【作者】 朱醒； 钟再敏； 孙泽昌；

【Author】 ZHU Xing, ZHONG Zai-min, SUN Ze-chang Automotive College, Tongji University

【机构】 同济大学汽车学院；

【摘要】 能够高效、快速、准确地检测出突变故障是在线故障诊断的难点。本文以燃料电池汽车(FCV)的辅助动力电池为诊断对象,首先对其建立动态解析模型并实现状态反馈控制．然后提出一种改进的参数估计自适应算法:遗忘因子自调整法．最后通过故障诊断仿真实验与常用的基于模型故障诊断法进行比较和分析。结果表明:基于遗忘因子自调整的故障诊断法可以有效检测和识别燃料电池汽车中的突变故障．为实车在线故障诊断奠定了基础。更多还原

【Abstract】 Efficiency, speediness and accurateness of abrupt fault detection are quite crucial for online fault diagnosis. Firstly the dynamical model of Fuel-Cell Vehicle power batteries is built with state-feedback control. Secondly an ameliorative adaptive algorithm named self-adjustable exponential forgetting factor recursive least square (SAEFRLS) algorithm is suggested. Finally the application of FCV power batteries fault diagnosis indicate SAEFRLS’s advantage of abrupt faults detection and isolation.更多还原