feat(behavior_path_planner): revise lane change module

[zulfaqar-azmi-t4]

Description

Previous lane change logic checks the distance between ego to the object using l2Norm. However, some information lost might occurs for example the object is on the left or right etc.

The lane change module is currently revised internally to better handle the dynamic environment. Instead of using l2norm, the distance is separated into lateral distance evaluation and longitudinal distance evaluation. the longitudinal distance evaluation takes into account of collision via predicted path.

This PR aims to realize the newly revised functions.

For configuration, add the following tier4/autoware_launch#355
For visualization, add the following #1140

Related links

Tests performed

1. rtc_auto_approval for right and left lane change needs to be disable
2. Run the PSIM.
3. Place ego starting position and goal. Goal must be on the adjacent lane to simulate lane change process.
4. Place an object on the target lane.
   
5. If ego is within object vicinity and there is potential of collision via predicted path comparison, no path appears.
   

For configuration, cherry-pick/pull the following tier4/autoware_launch#355

Documentation #1385

Notes for reviewers

1. Reviewer should also confirm that the solution will not break the normal avoidance behavior.
2. Please add the new parameters to the config file.
3. Force lane change is removed for now.

Pre-review checklist for the PR author

The PR author must check the checkboxes below when creating the PR.

• I've confirmed the contribution guidelines.
• The PR follows the pull request guidelines.

In-review checklist for the PR reviewers

The PR reviewers must check the checkboxes below before approval.

• The PR follows the pull request guidelines.
• The PR has been properly tested.
• The PR has been reviewed by the code owners.

Post-review checklist for the PR author

The PR author must check the checkboxes below before merging.

• There are no open discussions or they are tracked via tickets.
• The PR is ready for merge.

After all checkboxes are checked, anyone who has write access can merge the PR.

[codecov]

Codecov Report

Merging #1139 (5e2cd46) into main (04fe796) will increase coverage by 0.34%.
The diff coverage is n/a.

❗ Current head 5e2cd46 differs from pull request most recent head df26b33. Consider uploading reports for the commit df26b33 to get more accurate results

@@            Coverage Diff             @@
##             main    #1139      +/-   ##
==========================================
+ Coverage   10.34%   10.68%   +0.34%     
==========================================
  Files        1104     1105       +1     
  Lines       73650    78491    +4841     
  Branches    18155    18724     +569     
==========================================
+ Hits         7616     8387     +771     
- Misses      57449    61098    +3649     
- Partials     8585     9006     +421     

Flag	Coverage Δ
total	10.67% <0.00%> (+0.32%)	⬆️

Impacted Files	Coverage Δ
...lude/freespace_planning_algorithms/reeds_shepp.hpp	50.00% <0.00%> (-50.00%)	⬇️
...tic_evaluator/include/kinematic_evaluator/stat.hpp	62.50% <0.00%> (-19.32%)	⬇️
..._evaluator/include/localization_evaluator/stat.hpp	56.25% <0.00%> (-16.48%)	⬇️
...vehicle_model/sim_model_ideal_steer_acc_geared.cpp	73.58% <0.00%> (-9.44%)	⬇️
.../src/vehicle/test_vehicle_state_checker_helper.hpp	70.58% <0.00%> (-9.42%)	⬇️
planning/behavior_path_planner/test/input.cpp	60.00% <0.00%> (-7.86%)	⬇️
common/perception_utils/src/perception_utils.cpp	56.75% <0.00%> (-6.88%)	⬇️
...include/motion_utils/trajectory/tmp_conversion.hpp	60.00% <0.00%> (-6.67%)	⬇️
...oware_auto_geometry/include/geometry/common_2d.hpp	82.66% <0.00%> (-5.91%)	⬇️
...nning/behavior_velocity_planner/test/src/utils.hpp	82.22% <0.00%> (-5.59%)	⬇️
... and 830 more

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

[TakaHoribe]

Thank you for your great work!
My comments are just questions mainly.

And please update readme.

[TakaHoribe]

Just for my curios, what is the purpose of the enable_collision_check_at_prepare_phase parameter?

[TakaHoribe]

What is the use_butter for? In my understanding from the code, use buffer when executing lane change and zero buffer for abort check.

And is this a different one from lateral_distance_threshold in ros parameter?

[zulfaqar-azmi-t4]

the lateral_buffer is used to filter out dynamic object in the current lane.
Assuming the distance of the object to ego current path's line is d, then if

d < (0.5 vehicle_width + lateral_buffer)

then the object is considered in current lane.

lateral_distance_threshold is the threshold for safety check. Assuming distance between ego's polygon and object's polygon is h, if

h < lateral_distance_threshold

then we will check for the longitudinal distance.

[TakaHoribe]

Thanks. I understood the implementation, but still not clear about why there are two filtering for the object. Could you explain one situation where these two filters are needed?

[zulfaqar-azmi-t4]

One of the situation i can think of is for debug purpose, at least this is what my current use case are (I am not sure about the intention of the original author), and it will allow us to have more information about the object or failures.

[zulfaqar-azmi-t4]

@TakaHoribe @rej55
The readme/documentation is updated in another PR (#1385)

[TakaHoribe]

@zulfaqar-azmi-t4 The behavior looks good. My comments are mainly for the rationale of the implementation. If you update the readme and it is ok, then go merge.

[TakaHoribe]

LGTM