Design of Vehicle Unmanned Driving Navigation System

Building environment and energy application engineering, Beijing University of Technology, 100049, Beijing, China

^* Corresponding author: zyf2116@emails.bjut.edu.cn

Abstract

This paper studies the design of unmanned vehicle navigation system, aiming to improve the autonomous navigation capability of unmanned vehicles in complex environments. Based on multi-sensor information fusion technology, combined with positioning navigation algorithm and dynamic path planning method, a high-precision and high-real-time unmanned navigation system is constructed. During the system design process, key technologies such as visual SLAM, inertial navigation and GPS fusion positioning are studied, and a target recognition and environment perception module based on deep learning is designed to realize the autonomous decision-making and path planning of vehicles in dynamic environments. Through testing and verification in actual scenarios, the system performs well in navigation accuracy, obstacle avoidance ability and path planning efficiency in complex environments. The research results show that the proposed unmanned navigation system has good adaptability and reliability in scenarios such as urban roads and highways. What’s more, this paper shows that there are more and more theoretical support and practical reference for the application of unmanned driving technology in the future.