Online Flooding and Dehydration Diagnosis for PEM Fuel Cell Stacks via Generalized Residual Multiple Model Adaptive Estimation-Based Methodology 2019-01-0373
Improper water management such as flooding and dehydration of the membrane is one of the most critical issues in Proton Exchange Membrane Fuel Cell (PEMFC) stacks. Indeed, water management failure leads to a reduction in the efficiency and output power delivered by the fuel cell. hence its diagnose contributes to a reasonable water management, which will improve the reliability and durability of PEMFC. In the literature, previous researches established Electrochemical Impedance Spectroscopy (EIS) measurement as a powerful and useful characteristic tool to detect and identify different faults, but the sophistication and transforming operation from time to frequency domain lead to huge computational consume, or be unsuitable for commercial applications due to unaffordable dedicated instrumentation, therefore EIS is not considered to be a viable solution for the online and real-time diagnostic scheme. In this paper, an innovative approach based on EIS, is further developed for identifying some crucial PEMFC fault conditions online, in which generalized residual multiple model adaptive estimation (GRMMAE) methodology is implemented. The diagnosis process consists of multiple equivalent circuit models representing signature faults separately, such as flooding and dehydration, causing significant variation of model parameters. Applying a small sinusoidal AC current as a perturbation signal to various parallel models and measuring the potential response, whilst the noise from perturbation signal and measurement result in utilizing standard Kalman filters (KF) to generate residual signals. The residual signals are used in the GRMMAE methodology to evaluate probabilities in real-time that determine the type of fault occurrence. Simulation results show that the fault conditions can be detected and identified accurately and indicate the effectiveness of the proposed method.
Su Zhou, Shangwei Zhou, Jie Jin, Zejun Wen