Invited Speaker---Prof. Yi-Qing Ni
The Hong Kong Polytechnic University, Hong Kong.
Dr. Yi-Qing Ni is a Chair Professor of Smart Structures and Rail Transit at Department of Civil and Environmental Engineering, The Hong Kong Polytechnic University, Hong Kong, and the Director of National Engineering Research Center on Rail Transit Electrification and Automation (Hong Kong Branch), Hong Kong. He is a Fellow of Hong Kong Institution of Engineers (HKIE) and a Council Member of International Society for Structural Health Monitoring of Intelligent Infrastructure (ISHMII). His research areas include structural health monitoring (SHM), sensors and actuators, smart materials and structures, and monitoring and control in rail engineering. Professor Ni has published more than 180 SCI-cited journal papers with an H-index of 36 and over 4,300 citations in Web of Science Core Collection and an H-index of 48 and over 8,500 citations in Google Scholar, and over 290 international conference papers. He was selected as a Highly Cited Researcher in the Field of Civil Engineering by Shanghai Ranking Consultancy and Elsevier in 2016, and a recipient of “SHM Person of the Year Award” offered by Structural Health Monitoring: An International Journal in 2017. Professor Ni serves as an associate editor or editorial board member for 6 SCI-indexed journals, including Structural Control and Health Monitoring (John Wiley & Sons), Structural Monitoring and Maintenance (Techno-Press), and Journal of Vibration and Control (SAGE).
Innovative sensors and online monitoring for high-speed rail
Developing smart rail systems by integrating sensing, communication, computing and information technologies is becoming an urgent demand to satisfy the safety, punctuality and reliability requirements in modern rail industry. Sensory systems have been increasingly implemented on railway systems including rail infrastructure and trains for online and onboard monitoring to ensure the operational safety. In addition to innovative sensors, there is a need to develop advanced analytic tools which enable data-driven fault diagnosis and prognosis in a real-time manner. Probabilistic machine learning (PML) has currently emerged as one of the principal theoretical and practical approaches for designing ‘machines’ that learn from data acquired through sensing. In particular, probabilistic approaches developed in the framework of Bayesian machine learning (BML) provide an efficient means to interpret the heterogeneous monitoring data with diverse sources of uncertainty. The idea behind BML is that learning can be thought of as inferring plausible models in the Bayesian context to explain observed data in the presence of uncertainties. Not only does it allow for the consideration of uncertainties inherent in monitoring data in characterization and modelling, but it also enables the quantification of uncertainties in prediction and forecast. This presentation outlines innovative sensors and key BML methods successfully utilized for fault diagnosis and prognosis of high-speed rail.
High-speed rail, online and onboard monitoring, innovative sensors, fault detection, Bayesian machine learning.