故障诊断时间在目前地铁设备故障修复总时长中占比较高，如何有效缩短故障诊断时间成为提高地铁设备故障维修效率的关键。鉴于此，为快速分析地铁设备系统中的薄弱环节，以故障诊断时间为判定标准，提出基于贝叶斯网络的地铁设备故障诊断算法，将贝叶斯网络计算的故障概率与该种故障排查时间相结合作为预期故障诊断时间指标，并以该指标值从低到高依序进行故障排查诊断。然后基于调研获取的全国17家地铁公司列车客室门故障数据，利用蒙特卡洛仿真和3D数字孪生对比人工排查和本算法在故障诊断方面的效率差异。算例结果显示，利用本算法定位故障点时所消耗的时间是人工排查时长的43%~48%，表明基于故障诊断时间的贝叶斯网络地铁设备故障诊断算法能快速分析出系统中的薄弱环节，提高地铁设备故障的维修效率。

The time of fault diagnosis accounts for a large proportion in the process of fault repair of subway equipment at present. How to shorten the time of fault diagnosis effectively becomes the key point to improve the efficiency of fault repair of subway equipment. Given this, in order to quickly analyze the weak links in subway system, taking the fault diagnosis time as the judgment standard, a fault diagnosis algorithm of subway equipment based on Bayesian network was proposed. The fault probability calculated by the Bayesian network was combined with the troubleshooting time of the fault as the expected fault diagnosis time index, and the fault diagnosis was performed based on the fault diagnosis time index value from low to high. Based on the fault data of passenger doors of 17 metro companies in China, the efficiency difference between manual troubleshooting and this algorithm in fault diagnosis was compared by using Monte Carlo simulation and 3D digital twinning. The results show that the time consumed by the Bayesian network model fault diagnosis algorithm in locating the fault point is 43%~48% of time consumed by the manual troubleshooting. It indicates that the Bayesian network fault diagnosis algorithm based on fault diagnosis time can quickly analyze the weak links in the subway system and improve the maintenance efficiency.