University Events

Description: Topic: Predictive System Resilience Modeling

Location: Lester W. Cory Conference Room, Science & Engineering Building (SENG), Room 213A

Zoom Link: https://umassd.zoom.us/j/93299499624
Meeting ID: 932 9949 9624 Passcode: 673243

Abstract:
Resilience is the ability of a system to respond, absorb, adapt, and recover from a disruptive event. In order to predict system resilience and enhance human decision-making abilities to rapidly respond to unplanned events to avoid massive damage and catastrophic outcomes, the proposed dissertation focuses on predictive statistical techniques to model and predict the time at which a system will be restored to a high-performance level after experiencing degradation. Three alternative approaches to model and predict performance and resilience metrics with techniques from reliability engineering have been applied, including (i) bathtub-shaped hazard functions, (ii) mixture distributions, and (iii) a model incorporating covariates related to the intensity of events that degrade performance as well as efforts to restore performance. Historical data sets on job losses during seven different recessions in the United States are used to assess the predictive accuracy of these approaches, including the recession that began in 2020 due to COVID-19. Goodness of fit measures and confidence intervals as well as interval-based resilience metrics are computed to assess how well the models perform on the data sets considered. The results suggest that although classical reliability modeling techniques such as bathtub-shaped hazard functions and mixture distributions are suitable for modeling and prediction of some resilience curves possessing a single decrease and subsequent recovery, the covariate models are much more flexible and achieve higher goodness of fit and greater predictive accuracy. Next steps of this dissertation will (i) explore more advanced statistical techniques such as time series analysis, to remove the assumption that changes in performance are due to hazards and activities in the present time interval only, (ii) apply the methods in the context of cyber physical systems, which are subject to diverse threats that degrade performance, and (iii) optimally allocate limited resources among the covariates describing restoration activities to achieve a desired system performance within a specific time frame.

Advisor(s): Dr. Lance Fiondella, Associate Professor, Department of Electrical & Computer Engineering,
UMASS Dartmouth

Committee Members: Dr. Ruolin Zhou, Assistant Professor, Department of Electrical & Computer Engineering, UMASS Dartmouth; Dr. Gokham Kul, Assistant Professor, Department of Computer & Information Science, UMASS Dartmouth; Dr. Mindy Hotchkiss, Technical Statistics Specialist, Aerojet Rocketdyne

NOTE: All ECE Graduate Students are ENCOURAGED to attend.
All interested parties are invited to attend. Open to the public.

*For further information, please contact Dr. Lance Fiondella at 508.999.8596 or via email at lfiondel