Investigation of reliability of static deformation analysis for horizontal control networks

Deformation analysis is one of the most interest areas in the survey engineering. There aretwo kinds of network in deformation analysis, relative monitoring network and absolutemonitoring network. In this study, we investigate reliability of both relative and absolutedeformation networks with the conventional deformation analysis techniques using MonteCarlo method. For this aim we have simulated many different working samples. Thereliability of these techniques is measured by Mean Success Rate (MSR). Firstly, fourdifferent horizontal control networks are simulated for relative and absolute networks toinvestigate how the reliability of the method changes depending on the number of points andobservations, the magnitude of the displacements, kind of the displacements and the numberof the displaced points. The MSRs are computed in a given displacement interval for a certainnumber of displaced points. The MSRs of the method increase when the degrees of freedomand the magnitude of displacement are increased. They decrease when the number of thedisplaced points and the number of points are increased. In addition, the heteroscedastic andhomoscedastic horizontal control networks are also simulated. If random errors have acommon variance and they are independent and identically distributed, the model is calledhomoscedastic. In cases where the random errors' variances may vary depending on aparameter, the model is called heteroscedastic. If the model is homoscedastic, the reliability ofthe method increases respect to the ones of the heteroscedastic models.Keywords: Conventional Deformation Analysis Method, Static Model, Reliability, MeanSuccess Rate, Simulation, Homoscedastisity, Heteroscedasticity.