Simple on-board fault-detection method for proton exchange membrane fuel cell stacks using by semi-empirical curve fitting

Abstract

For the stable operation of proton-exchange membrane fuel cells, measurements are made using sensors, and the controls are designed to detect, identify, and avoid defects based on the measured values. These methods are expensive and increase the cost of fuel cell systems as they require many sensors and a substantial amount of data. Therefore, a method that ensures reliability and lowers the cost of proton-exchange membrane fuel cells is required. In this study, the control index using overpotential was calculated using the curve-fitting method. This method can increase the on-board operation and lower the cost because sensors and measurement system are not used. To verify the accuracy of the method, the overpotentials calculated from the proposed and the other methods were compared and it was found that the proposed method could maintain the same overpotential separation accuracy as previous study even with biased measured data. Flooding and dry-out in the high current range were reproduced as defects, and both were avoided using the proposed method. These results demonstrate the effectiveness of the proposed strategy and its viability under fault conditions.