Two Methods of Positioning

 

Relative Positioning using  map matching techniques based on Gage and cross level. All referenced to a reference file. This is Called Geometry Alignment

 

Geometry data from "measurement file" is compared to  that of stored reference file. The Errors between them are used to adjust the relative odometer correction between them, through a Kalman Filter, modeling odometer and measurement errors.

 

Key output is the Odometer Correction, used to correct measurement file annotation of MP and ft.  with that for reference.

 

It provides a file of regions where alignment may be suspect.

 

Quality plots are provided to asses if algorithms are working appropriately

 

Many integrity Tests to assure system is locked or correlated. And Sophisticated search algorithms for instances when the  algorithm gets lost.

 

 


 

                     GPS positioning a robust Augmentation

 

Aids in Geometry alignment:

 

Positions geometry data when geometry alignment is not possible

 

In a parallel computation with the Geometry Alignment: A Distance based Kalman filter inputs GPS, geometry curvature( a type of rate gyro), and odometer in a Kalman filter estimating position and distance along Track

 

It Can Reject bad GPS measurements , extrapolate through GPS outages and provide indication of the Quality of estimation.

 

It can reset itself if issues arise

 

It allows for Tie Down points to correct distance along track with the corridor track map when needed.

 

It provides data for geometry data positioning in the absence of Proper alignment

 

If Geometry alignment not possible, geometry data is positioned with this  within 3-4 ft!


 

 

Combining The Two Techniques

 

During running of the alignment program, GPS derived results are used to reinitialize the geometry alignment quickly and efficiently if alignment gets lost because of bad data or travel on alternate track

 

A Final program then combines the two measures of position.

 

If Alignment is deemed  good, position is corrected first by the estimated odometer offset, then according to track map, and MP& ft. are annotated though the Geodetic milepost file.

 

 .  If alignment is deemed poor  the position corrections are from the GPS derived Kaman filter.

 

 This competes the  annotation of the new.g01 file, where annotations are now synched to the reference file and referenced to ground truth.


 

Synergism's used in the approach include:

 

 

Fast recovery of track alignment when data does not correlate, using GPS positioning  in the search for a "restart".

 

If geometry alignment is possible ( that's most of the time) points can be geodetically positioned to within 2-3 ft.. even if GPS is not working in area

 

Linear positioning of data using GPS derived data when alignment for some reason cannot work is within 3-4 ft

 

Fewer "throw-away" Runs. Many of the GPS Faults, and Tachometer Outages, and large scale factors  that may have previously made a run nearly useless will be corrected