In this paper, a receding horizon (RH) controller is developed for tracking control of a nonholonomic mobile robot. The control stability is guaranteed by adding a terminal-state penalty to the cost function and constraining the terminal state to a terminal-state region. The stability analysis in the terminal-state region is investigated, and a virtual controller is found. The analysis results show that the RH tracking control has simultaneous tracking and regulation capability. Simulation results are provided to verify the proposed control strategy. It is shown that the control strategy is feasible.

This paper presents a differential game approach to formation control of mobile robots. The formation control is formulated as a linear-quadratic Nash differential game through the use of graph theory. Finite horizon cost function is discussed under the open-loop information structure. An open-loop Nash equilibrium solution is investigated by establishing existence and stability conditions of the solutions of coupled (asymmetrical) Riccati differential equations. Based on the finite horizon open-loop Nash equilibrium solution, a receding horizon approach is adopted to synthesize a state-feedback controller for the formation control. Mobile robots with double integrator dynamics are used in the formation control simulation. Simulation results are provided to justify the models and solutions.

In this paper, we propose DeepSLAM, a novel unsupervised deep learning-based visual Simultaneous Localization and Mapping (SLAM) system. The DeepSLAM training is fully unsupervised since it only requires stereo imagery instead of annotating ground-truth poses. Its testing takes a monocular image sequence as the input. Therefore, it is a monocular SLAM paradigm. DeepSLAM consists of several essential components, including Mapping-Net, Tracking-Net, Loop-Net and a graph optimization unit. Specifically, the Mapping-Net is an encoder and decoder architecture for describing the 3D structure of the environment while the Tracking-Net is a Recurrent Convolutional Neural Network (RCNN) architecture for capturing the camera motion. The Loop-Net is a pre-trained binary classifier for detecting loop closures. DeepSLAM can simultaneously generate pose estimate, depth map and outlier rejection mask. We evaluate its performance on various datasets, and find that DeepSLAM achieves good performance in terms of pose estimation accuracy, and is robust in some challenging scenes.

In this paper, we investigate a leader-follower flocking system where few members are group leaders who have global knowledge (a desired trajectory), while majority of the members are group followers who can communicate with neighbors but do not have global knowledge. The followers do not even know who the leaders are in the group. The flocking group is able to track a specific trajectory led by group leaders. In this system, all group members estimate the position of flocking center by using a consensus algorithm via local communication in order to keep the flocking group connected. Based on the estimated position of flocking center, a leader-follower flocking algorithm is proposed, and its stability is proved. A group of real robots ldquowifibotsrdquo are used to test the feasibility of the algorithm. Experiments show that this leader-follower flocking system can track the desired trajectory led by group leaders.

This paper focuses on the problem of moving target tracking with a group of mobile robots. Each robot in the group has a pan/tilt camera to detect the target and has limited communication capability to communicate with neighbor robots. The problem is solved by separating it into two parts. One part is the estimation of target position and another is the flocking control of multiple robots moving toward the estimated position. In the target estimation part, we propose to use a novel distributed Kalman filter to estimate the target position. The distributed Kalman filter is deduced based on a standard Kalman filter by modeling the neighbor's information as one of measurements. In the motion control part, a distributed flocking algorithm is developed. It is used to track the estimated target and avoid collision. In both parts, only local communication between neighbor robots is required. Finally, the tracking algorithms are simulated with 2-D and 3-D robots to verify their performance. The simulation results provide a firm conclusion that the proposed algorithms are able to track a moving target. A group of real ground mobile robots is used to test the proposed algorithm. The experiment results show that multiple robots are able to cooperate to track the target under the proposed algorithms and the tracking result outperforms the result produced by individual robots without cooperation.