Abstract

One of the most crucial issues in engineering of structure and investigating ground deformation is deformation monitoring. The only thing which is strongly required is to create microgeodesy networks. An essential issue in microgeodesy networks is detecting unstable points of network. L1-Norm minimization and the global congruency can be noted as one of the classical methods for identifying network unstable points. In all previously conducted studies regarding this issue, results distinctly demonstrates that when displacement point vector is small, the number of points which have really displaced is more than that of true detection of displaced points using common deformation analysis ways. The probable reason for that can refer to spreading nature of the least squares estimation. Considering the results of recent studies in the detecting the network unstable points, to tackle the limitation the idea of subnetwork analysis is offered. In this case, some subnetworks including a subject point and the other source points appeared from dividing the deformation monitoring network. According to the unstable points, subnetworks will be there. This method will enable us to investigate the stable and unstable points. Having divided whole network to subnetworks, each network would be adjusted and unstable points of it would be detected. So, unstable points and their relations are cutoff and spreading effect of the least squares is fallen. This paper is on effort to evaluate the method in a simulated and a real network. The results prove that in a better and correct detection of unstable point can be successfully achieved by using subnetwork analysis compared to global congruency test all stimulates states proved the 35% of improvement on average. One percent of improvement in the results of subnetwork method to L1-Norm minimization cannot be acceptable. The algorithms of detecting unstable points in common methods and the method of analyzing subnetwork were conducted on a real network and the results are in line with simulated network results.

This work is licensed under a Creative Commons Attribution 4.0 License.