Based On Data Fusion Technology To Analyze Multi-signals In A Gas-solid Fluidized Bed

Pressure fluctuation time series in gas-solid fluidized beds contain many dynamic information,such as particle characteristics,operating conditions,bubble behavior. In the paper,nonlinear analysis is adopted and implemented so as to understand the fluidization from a new viewpoint in depth.Based on a large amount of experimental data,nonlinear analysis,such as complexity theory,and Approximate Entropy (ApEn) which is adopted systematically in the fluidization research the first time,are implemented in the time series analysis of pressure fluctuation and particle concentration in gas-solid fluidized beds,and the nonlinear characteristics of fluidized beds are discussed comprehensively.After the discussion of the algorithm of complexity parameters,Lyapunov exponent and ApEn,the pressure fluctuation in fluidized beds is analyzed quantitatively. The effect of different operating conditions,such as variant gas superficial velocities,axial positions and static bed heights,is discussed. On the basis of mutual information and nonlinear theory,we analyzed the different time series of various channels at different regimes,The result show that this parameter can identify the different fluidized regimes and it is superior to the traditional method.Based on the mutual information,In the paper,Data fusion technology to analyze multi-signals in a gas-solid fluidized bed and select the new parameters to token nonlinear characteristics in fluidization,we analyzed the different time serials of various channels at different regimes and got the information. Then the information matrix was selected by some special rules.Finally,as a tentative engineering implementation,nonlinear characteristics parameters and data fusion technology are combined and adopted into flow regime identification the first time,Otherwise,several nonlinear characteristics parameters appearing constant over a wide range of operating conditions and this can help to apply the research results into industry applications.