Estimasi Parameter Model Nonlinear Menggunakan Analisis Sensitivitas dan Pengoptimalan Berbasis Turunan
Abstract
Model estimation based on the observation data of a system’s states is an important subject in the study of dynamical systems. Maximum likelihood (ML) estimation is a stochastic estimation method which can be used to obtain an optimal set of parameter based on noisy measurements. This paper describes the method and implementation of the ML estimator to identify an optimal parameter set in a discrete-time nonlinear state space model. In particular, the optimal parameter set is defined as the value that minimizes the error between the actual and estimated model outputs of the system. This paper discusses a gradient-based optimization that is equipped with sensitivity analysis method for searching such a parameter set. Simulation results which describe an implementation of the proposed estimation method in a nonlinear system model are also discussed.
References
J.R. Raol, G. Girija, dan J. Singh, Modelling & Parameter Estimation of Dynamic Systems, IET Control Engineering Series, Vol. 65, United Kingdom: The Institution of Engineering and Technology, 2004.
J.V. Beck dan K.J. Arnold, Parameter Estimation in Engineering and Science, Wiley Series in Probability & Mathematical Statistics, New York, USA: Wiley, 1977.
S.D. Silvey, Statistical Inference, Baltimore, USA: Penguin Books, 1970.
D.M. Malakoff, “Bayes Offers ‘New’ Way to Make Sense of Numbers,” Science, Vol. 286, hal. 1460–1464, 1999.
R. Aster dan C. Thurber, Parameter Estimation & Inverse Problems, Cambridge, USA: Academic Press, Vol. 90, 2011.
J.D. Hamilton, Time Series Analysis (Vol. 2), Princeton, USA: Princeton University Press, 1994.
W.H. Press, Numerical Recipes: The Art of Scientific Computing, 3rd Ed., New York, USA: Cambridge University Press, 2007.
M.A. Tanner, Tools for Statistical Inference: Methods for Exploration of Posterior Distributions and Likelihood Functions, 2nd Ed., New York, USA: Springer-Verlag, 1993.
A. Raue, C. Kreutz, T. Maiwald, J. Bachmann, M. Schilling, U. Klingmüller, dan J. Timmer, “Structural & Practical Identifiability Analysis of Partially Observed Dynamical Models by Exploiting the Profile Likelihood,” Bioinformatics, Vol. 25, No. 15, hal. 1923–1929, 2009.
S. Hengl, C. Kreutz, J. Timmer, dan T. Maiwald, “Data-Based Identifiability Analysis of Nonlinear Dynamical Models,” Bioinformatics, Vol. 23, No. 19, hal. 2612–2618, 2007.
F. Geier, G. Fengos, F. Felizzi, dan D. Iber., “Analyzing and Constraining Signaling Networks: Parameter Estimation for the User,” Computational Modeling of Signaling Networks, Totowa, NJ, USA: Humana Press, hal. 23–39 , 2012.
K. Jaqaman dan G. Danuser, “Linking Data to Models: Data Regression,” Nature Review Molecular Cellular Biology, Vol. 7, No. 11, hal. 813–819, 2006.
B. Efron dan R. Tibshirani, “Bootstrap Methods for Standard Errors, Confidence Intervals and Other Measures of Statistical Accuracy,” Statistical Science, Vol. 1, No. 1, hal. 54–75 , 1986.
MATLAB Optimization Toolbox Version 7.2 Release R2015a, The MathWorks, Inc., 2015.
G.F. Fussmann, S.P. Ellner, K.W. Shertzer, dan N.G. Hairston, “Crossing the Hopf Bifurcation in a Live Predator-Prey System,” Science, Vol. 290, hal. 1358 – 1360, 2000.
K.W. Shertzer, S.P. Ellner, G.F. Fussmann, dan N.G. Hairston, “Predator-Prey Cycles in an Aquatic Microcosm: Testing Hypotheses of Mechanism,” Journal of Animal Ecology, Vol. 71, hal. 802–815, 2002.
© Jurnal Nasional Teknik Elektro dan Teknologi Informasi, under the terms of the Creative Commons Attribution-ShareAlike 4.0 International License.
1.png)
