The empirical mode decomposition-recurrence analysis (EMD-RA)method was used to analyze the pressure fluctuation signal of biomass and silica sand particle flow.First,differential pressure fluctuation signal of two-component particle flow in the fluidized bed was collected.The IMFs (intrinsic mode functions)in the original signal were extracted with EMD method.The com-ponents of IMFs representing intermediate and high frequency areas of bubble and particle behaviors respectively,were used to recursive analysis.The results show that when the gas velocity is less than 0.50 m/s,the stage is changed from a fixed bed to a bubbling bed.The system behavior is affected by particle behavior with strong randomness,the variation regularity of the recursive characteristic is not obvious.When the gas velocity is 0.86 m/s,the bed is changed as a fluidized bed.The system is characterized by the strong periodicity of the bubble motion,and the recursive characteristic is ob-viously changed.The non-linear flow characteristics of two-component granular flow in the fluidized bed are further confirmed.EMD-RA method can be used as a new method to identify flow behaviors of the fluidized bed.%采用EMD-RA方法对生物质与石英砂颗粒混合流动的压差脉动信号进行分析．首先对稻壳与石英砂颗粒在流化床内混合流动的压差脉动信号进行数据采集，采用EMD方法提取原始信号中本身固有的内禀模态函数IMF，再分别对代表气泡、颗粒行为的中频区和高频区的IMF分量进行递归分析．结果表明：当气速u＜0．50 m／s时，床内由固定阶段向鼓泡阶段转变，系统受颗粒行为影响且随机性较强，各递归特征量变化规律不明显；当气速u＜0．86 m／s时，床内由鼓泡阶段向流化阶段转变，系统受气泡运动影响且呈现较强的周期性，各递归特征量发生明显变化，递归特性增强．研究进一步证实了流化床内双组分颗粒流动的非线性特性，所提出的EMD-RA方法可作为流化床流型识别的新方法．
展开▼
