Predictive Control for Autonomous Articulated Vehicles

A Bachelor's Thesis about self driving trucks done at Chalmers University of Technology in 2017. In collaboration with Volvo Trucks.

The result of the project was a control strategy for guiding a model semi-trailer truck through a narrow course consisting of several 90-degree turns and a roundabout. The truck is controlled with a RaspBerry Pi via WiFi. ROS (Robot Operating System) was used to build the system. Check out some videos at our Youtube playlist. Also check out the project presentation and the project report

Link to the 2018 project, that builds on this one: https://github.com/ArticulatedControl18

Try out the project

If you'd like to try out the project in a simulation environment, then follow the instructions below. System requirements: Ubuntu 16+

Installation

• Download this script (updated 2018-01-22), which downloads ROS, creates a new catkin workspace and clones down all the repositories you need. If you don't need to do all this, just comment out the lines you don't want to run

  • Put the file where you want to create the new workspace and source it (just running it won't work)

    cd path/to/installation     
    
    source ./installation.sh
    

How to run the simulator

• roslaunch truck_master master.launch
• Use the "2D Nav Goal" tool to set the goal.
  • Watch as the pathplanning does it's thing and the truck starts to move :)
• To set the position and direction of the truck, use the "2D Pose Estimate" tool
• To put subgoals use the "Publish Position" tool
• Add obstacles by focusing the window with the pyplot and press the number of the obstacle you want to add. (on the keyboard)
  • Watch as the rviz map is updated :)
• If the pathplanning seems to be stuck in an infinite loop you may have to kill the node (or just restart everything).
  • rosnode kill path_planning
  • Remove all obstacles
  • rosrun path_planning path_planning_node.py