基于非线性动态模型的质子交换膜燃料电池故障诊断  

Fault diagnosis of proton exchange membrane fuel cell based on nonlinear dynamic model

作　　者：雍加望 赵倩倩[2] 冯能莲[2] YONG Jiawang;ZHAO Qianqian;FENG Nenglian(Beijing Key Laboratory of Traffic Engineering,Beijing University of Technology,Beijing 100124,China;Faculty of Environment and Life,Beijing University of Technology,Beijing 100124,China)

机构地区：[1]北京工业大学北京市交通工程重点实验室,北京100124 [2]北京工业大学环境与生命学部,北京100124

出　　处：《化工学报》2022年第9期3983-3993,共11页CIESC Journal

基　　金：北京市自然科学基金面上项目(3222003);国家自然科学基金青年基金项目(52002009);汽车安全与节能国家重点实验室开放基金项目(KF2010)。

摘　　要：为了对质子交换膜燃料电池(proton exchange membrane fuel cell,PEMFC)系统进行故障诊断以提高系统的安全性和可靠性,针对PEMFC系统的强非线性,在九阶状态空间模型的基础上提出一种滑模观测器实时生成残差,利用故障阈值检测法建立故障特征矩阵检测故障,进而为了隔离故障,引入相对故障敏感度函数建立理论相对故障敏感度矩阵,在系统运行时实时计算各故障相对故障敏感度与理论相对故障敏感度的欧氏距离,最小欧氏距离对应的故障则为系统发生的故障,结果验证了所提出的基于模型的故障诊断方法的有效性,且所构建观测器可以估计PEMFC系统中难以直接测取的状态变量,平均相对误差在6%以内。In order to carry out fault diagnosis for proton exchange membrane fuel cell(PEMFC)system to improve the safety and reliability of the system,aiming at the strong nonlinearity of the PEMFC system,a sliding mode observer is proposed based on the ninth-order state space model to generate residuals in real time,and the fault threshold detection method is used to establish a fault feature matrix to detect faults.In order to isolate faults,a relative fault sensitivity function is introduced to establish a theoretical relative fault sensitivity matrix,and the relative faults of each fault are calculated in real time when the system is running.The Euclidean distance between the sensitivity and the theoretical relative fault sensitivity,the fault corresponding to the minimum Euclidean distance is the fault that occurs in the system.The results verify the effectiveness of the proposed model-based fault diagnosis method,and the constructed observer can estimate the state variables that are difficult to measure directly in the PEMFC system,the average relative error is within 6%.

关 键 词：质子交换膜燃料电池 滑模观测器 故障检测 故障隔离 

分 类 号：TQ028.8[化学工程]