EDM Calibration with GeoLab
By Dr. Robin Steeves
Let's go through an example of using "auxiliary parameters" in GeoLab. The usual parameters (unknowns) in a surveying adjustment are the station coordinates, but in GeoLab you can also define additional, or "auxiliary", parameters which model constants, scales, rotations, and translations. Have a look at GeoLab's AUX record for more detail.
For the best distance measurement accuracy using total station, or EDMs (Electronic Distance Measurement instruments) in general, we normally perform either a baseline or "in-network" calibration of our instruments.
Traditionally, distance measuring instruments are calibrated on a precise linear baseline, which is normally constructed of stable monuments (e.g. concrete pillars) at various distances apart. The example layout we'll use here is illustrated in the figure.
Note that, although this example uses a linear baseline for calibrating an EDM instrument, an "in-network" calibration may give better results because the resulting estimated calibration parameters may better model the instrument's actual systematic errors when based on the survey measurements in the network we measured.
To demonstrate how GeoLab can perform an EDM baseline calibration, we have set up an IOB file named "edmbaseline.iob", which you'll find in GeoLab as one of the installed sample networks.
This IOB file contains the following records. Note that some of the record names (like "LDEC" below) have been linked to the description of that record type, for your convenience - please see the comments below this IOB file for a discussion of some of its details:
* EDM Baseline Calibration Example
LDEC 4
PADJ NO NO NO NO NO YES NO NO
PGEO NO
PSOL YES NO
CONV 0.0001
VSCA 0.0508
* Define and weight the baseline coordinates. This is required because we have only
* distance measurements, and the 'network' stations will be undefined in azimuth without
* this information. The standard deviations of these coordinates should be larger than
* the standard deviations of the 'known' baseline distances.
GRP Obs 001
2DC
NE A 850000 250000 NB m
NE B 850150 250150 NB m
NE C 850225 250225 NB m
NE D 850250 250250 NB m
GENC LG DIAG 0.0000 1.00000 0.0001 1.00000 0.00000 0.00000 1.00000 m
* Define 'known' baseline distances and accuracies
GRP Obs 002
DIST A B 212.1441 0.0005 m
DIST A C 318.2159 0.0005 m
DIST A D 353.5743 0.0005 m
* Define auxiliary parameters for distance zero-correction and scale factor
AUX DIS EDM CONS SCAL
* Refer to the auxiliary parameters from a sigma-id (dis) for the distance
* observations, and set distance observation standard deviations:
SIGM dis 0.002 1.0 0.001 0.001 EDM m
* Enter the distances measured with the EDM for the calibration:
GRP Obs 003
DIST dis A B 212.154 m
DIST dis A C 318.227 m
DIST dis A D 353.585 m
DIST dis B A 212.154 m
DIST dis B C 106.082 m
DIST dis B D 141.439 m
DIST dis C A 318.227 m
DIST dis C B 106.082 m
DIST dis C D 35.367 m
DIST dis D A 353.585 m
DIST dis D B 141.439 m
DIST dis D C 35.367 m
All lines beginning with a "*" are comments, which are ignored by GeoLab, so the first line is a comment.
The next four lines (PADJ, PGEO, PSOL, and CONV) set specific options for this adjustment. The VSCA line sets a scale factor that GeoLab will use to scale the subsequent measurement covariance matrices (or variances).
The first group of measurements is named "Obs 001" using the "GRP Obs 001" record. This is a set of "weighted stations". The reason for this group of measurements is explained in the comments above the group.
The second group of measurements, named "Obs 002", specifies the accurately known (previously measured) distances of various lines of the baseline.
Note how the auxiliary parameters (CONS (constant bias), and SCAL (scale bias)) are defined by the AUX record. The following SIGM record is then used to assign this AUX group to any measurements that use the "sigma-record identifier" dis.
Finally we have the "Obs 003" group of measurements are the distances measured with the instrument we wish to calibrate. Note that all measured distances in this group are assigned the sigma-record identifier "dis", so the corresponding auxiliary parameters will be solved for by GeoLab as part of the observation equations for these measured distances.
Go ahead and try this adjustment for yourself in GeoLab. If you need help, just use the Contact Us page to let us know. Thanks!