基于小波包与神经网络的氢发动机异常燃烧研究  

Abnormal Combustion Research on Hydrogen-Fueled Engine based on Wavelet Packet and RBF Network

作　　者：杨振中[1] 祁儒明[1] 段俊法[1] 毕国进[1] 

机构地区：[1]华北水利水电学院机械学院,河南郑州450011

出　　处：《小型内燃机与摩托车》2011年第1期32-35,共4页Small Internal Combustion Engine and Motorcycle

基　　金：国家自然科学基金(50322262);郑州市科技创新人才专项(096SYJH25086);郑州市技术研究与开发项目(0910SGYG23261-3)

摘　　要：发动机压力信号包括很多能反映发动机异常燃烧征兆的有用信息,对压力信号进行小波包分解,提取异常燃烧特征,并用神经网络建立起从异常燃烧到特征参数的映射关系是诊断异常燃烧的有效技术手段。The engine pressure signal includes lots of useful information being able to reflect engine abnormal burning portent.Carrying out wavelet packet decomposition,extracting the characteristic of the abnormal combustion on the signal pressure,and using neural networks to build up mapping relation from abnormal combustion to characteristic parametric is the effective technology means to diagnose abnormal combustion.

关 键 词：氢发动机 异常燃烧 小波包 RBF神经网络 故障诊断 

分 类 号：TK463[动力工程及工程热物理—动力机械及工程]