Research On Multi-sensor Data Fusion Methods Of Unmanned Aerial Vehicle Based On Kalman Filtering

Small unmanned aerial vehicles(UAV)are widely used in military and civil fields,because of its advantages of low cost,strong maneuverability and the wide range of applications.In order to realize the autonomous flight and good control of the aircraft,it is necessary to use multiple sensor data fusion techniques to obtain accurate and complete data of the aircraft about motion state.Kalman filter is one of the most commonly used methods in data fusion,this thesis focuses on the multi-sensor data fusion of unmanned aerial vehicles based on Kalman filtering,the main work includes the following parts:Firstly,the basic principle of strapdown inertial navigation is introduced,the strapdown inertial navigation equations are given,the measurement models of inertial sensors and auxiliary navigation sensors including GPS,barometer and magnetometer are established respectively.The simulation system of UAV is set up.At the same time,the calculation methods of the strapdown inertial navigation equations are discussed.Secondly,the basic principle of SINS/GPS integrated navigation system and the correction methods of indirect estimation are introduced.The basic theory of linear Kalman filtering is described Then the error model of strapdown inertial navigation system is derived.On this basis,the filter model is established.Finally,the realization process of data fusion is discussed,and the simulations are carried out under different estimation conditions.The simulation results and theoretical analyses show that when the measurement precision of GPS receiver is worse,the initial error of the navigation parameters is larger and the refresh rate of the GPS receiver is slower,the estimation effects of the linear Kalman filter which is based on the output correction models are also worse.Then,in view of the above problems,the SR-CDKF method which is based on direct correction models is introduced and the filter model is set up.The simulation comparisons with linear Kalman filter are carried out when the estimated conditions were worse.The results indicate that the estimation effects of SR-CDKF method is better than linear Kalman filtering.Finally,the barometer and magnetometer are introduced on the basis of the previous work.The SINS/GPS/barometer scheme and SINS/GPS/barometer/magnetometer scheme are proposed and simulated.The results show that the estimation effects about the motion parameters of the aircraft under the SINS/GPS/barometer/magnetometer scheme has been further improved.