A review on model-based diagnosis methodologies for PEMFCs

• Model-based methodologies for PEMFC diagnosis are summarized based on 59 articles.
• The main approaches are introduced and discussed separately.
• Parameter identification, observer-based and parity space methods are presented.
• Models are classified in white-, grey- and black-box categories.
• Application areas, advantages and limitations are underlined for each approach.

The proton exchange membrane fuel cell systems (PEMFC)s are interesting devices for energy conversion. Recent researches are aimed at developing new monitoring and diagnosis techniques; a good management of these systems would allow optimizing the performance and reducing their degradation. The objective of a suitable diagnostic tool is to identify and isolate the different faults that may occur in the system being monitored in real time. Therefore, the main features of computational methods are accuracy, reliability and high computational speed. In order to perform the diagnosis, it is necessary to evaluate different approaches. In this work different model-based approaches are investigated as well as their validation and applications. An overview of different methodologies available in the literature is proposed, which is oriented to help in developing suitable diagnostic tool for PEMFC monitoring and fault detection and isolation (FDI).