Research Papers

Parameter Estimation of an Electrohydraulic Servo System Using a Markov Chain Monte Carlo Method

[+] Author and Article Information
Junhong Liu

e-mail: junhong.liu@lut.fi

Huapeng Wu

e-mail: Huapeng.wu@lut.fi

Heikki Handroos

e-mail: heikki.handroos @lut.fi

Heikki Haario

e-mail: heikki.haario @lut.fi
Faculty of Technology,
Lappeenranta University of Technology,
P.O. Box 20, FIN-53851, Lappeenranta, Finland

1Corresponding author.

Contributed by the Dynamic Systems Division of ASME for publication in the JOURNALOF DYNAMIC SYSTEMS, MEASUREMENT, AND CONTROL. Manuscript received September 8, 2011; final manuscript received May 9, 2012; published online October 30, 2012. Assoc. Editor: Won-jong Kim.

J. Dyn. Sys., Meas., Control 135(1), 011009 (Oct 30, 2012) (9 pages) Paper No: DS-11-1283; doi: 10.1115/1.4007054 History: Received September 08, 2011; Accepted May 09, 2012; Revised May 09, 2012

A parameter estimation method is presented by an example of an electrohydraulic position servo. The method is based on the Markov chain Monte Carlo approach. The method allows utilization of noisy measurement data in identification process, making use of original physical data possible without the requirement of a filter. The method seeks for the best fitting point estimate of the unknown model parameter vector, but the solution to the parameter estimation problem is given as a statistical distribution that contains “all” the possible parameter combinations. The robustness of the model developed with the proposed method is further demonstrated by verification in operating conditions that are independent of each other and the one used in the identification step. Results show that the system model with the hybrid leakage formula for the studied valve describes the system dynamics more precisely and matches the real responses better.

© 2013 by ASME
Your Session has timed out. Please sign back in to continue.



Grahic Jump Location
Fig. 1

Schematic diagram of servo hydraulic system

Grahic Jump Location
Fig. 5

Typical leakage curve [13]

Grahic Jump Location
Fig. 3

Block scheme of pressures and flows

Grahic Jump Location
Fig. 6

Fitting results: (a) The one-dimensional (1D) parameter chain of ωn of model 1; (b) the 2D marginal distribution for parameters of model 2: us1t and uss1 (legend: “ · ”—the MCMC chain points; the contour lines—50% and 95% regions of distributions constructed using a statistical kernel density function method [22]; distribution along axis—the 1D marginal density)

Grahic Jump Location
Fig. 7

Plots of fitting and uncertainties. (a) Valve input; (b) magnified curve section with figure legends for (c)–(h).

Grahic Jump Location
Fig. 8

Valve input in verification

Grahic Jump Location
Fig. 9

Verification of the fitted models with one physical experiment (legends: — — simulation; −  ·  −  · : real)




Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In