Design of a Drone System for Large Thermal Power Boiler Inspection

Drones virtually immune to human survival and physiological limitations and physical injuries, and are capable of performing its intended tasks properly in dangerous, contaminated environments. We propose a drone localization and navigation system in the thermal power boiler. Multiple sensor fusion algorithms, such as UWB, lidar, and IMU, are used to locate and establish maps. The system has outstanding advantages in dusty, dark and vertical feature degradation thermal power boiler drone inspection missions. Moreover, considering that there are no complex and dense obstacles within the boiler, this system only designs a local planner to implement autonomous cruise control within the boiler, which supports initial coordinate system calibration and real-time heading control. Finally, we built a simulation platform to test several performances of the system and compared with traditional localization algorithms. The experimental results verify the feasibility of the multi-sensor fusion algorithm and the reliability of the trajectory planning navigation algorithm.