Stochastic approximation of constrained systems with system and constraint noise
โ Scribed by H.J Kushner; E Sanvicente
- Publisher
- Elsevier Science
- Year
- 1975
- Tongue
- English
- Weight
- 520 KB
- Volume
- 11
- Category
- Article
- ISSN
- 0005-1098
No coin nor oath required. For personal study only.
โฆ Synopsis
Sequential Monte Carlo methods are frequently needed to optzmzze stochastic systems when the performance functwn is observed w~th additive nmse Analysis of convergent algorithms is especially zmportant when there are system parameter constraints.
Sumnmry--The paper considers the problem of mmlmmng f(x) over the constraint set C = {x q~(x)~<0, i m 1, , s}, but the f( ) and q'(.) are unknown and only nokqeperturbed samples of both the f( ) and the q'(.) are available, at selected parameter settings Under certain convexlty condiuons, a stochastic approximation algorithm is set up, and convergence wroth probabdity one to the optimum parameter value is proved Numerous practmcal examples fit the problem descnptmn *
๐ SIMILAR VOLUMES
A method to eliminate the fast modes in a singularly perturbed system driven by white noise is presented together with error analysis. An approximate model is derived by replacing the transfer functions of the fast elements with their truncated Maclaurin expansions. It is shown that if the slow elem
## Abstract The logarithmic matrix norm with respect to __l__~2~ matrix norm will be used to investigate mean square stability for a class of secondโorder weak schemes when applied to 2โdimensional linear stochastic differential systems with one multiplicative noise. (ยฉ 2004 WILEYโVCH Verlag GmbH &
This paper gives an extension of the Volterra-functional method to the analysis of nonlinear systems with both stochastic iqut and stochastic parameters. In Section II, $rst-order linear systems with stochastic input and stochastic parameter are analyzed. Certain desnitions and symbolic relations ar
## Abstract This paper is concerned with the reliable control problem against actuator failures for a class of uncertain discreteโtime stochastic nonlinear timeโdelay systems. The failures of actuators are quantified by a variable varying in a given interval. The stochastic nonlinearities described