Fixed Observations

by Micha ⌂, Bad Vilbel, Monday, March 15, 2021, 08:45 (1350 days ago) @ Angele

Hello Angele,

welcome to the foum.

it's possible to create fixed distances in the adjustment.

Yes, it is possible using a simple trick. In JAG3D, the distance is an observation having a specific measurement uncertainty. The uncertainty describes the possible dispersion/spreading of the measurement. For small uncertainties, the dispersion becomes small and vice versa. If the uncertainty tends to zero, i.e. $\sigma \to 0$, the observation is (quasi) fixed.

In JAG3D, it is not possible to set an uncertainty to zero but you can use a quite small number say 10 µm or something like that - the choice depends on the uncertainties of the remaining observations.

Please take a look to the first two figures that shows the normal/unmodified case.

Unmodified raw observation
[image]

Unmodified result observation
[image]

The selected row depict the distance from 1087 to 20. The ε-column indicates the residual of about ε = -11.5 mm.

To fix this distance, I modified the uncertainty of the distance and readjust the network.

Modified raw observation
[image]

Modified result observation
[image]

As you can see, the distance is quasi fixed and, thus, ε is close to zero.

kind regards
Micha

--
applied-geodesy.org - OpenSource Least-Squares Adjustment Software for Geodetic Sciences

Tags:
Observation, Uncertainty, Fixing

locked
1201 views

Complete thread:

 RSS Feed of thread