ABSTRACT
A resilient system is able to overcome an adversity as a consequence of the anticipation of the disruption, the resistance to the damage, and mainly the recovery of its original functions. Especially in the context of technical systems, the recovery process relies on diverse estimations about performance levels and functional states. Therefore, the purpose of this research is to theoretically explore and practically employ a measure of information entropy, as an optimization variable for the process of recovery. Two novel entropy-driven recovery approaches are developed, modeled, and comparatively evaluated with a performance-based approach under different conditions of disturbance via agent-based simulations. The results suggest that the recoveries using the information entropy are more effective in both the performance maintenance and the time to recovery of the examined system, compared to the traditional recovery approach. The findings provide a conceptual and practical background to further utilize the information entropy for the design of recovery strategies and the engineering of systems’ resilience.
