Research On Fusion Positioning And Navigation Strategies Of Intelligent Vehicles Under Specific Scenarios

With the continuous growth of car ownership at home and abroad,Intelligent vehicle have gradually become a research hotspot at home and abroad.Intelligent vehicle is a collection of advanced autonomous driving technologies such as environmental perception technology,positioning and navigation technology,path planning technology and decision control technology.Autonomous driving schemes in different environments are different.The autonomous driving of expressways requires high reliability and real-time performance,while the autonomous driving of urban roads requires the ability to deal with complex scenes.However,the autonomous driving schemes in specific scenes need to be improved according to the actual environment.This paper takes a specific scene,especially the semi-structured campus environment as the research object,and focuses on the multi-sensor fusion positioning technology and navigation strategy of the intelligent vehicle under this scene.The main research content of this paper can be summarized as the following aspects:(1)According to the key technology of existing automatic driving,put forward a set of intelligent vehicle platform architecture,design the framework of the hardware and the software framework of the intelligent car,to select the sensor for the fusion navigation positioning technology and programming technology,combining hardware framework,software framework is put forward under the applied to the scene of the autopilot.(2)Based on the proposed intelligent vehicle platform framework,a multi-sensor data fusion positioning technology based on UKF algorithm is proposed.Firstly,the existing positioning technology of intelligent vehicle is analyzed.According to the advantages and disadvantages of various positioning technologies and the advantages of each sensor,the multi-source sensor positioning technology is selected.The advantages and disadvantages of point cloud data matching algorithm ICP and NDT are analyzed,and NDT algorithm is selected as the algorithm for relative positioning of lidar.Finally,the positioning data of IMU,odometer and lidar are fused with UKF algorithm,and the experimental data in indoor environment are given.(3)After realizing fusion positioning,the global navigation and local navigation strategies of the intelligent vehicle are studied.In the aspect of global navigation,the optimal path from the current starting point to the end point is planned according to the third-party electronic map,so as to provide global path reference for the vehicle.In the aspect of local navigation,a scene-based local navigation method is proposed.In the scene of straight lane,the image processing method is used to detect the lane line in front of the intelligent vehicle to obtain the current feasible area and the expected driving route.In the intersection scene,the intersection is detected first,and then the coordinate system of the intersection is constructed to fit the target point,so as to obtain the driving path in the intersection scene.(4)Design the automatic driving experiment and outdoor test platform,verify the effectiveness of the fusion positioning method and navigation strategy in the paper,and point out the advantages and disadvantages of the method and the improvement direction.