Expand description
Deterministic corridor-loop scenario for LidarGraphSlam.
A robot drives slightly more than one lap of a 3 m wide rectangular
corridor (30 × 20 m outer walls, 24 × 14 m inner block) with pillars along
the walls — except the top corridor, which is left featureless so the
scan-to-map front end cannot observe forward motion there. Wheel odometry
over-reports distance and drifts in yaw; the simulated 2D LiDAR adds range
noise. run_corridor_loop records every step so examples, gallery
renderers, and the playground can replay the same run.
Structs§
- Corridor
Loop Config - Scenario parameters for
run_corridor_loop. - Corridor
Loop Frame - State recorded after one
LidarGraphSlam::update. - Corridor
Loop Run - A full recorded run of the corridor-loop scenario.
Enums§
- Corridor
Layout - Pillar arrangement along the bottom corridor.
Functions§
- aliased_
corridor_ config - Perceptual-aliasing variant: identical pillars every 2.5 m along the bottom corridor and a start in its middle, where no corner is in LiDAR range, so loop matches can lock onto the wrong pillar.
- corridor_
loop_ deltas - Body-frame step deltas along the corridor centerline (rounded rectangle),
starting
start_offsetmeters along it and covering one lap plusextra_distance. - corridor_
loop_ lap_ length - Length of one lap of the corridor centerline [m].
- corridor_
loop_ start - Start pose of the scenario (bottom corridor, heading +x).
- corridor_
loop_ start_ for - Start pose of
config:start_offsetmeters along the centerline. - corridor_
loop_ walls - Walls of the corridor loop: outer 30 × 20 m, inner block 24 × 14 m, and pillars at irregular spacing except along the top corridor.
- corridor_
loop_ walls_ for - Walls of the corridor loop with the given bottom-corridor
layout. - position_
rmse - Planar position RMSE between two equally indexed pose sequences [m].
- run_
corridor_ loop - Runs
LidarGraphSlamthrough the corridor loop and records every step.