Acta Scientiarum Polonorum

Scientific paper founded in 2001 year by Polish agricultural universities

| Informations | Reviewers | Advisory Council | Scientific Councils | Publisher’s addresses | Papers | Editorial requirements | Exemplary paper | Publication conditions | Reviewing procedure | Subscription | Abstracts | Search | Statistics |
Administratio Locorum
(Gospodarka Przestrzenna) 6 (4) 2007     ISSN: 1644-0741
Title
CONTINUOUS MODEL OF THE NEURAL NET OF THE HOPFIELD TYPE APPLIED TO ESTIMATION THE STABILITY OF POINTS OF VERTICAL GEODETIC CONTROL NETWORK
Autor
Józef Gil, Maria Mrówczyńska, Sławomir Gibowski
Keywords
a system dynamics, attractors, Lapunow’s exponents
Abstract
The idea of selection of points of a geodetic vertical measuring- controlling net was presented in the paper. These points remained stable in the course of the mea-surements. The procedure of points identification within their mutual stability consists in defining the energizing levels of Hopefield’s neural network having a minimum value. The network reaches an attractor as a final state of a dynamical system at the energizing minimum. Defining of time courses of approaching the individual height difference changes to an appropriate attractor and the research on behaviour of neighbouring trajectories of height difference changes on the basis of Lapunow’s determined exponents enables the points identification within their identified stability.
Pages
39-50
Cite
Gil, J., Mrówczyńska, M., Gibowski, S. (2007). CONTINUOUS MODEL OF THE NEURAL NET OF THE HOPFIELD TYPE APPLIED TO ESTIMATION THE STABILITY OF POINTS OF VERTICAL GEODETIC CONTROL NETWORK. Acta Sci. Pol. Geod. Descr. Terr., 6(4), 39-50.
Full text