Abstract

Deformation analysis is crucial to applications in geodesy, structural engineering, and geology, of which the main goal is to detect the behaviors of a deformed body. Traditional deformation analyses rely on a limited number of observations and thus give a relatively poor description of the strain field on the entire object. In this study, a method based on the displacement gradient model and unified least-squares adjustment is proposed to improve classical deformation analysis. Corresponding quality assessment and sensitivity analysis are derived accordingly to better assess significant deformation. Furthermore, by applying nearest neighbor searching and a triangulated irregular network, the efficiency of analyzing a vast number of observations is improved. Numerical experiments based on real data suggested that the proposed approach detected behaviors of a deformed body in an effective and efficient way. Consequently, the strain field on an object can be obtained rapidly and accurately using the proposed method and a large point dataset.

Key words: Deformation analysis, strain field, dispersed point data, structural health monitoring.