Abstract A method for generating the optimal base location of mobile manipulators is presented that considers the workspace characteristics in the presence of obstacles. Using the previously developed methods for generating singularity surface, manipulator workspace is firstly characterized based on task capability. Such a characterization process is realized through set operations on 2n-tree representations of the obstacles and subworkspaces in joint angle space. For a given task specification, optimal base location of mobile robot is then determined by a multidimensional optimal searching method that considers the configuration free space and their connectivity characteristics. The method is applicable to robots with all types of kinematic structures and task spaces of various dimensions. The method is demonstrated in an example for a 3 d.o.f. manipulator.
Park et al. (1995) studied this question.