摘要
为了对舰船电力推进系统设备发生的故障进行分析与诊断,利用可拓识别方法和贝叶斯判别方法,并结合Matlab仿真,提出了一种舰船综合电力推进系统设备故障损伤程度的贝叶斯-可拓诊断方法。该方法在考虑各故障总体的分布函数、先验概率及误判损失的基础上,既考虑了决定待诊断舰船电力推进系统设备故障损伤的各判别特征指标与各故障总体相应特征指标的"关联"程度,又考虑了故障总体各判别特征指标的权系数。通过对算例的分析,表明该方法集合了可拓识别方法和贝叶斯方法的优点,实用性和优越性较强,在舰船综合电力系统设备故障诊断中有较好的应用推广前景。
In order to accurately analyze and diagnose the damage degree of power equipment in ship electric propulsion system, it combined extension recognition method, bayed discriminant method and Matlab simulation, and put forward a Bayes-extension diagnosis method for damage degree of equipment in ship integrated electric propulsion system. The method was based on the fault distribution function, the prior probability and misclassification loss, and it not only consider the “correlation” degree between the distinguishing characteristic indexes and the corresponding characteristic indexes of the fault population, but also considered the weight coefficients of the distinguishing characteristic indexes of the fault population. Through the analysis of the example, it shows this method is more practical and superior, and has a good application and popularization prospect in equipment fault diagnosis of integrated power system.
作者
王丰
顾佼佼
王蓉
WANG Feng;GU Jiaojiao;WANG Rong(Naval Aviation University, Yantai 264001, China)
出处
《兵器装备工程学报》
CAS
北大核心
2019年第2期189-193,共5页
Journal of Ordnance Equipment Engineering
关键词
可拓识别
电力推进
故障
贝叶斯判别
extension recognition
power equipment
fault
Bayes discrimination
