Optimal Postprognostics Decision-Making for Multistack Fuel Cells in Transportation: Toward Joint Load Allocation and Maintenance Scheduling

Abstract

Durability, reliability, and efficiency remain the primary barriers that impede the commercialization of proton exchange membrane (PEM) fuel cells. Prognostics and health management (PHM) is a promising solution by providing effective operational decisions to enhance fuel cell lifetime and system reliability. However, the action phase of PHM is still under development in the sense that few maintenance scheduling-related strategies have been investigated. This work formulates a novel joint load allocation and maintenance scheduling problem for a stochastically deteriorating multistack fuel cell (MFC) system. MFCs have attracted growing interest in heavy-duty applications, such as heavy-duty vehicles. Under the framework of condition-based maintenance (CBM), a load allocation strategy based on the current health state and its evolution of the studied MFC is proposed to jointly optimize the global maintenance cost. The effectiveness of the proposed strategy is compared with the average split and daisy chain strategies through case studies. The obtained results verify that by defining a proper load allocation strategy, the joint strategy can help achieve a better tradeoff between preventive maintenance and corrective maintenance, reducing unavailable time, thus decreasing the global cost. An average of over 21% maintenance cost reduction is achieved by comparing with the average split strategy.