The　kinematic　properties　of　human　lower-limb　arthrosis,　the　gait　pattern　characteristics　of　human　up-stand　walking,　as　well　as　the　influences　of　the　kinematical　difference　between　human　arthrosis　and　machine　joint　on　human-machine　consistence　are　investigated　in　detail.　According　to　the　functional　demand　analysis　of　lower-limb　rehabilitative　training　and　by　adding　linking　joint　into　the　human-machine　connective　chains,　a　novel　exoskeleton　mechanism　used　to　detect　the　motion　information　of　lower-limb　sciatic　and　knee　arthosises　is　proposed.　The　motion　mapping　between　human　lower-limb　arthrosises　and　exoskeleton　mechanism＇s　joints　is　established,　and　kinematics　simulation　of　the　human-machine　closed　kinematic　chain　is　presented.　Results　show　that　the　motions　between　lower-limb　arthrosises　and　exoskeleton　mechanism＇s　joints　have　quantificational　relationships,　and　the　motion　information　of　human　lower-limb　arthrosises　corresponding　to　human　up-stand　walking　can　be　detected　from　the　sensors　installed　at　the　joints　of　exoskeleton　mechanism.