Attitude Free Position Control of a Quadcopter using Dynamic Inversion

This paper presents a novel controller architecture for a quadcopter. A two-loop controller using dynamic inversion is designed that finally allows direct commands for position. The inner loop is the body-fixed angular rate control and the outer loop is the position control. This derived controller is capable of decoupling the strongly coupled dynamics of the quadcopter and maximizing the transmission bandwidth of the position control without sacrificing the controller robustness, as well as eliminating the singularities caused by the attitude control (e.g. pitch angle at 90 degree). A distinguished difference of this two loop architecture and other designs is that the attitude of the quadcopter (be it Euler angle or quaternion) is not a controlled state. It only implicitly appears in the transformation matrix. Hence the position dynamics are controlled with relative degree 3, and rate control is applied with relative degree 1. Higher bandwidth can be achieved with this design compared with a more common two-loop structure 1-3 (position control with relative degree 2 plus attitude control with relative degree 2). Moreover, the dynamic equation appears in a simpler form and can be easily feedback linearized.In addition, the yaw dynamics is inherently decoupled from position dynamics; hence the heading can be well controlled by the inner loop. Fast update rate is essential for the high bandwidth controller. Control errors due to parameter uncertainties in the dynamic inversion and disturbances can be quickly compensated. Fast and robust sensor data fusion is also one key requirement to achieve high bandwidth of system dynamics. A specific Kalman based state observer is developed to fuse the IMU and vision sensor data, with consideration of the limited onboard memory and processing capacity of the onboard micro-processor. A Vicon system 4 is used to provide accurate position measurements for the state observer, which also accounts for the time delay due to the communication between the Vicon system and the quadcopter.

[1]  Florian Holzapfel,et al.  DYNAMIC INVERSION BASED CONTROL CONCEPT WITH APPLICATION TO AN UNMANNED AERIAL VEHICLE , 2004 .

[2]  Jian Wang,et al.  Markerless, vision-assisted flight control of a quadrocopter , 2010, 2010 IEEE/RSJ International Conference on Intelligent Robots and Systems.

[3]  Roland Siegwart,et al.  Backstepping and Sliding-mode Techniques Applied to an Indoor Micro Quadrotor , 2005, Proceedings of the 2005 IEEE International Conference on Robotics and Automation.

[4]  Florian Holzapfel,et al.  Nichtlineare adaptive Regelung eines unbemannten Fluggerätes , 2004 .

[5]  Holger Voos,et al.  Nonlinear control of a quadrotor micro-UAV using feedback-linearization , 2009, 2009 IEEE International Conference on Mechatronics.

[6]  Eric N. Johnson,et al.  Limited authority adaptive flight control , 2000 .