Joint State and Parameter Estimation for Stationary ARMA Model with Unknown Noise Model

详细信息    查看官网全文

• 英文论文题名：Joint State and Parameter Estimation for Stationary ARMA Model with Unknown Noise Model
• 论文作者：Shuhui Li ; Xiaoxue Feng ; Honghua Lin ; Feng Pan
• 英文论文作者：Shuhui Li ; Xiaoxue Feng ; Honghua Lin ; Feng Pan ; School of automation ; Beijing Institute of Technology ; School of mechanical engineering ; Beijing Institute of Technology ; Kunming-BIT Industry Technology Research Institute INC
• 年：2017
• 作者机构：School of automation,Beijing Institute of Technology;School of mechanical engineering,Beijing Institute of Technology;Kunming-BIT Industry Technology Research Institute INC;
• 英文论文关键词：EM algorithm ; Gaussian mixture model ; non-Gaussian noise ; Particle filter ; Stationary ARMA model
• 会议召开时间：2017-07-26
• 会议录名称：第36届中国控制会议论文集（B）
• 英文会议录名称：Proceedings of the 36th Chinese Control Conference（B）
• 语种：英文
• 分类号：O212.1
• 学会代码：KZLL
• 会议名称：第36届中国控制会议
• 会议地点：中国辽宁大连
• 主办单位：中国自动化学会控制理论专业委员会
• 学会名称：中国自动化学会控制理论专业委员会
• 页数：6
• 文件大小：569k
• 原文格式：D
• 会议级别：全国

摘要

The parameter estimation of a wide-sense auto-regressive moving-average(ARMA) model,which is widely applied into a variety of fields,is an extremely important research subject.Most research is conducted with the known driving environment noise or assuming that the driving noise consists unknown variance.Actually the driving noise is really complex in reality.Until now,less attention on parameter estimation for a wide-sense stationary hidden ARMA process with unknown noise is paid attention,although it is very common in the complex control system.The paper presents parameter estimation method for hidden wide-sense ARMA processes with the known model order.A dual particle filter-based method is adopted to estimate joint states and parameters.The method can be divided into two steps.The first step utilizes the particle filter algorithm to estimate the state of an ARMA model,then conduct the estimation of parameters in the PF algorithm on the basis of state estimation in the second step.For the noise model is extremely unknown,the Gaussian mixture model is adopted to approach the posterior probability function in the process of the above dual PF algorithm according to EM algorithm.Simulation results verify the effectiveness of the proposed scheme.
The parameter estimation of a wide-sense auto-regressive moving-average(ARMA) model,which is widely applied into a variety of fields,is an extremely important research subject.Most research is conducted with the known driving environment noise or assuming that the driving noise consists unknown variance.Actually the driving noise is really complex in reality.Until now,less attention on parameter estimation for a wide-sense stationary hidden ARMA process with unknown noise is paid attention,although it is very common in the complex control system.The paper presents parameter estimation method for hidden wide-sense ARMA processes with the known model order.A dual particle filter-based method is adopted to estimate joint states and parameters.The method can be divided into two steps.The first step utilizes the particle filter algorithm to estimate the state of an ARMA model,then conduct the estimation of parameters in the PF algorithm on the basis of state estimation in the second step.For the noise model is extremely unknown,the Gaussian mixture model is adopted to approach the posterior probability function in the process of the above dual PF algorithm according to EM algorithm.Simulation results verify the effectiveness of the proposed scheme.

