A　novel　method　of　voltage　sag　source　location　based　on　the　correlation　of　multipoint　positive　sequence　voltage　and　the　typical　pattern　matching　optimization　is　proposed　to　make　the　location　results　independent　of　fault　type　and　transition　resistance.　Firstly,　it　builds　the　fault　pattern　using　the　positive　sequence　voltage　variation　at　measuring　points　and　defines　the　pattern　similarity　based　on　the　Euclidean　distance　measure　and　the　Spearman　rank　correlation　coefficient.　Then　it　forms　the　set　of　typical　patterns　by　setting　up　a　preset　fault　at　the　middle　of　each　branch　and　clusters　it　into　several　subgroups　according　to　the　similarity　of　the　measuring　sequence　and　typical　patterns.　It　chooses　the　corresponding　lines　of　the　subgroup　with　the　highest　similarity　to　make　up　the　primary　fault　line　set,　which　can　effectively　narrow　the　searching　space.　Finally,　an　optimal　estimation　model　is　built　to　find　the　optimal　matching　fault　pattern.　The　object　function　is　to　maximize　the　similarity,　the　primary　fault　line　and　fault　distance　are　chosen　as　optimization　variables.　The　model　is　solved　by　particle　swarm　optimization　algorithm,　fault　line　and　fault　position　can　be　obtained.　The　numerical　examples　of　IEEE　30　and　IEEE　33　system　show　the　feasibility　and　validity　of　the　proposed　method.　The　results　show　that　the　method　can　effectively　eliminate　the　influence　of　fault　type　and　transition　resistance　and　has　higher　location　accuracy　than　previous　methods.　©　2017,　The　editorial　office　of　Transaction　of　China　Electrotechnical　Society.　All　right　reserved.