Intelligent diagnosis of proton exchange membrane fuel cell water states based on flooding-specificity experiment and deep learning method

Flood -related malfunctions stand out as a primary impediment, constraining the effective and consistent functioning of proton exchange membrane fuel cells (PEMFC). This study firstly confirmed the correlation between health characteristic and PEMFC watering based on flooding -specificity experiment of PEMFC. In addition, a new index of iterative power drop was calculated, which could reflect the effect of the set operation condition on the stack water states timely. Moreover, this study took into account that not only normal state and flooded state, but also the mutual transformation stages between the two have the monitoring significance. Finally, a data -driven method was deployed to further delineate the three -classification diagnosis of the water states inside the stack and the diagnostic accuracy had been reached to 99.5 %. The proposed new index and water states definition method could open up new ideas for improving the durability and hydrogen consumption economy of PEMFC.