| Motion planning algorithms have been of great interest in robotics, yet many developed algorithms have not been used in practice, partly due to the amount of time they require to obtain the desired path for most industrial robot arms. Computational complexity of path planning is exponential in the number of degrees of freedom (dof) of the robot. Therefore, although various fast planners have been developed for low dof robots, few methods offer close to real-time planning for higher dof robots. Use of parallelism is one clear way to speed-up the process. Different parallel formulations for current path planners have been proposed in the literature.; A parallel formulation for an exhaustive enumerative search in high dimensional spaces is developed. The method is based on a dynamic load-balancing scheme, inspired by ideas from Kumar and Rao, and tries to exploit all the processors at their maximum capacity at all times. (Abstract shortened by UMI.)... |
