Positioning Method for Unknown Confined Spaces Based on Tightly Coupled Iterative Kalman Filtering

In an unknown confined space environment with limited communication and perception, the self-positioning of unmanned aerial vehicles (UAVs) will be restricted, seriously affecting the subsequent mission execution. The fusion of multiple sensor perception information of UAVs is an effective method to improve the positioning and mapping accuracy. However, the discontinuous perception during the interaction process of UAVs with the environment and the dynamic and aperiodic changes of positions and postures at various locations in space introduce a large amount of noise to the positioning of UAVs, especially having adverse effects on the accurate measurement of sensors such as lidar and inertial measurement unit (IMU). In this paper, by applying the tightly coupled iterative Kalman filtering method to fuse the perception information of the onboard lidar and IMU with the vehicle kinematics to estimate the state of UAVs, the positioning accuracy of UAVs can be effectively improved.