A Research On Pose Estimation And Control Technology Of Multi-rotor UAVs

In recent years,multi-rotor UAVs have been successfully operated in both military and civil fields in consequence of their attractive advantages,such as low cost,low noise,strong maneuverability and convenient maintenance,what's more,they are faced with an even broader prospect of application.In this paper,a series of studies are conducted with the focus on the key problems of navigation and control of the multi-rotor UAV.The quaternion differential equation is applied to accomplish the attitude update process,and the EKF attitude determination algorithm is given on basis of the low-cost MEMS gyroscope,accelerometer and compass;The Gauss-Newton method is used to construct the quaternion-formed observations of the Kalman filter,and the calculating procee is further optimized to get the reduced-order Jacobian matrix,which significantly improves the computational efficiency.In this paper,the strapdown inertial navigation equation and the INS updating algorithm in the local navigation frame is studied,and the INS error equation in the local navigation frame is derived,after which the distance-formed INS error equation is given using the distance error of the vehicle from the local positon to the origin.Finally,the Kalman filter model of the loosely-coupled integrated navigation algorithm with multi sensors is constructed,so that we can get the optimal navigation estimation of the UAV.In this paper,the inner loop controller of the muti-rotor UAV is developed based on the robust Servomechanism LQR control theory,which includes the attitude stabilization controller and the heading controller.On this basis,the classical control theory and the integral backstepping control theory were applied to design the outer loop controller of the hexa-copter,respectively,which includes the height stabilization controller and the horizontal trajectory tracking controller.A hardware scheme of triplex hot-redundancy flight control computer based on a distributed architecture is presented,the schematic circuit diagram as well as the PCB diagram is designed based on the low-cost MEMS devices.Finally,the performance of attitude determination algorithm,the integrated navigation algorithm,the attitude stabilization controller and the trajectory tracking controller are tested and verified on the self-designed hardware platform.