Design Of Multi-rotor UAV Control System Based On Multi-sensor Fusion

In an environment with stable outdoor satellite signals,UAVs can perform autonomous flight tasks such as material distribution and street patrols by using the GNSS positioning system.However,in some indoor or surrounding buildings with more complex environments,the flight of the UAV mainly relies on professional pilots for operation.Therefore,the design of a UAV control system that can achieve fixed-point hovering and autonomous flight in a variety of environments can meet the needs of UAV development,which is important for improving the safety of UAV flight and the expansion of application fields significance.This paper improves the flight control system independently developed by the laboratory,and designs a UAV control system based on multi-sensor fusion,which enables UAVs to achieve autonomous flight functions in environments with large changes in satellite signal strength.The specific work content is as follows:(1)Implementation of SLAM-based positioning algorithm: Cartographer algorithm based on 2D lidar and ORB-SLAM algorithm based on binocular camera are used to complete the position estimation in indoor environment respectively,and the position estimation accuracy of each algorithm is verified in indoor environment.(2)Multi-sensor fusion system design: The fusion system designed in this paper adopts one two-level fusion architecture.In the first level fusion,the fusion node 1 uses an adaptive weighting algorithm to estimate the position of the UAV in multiple environments,and handles the failure and recovery of the GNSS receiver;the fusion node 2 uses the Kalman filter scheme to extend the various speed data processed at the time are fused to obtain the speed information of the UAV.The second-level fusion is to use the extended Kalman filter to fuse the estimated position and speed in the first-level fusion with the inertial navigation data to estimate the real-time status information of the UAV.(3)Flight controller design: Aiming at uncertain factors such as battery voltage changes and load changes,this paper uses a reference model-based sliding mode control algorithm to design a speed controller.The control performance is determined by the design of the reference model,and the optimal theory is used for hyperplane design,Improve the robustness of the speed controller and target tracking performance.At the end of this article,the UAV control system based on multi-sensor fusion proposed in this article is verified through actual flight experiments in indoor and outdoor environments.The experimental results show that the system designed in this paper can complete the autonomous flight mission stably and accurately under the environment where the signal strength of the satellite changes greatly,and achieves the expected design goal.