The Use of Artificial Neural Networks in Predicting Vertical Displacements of Structures
George Pantazis, Eleni-Georgia Alevizakou

Abstract
Geodesy can make a significant contribution to the monitoring of structures. The geodetic methods that have been developed can give reliable results. The aim of this article is to use the results obtained through the monitoring of a structure to predict its position in the future, using ANNs. It presents a detailed study on the development of an ANN that can be used to predict vertical displacements in a cultural heritage monument, with the ultimate aim of preventing it from falling apart. To this end, a geodetic network of 15 control points was established. The results of twelve series of geodetic measurements and adjustments to this network are used in this study. Using the trained ANN, the vertical displacement (ΔΗ) of any specific point in the monument’s geodetic network can be predicted for a certain time in the future, with an uncertainty of ±0.5mm.

Full Text: PDF