Autonomous Driving Model with Collision Prediction for Urban and Extra-Urban Environments

This study introduces an architecture for an autonomous vehicle control system based on a collision detector and geometric modeling of trajectories. The goal is to develop a robust and reliable control model that can navigate metropolitan environments, often crowded with pedestrians and bicycles, as well as suburban areas, where traffic patterns can fluctuate. We have created a modular control unit that includes a collision predictor, which interacts closely with the decision module. The executed algorithm demonstrates the effectiveness of our system by ensuring the safety and comfort of the passengers. It can identify potential collisions from a distance and initiate braking preventively, following precise guidelines for deceleration and acceleration. To validate our methods, we are looking at simulations of realistic case studies. The research conducted underscores a crucial advancement in the development of safer and more flexible autonomous driving technologies.