基于广义多重分形维数算法的汽车变速箱故障诊断研究

提出了一种基于扩展广义多重分形维数算法的汽车变速箱故障诊断方法。该算法是基于传统的G-P关联维数算法扩展而形成的,通过该算法对变速箱上采集的不同工作状态下的振动信号进行处理,提取变速箱齿轮的振动信号的分数维,观察及分析分形维数与变速箱齿轮的磨损规律的关系,发现其反映变速箱齿轮的真实运行状态,故可以此作为齿轮磨损预测和诊断的有效依据。

Study on Multiple Fractal Fault Diagnosis of Gear Wear

Proposed auto gearbox fault diagnosis method based on generalized multi-fractal dimension algorithm.Using an exten-sion from the traditional G-P correlation dimension algorithm from generalized multi-fractal dimension algorithm can be com-pleted simply and reliably calculate the fractal dimension of the generalized multiple vibration signal.The algorithm of vibra-tion signals through different working conditions under the gearbox col ected are processed to extract the fractal dimension of the gearbox vibration signal,the fractal dimension is found to reflect the laws of gear wear can reflect the true state of the transmission operation,and gear wear can be predicted and diagnosis.