基于Hilbert-Huang变换和多元状态估计技术的内燃机气缸压力识别

提出了一种新的内燃机气缸压力识别方法。为了实现对内燃机运行状态的实时监测与故障诊断,通过缸盖振动信号反演内燃机气缸压力曲线是一种行之有效的重要手段;本文在对缸盖振动信号特性进行分析的基础上,应用Hilbert-Huang变换对缸盖振动信号进行模态分解,提取最能反映内燃机气缸压力变化特性的振动本征模态分量;使用多元状态估计技术(MSET)建立该振动本征模态分量的Hilbert谱与气缸压力信号之间的非线性非参数回归模型;利用所建立的模型从缸盖振动信号中重构出内燃机气缸压力曲线。试验结果表明,基于Hilbert-Huang变换和MSET的气缸压力识别方法简单有效,可满足对内燃机的实时监测需求。

On Cylinder Pressure Recognition of Internal Combustion Engines by Hilbert-Huang Transform and Multivariate State Estimation Technique

A novel cylinder pressure recognition method for internal combustion engines is proposed.In order to realize real-time condition monitoring and fault diagnosis for internal combustion engines,it is a very important to retrieve cylinder pressure waveform from the vibration signal measured on the cylinder head.On the basis of characteristic analysis for the vibration signal measured on the cylinder head,the vibration signal measured on the cylinder head is decomposed by applying Hilbert-Huang transform,and the intrinsic mode function(IMF) that characterizes the response of cylinder pressure variation on the cylinder head is extracted.Then,a nonlinear and nonparametric regression model between the Hilbert spectrum of the vibration IMF and the cylinder pressure signal is built by using multivariate state estimation technique(MSET).Finally,cylinder pressure waveform is reconstructed from the vibration signal according to the developed model.Experiment results demonstrate that the proposed method is simple and effective.It can meet the requirement of real-time surveillance for internal combustion engines.