气固流化床声发射信号的多尺度解析

Multi-scale resolution of acoustic emission in gas-solid fluidized bed

为了研究声波信号结构与气固流化床内流动行为的对应关系，对气固流化床的声波信号运用Daubechies二阶小波在1～9尺度下进行分解，用R／S分析分别研究各尺度下声波信号的分形特征．研究发现，气固流化床的声波信号具有多尺度特征．根据各小波尺度下波动信号的Hurst指数变化规律（均小于0．5、均大于0．5或两者皆有），建立了声波信号的尺度划分标准，认为声波信号可划分为微尺度、介尺度和宏尺度，分别反映颗粒行为、气泡和颗粒团聚物行为和平均流动行为等．与压力脉动信号多尺度结构的比较表明，声波信号主要反映微尺度的颗粒信息，而压力信号主要反映介尺度的气泡信息．

The Daubechies second order wavelet was used to decompose acoustic emission （AE） signals in gas-solid fluidized bed to 1-9 scale detail signals and scale approximation signals in order to investigate the relationship between the acoustic signal structure and the flow behavior in the fluidized bed. Multi-fractal characteristics of different scale signals were analyzed by the Hurst analysis. Results showed that the AE signals had multi-scale character. According to the Hurst analysis of the wavelet decomposed signals, the criterion of dividing the signal information to three characteristic scales was established and the AE signals were divided to micro-scale signals with all Hurst exponents less than 0. 5, meso-scale signals with all Hurst exponents more than 0.5, and macro-scale with two characteristic Hurst exponents. And the three scale signals reflected the action of particles, average dynamics and bubbles and agglomeration respectively. Compared with the pressure signals of multi-scale structure character, the AE signal energy distribution in the three scale components confirms that the AE signals mainly reflect the particle information in micro-scale, while the pressure signals mainly reflect the bubble information in meso-scale.