Performing multi-track Alignment
See figure 5.1, below. It shows the process of the multi-track alignment . Notice the similarities in the processes in the primary alignment and the those used to also align to a secondary reference file.
Option *F.g01
Figure 5.1
What is depicted
above as stage I is the “conventional’ alignment. This is all performed
in the data collection directory such as
\NYWASR2. The output is the new .g01 file *A.g01. Note the *A.g01
file (generated for the major reference file), has typical track number placed in column 42 of the binary data. If Alignment
is not successful at a given location,
a “0” is entered in that location, and MP ft corrections are based on
geodetic data only.
Stage II is the
secondary alignment. Note the
“Alternate reference file” is for a path taken, usually on a track other than
that on the primary reference file.
There may be some overlap. But if there
is, and the alignment was
successful for the Primary reference file, no action is taken for further
correction on the second stage for that section of track
The second stage is
performed in a multi-track data collection
directory such as \offtkNYWAs . When correcting MP and ft. for the secondary
alignment, corrections are only made
where:
- Entry on col. 42 is “0” and,
- Where alignment is indicated for the
secondary reference. Then the typical track number of secondary alignment
is placed in col. 42.
Note that the original measurement .g01 file is used in the alignment is matched against a secondary reference file. The Original GPS file is used. Note that the new .g01 file, resulting from
the primary alignment, *A.G01, is not used or modified until a new .g01 file generator ( different than that used for the primary
alignment) is executed based on the outputs of the secondary
alignment. The output is the file: *B.g01 ( “B” can be any letter except “A“or “F”). Note that the only difference between original .g01 file, *A.g01, and *B.g01 is
that the milepost and ft. annotation is
changed appropriately ( and the inserted numbers onto “col. 42). Also note that
at option A “square new.g01 file” ,*F.g01 can also be created
Scenario of the Secondary
Alignment Process
The scenario is the
following in the directory of the data collection of the secondary alignment.
The original .g01 file is processed with an
Alignment program such as “tran94znAMtot6gg.exe”, the same program that
could have been used for the Primary alignment. The outputs include: odometer correction file,bda*.dat, alignment
skip file, skipd*.dat all of which are
required for secondary alignment. If
one were to observe the quality plot , there would only be alignment ( if any
at all) at a few intervals during the
run. The skipd*.dat file indicates
where alignment did not work
Note that the secondary alignment porcess can be repeated several times if necessary.
Brief Summary of an Example
of process
Steps in getting a new .g01 file for run 210032:
Run Primary Alignment for NY-WAS Corridor
Obtain Quality plot Shown Below.
The Green Line represents the smoothed odometer correction estimate, that is the value used. The Blue line are the raw corrections before smoothing
The Red dots indicate correlation coefficients with the reference file. It is seen that the Correlation is close to 90% Most of the time, except between approximately MP 8 ,_30000 ft, to Milepost 23 ~110000 ft. Actually in this region the train is on the alternate track,4. And after Milepost 112 at ~1,122,000 ft. , where actually it is on track 2 , That usually reserved to opposite directed traffic.
A new G01 file, 210032A.g01 is created. Correcting data in regions where matching worked, with the odometer correction information. Column 42 of the Binary file has a “3” ( For track 3) placed there. In the other regions where correlation was not high , position was corrected with GPS based data, and a “0” was placed in column 42.
Then A Secondary alignment is performed, in the appropriate secondary alignment directory, for travel on Track 4. The Quality plot generated is that shown below
Quality plot for secondary alignment
Note the correlation points out to about 100000 ft are high. It is also high at other places because in fact the alternate track sometimes disappears and merges with the primary track, thus the secondary reference has the same data as the secondary in these regions.
Also note that even though there are long periods of no correlation , the GPS allows suitable slewing of the algorithms so when there is correlation it will be found ( algorithm does not get lost).
The New .g01 file 210032A.g01 is then operated on by odometer corrections in those locations where there is high correlation AND data has not been successfully aligned to the primary reference. That is between 30000 ft and 100000 ft., only. This is used to Create 210032B.g01. Where, in that region, Col. 42 of the binary data has a “4” inserted. This file will then be aligned to other files on track 4 during that region, and in other regions to files that are on Track 3.
At the end of the run, the vehicle traveled on the opposite directed track, track 2. Another secondary alignment is performed using the opposite directed track data ( in the current direction of motion) . The resultant Quality plot is shown next.
It is seen high correlation only is present at the very end at ~ 1122000ft. till the end. Then the new .g01 file, 210032B.g01 is operated on with the estimated odometer corrections to adjust MP and ft for this region, lined up with other runs also in that track (2). 210032C.g01 is created with appropriate lines of data containing “2” in column 42.
At user option this resulting file can then be converted to a 210032F.g01 file which has one line entry for every foot as defined by the Geodetic Milepost File.
All *F.g01 files have the same numbr of entries in a given milepost interval.
It should be noted that, Although the are operator input indicating which track the car is on, this data is generally not used in the process, the program automatically makes the decisions based on the correlating of the data.