摘要
为实时掌握发动机及其关键零部件的健康状态和剩余寿命,文章围绕航空发动机寿命监测系统硬件与软件进行了设计分析,并针对传统的极限学习机处理大量数据或者繁杂的学习任务时导致泛化能力、精度下降的问题,提出了一种基于经验模态分解和差分进化算法改进的极限学习机模型。以发动机关键部件滚动轴承为例,提出了基于经验模态分解和差分进化算法-极限学习机的健康状态评估模型,与传统的极限学习机健康状态评估模型相比,该优化的模型分类精度明显提高。
In order to grasp the health status and remaining life of the engine and its key components in real time,maximize its service potential and avoid major accidents,the aero-engine life monitoring system is designed.Aiming at the problem that the generalization ability and accuracy of traditional extreme learning machines are reduced when processing large amounts of data or complicated learning tasks,this paper proposes an improved limit learning machine model based on empirical modal decomposition and differential evolution algorithms.Taking the rolling bearing of key engine components as an example,a health state evaluation model based on EMD and DE-ELM is proposed,and the classification accuracy of the optimized model is significantly improved compared with the traditional health state assessment model of the limit learning machine.Through the simulation results,failure mode analysis is carried out for different parts of the engine,and the fatigue life of the parts is determined to provide decision-making basis for the realization of predictive maintenance and accurate assurance,so as to truly prevent problems before they occur.
作者
李思雨
程中华
刘子昌
韩凯
闫云斌
黄少罗
LI Siyu;CHENG Zhonghua;LIU Zichang;HAN Kai;YAN Yunbin;HUANG Shaoluo(Shijiazhuang Campus,Army Engineering University,Shijiazhuang 050003,China)
基金
军内科研课题项目(212LJ44004)。
关键词
航空发动机
寿命监测系统
健康状态评估
仿真
areo-engine
life monitoring system
health status assessment
simulation
