In recent years, there has been intense progress in the autonomous car field. In large part of the publication, the problem of controlling the movement of an autonomous vehicle is solved based on machine learning algorithms that require training data, significant computing power, and memory resources. The work aimed to find an alternative methodology that would allow adaptive car control in various road scenarios. The following tasks were analyzed: driving in a column of cars with a safe distance, overtaking maneuver and returning to the correct lane, keeping the vehicle in the middle of the lane while moving along a winding track with automatic speed adjustment. During the simulation, data from sensors (cameras, LIDAR) were used by adaptive controllers. A state machine was used to switch between controllers. The visualization of the system's operation was realized thanks to the newly introduced possibility of integrating the graphics engine “Unreal Engine” and “Matlab 2021a”. Based on the algorithm's operation, it was found that adaptive controllers can effectively cope with the control of autonomous vehicles in various situations.
Adaptive Control of the Autonomous Car
Lect. Notes in Networks, Syst.
International Conference Mechatronics ; 2020 ; Międzyzdroje, Poland September 09, 2020 - September 11, 2020
2022-02-05
10 pages
Article/Chapter (Book)
Electronic Resource
English
Adaptive control of autonomous or semi-autonomous vehicle
European Patent Office | 2023
|ADAPTIVE CONTROL OF AUTONOMOUS OR SEMI-AUTONOMOUS VEHICLE
European Patent Office | 2022
|Adaptive control of autonomous or semi-autonomous vehicles
European Patent Office | 2022
|ADAPTIVE CONTROL OF AUTONOMOUS OR SEMI-AUTONOMOUS VEHICLE
European Patent Office | 2021
|ADAPTIVE CONTROL OF AUTONOMOUS OR SEMI-AUTONOMOUS VEHICLE
European Patent Office | 2023
|