Distributed Asset Fleet Maintenance Rescheduling

Abstract

We present an approach for dynamic asset fleet maintenance scheduling that allows for rescheduling of planned maintenance based on the trajectory of asset health degradation. For scenarios wherein asset health is continually monitored, we propose a Bayesian framework to characterize the stochastic degradation process. Based on temporal asset performance degradation and failure probabilities, the scheduling optimization problem is decomposed into a master problem that solves the dynamic rescheduling problem, and a subproblem that addresses power network generation and demand fulfillment optimization.