Based　on　extreme　learning　machine　(ELM),　an　adaptive　neural　control　method　for　a　class　of　uncertain　continuous-times　multiple-input-multiple-output　(MIMO)　affine　nonlinear　dynamic　systems　is　developed.　The　ELM　for　signal-hidden　layer　feedforward　neural　networks　(SLFNs)　can　analytically　determine　the　output　weights　of　SLFNs　and　randomly　choose　the　parameters　of　hidden　nodes,　which　provide　good　generalized　performance　at　extremely　fast　learning　speed.　In　the　proposed　control,　ELM　is　employed　to　approximate　the　plant’s　unknown　nonlinear　functions　and　additive　robustifying　control　terms　are　used　to　compensate　for　approximation　errors.　To　obtain　an　adaptive　controller　not　depending　on　any　parameter　initialization　conditions　and　to　relax　the　requirement　of　bounding　parameter　values,　all　parameter　adaptive　laws　for　the　controller　and　robustifying　control　term　are　derived　based　on　Lyapunov　stability　analysis　so　that　semi-global　stability　of　the　closed-loop　system　can　be　guaranteed　and　the　tracking　error　can　asymptotically　converge　to　a　small　neighborhood　of　around　zero.　The　performance　of　the　proposed　approach,　which　is　compared　with　that　of　the　existing　methods　such　as　Radial　basis　function(RBF)　neural　adaptive　controller　under　the　same　condition,　is　demonstrated　through　simulation　example.　Experiment　results　confirm　that　the　designed　ELM　controller　shows　better　performance　validity.　©　2019,　Chinese　Institute　of　Electrical　Engineering.　All　rights　reserved.