A multi objective genetic algorithm approach to a design parameter generation for a robot platform on three omnidirectional wheels. (17th November 2017)

• Record Type: Journal Article
• Authors: Lin, Chyi-Yeu; Gussu, Tesfaye Wakessa; Tsai, Yao-Nien
• Is Part Of: Journal of the Chinese Institute of Engineers. Volume 40:Number 8(2017)
• Page Start: 659
• Page End: 668
• Publication Date: 2017-11-17
• Subjects: Mobile platform -- omnidirectional wheel -- wheelbase -- stability analysis -- optimal design; Technology -- Periodicals; Engineering -- Periodicals; 620.005
• DOI: 10.1080/02533839.2017.1384327
• Languages: English
• ISSNs: 0253-3839

Abstract

Abstract: Superior mobility while ensuring maximum stability is a key requirement for most mobile robots. These requirements are fulfilled if the robot has a well-designed robotic platform that serves as a physical interface between its body and the floor. Although there are different design and development approaches for car-like robotic platforms, there is little design guidance for robot platform on three omnidirectional wheels. Hence, a method for generating optimal design parameters of such robotic platforms is proposed in this work. An objective function relevant to platform stability, the robot mass, and driving torque is formulated to form a min–max optimization problem. Finally, a trade-off between the design variables of the robotic platform is carried out from the result of Multi Objective Genetic Algorithm in Matlab. An omnidirectional wheeled robotic platform is developed to study the stability and mobility analysis of this robotic platform for various maneuver modes. This approach is seen to be effective for generating design variables of omnidirectional three wheels robotic platform that has maximal stability and minimal driving torque while traversing across inclined and flat floors. As a result, this optimization approach can be used to yield better design analysis with reduced reworking cost and manufacturing time.