A Quasi-Newton Method for Estimating the Parameter in a Nonlinear Hyperbolic System
โ Scribed by Wenhuan Yu
- Publisher
- Elsevier Science
- Year
- 1999
- Tongue
- English
- Weight
- 179 KB
- Volume
- 231
- Category
- Article
- ISSN
- 0022-247X
No coin nor oath required. For personal study only.
โฆ Synopsis
A quasi-Newton method QNM in infinite-dimensional spaces for estimating parameters involved in distributed parameter systems is presented in this paper. Next, the global convergence of a sequence generated by the algorithm QNM is also proved. We apply the algorithm QNM to an identification problem of a distributed parameter system governed by a nonlinear hyperbolic partial differential equation.
๐ SIMILAR VOLUMES
An 'average' approximation for calculating the zero-field splitting parameter, D, gives reasonable results when used to calculate D values for non-disjoint delocalized organic biradicals. When used to calculate disjoint localized organic biradicals the D values are approximately half the experimenta
## Communicated by J. Banasiak We propose a new quasi-linearization technique for solving systems of nonlinear equations. The method finds recursive formulae for higher order deformation equations which are then solved using the Chebyshev spectral collocation method. The implementation of the meth
## Abstract A new objective function for estimating parameters in differential equations, based upon a weighted least squares criterion for the residuals of these equations, is presented. The use of Lobatto quadrature in combination with the collocation technique reduces the original problem to one