针对质子交换膜燃料电池(proton exchange membrane fuel cell, PEMFC)系统可靠性较低的问题，提出了一种基于变分模态分解(variational mode decomposition, VMD)和蝴蝶 优化算法(butterfly optimization algorithm, BOA)的最小二乘支持向量机(least square support vector machine, LSSVM)故障诊断方法。首先，利用传感器获取系统运行的时序信号，并对采集到的时序信号进行 VMD；然后，计算各模态分量的模糊熵，将模糊熵作为故障检测特征值构建数据集；最后，利用数据集构建 BOA 优化的 LSSVM 故障分类器。通过测试集对构建的故障分类器进行实验验证，结果表明，所提方法能够有效检测出故障信号， 并对故障类型进行准确辨识，具有一定的工程应用价值。

In response to the low reliability of proton exchange membrane fuel cell (PEMFC) system, a fault diagnosis method of least square support vector machine (LSSVM) based on variational mode decomposition (VMD) and butterfly optimization algorithm (BOA) is presented. Firstly, the time series signal of the system operation is obtained by using the sensors, and decomposed by VMD technique. Then, the fuzzy entropy of each mode component is calculated, and the data set is constructed by using the fuzzy entropy as the fault detection eigenvalue. Finally, the LSSVM fault classifier which is optimized by BOA is constructed by using the data set. The test set is used to validate the obtained fault classifier. The results show that the proposed method can detect the fault signal effectively and identify the fault types accurately, and has certain engineering application value.