Research On Cooperative Consensus For Nonholonomic Wheeled Mobile Robot Systems Over The Network Constraint Environments

With the rapid advance of the network science theory and its applications in the past two decades,the multi-robot coordination under the network environments.as one of novelty form of robot applications,has attracted more and more interest and attention in many fields of science and engineering,in particular the coordinated control of multiple robot ic systems described by non-holonomic Lagrange dynamics has become one of the most significant research topics.On one hand,this is main-ly because it can characterize a large amount of physic al and mechanical objects in complex and integrated production process,such as mobile robots.flexible ma-nipulators and automatic vehicles,where the flexibility,manipulability,reliability and maneuverability are highly desirable or even necessary features.Moreover,theobtained result has a wide range of engineering applications including automatic ve-hicle manufacturing,lunar exploration and manned spaceflight,underwater surveys and deep space exploration,etc.On the other hand.the multi-wheeled mobile robot system over the network constrained environments is viewed as a.highly nonlinear and strongly coupled high-dimensional dynamic system with multi-levels,multi-structure,and multi-scales,which also possess nonholonornic and under-actuated dynamic characteristics.As a consequence,the swarm cooperative control problem for a large-scale nonholonomic mobile robots under complex network constraint on-vironments is more challenging compared with traditional precision motion control forsingle-wheel robots such as fixed-point regulation,trajectory tracking and path plaming.Therefore,the research on coordinated control of networked wheeled mo-bile robot systems formulated by non-holonomic Lagrange dynamics not only is of great scientific significance in theory,but also has very important practical value in engineering application.Motivated by the two key factors of time-delay and stochastic noises in the process of communication interaction and information fusion among agents for col-laborative multi-robot systems under complex network constraint environments,in combination with classical analytical dynamics and modern network science,this dissertation is devotes to study the dynamics and control of swarm cooperative dis-tributed consensus problems of networked wheeled mobile robot systems formulated by nonholonomic Lagrange dynamics from the viewpoint of distributed control the-ory on multi-agent systems.The main work of the thesis can be summarized as the followirg three aspects:1.Adaptive consensus for networked nonholonomic mobile robot systems By aiming at the represent two-wheel drive vehicle with nonholonomic or under-actuated dynamics,in combination with backstepping approach and sliding mode control technique,an adaptive consensus tracking control algorithm by integrating with the kinematic controller and dynamic torque controller is proposed.A unified analysis methodology is then presented for the convergence analysis of the closed-loop system by the Lyapunov function technique.A key feature of the pro-posed adaptive consensus algorithm is appropriately introduced the communication network topology characterizing the information exchange among robots,which is consistent.with the structures and features of the realistic multi-robotic systems.Finally,the correctness of the proposed adaptive algorithm is verified by numeri-cal simulation,and the nonlinear responses of network structures including various network topologies,different coupling strength and connectivity,on the consensus tracking performance of multi-wheeled mobile robots is further analyzed.2.Practical consensus of networked nonholonomic mobile robot sys-tems By focusing on the nonholonomic wheeled mobile robot with input distur-bance in polar coordinates,a practical consensus control algorithm integrating a position controller and a heading controller is proposed based on sliding mode control technique.The proposed practical consensus scheme can make full use of either nonholonomic or underactuated dynamics for wheeled mobile robots to eliminate the geometric constraints of the design sliding mode surface,and thus the system-atical integration of the kinematics and dynamics controllers is realized.It is shown that the developed practical consensus strategy can demonstrates some advantages of control performance including stability,robustness,and effectivencss over the existing control method proposed even for their single-robot.counterparts.To this end,the effectiveness of the proposed practical consensus algorithm is further confirmed by the comparing numerical simulations with the typical network topologies and different inputs.3.Stochastic consensus of networked mobile robot systems:By considering the inevitable phenomena.of time-delay and stochastic noise in the process of communication interaction among robots,an auxiliary time-varying reference tracking velocity describing the local communication interaction among the networked robots is properly presented such that both the communication delay and input delay can be suitably introduced into the design of the kinematics controller for wheeled mobile robots.A mean square consensus tracking algorithm which integrates kinematics and dynamics is proposed,and the corresponding convergence analysis of the closed-loop control systems are then performed based on the well-known developed stochastic delayed Halanay inequality.Moreover,the up-per bound analysis expressions of communication delay and input delay to ensure the stability of the corresponding control algorithm are finally given respectively.Subsequently,the simulation results are provided to demonstrate the theoretical results,and the nonlinear effects of communication delay and input delay on stochastic consensus strategy are further discussed.