In　this　paper,　based　on　an　improved　radial　basis　function　(RBF)　kernel　extreme　learning　machine　(ELM)　optimized　by　simulated　annealing　algorithm,　a　novel　intelligent　fault　diagnosis　approach　for　photovoltaic　(PV)　array　is　proposed.　Firstly,　three　common　PV　array　faults　are　analyzed　in　detailed.　And　then,　the　ELM　is　proposed　to　automatically　detect　the　faults　of　PV　array.　Moreover,　simulated　annealing　(SA)　algorithm　is　exploited　to　optimize　the　parameters　of　RBF-ELM　model.　Finally,　a　simulation　experiment　is　carried　out　to　verify　the　proposed　SA-RBF-ELM　and　the　result　shows　that　the　proposed　SA-RBF-ELM　approach　can　quickly　and　accurately　identify　the　typical　PV　faults　including　short　circuit,　aging　and　partial　shadow.　(C)　2016　The　Authors.　Published　by　Elsevier　Ltd.