Multi-Sensor Fusion for Unmanned Vessel Obstacle Avoidance System: Intelligent Navigation for Waters Around Nuclear Power Plants

The aim of this study is to develop an unmanned vessel obstacle avoidance system based on multi-sensor fusion, which is specifically designed for intelligent navigation in complex waters around nuclear power plants. By integrating data from multiple sensors such as LIDAR, sonar, camera, GPS and inertial navigation system (INS), the system is able to comprehensively sense the environment and achieve accurate path planning and real-time obstacle avoidance. Experimental results show that the system is able to efficiently identify and avoid potential threats in the face of both fixed and moving obstacles, with an obstacle avoidance success rate of more than 97% and excellent performance in real-time. Nevertheless, the study also points out room for improvement in sensor fault handling and dynamic environment response. The research results in this paper provide reliable technical support for safety inspection and environmental monitoring in the waters around nuclear power plants and lay the foundation for the application of unmanned vessel technology in high-risk waters.