Cyber Resilience Assessment of Unbalanced Distribution System Restoration under Sparse Load Forecasting Attacks

Abstract

System restoration is critical for power-system resilience, but its growing reliance on artificial intelligence (AI)-based load forecasting creates a cyber-physical vulnerability in the restoration decision loop. Manipulated forecasts can cause infeasible restoration schedules, insufficient inverter-based-resource ramping margins, and unsuccessful recovery of de-energized segments, yet the resilience of restoration processes to such attacks remains largely unexplored. This paper evaluates restoration vulnerability at the system level rather than only measuring forecasting error. A gradient-based sparse perturbation method is developed as a stress-testing tool to identify influential forecasting inputs. We further create a restoration-aware validation framework that embeds these compromised forecasts into a sequential restoration model and evaluates operational feasibility using an unbalanced three-phase optimal power flow formulation. Case studies on a modified IEEE 123-bus feeder show that sparse input perturbations can substantially increase forecasting error and make selected microgrid restoration stages infeasible. The results reveal system-level failures caused by active-power-balance infeasibility and power ramping violations, which can prevent the restoration of critical loads. These findings provide actionable insights for designing cybersecurity-aware restoration planning frameworks.