基于分形理论的内燃机噪声信号分析  

Study on Internal Combustion Engine Noise Based on Fractal Theory

作　　者：廖世勇[1,2] 米林[2] 石晓辉[2] 陈强[1] 

机构地区：[1]重庆通信学院军事电力工程系,重庆400035 [2]重庆工学院汽车零部件制造及检测技术重点实验室,重庆400050

出　　处：《内燃机工程》2010年第2期59-63,共5页Chinese Internal Combustion Engine Engineering

基　　金：国家自然科学基金资助项目(50706058);汽车零部件制造及检测技术教育部重点实验室开放基金项目

摘　　要：提出了基于分形的内燃机噪声信号的内燃机故障诊断方法。根据分形理论的G-P算法,阐述了内燃机噪声信号分形关联维数的计算方法,并对维数计算过程中参数的选取进行了详细的论述。通过对实际测量的内燃机噪声信号的关联维数计算分析,证明了内燃机噪声信号的分形特性。计算结果表明:内燃机噪声信号关联维数随内燃机工作状况改变有明显的变化,可作为内燃机工作状态监测和故障诊断的一个特征量。A method based on fraetal theory was proposed to study the noise signals of internal combustion engine. According to the G-P computational method,the calculation of the fractal correlation dimension of internal combustion engine noise signals was conducted, and then the selection for some important parameters in this calculation process was discussed. From the correlation dimension analysis of noise signals of internal combustion engine,it was proved that engine noise signals have an obvious fractal characteristic. The results show that the correlation dimension of engine noise signals varies with the change of engine operating conditions, and it is correlative with engine operating condition, and it can be used as a characteristic parameter in engine monitoring and fault diagnosis to some extent.

关 键 词：内燃机 噪声 分形 关联维数 故障诊断 

分 类 号：TK427[动力工程及工程热物理—动力机械及工程]