质子交换膜燃料电池故障检测研究

Fault detection research based on proton exchange membrane fuel cell

针对质子交换膜燃料电池(PEMFC)发电系统的故障检测和系统稳定性问题,结合其多传感器特性,采用基于实时PCA的质子交换膜燃料电池故障检测方法,根据燃料电池反应信号数据建立PCA模型,通过窗口过滤方式和遗忘因子算法实时更新模型,并将降维后获得的数据用统计方法进行处理,从而检测出故障。有效地简化了燃料电池系统故障检测的过程,改善了故障检测的实时性,提高了燃料电池系统工作的稳定性和可靠性。

For the fault detection and system stability issues of proton exchange membrane fuel cell （PEMFC） power system, used the proton exchange membrane fuel ceils fault detection method based on real-time PCA combining with its multi-sensor characteristics, created the PCA model on the data of fuel cell reaction signals, and updated the model in real time by filtering through the window and forget- ting factor algorithm, processed the data after dimensionality reduction by statistical methods, so the fault was detected. This method simplified the fault detection process of fuel cell systems effectively, improved the real-time of fault detection, and enhanced the stability and reliability of the fuel cell sys- tem operation.