ROS Resources: Documentation | Support | Discussion Forum | Index | Service Status | ros @ Robotics Stack Exchange
Ask Your Question

Revision history [back]

click to hide/show revision 1
initial version

Hi thibs, I've encountered something similar to this issue myself. The thing about teb planner is that it doesn't consider the inflation layer like the global planner does it's more concerned with the lethal cost of a maps boundaries (teb_tutorial).

The ways I tuned for areas with narrower free space is by using tebs in built inflation calculation [inflation_dist] [weight_inflation], [feasibility_check_no_poses] which checks against the foot print along the current teb poses for viability, and [global_plan_viapoint_sep] which sets points along the global plan that teb tries to stay close to. If you have some processing power to spare I would also recommend switching from line to "two circles" as your footprint, this should help with the pose calculations

In terms of moving slightly backward when the robot is stuck, that's better handled by a recovery behavior.