Automated Anomalies Detection and Analysis for Remote Monitoring
Abstract: One of the numerous responsibilities of nuclear power plant (NPP) operators in ensuring safe, reliable plant operations is to monitor and respond to plant alarms. Alarms are generated when plant process anomalies occur and are significant enough to exceed the particular threshold specified in the plant designs. Operators do not search for and detect subtle anomalies in the plant data. However, anomaly detection could help plants prevent such anomalies from escalating into unexpected equipment failure, especially since alarms may not provide sufficient lead time for the plant to act. Intelligent tools could detect subtle signs of anomalies in NPP components and alert operators to mitigate any consequences.
To aid the nuclear power industry, the U.S. Department of Energy Light Water Reactor Sustainability (LWRS) program has been investigating machine learning methods for automated anomaly detection based on time-series data. This has included studies conducted on NPP test cases; studies investigating extraction and incorporation of sparsely labeled known anomalous events into the anomaly detection methods; studies on means to determine the cause of anomalies as they are detected; and studies to scale anomaly detection models at various power levels. This talk aims to provide an overview of the LWRS program efforts in general and anomaly detection research in specific.
Bio: Ahmad Al Rashdan, Ph.D., is currently a senior research and development scientist in the nuclear science and technology directorate at INL. Dr. Al Rashdan holds a Ph.D. in Nuclear Engineering from Texas A&M University, a M.Sc. in Information Technology and Automation Systems from Esslingen University of Applied Science from Germany, and a B.Sc. in Mechanical Engineering from Jordan University of Science and Technology in Jordan. He has around 15 years of industrial and research experience in automation, instrumentation, and control, including experience at INL, the ABB Group, Texas A&M University, the International Atomic Energy Agency, Daimler Chrysler-Mercedes Group, and Fraunhofer Institute for Production and Automation. His experience includes automated work processes using artificial intelligence methods and advanced analytics, online condition monitoring of nuclear systems, control systems design and development, anomalies detection, and automated modeling and simulation. Dr. Al Rashdan is an active contributor to and organizer of several DOE events and scientific conferences. He authored or co-authored more than 40 technical reports and journal papers and five patent applications and is an active reviewer for several nuclear energy and Institute of Electrical and Electronics Engineers (IEEE) journals, as well as many DOE grants. He is currently a guest editor for the Big Data Analytics for the Nuclear Power Plants Special Issue of the Progress in Nuclear Energy Journal. He is the recipient and co-recipient of more than 15 recognition and funding awards. Dr. Al Rashdan is a senior member of the IEEE and a member of the American Nuclear Society.