Priority-Based Trajectory Planning for Networked Vehicles Using Motion Primitives

• Priority-Based Trajectory Planning for Networked Vehicles Using Motion Primitives
• Setup
  • MATLAB R2023a
  • System Requirements for MATLAB ROS 2 Toolbox
  • System Requirements for using Lanelet2
• Run Exeriments
• References
• Acknowledgements

Setup

MATLAB R2023a

Install MATLAB R2023a with the following toolboxes:

• Statistics and Machine Learning Toolbox
• ROS Toolbox
• Parallel Computing Toolbox

System Requirements for MATLAB ROS 2 Toolbox

In our priority-based trajectory planning, vehicles communicate using the MATLAB ROS 2 toolbox. Their custom messages are compiled with the MATLAB built-in function ros2genmsg(), for which you must have Python software, CMake software, and a C++ compiler for your platform (ROS Toolbox Requirements). For MATLAB R2023a

1. Python 3.9
   1. Install Python 3.9 and add it to your system path variable.
   2. Go to the MATLAB command window and execute pyenv('Version','version') to set up Python with MATLAB. For example, pyenv('Version','3.9') will let MATLAB use Python 3.9. If MATLAB cannot find the version, provide its path with pyenv('Version','fullPathOfYourPythonInstallFolder\YourPythonExecutable').
2. CMake 3.16.3+
   1. Install from https://cmake.org/download/
3. C++ compiler
   1. Installation
      • Windows: Install the Visual Studio 2019 Community Edition. The (free) community version is sufficient. Select "Desktop development with C++" during the installation.
   2. Configure your C++ compiler using mex -setup in the MATLAB command window.

System Requirements for using Lanelet2

As one map representation we use lanelet2, because it offers libraries with lots of useful functionality. Since the libraries only exist in Python and C++, mex functions are used to benefit from that. In order to compile these mex functions the following dependencies need to be met.

1. Lanelet2
   1. Install according to https://github.com/fzi-forschungszentrum-informatik/Lanelet2#installation . Under Ubuntu22 you probably want to use it as part of the ROS2 humble version, i.e., sudo apt install ros-humble-lanelet2
2. Eigen3
   1. Install Eigen3 as dependency of Lanelet2. In Ubuntu, the following command should be sufficient: sudo apt install libeigen3-dev
3. Git Submodules
   1. Since some of our functionality is added by using git submodules, use git submodule init after cloning to retrieve the content of these repos.

Run Exeriments

See here

References

Please refer to the respective publication if you are using it for your work. Thank you very much!

P. Scheffe, J. Xu and B. Alrifaee, "Limiting Computation Levels in Prioritized Trajectory Planning with Safety Guarantees", ResearchGate, Preprint, 2023, doi: 10.13140/RG.2.2.32731.03368

Checkout the the software version 5.0. The results of the publication can be reproduced by running

open graph_based_planning.prj
eval_parallel_computation_prediction_inconsistency()
eval_parallel_computation_CLs()

The results are saved in the folder "results".

P. Scheffe, J. Kahle and B. Alrifaee, "Reducing Computation Time with Priority Assignment in Distributed MPC," TechRxiv, Preprint, 2023, doi: 10.36227/techrxiv.20304015.v2

Checkout the the software version 4.0. The results of the publication can be reproduced by running

eval_coloring_paper()

This evaluation comprises 720 simulations, so it will take days until completion. The results are saved in the folder "results".

P. Scheffe and B. Alrifaee, "A Scaled Experiment Platform to Study Interactions Between Humans and CAVs," 2023 IEEE Intelligent Vehicles Symposium (IV), Anchorage, AK, USA, 2023, doi: 10.1109/IV55152.2023.10186623.

Checkout the the software version 2.0. The results of the publication can be reproduced by running

hdv_reachable_set_experiment()

The results are saved in the folder "results".

P. Scheffe, M. V. A. Pedrosa, K. Flaßkamp and B. Alrifaee, "Receding Horizon Control Using Graph Search for Multi-Agent Trajectory Planning", in IEEE Transactions on Control Systems Technology, 2022, doi: 10.1109/TCST.2022.3214718.

Checkout the the software version 1.0. The results of the publication can be reproduced by running

eval_rhgs()

The results are saved in the folder "results".

P. Scheffe, G. Dorndorf, and B. Alrifaee, “Increasing Feasibility with Dynamic Priority Assignment in Distributed Trajectory Planning for Road Vehicles,” in IEEE International Conference on Intelligent Transportation Systems (ITSC), 2022, pp. 3873–3879. doi: 10.1109/ITSC55140.2022.9922028.

The code is implemented in the CPM Lab software repository. The High Level Controller is implemented in C++ and is named "dynamic_priorities".

References in Bibtex format

@article{scheffe2023limiting,
    author  = {Patrick Scheffe and Jianye Xu and Bassam Alrifaee},
    title   = {Limiting Computation Levels in Prioritized Trajectory Planning with Safety Guarantees},
    year    = {2023},
    doi     = {10.13140/RG.2.2.32731.03368}}
}

@article{scheffe2023reducing,
    author = {Patrick Scheffe and Julius Kahle and Bassam Alrifaee},
    title  = {Reducing Computation Time with Priority Assignment in Distributed MPC},
    year   = {2023},
    month  = {2},
    doi    = {10.36227/techrxiv.20304015.v2}
}

@article{scheffe2023scaled,
    author = {Patrick Scheffe and Bassam Alrifaee},
    title  = {A Scaled Experiment Platform to Study Interactions Between Humans and CAVs},
    year   = {2023},
    month  = {2},
    doi    = {10.13140/RG.2.2.24697.13923}
}

@article{scheffe2022receding,
    author  = {Patrick Scheffe and Matheus Vitor de Andrade Pedrosa and Kathrin Flaßkamp and Bassam Alrifaee},
    journal = {IEEE Transactions on Control Systems Technology},
    title   = {Receding Horizon Control Using Graph Search for Multi-Agent Trajectory Planning},
    year    = {2022},
    volume  = {},
    number  = {},
    pages   = {1-14},
    doi     = {10.1109/TCST.2022.3214718}
}

@inproceedings{scheffe2022increasing,
  title = {Increasing {{Feasibility}} with {{Dynamic Priority Assignment}} in {{Distributed Trajectory Planning}} for {{Road Vehicles}}},
  booktitle = {{{IEEE International Conference}} on {{Intelligent Transportation Systems}} ({{ITSC}})},
  author = {Scheffe, Patrick and Dorndorf, Georg and Alrifaee, Bassam},
  year = {2022},
  pages = {3873--3879},
  doi = {10.1109/ITSC55140.2022.9922028}
}

Acknowledgements

This research is supported by the Deutsche Forschungsgemeinschaft (German Research Foundation) within the Priority Program SPP 1835 "Cooperative Interacting Automobiles" (grant number: KO 1430/17-1).