The paper presents a predefined-time Vector-Field-Orientation (VFO) control law for unicycle-like nonholonomic mobile robots. We consider a set-point control problem in the presence of strict time constraints, which has to guarantee satisfaction of a prescribed upper bound of the settling time for the configuration errors. The control law is based on the VFO methodology, which is characterized by non-oscillatory and well-predictable time evolution of transient states for unicycle-like robots. A formal stability analysis based on the Lyapunov theory is provided for the closed-loop dynamics. Then, the results of extensive numerical simulations as well as experimental tests illustrate the resultant control performance where time constraints are considered. The proposed approach is compared with an alternative predefined-time stabilizer, recently introduced for nonholonomic chained-form systems.
This article presents a fixed-time vector-field-orientation (VFO) control law for unicycle-like nonholonomic mobile robots. We consider a set-point control problem in the presence of control inputs constraints and task-execution time constraints. The control law is based on the VFO methodology, which is characterized by nonoscillatory and well-predictable time evolution of transient states for unicycle-like robots. A formal stability analysis for the closed-loop dynamics is provided based on Lyapunov theory, and moreover, a method for a priori estimation of the upper bounds of the convergence time, in the presence of control input constraints, is presented. Finally, the control performance is illustrated by results of numerical simulations and experimental tests.
The aim of this work is to present an automatic solution to control the surveillance camera merely by the movements of the operator’s head. The method uses convolutional neural networks that work in a course-to-fine manner to estimate head orientation in image data. First, the image frame of the operator’s head is acquired from the camera on the operator’s side of the system. The exact position of a head, given by its bounding box, is estimated by a Multitask Cascaded Convolutional Network. Second, the customized network for a given scenario is used to classify the orientation of the head-on image data. In particular, the dedicated image dataset was collected for training purposes and was given a discrete set of possible orientations in the vertical and horizontal planes. The accuracy of the estimators is higher than , with an average of fps of validation time. Finally, the current head orientation data are converted into a control signal for two degrees of freedom surveillance camera mounting. The feedback response time is s, which is sufficient for most real-life surveillance applications.