Abstract: A dynamical model of multi-rotor UAVs was established, and a positioning controller of multi-rotor UAVs in an indoor environment was designed. In order to improve the accuracy of positioning, the specific tag was set to provide visual data for the UAV in interior space. Fusing the inertial data and the height data obtained from sensors of the UAV with visual data through the Extended Kalman Filter, and the UAV’s position was obtained from the estimation of flight state in real time. Simulation results show that the error of the UAV’s position reduces significantly after fusing the visual data, and the positioning accuracy of the UAV improved a lot, which shows that the controller works well.

Key words: dynamical model, visual data, extended kalman filter, data fusion, positioning controller