引文

[1]Junho Park,Geon Lee,Youngin Choi,"Determining ARMA model parameters for biomedical signal applications based on estimation of the equivalent AR model,"the 12 international conference on Ubiquitous Robots and Ambient Intelligence(URAI2015)October28-30,2015/KINTEX,Goyang city,korea.
    [2]Zhang Shufang,Liu Renjie,"A rapid algorithm for on-line and real-time ARMA modeling,"Proceedings of ICSP2000.
    [3]Adnan Al-smadi,"ARMA model parameters estimation using SVD,"the international conference on sciences of electronics,technologies of information and telecommunications 6(2012)814-816.
    [4]Jianzhou Wang,Jinzhao Liang,Jinxing Che,Donghuai Sun,"ARMA model identification using particle swarm optimization Algorithm,"International conference on computer science and information technology 2008,pp.223-227.
    [5]Juqing Hang,Random signal processing,press of beihang university,1990.
    [6]K.Steiglitz,"On the simultaneous estimation of poles and zeros in speech analysis",IEEE Trans.Acoust,speech,Signal Process,vol.ASSP-25,pp.229-234,June 1977.
    [7]Fattah,S.A.,Zhu,W.P.,Ahmad,M.O.:'Identification of autoregressive moving average systems from noise-corrupted observations'.Proc.IEEE Int.Northeast Workshop on Circuits Systems(NEWCAS),Montreal,Canada,June 2008,pp.69-72.
    [8]Fattah,S.A.,Zhu,W.P.,Ahmad,M.O.:'An algorithm for ARMA model parameter estimation from noisy observations'.Proc.IEEE Int.Symp.Circuits and Systems(ISCAS),Seattle,WA,May 2008,pp.3202-3205.
    [9]Fattah,S.A.,Zhu,W.-P.,Ahmad,M.O.:'A novel technique for the identification of an ARMA system under very low levels of SNR',IEEE Trans.Circuits Syst.1,2008,55,(7),pp.1988-2001.
    [10]Fattah,S.A.,Zhu,W.-P.,Ahmad,M.O.:'An approach to ARMA system identification at a very low signal-to-noise ratio'.Proc.IEEE Int.Conf.Acoustics,Speech,Signal Processing(ICASSP),Philadelphia,PA,March 2005,vol.4,pp.259-261.
    [11]J.Durbin S.J.Koopman Time Series Analysis by State-Space Methods ser.Oxford Statistical Science Series.Oxford U.K.:Oxford University Press 2001.
    [12]F.Gustafsson,"Particle theroy and practice with positioning applications,"IEEE Aerosp.Electron.Syst.Mag.,vol.25,no.7,pp.53-81,Mar.2010.
    [13]P.M.Djuric'and M.F.Bugallo,Particle Filtering.Hoboken,NJ,USA:Wiley,2010,ch.5,pp.271-333.
    [14]Zhiyu Zhu ect.Particle filter method and application.Beijing:Sciencep.2010.
    [15]Christophe Andrieu and Arnaud Doucet.Particle methods for change detection.System identification and Control[C].proceedings of the IEEE,2004(92):423-436.
    [16]L.Burget,P.Matejka,P.Schwarz,"Analysis of feature extraction and channel compensation in a GMM speaker recognition system,"IEEE Transactions on Audio Speech and Language Processing,vol.15,no.7,pp.1979-1986,2007.
    [17]Richard Blundell,Stephen Bond,"GMM estimation with persistent panel data:an application to production functions",Econometric Reviews,vol.19,no.3,pp.321-340,2000.
    [18]BL Pellom,J H L Hansen,"An efficient scoring algorithm for Gaussian Mixture Model based speaker identification,"IEEE Signal Processing Letter,vol.5,no.11,pp.281-284,1998.
    [19]A.P.Dempster,N.M.Laird,D.B.Rubin,"Maximum likelihood from incomplete data via the EM algorithm,"J.R.Stat.Soc,vol.39,no.1,pp.1-38,1997.
    [20]J.Liu and M.West,"Combined parameter and state estimation in simulation-based filtering,"in Sequential Monte Carlo in practice,A.Doucet,N.de Freitas and N.Gordon,Eds.New York:Springer-verlag,in2001,pp.197-223.
    [21]Gencaga D,Ertuzun A,Kuruoglu E.Modeling of non-stationary autogressive alph-stable process by particle filters.Digit.Signal Process18(3),465-478(2008).
    [22]M.Jordan.Pattern Recognition And Machine learning.
    [23]Yasuo Matsuyama,"Hidden Markov model estimation based on alpha-EM algorithm:Discrete and continuous alpha-HMMs",Neural Networks(IJCNN)The 2011 International Joint Conference on,pp.808-816,2011,ISSN 2161-4407.
    [24]I.Urteaga and P.M.Djuric,"Sequential estimation of hidden ARMA processes by particle filtering-Part I,"IEEE Trans.Signal Process.,vol.65,no.2,pp.482-493,2017.
    [25]R.Huang,H.Zheng,and E.E.Kuruoglu,"Time varying ARMA stable process estimation using sequential Monte Carlo,"Signal,Image and Video Processing,pp.1-8,2011.