This　paper　focuses　on　the　adaptive　fuzzy　practical　predefined-time　(PT)　bipartite　consensus　tracking　control　(BCTC)　problem　for　heterogeneous　nonlinear　multi-agent　systems　(HNMASs)　with　actuator　faults.　First,　the　fuzzy　logic　systems　(FLSs)　are　used　to　approximate　the　unknown　nonlinear　functions.　Then,　the　partial　loss　of　effectiveness　and　bias　fault　of　actuator　are　considered　simultaneously　in　the　HNMASs,　which　is　effectively　handled　by　using　adaptive　compensation　technology.　In　addition,　the　developed　bipartite　consensus　tracking　control　protocol　based　on　a　practical　predefined-time　strategy　not　only　ensures　the　fast　convergence　of　the　studied　systems,　but　also　predetermines　the　convergence　time　not　relied　on　the　initial　conditions.　The　theoretical　result　shows　that　all　signals　of　the　closed-loop　system　are　semiglobally　uniformly　predefined-time　bounded　(SGUPTB),　and　the　BCTC　performance　is　guaranteed　within　the　predefined　time.　Finally,　the　simulation　example　based　on　the　agents　of　different　orders　shows　the　validity　of　the　obtained　results.　IEEE