الفهرس 
The Concept of Least Squares Network AdjustmentOnly 14 pages are availabe for public view
 |  | from | 234 |  |  |
| | | from | 234 | | |
Abstract
Geodetic network constraints provide information to the adjustment process about the absolute position and orientation of the network. In practice, all of the stations in the network will have some form of position constraints (by their approximate coordinates) that define their relationship to the project datum. Geodetic networks can be classified into the following types:
1) Free networks, for which all the coordinates of its points need to be calculated.
2) Minimum constraint networks, which contains the ordinary minimum constraints for adjustment while other parameters are free.
3) Over-constrained networks, which contains more constraints than necessary to remove the network datum defects which have been defined and there are points need to be determined.
4) Fully-constrained networks: where adjustments to all stations in the reference network are assumed to have well-known coordinates (i.e., stable points), and these are fixed with zero error in the adjustment process. With fully-constrained adjustments, for the monitoring networks, only the monitoring point stations are allowed to float and adjust its position. The drawback of a fully-constrained adjustment is that any errors due to inaccurate reference station coordinates will be transferred to the monitoring points. Finding an optimal configuration and an observation plan are the most important steps in the design and establishment the geodetic networks (especially a deformation monitoring network). The main goals of an optimal network design process include finding proper locations of control stations (First order Design) as well as proper weights of observations (Second order Design) in a way that satisfies all the criteria considered for quality of the network which is itself evaluated by the network accuracy. There are two methods for designing a network: trial and error and analytical methods. In trial and error method the objective function is computed with a suggested solution for the problem. If the suggested solution does not satisfy the objective function, the solution is changed a bit and the objective function is re-computed. This process is repeated until the requirement is satisfied. The solution in the trial and error method depends strongly on the experience of the designer. However, in some cases the solution might not even be found. On the other hand, the analytical approach takes advantage of a mathematical algorithm and designs the network in a way that the quality requirement of the network is satisfied while the network is optimum from mathematical point of view.