基于云关联度的航空发动机传感器、部件故障识别系统设计  

Fault diagnosis system design for the sensors and components of aircraft engine based on cloud relational analysis

作　　者：仇小杰[1] 黄金泉[1] 鲁峰[1] 刘楠[1] 

机构地区：[1]南京航空航天大学能源与动力学院,南京210016

出　　处：《航空动力学报》2011年第11期2584-2592,共9页Journal of Aerospace Power

基　　金：江苏省2010年度普通高校研究生科研创新计划(CX10B_108Z);南京航空航天大学科研业务基金(56v1042)

摘　　要：针对灰色关联度方法用于发动机故障诊断精度低的问题,结合云理论和关联度分析方法,提出了云关联度方法,利用"云滴"能够反映映射的模糊性和随机性与整体"形状"变化分布这一特性,克服数据挖掘的模糊性和随机性问题,采用综合隶属度方法能够充分对数据进行挖掘,最终计算出云关联度.为了实现航空发动机传感器、部件单一故障的实时识别和诊断,在分析航空发动机传感器故障和部件故障特点的基础上,利用云关联度方法,设计了航空发动机传感器、部件故障的识别系统.仿真结果表明该方法不仅能够实时正确区分航空发动机传感器、部件故障,还能准确诊断出故障发生的传感器或者部件的位置,有效地改善了航空发动机故障诊断能力.Considering low accuracy problem of fault diagnosis for aircraft engine by gray relational analysis,the method of cloud relational analysis was proposed by integrating the cloud theory and the relational analysis.The characteristics of ＂cloud droplets＂ reflect fuzziness and randomicity of the mapping,and solve the problem in data mining while representing the distribution change of the integral ＂shape＂.The comprehensive subjection method could mine information fully,then the cloud relation can be calculated.In order to diagnose the single fault for the sensors and components of aircraft engine,a fault diagnosis system based on cloud relational analysis was designed based on analysis of the fault characteristics of the sensors and the components.The simulation results show that the proposed method can not only distinguish the faults of sensors and components,but also locate the fault position of the sensors or the components.The methods can effectively improve the fault diagnosis ability for aircraft engine.

关 键 词：航空发动机 传感器故障 部件故障 故障诊断 云理论 关联度 

分 类 号：V233.7[航空宇航科学与技术—航空宇航推进理论与工程]