Automated health monitoring and fault diagnosis in complex dynamical systems-- theory and applications

提出的研究是关于基础研究问题的研究，这些研究问题将为设计下一代工程系统铺平道路，这些系统将传统的控制系统设计与故障检测、隔离、估计和调节（FDIEA）目标相结合，从而使系统能够长时间、安全、有效地运行，具有更多的自主性，并且在服务中中断更少。在安全和关键任务领域，如空间和航天应用中，FDIEA问题是一个至关重要的问题。在某些其他领域，如化学/石化、制造业、电力和通信网络，安全在某种程度上，但主要是经济是拥有这些能力的主要动机。然而，出于环保考虑和对燃油效率的渴望，政府要求新车配备车载发动机诊断功能（OBD II）。我在SFU的诊断和控制（DCL）实验室有一个成熟的研究项目。多年来，这个实验室培养了许多高素质的研究人员。我们已经与政府组织，以及各种行业，如美国宇航局刘易斯（现在格伦），美国宇航局艾姆斯研究实验室，加拿大航天局（CSA），通用汽车研究实验室，帝国石油有限公司，卑诗水电，PowerTech实验室等，在现实世界的工程应用。我最大也是最值得注意的实际贡献是在2004年实现的，当时我在1993年至1994年期间发起了基于模型的发动机诊断策略，并在之后与通用汽车的研究人员合作了数年，最终投入了生产。尽管我们在技术转让方面取得了成功，但仍有许多具有挑战性和开放性的研究问题有待解决。本文针对基于模型的故障诊断中一些尚未解决的基本问题进行了研究。

The proposed research is concerned with the study of fundamental research problems that would pave the way for designing the next generation of engineering systems that integrate conventional control system design with Fault Detection, Isolation, Estimation, and Accommodation (FDIEA) objectives, thus enabling the system to function for extended periods of time, safely, efficiently, with more autonomy, and less interruption in service. FDIEA problem is of crucial importance in safety and mission critical domains such as space and aerospace applications. In certain other domains, such as chemical/petrochemical, manufacturing, power and communication networks, safety to some extent, but mainly economics is the main motivation for having such capabilities.  Yet environmental concerns and the desire for fuel efficiency drive the governments requirement for new vehicles to have On-Board engine Diagnostics capabilities (OBD II).I have an established and mature research program that I run in the Diagnostics and Control (DCL) lab at SFU. Over the years, many highly qualified research personnel have been trained in this lab. We have worked with government organizations, as well as various industries such as NASA Lewis (now Glenn), NASA Ames Research Labs, Canadian Space Agency (CSA), GM Research Lab, Imperial Oil Ltd., BC Hydro, PowerTech Labs,  among others, on real world engineering applications. My greatest and most noteworthy practical contribution came to fruition in 2004, when the model based engine diagnostic strategy that I initiated over 1993-1994, and worked with GM researchers for a number of years after, was finally rolled into production. Despite our success in technology transfer, many challenging and open research problems remain to be addressed. This proposal is about some of the fundamental unresolved research problems in model based fault diagnosis.