Parameter estimation with constraints based on variational method

Abstract

The accuracy of parameter estimation is critical when digitally modeling a ship. A parameter estimation method with constraints was developed, based on the variational method. Performance functions and constraint equations in the variational method are constructed by analyzing input and output equations of the system. The problem of parameter estimation was transformed into a problem of least squares estimation. The parameter estimation equation was analyzed in order to get an optimized estimation of parameters based on the Lagrange multiplication operator. Simulation results showed that this method is better than the traditional least squares estimation, producing a higher precision when identifying parameters. It has very important practical value in areas of application such as system identification and parameter estimation.

Keywords

least squares estimation / parameter estimation / variational method / constraint

References

[1] Chen X., Wang S. Least-squares methods in linear model, 2003, Shanghai: Shanghai Science and Technology Press, 82-95

[2] Ding F., Yang J. Convergence analysis of stochastic gradient algorithms. Journal of Tsinghua University(Sci & Tech), 1999, 39(1): 83-86

[3] Ding F., Yang J. Remarks on martingale hyperconvergence theorem and the convergence analysis of the forgetting factor least squares algorithm. Control Theory and Application, 1999, 16(4): 569-572

[4] Hasting-James R., Sage M.W. Recursive generalized least squares procedure for on-line identification of process parameters. Proc. of the IEE, 1969, 116(12): 2057-2062

[5] Hsia T.C. System identification, least-squares methods, 1977, Utrecht, the Netherlands: VNU Science Press, 65-70

[6] Liu B. Modern control theory, 2002, Beijing: Publishing House of Electronics Industry, 65-69

[7] Lu W.P., Fisher G. Least-squares output estimation with multirate sampling. IEEE Transactions on Automatic Control, 1989, 34(6): 669-672

[8] Xu N. System identification, 1986, Beijing: Publishing House of Electronics Industry, 60-63

[9] Zhang J. Parameter estimation and improvement of linear modeling, 1999, Changsha: Press of National University of Defense Technology, 102-106

[10] Zhang X. Modern signal processing, 2002, 2nd Ed., Beijing: Tsinghua University Press, 49-59