Safe and Non-Conservative Trajectory Planning for Autonomous Driving Handling Unanticipated Behaviors of Traffic Participants
This repository contains the code to reproduce the simulations of the manuscript "Safe and Non-Conservative Trajectory Planning for Autonomous Driving Handling Unanticipated Behaviors of Traffic Participants" by Tommaso Benciolini, Michael Fink, Nehir Güzelkaya, Dirk Wollherr, Marion Leibold.
Run pip install -r requirements.txt to intall required libraries.
Change in the main.py file the Start-up section depending on your goals and run the main script with python3.
The file plot_all.py plots stored data.
With plots_for_paper.py plots in pdf and png format of defined time steps are generated.
The file print_statistics_multi_runs.py computes the mean and standard deviation of the run time for multiple runs.
You can find the used parameters and settings in the following files:
main.py
line 21: MPC Horizon
line 22: Flag, iterative recomputation of cosy
line 24: Show plot after run
line 25-26: Number of Simulations to evaluate statistical properties of the computation time.
line 30-31: Select the file name of the scenario
line 63-64: Linear dynamics of the obstacles
line 81-85: Initialize controller, set reference trajectory and weight matrices for MPC controller
controller.py
line 7-8: Use debug plot, plot for each predictet time step with feasible area.
line 9-10: use plot, Plot of the current situation
ev.py
line 12: Maximum velocity
line 13: Minimum velocity
line 14: Maximum acceleration
line 15: Minimum acceleration
line 16: Maximum change in acceleration