基于动态贝叶斯网络的动车组牵引传动系统可靠性分析

Reliability analysis of EMUs traction drive system based on Dynamic Bayesian Network

针对传统可靠性分析方法对动车组牵引传动系统可靠性分析时存在的局限性,采用动态贝叶斯网络(DBN,Dynamic Bayesian Network)对其进行可靠性分析。建立动车组牵引传动系统的动态故障树,按照DBN转换规则,将动态故障树映射为DBN;综合考虑动车组牵引传动系统的动态特性和可维修性,利用DBN的正向推理得到系统可靠度和可用度随服役时间动态变化的规律,利用DBN的反向推理识别系统薄弱环节。对实例进行分析,结果表明:DBN能够全面刻画动车组牵引传动系统的动态特性和可维修性,有效地识别系统薄弱环节,可为运行风险评估和可靠性评估提供参考依据。

To address the limitations of the traditional reliability analysis methods in analyzing the reliability of EMUs traction drive system,this paper performed reliability analysis based on Dynamic Bayesian Network(DBN).The paper established the dynamic fault tree of EMUs traction drive system,and mapped it into a DBN according to the DBN transformation rules,considered the dynamic characteristics and maintainability of EMUs traction drive system,used the forward inference of the DBN to obtain the dynamic change law of system reliability and availability with service time,and identified the weak links of the system by reverse inference.The analysis results showed that the DBN can comprehensively portray the dynamic characteristics and maintainability of the traction transmission system,effectively identify the system weaknesses,and provide a reference basis for operational risk assessment and reliability evaluation.