Reliability, Resilience and Digital Twins
Keeping power systems and their components reliable and resilient, from PV inverters to extreme weather, with digital twins and prognostics.

Overview
The EPSi Group studies reliability and resilience across power system components, distribution networks, and interconnected infrastructure. Our research asks how operating conditions, weather, component aging, and resource placement affect dependable electricity service. Early work modeled weather-related distribution interruptions and assessed smart grid reliability. This foundation developed into Markov and Boolean-driven Markov models, photovoltaic component reliability analysis, and battery life prediction using Weibull models and discharge characteristics.
Recent research extends these methods to underground distribution systems, including temperature- and load-dependent failure rates and Monte Carlo analysis of transformer vaults. Other studies examine photovoltaic inverter reliability using classified field data, fault-tolerant inverter designs, and coordinated distributed energy resource control during extreme events. Digital twin research adds the validation and repair of operational distribution models so they can support simulation. The group has also developed patented methods for boosting distribution network resilience and estimating inverter component degradation. Our projects connect these themes through automated testing of distribution management software, cybersecurity validation, and microgrid research. Together, these activities support investigation of both component condition and system behavior, including the uncertainty in generation and demand that shapes resilience planning.
Projects in this area
- Center for Agile and Intelligent Power Systems (CAIPS): Cybersecurity Research, Development, and Workforce Training Sponsored by the U.S. Department of Energy, CAIPS develops agile cyber-physical defenses across power system layers, combining artificial intelligence, formal verification, testbed validation, education, and workforce training.
- Advanced Research on Integrating Emerging & Existing Systems (ARIES) Sponsored by FPL, ARIES develops automated graphical interface tests for advanced distribution management software, verifies command-and-control functions, and gives students hands-on experience with utility software, test development, and issue reporting.
- Artificial Intelligence-Based Microgrid FIU's renewable microgrid combines photovoltaic generation, battery storage, and intelligent controls to support critical campus loads and research on resilience, battery health, frequency response, inverter dynamics, and hardware-in-the-loop equipment validation.
Key publications
- "Automated Validation and Repair Framework for Converting Operational Distribution System Into Digital-Twin-Ready Simulation Models," H. Iqbal, J. Mohammed and A. Sarwat, in IEEE Access, 2026.
- "Optimal and Coordinated DER Control for Resiliency Improvement Considering Generation and Load Uncertainties During Extreme Events", T O Olowu et al., IEEE Transactions on Industry Applications, 2025.
- "Continuous Markov-Based Reliability Analysis of Underground Distribution Power Systems Under Temperature and Load-Dependent Failure Rates", U Khan et al., IEEE Access, 2025.
- "Algorithmic formulation for network resilience enhancement by optimal der hosting and placement" S. Dharmasena, T.O. Olowu, A.I. Sarwat – IEEE Access, 2022
- "Life cycle prediction of Sealed Lead Acid batteries based on a Weibull model.", Mekonnen, Yemeserach, Haneen Aburbu, and A.I. Sarwat. Journal of energy storage 18 (2018): 467–475.
