Modeling and multiobjective optimization of traction performance for autonomous wheeled mobile robot in rough terrain  

作　　者：Ozoemena Anthony ANI He XU Yi-ping SHEN Shao-gang LIU Kai XUE 

机构地区：[1]College of Mechanical and Electrical Engineering,Harbin Engineering University

出　　处：《Journal of Zhejiang University-Science C(Computers and Electronics)》2013年第1期11-29,共19页浙江大学学报C辑（计算机与电子（英文版）

基　　金：Project supported by the National Natural Science Foundation of China(No. 60775060);the Natural Science Foundation of Heilongjiang Province of China (No. F200801);the Specialized Research Fund for the Doctoral Program of Higher Education (Nos. 200802171053 and 20102304110006);the Harbin Science and Technology Innovation Talents Special Fund (No. 2012RFXXG059),China

摘　　要：Application of terrain-vehicle mechanics for determination and prediction of mobility performance of autonomous wheeled mobile robot （AWMR） in rough terrain is a new research area currently receiving much attention for both terrestrial and planetary missions due to its significant role in design, evaluation, optimization, and motion control of AWMRs. In this paper, decoupled closed form terramechanics considering important wheel-terrain parameters is applied to model and predict traction. Numerical analysis of traction performance in terms of drawbar pull, tractive efficiency, and driving torque is carried out for wheels of different radii, widths, and lug heights, under different wheel slips. Effects of normal forces on wheels are analyzed. Results presented in figures are discussed and used to draw some conclusions. Furthermore, a multiobjective optimization （MOO） method for achieving optimal mobility is presented. The MOO problem is formulated based on five independent variables in- eluding wheel radius r, width b, lug height h, wheel slip s, and wheel rotation angle 0 with three objectives to maximize drawbar pull and tractive efficiency while minimizing the dynamic traction ratio. Genetic algorithm in MATLAB is used to obtain opti- mized wheel design and traction control parameters such as drawbar pull, tractive efficiency, and dynamic traction ratio required for good mobility performance. Comparison of MOO results with experimental results shows a good agreement. A method to apply the MOO results for online traction and mobility prediction and control is discussed.Application of terrain-vehicle mechanics for determination and prediction of mobility performance of autonomous wheeled mobile robot(AWMR) in rough terrain is a new research area currently receiving much attention for both terrestrial and planetary missions due to its significant role in design,evaluation,optimization,and motion control of AWMRs.In this paper,decoupled closed form terramechanics considering important wheel-terrain parameters is applied to model and predict traction.Numerical analysis of traction performance in terms of drawbar pull,tractive efficiency,and driving torque is carried out for wheels of different radii,widths,and lug heights,under different wheel slips.Effects of normal forces on wheels are analyzed.Results presented in figures are discussed and used to draw some conclusions.Furthermore,a multiobjective optimization(MOO) method for achieving optimal mobility is presented.The MOO problem is formulated based on five independent variables including wheel radius r,width b,lug height h,wheel slip s,and wheel rotation angle θ with three objectives to maximize drawbar pull and tractive efficiency while minimizing the dynamic traction ratio.Genetic algorithm in MATLAB is used to obtain optimized wheel design and traction control parameters such as drawbar pull,tractive efficiency,and dynamic traction ratio required for good mobility performance.Comparison of MOO results with experimental results shows a good agreement.A method to apply the MOO results for online traction and mobility prediction and control is discussed.

关 键 词：Autonomous wheeled mobile robot （AWMR） Terramechanics TRACTION Motion control Multiobjectiveoptimization （MOO） Genetic algorithm （GA） 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]