Disturbance observer-based visual servoing for multirotor unmanned aerial vehicles

Abstract

This paper presents a disturbance observer based input saturated visual servoing law for a quadrotor unmanned aerial vehicle (UAV). The controller regulates the 4D relative pose, i. e., 3D translational and yaw motion, between the vehicle and a planar horizontal visual target in an environment with external disturbances. A feedforward control is used to compensate the lumped disturbance consisting of both system uncertainties and external disturbances. The feedback control part is based on a nested saturation control, which is used to bound the orientation of the UAV and therefore helps to keep the visual target in the camera’s field of view. Simulation results are provided to demonstrate controller performance.

Nested saturation control of multiple vector integrators and its application to motion control of UAVs

“…Proof The proofs of properties (1), (2), and (3) can be found in the work of Xie et al Consider property (4). When ‖ x ‖ ≤ l i , it holds trivially.…”

“…When ‖ x ‖ ≤ l i , it holds trivially. When ‖ x ‖> l i , the partial derivative of Σ i ( x ) defined in can be written as (see the work of Xie et al) where Using and the property x T Sx =0, we can show …”

“…When ||x|| ≤ l i , it holds trivially. When ||x|| > l i , the partial derivative of Σ i (x) defined in (1) can be written as (see the work of Xie et al 36 )…”

“…In the work of Xie and Lynch, with a backstepping design scheme and a new set of saturation functions, a nested saturation control for visual servoing of UAVs is proposed; however, only a constant disturbance to the vertical motion is considered. This work is extended in the work of Xie et al, where the effect of the disturbance is removed by using an exact disturbance observer.…”

Image-Based Visual Servoing of Rotorcrafts to Planar Visual Targets of Arbitrary Orientation

“…Because of the above-mentioned benefits, in recent years many virtual camera-based DIBVS laws are proposed [18]- [29]. These works adopt the same perspective image moments features as in [16] and propose various control laws to address specific issues, such as system uncertainties [18]- [21], removing the requirement of velocity estimation [19], [21]- [24], field of view (FoV) constraint [25], disturbance rejection [21], [26], tracking of moving targets [27], and global stability of IBVS [28], [29]. These IBVS laws in [18]- [29] are based on the assumption that the visual target lies on a horizontal plane.…”