نوع مقاله : مقاله پژوهشی
نویسندگان
گروه مهندسی برق، دانشگاه صنعتی همدان، همدان، ایران
چکیده
کلیدواژهها
موضوعات
عنوان مقاله [English]
نویسندگان [English]
Proton exchange membrane fuel cells are considered one of the most important options for clean energy production in the automotive industry, but their performance is strongly influenced by water management within the system. In this study, a novel approach for detecting and diagnosing the water balance state in a fuel cell is presented, based on the combination of multi-valued fuzzy logic and geometric parameters of electrochemical impedance spectroscopy. Multi-valued fuzzy logic enables modeling of intermediate and boundary conditions between dehydration, normal, and flooding states. Geometric analysis of impedance spectra extracts three key parameters: ohmic resistance, the diffusion semicircle radius, and the real component of its center, which are used as inputs to a fuzzy system with 640 rules. Simulations in MATLAB on 50 operational states show that the proposed method can diagnose the fault type with an overall accuracy of 98% and estimate the severity of phenomena with a mean absolute error of 1.45%. The mean severity errors are 0.17% for normal state, 2.1% for dehydration state, and 1.7% for flooding state. Compared to conventional methods such as long short-term memory recurrent neural network with 94% accuracy, deep convolutional neural network with 90% accuracy, and type-1 fuzzy logic with 80% accuracy, this method demonstrates significant superiority and can be employed as an efficient tool for online monitoring and improving the reliability of fuel cells. This achievement can pave the way for the development of intelligent control systems and extend the operational life of fuel cells in industrial and transportation applications.
کلیدواژهها [English]
)