Robust Image-Based Visual Servoing for Accurate Target Trajectory Tracking of Omnidirectional Mobile Robot With a Single Camera
提出一种基于图像的视觉伺服方法，利用单相机使全向移动机器人精确跟踪目标轨迹，通过自适应积分滑模观测器补偿不确定性，无需GPS或复杂位姿估计。
We propose an image-based visual servoing (IBVS) method to accurately track the trajectory of a target using a single camera mounted on an omnidirectional mobile robot (OMR). Existing IBVS methods for target tracking mainly focused on keeping the target centered in the camera’s field of view, hindering accurate trajectory tracking of the target along curves. To overcome this limitation, the proposed IBVS method for target trajectory tracking introduces a novel approach for generating the desired feature points while considering practical uncertainties caused by slipping, uncertain interaction matrix, and target motion. Specifically, an adaptive integral sliding mode observer (AISMO) is proposed to compensate for uncertainties, eliminating the limitations of previous integral sliding mode observers (ISMOs) that require prior knowledge of the maximum uncertainty values. In addition, using the AISMO estimates, an integral sliding mode control (ISMC) is proposed for robust IBVS of the OMR to guarantee finite-time convergence of the trajectory tracking error to zero. Notably, the proposed IBVS method aids the OMR in accurately tracking the target trajectory using a single camera instead of global external sensors (e.g., global positioning system (GPS) receivers) or applying complex pose estimation techniques. The tracking performance of the proposed method is demonstrated through Lyapunov stability analysis and confirmed with simulation and experimental results.