Approach for singularity avoidance path planning of parallel robots

In contrast to Serial Robots, Most Parallel Robots have singular configurations in workspace. At singular configuration, both kinematics transformation and force transformation of Parallel Robots become poor, and on the other hand, end-effector become uncontrollable due to gaining one or more extra Degree of Freedom (DOF). In this paper, a path plan algorithm for parallel robots to avoid singularity is presented which is based on classical algorithm A^* and takes condition number and determination value of Jacobian matrix as indices for describing singular configuration and ill-conditioning region. When initial pose and goal pose are specified, the algorithm can plan the shortest continuous path avoiding ill-conditioning region if a valid path exists, and moreover, it can give out a warning in case of the nonexistence of valid path.