Utilizing　the　islanding　operation　ability　of　the　renewable　energy　generators　to　restore　as　much　power　supply　after　faults,　can　not　only　increase　the　consumption　of　renewable　energy,　but　also　enhance　the　power　supply　reliability　of　the　power　system,　thereby　improve　the　resilience　of　the　power　system.　This　paper　proposes　a　novel　coordinated　restoration　model　to　optimize　the　islanding　partition　stage　for　a　total　or　partial　blackout　system.　Firstly　the　restoration　principles　of　generators　in　the　system　restoration　process　after　fault,　are　proposed　in　terms　of　factors　that　affect　the　restoration　priority.　An　analytic　hierarchy　model　is　then　proposed　to　determine　the　main　generators　for　islands　based　on　the　principles.　Secondly,　an　islanding　model　is　established　to　divide　the　total　or　partial　blackout　system　into　optimal　electricity　supply　islands　with　the　main　generators　obtained　as　root　nodes.　The　islanding　optimization　model　is　solved　using　an　improved　genetic　algorithm.　Finally,　case　studies　are　performed　to　verify　the　effectiveness　of　the　proposed　coordinated　restoration　model　and　solving　method.　©　2019　IEEE.