Bayesian Train Localization with Particle Filter, Loosely Coupled GNSS, IMU, and a Track Map

Algorithm 2

Track-selective evaluation over distances.

Algorithm: Track Selective Evaluation

Input: Train state , true track IDs , map

Output: Evaluation: cumulative ,

() : all track IDs of wrong switch ways from , map

() : switch positions from , map

() : clearance length of each switch

() for all train states   do

()   

()    if track is not in (true track ID list) then

()      if   is in   &  distance to switch <   then

()        

()      else

()       

()     end if

()   end if

()   if other track in the vicinity of , (20 m) then

()    

()   end if

() end for

() compute cumulative evaluation (40)

() compute cum. eval. of parallel tracks (41)

() return  ,