With　the　increasingly　prominent　traditional　fossil　energy　crisis　and　environmental　pollution　problems,　renewable　distributed　generation　(RDG)　has　been　widely　used　by　virtue　of　its　characteristics　of　resource　conservation　and　environmental　friendliness.　In　order　to　reduce　the　influence　of　uncertainty　of　the　output　of　wind　power　and　photovoltaic　in　the　distribution　network　planning,　a　multi-scene　interval　optimization　model　based　on　an　improved　K-Means　clustering　method　and　NSGA-II　algorithm　is　proposed.　By　clustering　the　output　curves,　the　typical　output　interval　scene　set　of　RDG　is　constructed,　and　the　interval　power　flow　of　distribution　network　is　calculated　separately.　The　tolerance　of　the　uncertainty　for　RDG　is　defined.　The　multi-objective　optimization　model　is　established　based　on　NSGA-II　algorithm　aiming　at　minimum　system　network　loss,　minimum　voltage　deviation　and　best　tolerance　of　uncertainty.　Finally,　the　result　of　the　IEEE　33-node　test　system　verified　the　correctness　and　effectiveness　of　the　proposed　model　and　method.　©　2019　IEEE.