Research On Motion Control And Coordination For Autonomous Mobile Robots

An autonomous mobile robot is an integrated system which incorporates functions of environmental perception, decision making and planning, and behavior controlling and execution. Due to its ability of locomotion, mobile robots have been widely used in many fields, like industrial manufacturing, national defense and space exploration. Motion control, as a fundamental problem in the research of mobile robots, is a prerequisite for mobile robot to complete other complex tasks. Motion control refers to designing a control law using the state feedback so that the robot can track a desired trajectory or reach a specified position. In addition, with the expansion of the application field and scope of mobile robots,the applications of multiple mobile robots to complete some complex tasks have promising prospect. Motion coordination of multiple mobile robots aims to define the motion directives for each robots of the team to achieve missions in a coordinated way. In this thesis,the motion control coordination of multiple mobile robots is studied with nonlinear control theory and Lyapunov stability theory. The main body of the thesis can be divided into two parts, the point stabilization and trajectory tracking problem of mobile robots under the nonholonomic constraints is studied in the first part. The second part addresses the formation control and coordinated target tracking problem for multiple mobile robots. The main contributions and innovations of the thesis are listed as follows:1) To simultaneously solve the stabilization and tracking problem for nonholonomic mobile robots, smooth time-varying controllers at the kinematics are proposed to simultaneously stabilize a given reference point or track a given reference trajectory. The controllers are developed with the aid of a delicately designed time-varying signal and Lyapunov method. Asymptotic convergence of regulation or tracking errors is achieved by the proposed controller without switching. Our approach provides an interesting way to unify the existing results on point stabilization and trajectory tracking of mobile robots. Results of simulations and experiments on a wheeled mobile robot are presented to demonstrate the effectiveness of the proposed controller.2) To address the simultaneous stabilization and tracking problem as torque level, an adaptive backstepping control strategy is proposed for nonholonomic mobile robots with parametric uncertainties. The backstepping design technique is used to derive the torque control law from the kinematic control law, and an adaptive control law is design to deal with the unknown parameters. The stability and convergence analysis are presented us-ing Lyapunov tools, and the effectiveness of the proposed control strategy is demonstrated through simulations and experiments.3) An adaptive neural network control system for tracking control of nonholonomic robotic systems in the presence of uncertainties and disturbances is developed using the backstepping design and RBF(Radial Basis Function) neural networks. First, a virtual velocity command for kinematic model is presented, and then backstepping design method is adopted to develop a torque law. In order to deal with uncertainties, a RBF neural network identifier is utilized to learn the dynamics of nonholonomic system on-line, and a compensation controller to eliminate the influence of approximation error and external disturbances so that the asymptotically convergence of the tracking error can be guaranteed. The proposed controller is non-regressor based, and requires no information on system dynamics.4) To preserve network connectivity in formation control of mobile agents under limited sensing ranges, distributed potential field-based controllers are proposed for agent group without and with a virtual leader to achieve the desired formation and meanwhile maintain the connectivity of distance-based communication graph. In the proposed strategy, the network connectivity constraint is modeled as imaginary barriers in the free space, and each agent’s configuration is represented as a point in a potential field that combines attraction to the desired formation, and repulsion from barriers. Agents move in the direction of the negative gradient of the potential fields, and thus stay within the interior of feasible or safe area. In this way, the communication graph remains connected throughout the evolution if it is initially connected.5) The coordinated standoff tracking problem which involves multiple nonholonomic robots and a moving target is addressed. The goal for the multi-robot system is to circumnavigate the moving target with prescribed radius, circular velocity, and inter-robot angular spacing. Two distributed control strategies respectively in local Cartesian coordinates and polar coordinates are proposed using backstepping techniques, in which some virtual signals are introduced to facilitate the control design. Asymptotic stability has been established under some mild assumptions on the velocity of the target. Explicit stability and convergence analyses are presented using Lyapunov tools, and effectiveness of the proposed approach are verified by the simulations and experiments.Finally, the main contributions and innovations of this thesis are summarized and the further works are prospected.