FAST GENERATION OF EFFICIENT MOTIONS FOR DIFFERENTIAL DRIVE ROBOTS

摘要: Abstract This paper presents a methodology for generating, in real-time, efficient trajectories differential drive wheeled robots. Given dynamic model of robot, reasonably large training set minimum torque are first obtained various final configurations. The then organized into categories, and principal component analysis performed each category. components used as basis functions rapidly interpolating to arbitrary ending Through case study we examine some the issues underlying numerical optimization classification trajectories, report on performance. is easily generalizable other robotic platforms, well motion data different physical notions efficiency.