Perception and Decision Optimization of Autonomous Driving System Driven by Internet of Things and Artificial Intelligence Algorithm

AI algorithm can also carry out self-learning and self-optimization through Machine learning (ML) and deep learning (DL), and continuously improve the perception and decision-making ability of self-driving cars. In order to build a more intelligent computer-aided driving system, this article applies the algorithms to the perception and decision optimization of autonomous driving system. In the experiment, the algorithm is verified and tested, and pictures of various scenes are used for experiments, which verifies the adaptability, stability and accuracy of the algorithm to different scenes. The algorithm in this article has a significant advantage in the accuracy of vehicle or obstacle feature detection, which is 28.64% higher than the contrast algorithm, which means that the algorithm in this article can identify the features of vehicles and obstacles more accurately and locate their edge contours more accurately. The system can realize real-time monitoring and target recognition of the road in front of the vehicle, and make decision and control of vehicle-assisted driving according to the target information and the vehicle's own information and instructions. © 2024 U-turn Press LLC,.