分形维数和近似熵用于度量信号复杂性的比较研究

RESEARCH ON COMPARISON BETWEEN APPROXIMATE ENTROPY AND FRACTAL DIMENSION FOR COMPLEXITY MEASURE OF SIGNALS

本文介绍了两种度量时间序列复杂性的方法 ,一种是建立在分形理论基础上的分形维数 ,一种是最近新发展起来的度量序列复杂性的统计方法———近似熵。应用这两种方法对机械设备振动信号进行了复杂性特征的提取 ,对比研究表明 ,它们都可以作为表征信号复杂性的一种度量 ,各具特点。但近似熵具有一定的抗噪、抗野点的能力 ,包含了更多的时间模式的信息 ,是一种值得重视的、很有应用前景的故障诊断新方法。

Two algorithms of complexity measure for time series are introduced in the paper.One is the fractal dimension based on the fractal theory and the other is a recently developed statistic method quantifying regularity and complexity—Approximate Entrop(ApEn).They both have been applied to the complexity feature extraction from the vibration signals of mechanical equipment,the result shows that they both can be used to measure the complexity of the signals.ApEn has high antiinterference ability,contains more information of the time pattern,and is a newly remarkable method for the fault diagnosis of mechanical equipment.