Holistic Vibration Condition Monitoring for Heterogeneous Airport Assets

多伦多的皮尔逊国际机场是加拿大最大的机场，2018年处理了近5000万名旅客，预计将大幅增长。机场的旅客捷运系统、客运隧道和自动梯等设施，对机场的畅顺运作至为重要，而要确保这些设施运作畅顺，便须不断或定期进行维修，以便及时察觉情况的任何变化，例如损坏或损坏。许多容易发生故障的关键部件有利于振动监测，但其动态特性的异质性以及许多情况下故障数据的缺乏使这项工作具有挑战性。该项目旨在开发一种全面的状态监测策略，该策略可用于检测此类异类关键部件的早期故障，以便更有效地做出维护决策。在诊断方面，新的机器学习算法将与复杂的时频变换相融合，以可靠地检测瞬时事件，即使在非平稳环境中也是如此，并开发故障敏感状态指示器。然后，这些条件指标将被用于通知下游随机退化模型，以便能够在基于贝叶斯推理的单一数学框架内进行故障检测和预测。 该项目是申请者研究团队制定维护战略的更广泛努力的一部分，以减少总体维护成本和关键机场设施的计划外停机。这将使大多伦多机场管理局能够更好地实现其客运量增长目标(预计到2030年将增长6000万以上)，并与该地区的其他美国主要机场(如芝加哥)竞争。该项目还将为HQP在仪器、信号处理和可靠性理论领域的培训提供宝贵的机会。

Pearson International Airport in Toronto is the largest airport in Canada, having handled nearly 50 million passengers in 2018 and projected to grow significantly. Assets such as the automated people mover, passenger tunnels and escalators at the airport are vital to the smooth functioning of the airport and ensuring their smooth operation requires maintenance, either continuously or at regular intervals, so that any change in the condition such as deterioration or damage to its components can be detected in a timely manner. Many of the critical components prone to failure are conducive to vibration monitoring, but the heterogeneity in their dynamic characteristics and lack of failure data in many cases makes this exercise challenging. This project is aimed at developing a holistic condition monitoring strategy, which can be used to detect incipient failures in such heterogeneous critical components such that maintenance decisions can be undertaken more efficiently. For diagnostics, novel machine learning algorithms will be fused with sophisticated time-frequency transforms to reliably detect transient events, even in non-stationary environments, and to develop fault-sensitive condition indicators. Such condition indicators will then be used to inform downstream stochastic degradation models to enable both fault detection and prognosis within a single mathematical framework based on Bayesian inference. This project is a part of a broader effort by the applicant's research team in developing maintenance strategies to reduce the overall maintenance cost and unplanned outages to critical airport facilities. This will enable Greater Toronto Airports Authority to better cope with their passenger growth targets (projected to grow over 60 million by 2030) and compete with other major US airports in the region (e.g., Chicago). This project will also provide valuable opportunities for HQP training in the areas of instrumentation, signal processing and reliability theory.