基于缸盖振动信号的活塞环腔压力径向基神经网络识别研究  

RBF Neural Network Identification for Inner-Rings Gas Pressure Based on Cylinder Head Vibration Signal

作　　者：孟凡明[1] 张优云[2] 

机构地区：[1]清华大学摩擦学国家重点实验室,北京100084 [2]西安交通大学润滑理论及轴承研究所

出　　处：《内燃机工程》2005年第6期44-47,共4页Chinese Internal Combustion Engine Engineering

基　　金：国家自然科学基金资助(50375115)

摘　　要：介绍了利用气缸盖振动信号,借助径向基神经网络(RBFNN),进行活塞环腔气体压力识别的方法。以1100柴油机为试验对象,测得其缸盖振动位移和气缸内气体燃烧压力,将缸盖振动信号作为识别的输入信号,利用径向基神经网络和ARMA时间序列分析法对气缸燃烧压力和环腔内气体压力进行了识别。结果表明:利用径向基网络和ARMA时间序列分析法,均能较为准确地识别活塞环环腔气体压力和气缸内气体燃烧压力;径向基神经网络的识别方法比ARMA时间序列识别方法更加准确。A new method identifying gas pressure in inner rings in an ICE was presented. An experiment was carried out on a 1100-typediesed engine, and the vibration signal from the cylinder head and combustion gas pressure in the cylinder were obtained. By the use of radial basic function neural network （RBFNN）, the gas pressure in inner rings was identified. The validity of the new method, compared with that by the ARMA time series analysis method, was demonstrated. The results show that the proposed method is valid in identifying gas pressure in inner rings, and will become a useful one in this way.

关 键 词：内燃机 活塞环 气体压力 识别 缸盖振动 径向基网络 ARMA时间序列 

分 类 号：TK421[动力工程及工程热物理—动力机械及工程]