UAV Position And Attitude Measurement Based On Optical Flow And MARG Sensors

Attitude measurement,positioning and navigation are the core technologies of unmanned aerial vehicle(UAV).The combination of three-axis magnetometer,three-axis accelerometer and three-axis gyroscope(MARG)is widely used for three-dimensional attitude measurement of UAV,while the optical flow sensor can detect the motion of UAVs relative to the ground.In order to suppress sensor errors and further improve the measurement accuracy of UAV position and attitude,a Dual vector Discrete time complementary filter(DV-DTCF)is proposed in this paper,which can achieve data fusion between optical flow sensor and MARG sensor,and can also greatly reduce the complexity of attitude calculation.Besides,a discrete time optical flow model and an optical flow/MARG sensor data fusion algorithm based on Cubature Transform(CT)are proposed to improve the accuracy of position and attitude solution.The main research content and innovation points of this paper can be summarized as follows:Firstly,a DV-DTCF and its parameter adjusting method are proposed,The proposed DV-DTCF estimates the gravity and geomagnetic vectors in parallel,so as to implement data fusion and attitude calculation via MARG sensor.The theoretical analysis shows that the DV-DTCF algorithm proposed in this paper is essentially equivalent to the fixed gain Kalman Filter(KF),and then a parameter tuning method for DV-DTCF using the fixed gain KF algorithm and Riccati equation is derived.The experimental results show that the DV-DTCF algorithm can achieve the same estimation accuracy as that of KF,but with much lower time consumption,and thus it is more suitable to run on the single chip microcomputer that has limited computing power.Secondly,the commonly used optical flow sensor model is derived under continuous time condition,and it is not suitable for discrete time system.To solve this problem,a discrete time optical flow measurement model is proposed,which directly gives the relationship between the optical flow measurement value and the linear and angular motion of the UAV in each sampling period,so that it is easier to use the optical flow to improve the accuracy of the UAV position estimation.Finally,an optical flow/MARG sensor data fusion algorithm based on cubature transform is designed to improve the position accuracy of UAV using optical flow data.The effectiveness of this method is verified by UAV flight test.