The algorithms of the organized search for identification of parameters of models of nonlinear dynamic system

作者:

Highlights:

摘要

The numerical methods for an estimation of parameters based on use of function fmins of the software package MATLAB 5.2 and on use of Bayes interpretation are stated. The parametrical mathematical model of dynamic system represented by system of the ordinary nonlinear differential equations depends on parameters. For dynamic system in discrete time the functions from the current state vectors are observed. The stated methods do not require calculation of gradients and are applicable, if into representation of mathematical model enter not differentiated function.

论文关键词:Identification,Estimation,Search,Local minimum,Global minimum,Matrix of correlations,Bayes formalism

论文评审过程:Available online 30 April 2004.

论文官网地址:https://doi.org/10.1016/j.amc.2004.02.012