基于贝叶斯网络的电控发动机可靠性分配

Reliability Distribution of Electric Control Engine Based on Bayesian Network

随着汽车发动机各系统部件的增多,电控程度越来越复杂,各部件之间的相互影响逐渐增多,存在很多不确定性问题。基于以上问题,提出一种基于贝叶斯网络的可靠性分配方法,以电控发动机不能起动为例,在考虑系统结构和故障原因的基础上,结合相关数据建立贝叶斯网络模型,计算各个根节点的后验概率,利用贝叶斯网络的诊断结果对系统进行可靠性分配,分配结果符合设计要求,证明了基于诊断结果的可靠性分配方法的有效性。

With the increase of car engine system components,the increasingly complex electric control and the increasing mutual influence of components lead to many uncertainties.To address these problems,the paper puts forward a reliability distribution method based on Bayesian networks.The failure of an electrically-controlled engine is taken as an example.A Bayesian network model is established on account of the system structure,the cause of the failure and relevant data to calculate a posteriori probability of each root node,and the diagnosis results based on Bayesian networks are used to conduct reliability distribution to the system.The distribution results comply with the requirement of the design and the effectiveness of such a reliability distribution method based on diagnosis results is testified.