Research On The Navigation Environment Perception System Of Inland River Unmanned Surface Vehicles Based On Multi-sensor

With the rapid development of deep learning and artificial intelligence,unmanned surface vehicle(USV)will become an important development direction in the field of ship automation and intelligence.However,the current research on the USV is in its infancy,and the study of USV sailing under the complex navigation environment of the inland river is more slow.One of the main reasons for this dilemma is that the USV cannot make an accurate and real-time perception of the surrounding environment during the navigation.In this paper,a navigation environment perception system for Inland River USV based on mult-sensor is proposed in view of the existing problems in the environment information perception of USV.The main contents are as follows:(1)A navigation environment perception system platform based on multi-sensor for Inland River USV is designed,which includes three parts: Multi-sensor perception subsystem,ship state perception subsystem and ship auxiliary subsystem.The working principle,parameters,working connection mode and data processing method of each sensor are analyzed,which lays the foundation for the subsequent data acquisition and processing.(2)The environmental perception information processing route based on 4G solid radar and laser radar is analyzed.For 4G solid radars,OpenCV is mainly used to perform a series of processing on radar images to remove noise and extract obstacle information.For lidar data,obstacle information is mainly processed according to the features,combined with filtering,adaptive distance threshold clustering and outlier set elimination.For the radar image data and lidar data after processing,the raster processing method is used to further extract the information of the obstacles.(3)A navigation environment perception system software based on multi sensor is designed and developed.And the operation modes and functions of each module in the system software are introduced in detail.(4)For the proposed multi-sensor-based inland river USV navigation environment perception system,performance tests were conducted in the experimental environment.Firstly,the experimental environment and the experimental platform are introduced,and the stability of the ship is adjusted in the experimental environment and the coordinate system of the experimental map is calibrated.Then,each sensor was subjected to a sensing experiment in an actual scene,and the performance of the sensor was tested and analyzed.Finally,the navigation security of the system platform is analyzed in the experimental environment,and the environment perception system is tested.The actual experimental results show that the proposed multi-sensor-based USV environment perception system can combine the advantages of the two sensors,improve the ship's environmental information perception capability and accuracy,and improve the navigation safety of USV to a certain extent.