Instability Analysis Of Dense Gas Solid Separation Fluidization

Nowadays,the dense gas solid fluidized dry coal separation technology has received extensive attention comprehensively due to the advantages of high separation precision,high capacity and no consumption of water,etc.However,the existing instability and complexity mechanisms in the dense gas solid fluidization have restricted the further development of the industrialization and scaling up of the fluidized bed.Hence,this paper studies on the instability and complexity mechanism in the dense gas solid fluidization process.A variety of analysis methods,namely,experimental measuring analysis,statistical theory analysis,entropy derivation and instability index derivation methods,chaotic analysis methods and quantitative evaluation of the instability,have been implemented in the proposed research work to investigate and achieve insights into the instability of dense gas solid separation fluidization comprehensively.The novel real time pressure drop measuring and analyzing system was established to measure and analyze the pressure drop fluctuation of dense gas solid separation fluidization.The static packing of magnetite powder is analyzed in order to study the influence of packing pattern on the crucial fluidization status.Meanwhile,the influence of maximum static friction(friction between bed particle and wall,particle and particle)on the fluidization characteristics of magnetite powder is investigated.Results demonstrate that the static friction has significant impact on the fluidization of bed particles.With the tracking ball,the irregular eruption of bubbles is investigated.The paper obtains the velocity probability distribution function for the single ball by fitting velocity data acquired by high speed camera measurement.The ball velocity distribution function reveals the heterogeneous eruption mechanisms for bubbles.The particle status in the fluidized bed is classified into three categories: particle solid,particle liquid and particle gas,which are attempted to be described by the statistical ensemble theoretically.Particles in the dense gas solid separation fluidization process are classified as particle liquid status.The instability evaluation index was derived based on the voidage distribution in the dense gas solid separation fluidization and presented in the proposed work.Furthermore,the instability index was extended to evaluate the separation effect of separating coal with dense gas solid fluidized bed and was confirmed as the efficacious evaluation method.The chaotic information and information loss rate in the dense gas solid separation fluidization are quantified by analyzing the spatial pressure drop signals.Combined with the concept of thermal entropy and information entropy,the chaotic characteristics of dense gas solid separation fluidization is investigated.The variation of entropy in terms of bed height and air flow is explored.With the summary of literature work,the Kolmogorov entropy as a function of bubble eruption and thermal entropy as a function of gas velocity are introduced in the paper,which are proposed to be a reference of describing the information loss and energy dissipation for dense gas solid separation fluidization.The pressure drop measuring and analysis system is established and implemented to characterize the instability structure in the dense gas solid fluidized bed.The phase space attractors are reconstructed based on pressure drop signals measured under different air flows and bed heights.Furthermore,the maximum Lyapunov exponents and entropy values are derived from attractors.It is conformed that dense gas solid separation fluidization exhibits chaotic characteristics,which are quantified by entropy,approximate entropy and sample entropy.Ultimately,the chaotic and instability characteristics of dense gas solid separation fluidization are quantified,which offers insight into further control of the stability of dense gas solid separation fluidization.With the pilot dense gas solid fluidized separator,the feed coal(-50+10mm)was separated efficiently.Two parallel separation experiments were conducted.The distribution rates for two separation experiments were calculated to evaluate the separation efficiency with pilot dense gas soild fluidized separator.The separation results show that E values are 0.055g/cm~3 and 0.060g/cm~3,respectively,which indicates efficient separation results were achieved.The implementation of machine learning,especially Long Short-Term Memory(LSTM)Recurrent Neural Networks in the prediction of the instability in dense gas solid separation fluidization has been explored and investigated.Based on the measured pressure drop signals from both pilot and laboratory dense gas solid fluidization,the LSTM prediction model was achieved.The prediction results reveal that LSTM is an efficient and robust model for the prediction of the instability for dense gas solid separation fluidization.