The　increase　in　the　number　of　people　with　ankle　injuries　and　the　shortage　of　rehabilitation　physicians　is　an　important　social　problem　at　this　stage.　In　this　regard,　a　combined　active-passive　assisted　rehabilitation　strategy　is　proposed　based　on　a　2-SPU/RR　ankle　rehabilitation　robot.　Firstly,　the　virtual　prototype　model　and　the　corresponding　structural　sketch　are　established,　and　the　kinematic　analysis　is　carried　out　to　solve　the　workspace　of　the　robot.　The　results　show　that　the　robot　is　capable　of　meeting　the　demands　of　plantarflexion-dorsiflexion　and　inversion-eversion　compound　movements　of　the　ankle　joint.　Then,　aiming　at　the　problem　that　the　current　research　content　of　rehabilitation　robot　is　only　passive　rehabilitation,　failing　to　rehabilitate　patients　according　to　their　actual　ankle　injuries,　a　method　of　combining　active　and　passive　rehabilitation　of　the　ankle　robot　is　proposed.　First　preset　the　rehabilitation　movement　trajectory,　the　patient＇s　movement　intention　is　obtained　based　on　the　plantar　pressure　data　and　force　sensing　algorithm.　Robot　follows　actively　through　software　and　hardware　control　system,　the　movement　trajectory　is　recorded　and　datalized　synchronously.　Finally　optimizes　the　trajectory　tracing,　according　to　the　optimal　trajectory　using　a　5-stage　S-type　acceleration　and　deceleration　control　method　for　customized　passive　recovery.　The　active-passive　rehabilitation　system　combines　robotic　following　control　with　active　patient　behavior　perception,　which　helps　physicians　diagnose　and　treat　patients　with　functional　ankle　injuries.　©　2022　Editorial　Office　of　Chinese　Journal　of　Mechanical　Engineering.　All　rights　reserved.