Multi-sensor Fusion Of Unmanned Surface Vessel Positioning And Navigation

With the advent of the robot age,more and more intelligent robots have appeared,and these robots have brought huge benefits to human society.With the large-scale commercialization of unmanned delivery vehicles and drones,and the rapid development of computer technology and communication technology,unmanned ships will also usher in a golden period of development.The foreign unmanned ship technology is relatively developed,and the domestic unmanned ship technology has been developed late.Since the emergence of the "Sky Elephant One" unmanned ship in 2008,although the domestic unmanned ship technology has made great progress,some of them are unmanned Ships have been commercialized,but manned ships are still the mainstream of the market due to the immaturity of unmanned ship technology.The autonomous positioning of unmanned ships is the foundation of unmanned unmanned ships,and a good autonomous navigation method is the first step for unmanned ships to realize their own value.Traditio nal unmanned ships use GPS to complete their own positioning.This method has limited positioning accuracy and will cause positioning failure due to poor satellite signals from unmanned ships.To solve the problem of unmanned ship positioning and navigatio n,in the actual project,this paper uses multi-sensor fusion to achieve accurate positioning of the unmanned ship.After GPS / IMU fusion positioning,SLAM is used to fuse GPS / IMU and laser data.Do further positioning of unmanned ships and create water maps.On the basis of accurate positioning and high-precision water area maps,multi-target points are set for unmanned ships,and autonomous cruise of unmanned ships is realized through multiple mixed path planning.The main work of this article is as fo llows:(1)Use GPS / IMU loose coupling to improve the stability and accuracy of satellite positioning.(2)On the basis of GPS / IMU positioning,SLAM is used to further improve the accuracy of unmanned ship positioning and build an unmanned ship environm ent map.(3)In the SLAM construction map,based on the environment of the unmanned ship’s actual working area,the environment is empty,the feature points are single and repeated,design experiments,and test the limitations of the traditional two-dimensional laser SLAM in the creation of unmanned water map.Use 3D laser to build environment map.(4)In the 3D laser SLAM algorithm,the map-optimized construction method is used,and the information obtained by GPS / IMU fusion is encapsulated as an odom edge and integrated into the graph.The error is obtained by solving the loopback edge with the 3D laser to obtain the unmanned ship height.Accurate pose estimation,and realize the creation of a large-scale outdoor high-precision 3D point cloud map.According to the actual working area of the unmanned ship,the three-dimensional point cloud map is converted into a two-dimensional grid map to prepare for the unmanned ship navigation.(5)Relying on high-precision environment maps and unmanned ship positio ning,and independently designing the tracking reducer based on the unmanned ship’s dynamic characteristics,together with global path planning and local path planning,to carry out mixed path planning for unmanned ships,and Adopt the strategy of fuzzy arrival point,set multiple target points for the unmanned ship,complete the autonomous cruise on the surface of the unmanned ship.