Conference Paper Position and Attitude Control of Two-wheeled Mobile Robot using Multilayer Minimum Projection Method

木村, 駿介  ,  Kimura, Shunsuke  ,  中村, 文一  ,  Nakamura, Hisakazu  ,  伊吹, 竜也  ,  Ibuki, Tatsuya  ,  三平, 満司  ,  SAMPEI, MITSUJI

For robotics navigation, artificial potential functionsare commonly utilized. Among these functions, controlLyapunov functions (CLFs) guarantee stability of nonlinear autonomoussystems. Particularly, a vehicle such as a two-wheeledmobile robot is a major application for a navigation problem.However, it is difficult to stabilize due to the its nonholonomicconstraint. This paper presents a controller design procedurebased on a non-smooth CLF. The controller achieves asymptoticstability of the origin of the two-wheeled mobile robot in acomplex workspace. For the complex workspace, the multilayerminimum projection method can generate a CLF. The methodrequires a CLF for an unconstrained system and a smooth mapping.This paper proposes a composite mapping constructedby smooth mappings. The composite mapping is adopted tobe combined with a non-smooth CLF. The non-smooth CLFmanages theoretical difficulty from the nonholonomic constraintof the two-wheeled mobile robot. The approach is validated incomputer simulation. The result demonstrates the effectivenessof the presented method.

Number of accesses :  

Other information