This thesis uses the Minimum Norm Quadratic Unbiased Estimation (MINQUE) to estimate standard deviation of observations of a total station. Different setups are created by altering the number of stations and targets and their relative position in the network to study the effect that different setups have to the estimation and define what are important to minimize the effect of the setup to the estimation.
A lot of research has been done around methods for estimation of variance and covariance components, since it is useful in many fields. Various approaches exists to solve the problem of variance components estimation. Geodesy is a special case, were their often is a a prior knowledge of how well an instrument is able to record measurements. There is an ISO-standard for testing and verification of geodetic instrument but also an alternative approach the KTH-Total Station Check.
For the estimation three main types of setups were defined and used in the simulation. These main types were then altered to see how different changes to the setup effect the overall estimation. The alterations were changes in distance between station and targets, changes in vertical distance between stations and targets and the amount of observations carried out by adding more stations and targets to the setups.
The result of the simulations shows that the tested changes in the setups do effect the estimation. It was not possible to determine by how much for each change, because a change in vertical displacement also meant a change in angles and distance between the station and the target. Increasing the amount of stations and targets or one of them shows that standard deviation of the estimation becomes smaller. The effect can be seen independent of which type of setup that is used. The most important factor to how good the estimation will be is the amount of observations.
Source: KTH
Author: Bergkvist, Joel