水上应急救援无人船关键技术应用及实验研究OA

In order to ensure the timeliness and reliability of water emergency rescue and improve the intelligent level of water emergency rescue equipment,an unmanned ship system for water emergency rescue is proposed.Firstly,the identification module of people who fell into the water is proposed,which is based on the single-stage detection algorithm of deep learning,and YOLOv5 is selected as the network model of rescue target detection,which can achieve accurate target detection,and the location algorithm of pinhole camera model is used to locate the location of people who fall into the water.On this basis,the autonomous navigation module is studied,and the autonomous navigation and obstacle avoidance of the emergency rescue unmanned ship are realized through the LOS guidance algorithm,and a better tracking effect is achieved through the trajectory tracking experiment.Finally,the automatic rescue module is designed,when the lifebuoy is thrown by the life-saving system within 1 meter from the target point.The experimental results show that it can meet the needs of emergency rescue of rivers and lakes and improve the success rate of rescue operations.