利用局部关联维数分析机械系统故障信号

The Fault Signal Analysis of Mechanical System Based on Local Correlation Dimension

对机械系统运行中测量的振动信号进行处理并识别系统的状态。针对非线性系统的特点,用分形维数定量刻画机械系统运行状态。使用分形维数中能直接反映转子系统运行特点的关联维数来获得转子运行故障以及故障发展程度的定量描述。定义并使用了局部关联维数算法,克服了使用单一维数对系统描述的局限性,同时提高了描述转子系统故障的精度,并得到系统状态预测的可靠保证。

The system feature is identified based on the processing of signal that is got from the running mechanical system. According to the nonlinear system, the fractal dimension is adapted to quantificationally descript the system. The correlation dimension of the fractal dimension is the good one to directly reflect the rotator system, which could give the information of the fault and the extent of the fault in quantity. The method of local correlation dimension computation in the paper is high efficient in compute time and the description of the fault extent of the rotator system. The method is used to the status predicting of system.