Vývoj a simulace asistenčního parkovacího systému pomocí MATLAB/Simulink

Abstract

This master thesis deals with the development and simulation of a parking assist system using co-simulation between the MATLAB/Simulink environment and the CARLA simulator. The proposed system processes image data from cameras to create a Bird’s Eye View (BEV), which is then used to detect available parking spots. Collision-free trajectory planning is handled by the RRT algorithm, and the actual vehicle control is performed by Stanley controllers from Automated Driving Toolbox. The overall decision-making logic, including the handling of error states, is managed by a state machine implemented in Stateflow. The functionality of the complete solution was successfully verified using a Nissan Patrol 2021 vehicle model across three testing scenarios: parallel, angled, and perpendicular parking.