Adaptive Image-Based Moving-Target Tracking Control of Wheeled Mobile Robots With Visibility Maintenance and Obstacle Avoidance
Shi-Lu Dai,Jianjun Liang,Ke Lu,Xu Jin
DOI: https://doi.org/10.1109/tcst.2023.3331553
IF: 4.8
2023-01-01
IEEE Transactions on Control Systems Technology
Abstract:This article presents an adaptive image-based visual servoing (IBVS) for nonholonomic wheeled mobile robots to solve the moving-target tracking problem in obstacle environments. A pinhole camera that is equipped with the following robot to monitor the target’ motion is limited field of view (FOV), which relates to the issue of visibility maintenance. Safety navigation is another concern, which requires that the following robot can avoid collisions with the target and static obstacles. Under the IBVS framework, we design novel constrained boundary functions based on pixel coordinate, which can be deviated away from zero to ensure that the following robot drives away from obstacles because obstacle avoidance is a higher priority rather than the tracking task. When there is no obstacle detected, the constrained boundary functions are taken as exponentially decaying functions of time. Using fixed-time stability and control Lyapunov synthesis, the tracking errors are shown to converge in fixed time to a small neighborhood of the desired obstacle-avoidance trajectory generated from the centerline of constrained boundaries while guaranteeing visibility maintenance and obstacle/collision avoidance. The proposed fixed-time IBVS controller (FTIBVSC) only depends on locally relative information acquired by onboard sensors without the need of knowing the feature height and target’s velocity. Simulation and experiment studies are carried out to show the efficacy of the proposed FTIBVSC.
automation & control systems,engineering, electrical & electronic