Trajectory Planning Of Wheeled Mobile Robet On A Moving Platform Based On Differential Flatness

The classical wheeled mobile robot is a kind of typical nonholonomic system,such as the unicycle,car-like robot,n-trailer system and so on.It is also an important research object in the nonlinear control systems.The object of this thesis is to study the wheeled robot systems on a moving platform.Different from the classical wheeled mobile robot system,the constraints of this type of systems are in affine nonholonomic forms.Consequently,the corresponding kinematic models can be expressed as nonlinear control-affine systems.In this thesis,the mathematical model.differential flatness analysis,trajectory planning and optimal trajectory planning based on differential flatness of wheeled robot systems on moving platform will be studied.Kai studied firstly the nonholonomic constraints and the kinematic model of a coin on rotating table and then studied its motion planning problem.The author finds that Kai confused the relative movement and the absolute movement of the coin and consequently,the kinematic model that he got is not correct.First,based on the analysis of the error appeared in the kinematic model of a coin on rotating table given by Kai,the correct affine nonholonomic constraints and the kinematic model of this system are proposed.Then the differential flatness property of this system is proved and one of its flat outputs is given.And the corresponding flatness-based point-to-point trajectory generation problem is studied and the simulation results by MATLAB demonstrate the efficiency of the proposed method.Secondly,the w-trailer system on a moving platfom is analyzed and its kinematic model is established.Moreover,we studied the special case thatn= 2 and the platform rotate at a constant speed,that is,Car-like robot on a rotating table.After proving its differential flatness property and deducing one of its flat output,the flatness-based optimal poirnt-to-point trajectory planning problem is studied and resolved.The essential idea is to transform the optimal point-to-point trajectory planning problem into a Nonlinear programming problem.Finally the simulation are presented by MATLAB to demonstrate the efficiency of the proposed method.