基于数据驱动的质子交换膜燃料电池寿命预测

质子交换膜燃料电池发电技术是极具应用潜力和工业前景的技术，对质子交换膜燃料电池进行寿命预测是其走向商业推广应用的重要一环. 从质子交换膜燃料电池的退化特性及输出特性出发，分析系统及环境因素是如何影响其退化的；阐述了基于数据驱动的寿命预测方法的研究现状，着重对基于神经网络算法的寿命预测进行了综述；分析了寿命预测算法存在的不确定性来源；最后，对质子交换膜燃料电池的寿命预测研究进行了展望，阐明当前存在的经验数据有限、缺乏对瞬态过程的建模、难以实现在线预测等问题，尤其机车用大功率质子交换膜燃料电池的寿命预测仍存在诸多难点. 

Data-Driven Based Remaining Useful Life Prediction for Proton Exchange Membrane Fuel Cells

Proton exchange membrane fuel cell (PEMFC) is a power generation technology with promising application prospects. The prognosis for the remaining useful life of PEMFCs plays an important role in its commercial use. In this work, the degradation mechanism and output characteristics of PEMFCs are reviewed to explore how systems and environmental factors affect the degradation. Then, the status quo of data-driven based RUL prediction methods is summarized, while the neural network prognostics algorithms are highlighted. Furthermore, the sources of uncertainty in prediction algorithms are analyzed. Finally, the future research of RUL prediction is discussed, which focuses on the problems such as limited empirical data, lack of modeling transient processes, and being hard for online prediction. In particular, there remains many difficulties in the remaining useful life prediction of large-power PEMFCs..