Redundant　and　dual-arm　manipulators　exhibit　good　flexibility　and　excellent　cooperation　ability,　respectively.　Therefore,　both　these　manipulators　are　used　to　design　a　redundant　dual-chain　manipulator　with　two　kinematic　chains　and　a　fixed　base.　In　addition,　a　motion　planning　method　is　proposed　using　a　manifold　analysis　based　on　the　characteristics　of　this　designed　manipulator　to　resolve　the　difficulty　associated　with　motion　planning　for　a　dual-chain　manipulator　under　multiple　constraints.　Firstly,　the　paths　of　the　two　end-effectors　of　the　dual-chain　manipulator　are　planned　synchronously　by　the　Rapidly-exploring　Random　Tree　Star　(RRT*)　in　the　Cartesian　space.　Then,　joint　trajectories　that　correspond　to　the　paths　are　generated　by　the　manifold　analysis　in　the　joint　space.　In　this　method,　the　manifold　continuity　constraint　is　added　to　the　end-effector　path　planning　of　RRT*　to　ensure　that　the　planned　paths　have　continuous　joint　trajectories.　The　joint　trajectories　are　optimised　by　analysing　the　performance　of　the　manipulator　on　the　manifolds　to　improve　their　motion　performance.　The　prototype　experiments　of　loose　coordination　and　tight　coordination　are　provided　in　the　experimental　section;　these　experiments　prove　the　feasibility　and　practicability　of　the　motion　planning　method　proposed　in　this　study.