Skip to content

Hierarchical Model Predictive Control for the F1/10 autonmous car

Notifications You must be signed in to change notification settings

mlab-upenn/Hierarchical-MPC

 
 

Repository files navigation

Hierarchical-MPC

Hierarchical Model Predictive Control for the F1/10 autonmous car

High-level Trajectory Planner: At every time instance, select from a trajectory that maximizes progress along track centerline from a precomputed table of feasible trajectories.

Low-level MPC tracking controller: The MPC generates a dynamically feasible trajectory that tracks the reference trajectory given by the high level planner while satisfying all constraints.

For obstacle avoidance, RRT* is incorparated to provide convex feasible space constraints in real time to MPC based on LiDAR feedbacks.

Testing on the car:

Head-to-head Racing:

About

Hierarchical Model Predictive Control for the F1/10 autonmous car

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • C++ 88.5%
  • CMake 11.5%