Local planning of an autonomous driving car prototype

Abstract

Autonomous vehicles are mobile robots that integrate advanced technology for navigation, decision-making, and control. These vehicles have been developed in response to the growing concerns around accidents and injuries caused by human drivers, as well as to reduce the negative impact of cars on the environment, including energy consumption, pollution, and congestion. A combination of sensors and algorithms gives autonomous cars a comprehensive understanding of their surroundings, allowing them to safely navigate roads, detect other vehicles, pedestrians, and traffic lights, and reach their destination. This research is intended plan the path with obstacle avoidance while restricting to the constraints imposed by the maximum lateral acceleration and turning angles along lane.The proposed approach is capable of real-time detection and recognition of obstacles and tracking lanes. Several obstacle detection sensors are used to detect and avoid obstacles coming from the front, left and right directions with the detection accuracies of 96%, 87% and 81% respectively. Lane line detection and tracking are performed by the real-time acquisition of images followed by realtime image processing. Evaluation results prove that the proposed method can efficiently detect and avoid obstacles while following the lane lines on the road.