Research On Motion Control In The Complex Environment For The Wheeled Mobile Robots With Uncertainties

Motion control of robots uses real-time feedback environment and their own statusinformation by sensors to achieve robot trajectory tracking or stabilization control tasks.In recent years, wheeled mobile robots have vast applications in real life such securityand surveillance, dangerous sites inspection, other planetary exploration, family service,and so on. Different from traditional industrial robots, in the actual complexenvironments, wheeled mobile robots have very complicated uncertainty andenvironmental interferences. Therefore, the feedback control laws based on the accuratemodel have a lot of limitations in the practical applications, and thus the correspondingcontrol laws should be re-designed in a completely different way. Therefore, how todesign motion control systems for wheeled mobile robots in presence of variousuncertainties in complex environments, is still a challenging problem in the fields ofcontrol and robotics.This dissertation focuses on the motion control of the wheeled mobile robots inpresence of various uncertainties, including kinematic model parameter uncertainties,dynamic model parameter uncertainties, external disturbances, and wheel slippinginterferences in actual environments. Specially, based on kinematic model and dynamicmodel of the wheeled mobile robots with uncertainties in absence of slipping, theadaptive tracking controllers are designed respectively. In absence of slipping, on onehand, on the basis of kinematic model an adaptive sliding mode controller based ondisturbance observer is presented for selecting the appropriate PID-type sliding surface,on the other hand, the tracking controller of the mobile robot is designed byBackstepping technique and nolinear digital filter technology based on the dynamicmodel. In presence of slipping, when longitudinal slipping occurs, correspondingmobile robot kinematics model is established, Backstepping technique and Lyapunovdirect method is used to design trajectory tracking controller of the wheeled mobilerobots and adaptive updating laws of the unknown slipping parameters. When bothlongitudinal slipping and lateral slipping happen at the same time, based on thekinematics model of mobile robot with unknown slipping parameters, the smoothtime-varying control algorithm and on-line parameters adjusting method are proposedbased on nonlinear sliding mode observer.In the end, unified controller is designed tosolve both tracking and obstacle avoidance for wheeled mobile robots in environments with obstacles.The main works of this dissertation are as follows:①Adaptive tracking controller for wheeled mobile robots with uncertainties inabsence of slipping based on kinematic model.When the mobile robot works in complex environments, the influences of theparameters perturbation and external disturbances on performance of the system isinevitable. Based on the uncertain nonlinear kinematic model of wheeled mobile robots,an adaptive sliding model control method is used to design a controller for trajectorytracking of wheeled mobile robots with unknown parameter variations and externaldisturbances. The total uncertainties of robot system are estimated online by improvedlinear extended state observer. After nonlinear kinematic model of wheeled mobilerobots is transformed by input-output transformation technology of state feedbackcontrol theory, the adaptive sliding model controller with switching gain is adjustablereal-time online is developed for selecting the appropriate PID-type sliding surface. Theconvergence of the complete equations of motion of wheeled mobile robots is proved bythe Lyapunov stability theory. Moreover, the simulation results show that the proposedmethod greatly compensates the effects of parameter perturbation and externaldisturbances and improves the system tracking accuracy and robustness, in comparisonwith traditional sliding mode control laws.②Adaptive tracking controller for wheeled mobile robots with uncertainties inabsence of slipping based on dynamic model.Based on the dynamics model of wheeled mobile robots, an adaptive Backsteppingmethod is used to design a controller for trajectory tracking of wheeled mobile robotswith parameter uncertainties. The adaptive control laws based on state feedback aredeveloped with Backstepping approach for selecting the appropriate Lyapunov functionstep by step. The convergence of the complete equations of motion of wheeled mobilerobots is proved by the Lyapunov stability theory. Moreover，the simulation resultsshow that the proposed method greatly compensates the effects of parameterperturbation and improves the system tracking accuracy and robustness, in comparisonwith traditional PID control laws.③Adaptive tracking control of wheeled mobile robots in presence of longitudinalslipping.Based on wheeled mobile robots with unknown longitudinal slip parameters, anadaptive control strategy for a tracked mobile robot is presented, in which the longitudinal slip of the left and right wheels are described by two unknown parameters.It is assumed that the kinematic model of the tracked robot is approximated by the oneof a differential wheeled mobile robot. An adaptive nonlinear feedback control law thatcompensates for the longitudinal slip is proposed to achieve a given trajectory trackingobjective. Asymptotic stability of the close-loop system is ensured using an appropriateLyapunov function. The controller gains are determined online by poles placementmethod. Numerical results show the effectiveness of the proposed approach.④Adaptive tracking control of wheeled mobile robots with unknown longitudinaland lateral slipping parameters.Considering the wheeled mobile with longitudinal slipping and lateral slipping, onthe basis of③,an adaptive control approach is proposed for trajectory tracking ofwheeled mobile robot with unknown longitudinal and lateral slipping. A kinematicmodel of tracked wheeled mobile robot is established in this paper, in which bothlongitudinal and lateral slipping is considered and processed as three time-varyingparameters. Sliding Mode Observer is then introduced to real time estimate the slippingparameters online. A stable tracking control law for this robot system is proposed bybackstepping method and the asymptotic stability is guaranteed by Lyapunov theory.Meanwhile, the controller gains are determined online by poles placement method.Simulation results show the effectiveness and robustness of the proposed method.⑤The unified adaptive controller is designed to solve both tracking and obstacleavoidance for the wheeled mobile robots with unknown sliding in environments withobstacles.On the basis of④, an adaptive control approach is proposed for trajectory trackingand obstacle avoidance of mobile robots considering unknown longitudinal slipping andlateral slipping. Based on the kinematic model of wheeled mobile robots, sliding modeobserver is introduced to estimate the slid parameters online. A stable tracking controllaw for this nonholonomic system is proposed to compensate the unknown sliding effect.From Lyapunov-stability analysis, it is proved regardless of unknown sliding thattracking errors of the controlled closed-loop system are asymptotically stable, thetracking errors converge to the zero outside the obstacle detection region and theobstacle avoidance is guaranteed inside the obstacle detection region. The efficiencyand robustness of the proposed control system are verified from simulation results.In conclusion, the motion control methods are studied deeply for wheeled mobilerobots with uncertainties in complex environments, moreover, systematic discussions of the modeling approaches, on-line identification of sliding parameters, design of trackingcontroller, stability analysis are conducted, and in the end, several adaptive trackingcontrollers are proposed and theoretical analysises and simulation results validate theeffectiveness of the work.