基于进气总压畸变指数的发动机实时监控  

Aeroengine realtime monitoring based on inlet total pressure distortion

作　　者：丁未[1] 何立明[1] 伍友利[1] 

机构地区：[1]空军工程大学工程学院,陕西西安710038

出　　处：《推进技术》2008年第5期609-613,636,共6页Journal of Propulsion Technology

摘　　要：提出了一种基于进气总压畸变指数智能预测的发动机实时监控方法。在对进气总压畸变试验数据进行混沌特性分析的基础上,采用饱和嵌入维数作为支持向量机输入节点个数的选择,建立支持向量机预测模型,并将该模型应用于对发动机进气总压畸变指数的预测。结果表明,该模型可以对进气总压畸变指数随时间的变化规律做出有效预报,从而为进一步对发动机实施有效的、实时的监控提供有力的保障。A method of aeroengine real time monitoring based on inlet total pressure distortion intelligent forecasting was presented. Firstly, chaotic characteristic parameters of inlet total pressure distortion parameter series were analyzed. Then, saturated embedding dimension was selected as the number of input nodes. Application results in aeroengine compressor show that this method has better precision, which proves that the method is feasible and effective for keeping aeroengine from entering into unstable status. This method is contributive to aeroengine effective and real-time monitoring.

关 键 词：航空发动机 进气总压畸变 支持向量机 

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]