| The Canadian Mobile Servicing System (MSS) is a 22 DOF robot system that is composed of a base, the SSRMS (Space Station Remote Manipulator System) and SPDM (Special Purpose Dexterous Manipulator System). Current motion planning for MSS is divided into two separate phases: posture planning and path planning. Posture planning is done by trial-and-error, which is highly human-involved and labor-intensive. This thesis treats the two separate phases above as one integrated problem, and develops automatic off-line motion planning methods for MSS to perform repair tasks in the International Space Station, while guaranteeing that the worksite of the manipulators are visible to various cameras mounted on MSS.; We modified existing algorithms to take into account camera constraints, and improved some algorithms to speed up the planning process. These automated methods can solve motion planning for MSS in most tasks in about thirty minutes, leading to great savings in time. |
