Optimal Force Distribution And Robustness Analysis Of Wheeled Mobile Robot Based On Multi-objective Optimization

Force distribution and robustness analysis is an integral part of research field of wheeledmobile robot research in rough terrain. A set of quasi-static force analysis method of forcedistribution is proposed based on three-dimensional force of the wheeled robot on the roughground surface. Analysis has been conducted for the wheeled mobile robot on rough groundwith consideration of the maximum allowable shear force and the internal friction angle of theterrain. The algorithm tries to avoid excessive slip. A simulation study in Matlab software iscarried out in a typical three dimensional terrain environment with regard to the powerconsumptionof motorsand forces of robotwith related constraints.Multi-objective optimization method is used to obtain the solution of the simultaneousequations of under-mined quasi-static force distribution in terms of robot’s traction and energyconsumption. Terramechanics model is introducedbetween the wheel and the groundincluding different types of constraints.Simulation is employed on the typical three-dimensional terrain model. Forcedistribution emerged as a major concern in the case of autonomous robot due to its importanceof control in roughterrain. A kind of multi-objective evolutionary algorithm such asminimum-maximum approaches was used to obtain trade-off solution set, which minimizedmotor energy along with the maximization of the smallest meanequivalent friction coefficientwhichcan ensure the smallest slippage.Robustness towards Pareto optimal solution set was analyzed as a major concern.Furthermore different Pareto fronts were obtained with different percentage of noise inducedinto the terrain characteristics mainly caused by soil cohesion (c), shear deformation modulus(k) and internal friction angle (o) respectively. It is found that the noise intensity from terraincharacteristics namely c, k and o of>2%impacts greater in the solution quality should betaken as less robust solution and can be implemented to the control algorithm to designcontrol system design and it can be a metric to select one non-dominated solutions among thePareto solutions set along the Pareto fronts.