基于混沌理论与Elman神经网络的气固流化床流型识别  

Flow Pattern Identification of Gas-solid Fluidized Bed Based on Chaos Theory and Elman Neural Network

作　　者：周云龙[1] 何强勇[2] 

机构地区：[1]东北电力大学能源与机械工程学院,吉林吉林132012 [2]东北电力大学自动化工程学院,吉林吉林132012

出　　处：《化工自动化及仪表》2009年第5期50-55,共6页Control and Instruments in Chemical Industry

摘　　要：采用混沌理论对气固流化床压力脉动信号进行混沌特性分析,包括求延迟时间τ、Hurst指数、关联维数、Kolmogorov熵、Lyapunov等混沌参数,并结合统计参数作为流型辨别的特征输入量,利用Elman神经网络对其训练。试验结果表明,气固流化床压力脉动信号具有混沌特性,Elman神经网络能够有效地快速地识别流化床的五种流型,识别率达95%,为在线识别气固流化床流型提供了一种新的有效方法。Chaos theory was applied for analyzing characteristics of chaotic on the pressure fluctuation signals of the gas-solid fluidized bed,including chaotic parameters such as the delay time τ,Hurst index,correlation dimension,Kolmogorov entropy,Lyapunov index,etc,and the statistical parameters were combined as input characteristics of distinguishing the flow pattern.Elman neural network was applied for their training.The results show that the pressure fluctuation signals of the gas-solid fluidized bed have chaotic characteristics,Elman neural network can identify 5 kinds flow of the fluidized bed effectively and rapidly,and the identification rate reach 97%.It provides a new effective way to identify gas-solid flow regime of the fluidized bed online.

关 键 词：混沌理论 ELMAN神经网络 气固流化床 统计参数 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]