To　ensure　that　the　dual-arm　robots　can　complete　the　coordination　work　excellently,　this　paper　studies　the　practical　tracking　control　problem　for　the　object　position　with　output　constraints.　Firstly,　position　tracking　error　is　transformed　through　a　designed　constraint　function.　By　guaranteeing　that　the　new　error　state　is　bounded,　the　original　output　constraint　can　be　satisfied,　whose　advantage　is　that　it　can　avoid　meeting　the　feasibility　conditions　when　combined　with　barrier　Lyapunov　function　and　backstepping　technology.　Secondly,　the　newly　formed　error　state　variable　is　transformed　by　an　exponential　function　to　improve　the　tracking　accuracy.　Then,　a　robust　adaptive　control　scheme　is　proposed,　which　can　make　the　trajectory　tracking　error　of　the　object　achieve　the　specified　accuracy　and　the　output　is　always　kept　within　the　constraint　range.　Furthermore,　through　Lyapunov　stability　theory,　it　is　proved　that　the　tracking　error　of　positions　can　converge　to　a　smaller　expected　value.　Finally,　numerical　simulations　indicate　the　good　performance　of　the　designed　controller.