Formation Control Of Mobile Robots Research Articles

Formation Control of Mobile Robots with Active Obstacle Avoidance

In this paper, the formation control and obstacle avoidance problems are dealt with a unified control algorithm, which allows the follower to avoid obstacle while maintaining desired relative bearing or relative distance from the leader. In the known leader-follower robot formation control literature, absolute motion states of the leader robot are required to control the followers, which may not be available in some environments. In this research, the leader-follower robot formation is modelled and controlled in terms of the relative motion states between the leader and follower robots. The absolute motion states of the leader robot are not required in the proposed formation controller. Furthermore, the research has been extended to a novel obstacle avoidance scheme based on sensing the relative motion between robot and obstacle. Experimental investigation has been conducted using the platform consisted of three nonholonomic mobile robots and computer vision system, and the results have demonstrated the effectiveness of the proposed methods.

Formation control of multiple nonholonomic mobile robots

AbstractTracking control problem of multiple mobile robots is considered. Our system is composed of a reference and follower robots of unicycle type. The robots are assumed to satisfy pure rolling and nonslipping conditions, which lead to nonholonomic constraints. The purpose is to control the followers so that the reference is tracked with arbitrary desired clearance and also to avoid interrobot collisions. To accomplish this goal, we first introduced a virtual robot (VR) tracking control to establish and maintain the formation when no collision is detected. VR is an ideal robot fixed with each follower and helps in simplifying the control, as the tracking error becomes zero in the final stage. During the process, if the possibility of collision is detected, a collision avoidance technique (l−ψ or l−l control) will be applied to the lower priority robot. The result of this control gives a monotonic convergence in an internal shape of distance variables, which enables us to predict the movement of the robots during this control and leads to collision‐free movement. These processes are repeated until the desired motion is accomplished. Simulation results prove the efficiency of our control techniques. © 2006 Wiley Periodicals, Inc. Electr Eng Jpn, 157(3): 81–88, 2006; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/eej.20192 Copyright © 2006 Wiley Periodicals, Inc.

Formation Control of Mobile Robots

A constructive method is presented to design cooperative controllers that force a group of N mobile robots to achieve a particular formation in terms of shape and orientation while avoiding collisions between themselves. The control development is based on new local potential functions, which attain the minimum value when the desired formation is achieved, and are equal to infinity when a collision occurs. The proposed controller development is also extended to formation control of nonholonomic mobile robots.

Vision-based formation control of mobile robots

In this paper, a formation control algorithm and an obstacle avoidance control algorithm for mobile robots are developed based on a relative motion sensory system such as a pan/tilt camera vision system, without the need for global sensing and communication between robots. This is achieved by employing the velocity variation, instead of actual velocities, as the control inputs. Simulation and experimental results have demonstrated the effectiveness of the proposed control methods.

Multiagent-Based Multi-team Formation Control for Mobile Robots

In the field of formation control, researchers generally control multiple robots in only one team, and little research focuses on multi-team formation control. In this paper, we propose an architecture, called Virtual Operator MultiAgent System (VOMAS), to perform formation control for multiple teams of mobile robots with the capabilities and advantages of scalability and autonomy. VOMAS is a hybrid architecture with two main agents. The virtual operator agent handles high level missions and team control, and the robot agent deals with low level formation control. The virtual operator uses four basic services including join, remove, split, and merge requests to perform multi-team control. A new robot can be easily added to a team by cloning a new virtual operator to control it. The robot agent uses a simple formation representation method to show formation to a large number of robots, and it uses the concept of potential field and behavior-based control to perform kinematic control to keep formation both in holonomic and nonholonomic mobile robots. In addition, we also test the stability, robustness, and uncertainty in the simulation.

Following the flock

A new approach to formation control of nonholonomic mobile robots equipped with central panoramic cameras is presented. The approach uses motion segmentation techniques to estimate the position of each leader and omnidirectional visual servoing for tracking and collision avoidance. The paper showed that direct feedback-linearization of the leader-follower dynamics leads to asymptotic tracking but suffers from degenerate configurations. A nonlinear controller was presented that avoids such singularities but can only guarantee input-to-state stability of the formation.

Formation Control of Multiple Nonholonomic Mobile Robots

Tracking control problem of multiple mobile robots is considered. Our system is composed of a reference and follower robots of unicycle type. The robots are assumed to satisfy pure rolling and non-slipping conditions, which lead to nonholonomic velocity constraints. The purpose is to control the followers so that the reference is tracked with arbitrary desired clearance and also to avoid an inter-robot collision. To accomplish this goal, we first introduce a virtual robot (VR) tracking control to establish and maintain the formation when no collision is detected. VR is an ideal robot fixed with each follower and helps in simplifying the control, as the tracking errors become zero in final stage. During the process, if a possibility of collision is detected, a collision avoidance technique (l-ψ or l-l control) will be applied to the lower priority robot. The result of this control gives a monotonic convergence in an internal shape of distance variables, which enables us to predict the movement of the robots during this control and leads to the collision-free movement. These processes are repeated until the desired motion is accomplished. Simulation results prove effciency of our control techniques.