An Optimal Local Path Planning in High Speed Scenario

This repo is for an optimal local path planning algorithm deployed in F1tenth Rviz.

Run the Rviz simulator

roslaunch f110_simulator simulator.launch

Run your Follow the wall

roslaunch f110_simulator wall_follow.launch

Run your Global Path tracking

roslaunch f110_simulator stanlley_controller.launch

Run your Local Path tracking

roslaunch f110_simulator frenet_controller.launch

Follow the wall

The wall following algorithm is to implement a Proportional Integral Derivative (PID) controller to make the F1tenth drive parallel to the walls of a corridor at a fixed distance.

Global Path planning

Based on the predefined way points, the Stanley Controller is deployed to track the global path.

Local Path planning

An optimal local path planning has been designed in the Frenet frame, then the best path is selected by the cost function.

Perception

The max gap is to find the maxmium gap range ahead of the vehicle.

Integration

This part integrate Auto Emergency Brake (AEB), path planning and Perception together.