diagnostics, and new techniques for prognostics which involve the inclusion of heuristic, model-based and experimentally-derived information. These concepts will be proven on actual hardware test platforms and using design data from the manufacturer.Broader ImpactsResearch benefits include the timely and accurate detection of alternator problems which will decrease stranded motorist incidences and significantly decrease misdiagnoses, thereby resulting in warranty cost savings. Also, the project results will improved automotive reliability affects the entire economy, and reduce maintenance costs and warranty costs which directly affect consumer prices. This represents many millions of dollars per year. This technology applied in US cars will improve our countrys competitiveness. Finally, the work on prognostics will benefit all types and applications of electric machines. The research will also contribute to new graduate course content as well as to training of undergraduate and graduate students. Building on existing collaboration with the institution's Opportunity Scholars Program, an effort will be made to recruit students from underrepresented groups.