Dynamic　VAR　compensation　can　effectively　improve　the　voltage　stability　in　the　environment　of　the　increased　induction　loads　and　renewable　energy　penetration　with　the　rapid　development　of　industry　and　economy.　This　paper　proposes　a　multi-objective　optimization　for　dynamic　VAR　planning　based　on　the　power　system　partitioning　method.　Firstly,　a　novel　power　system　partitioning　approach　considering　the　electrical　distance　and　transient　characteristics　is　proposed　to　reduce　the　dimensions　of　installation　locations　and　capacity　of　STATCOM.　Then,　the　optimal　candidate　buses　installed　the　STATCOMs　are　determined　by　the　voltage　control　index　at　every　zone.　Thirdly,　a　multi-objective　dynamic　VAR　planning　model　is　constructed　with　three　sub-objectives:　1)　total　planning　cost;　2)　L　index;　3)　transient　voltage　severity　index　TVSI.　Finally,　a　multi-objective　group　search　optimizer　with　an　adaptive　covariance　matrix　and　chaotic　search　(MGSOACC)　is　developed　to　solve　the　aforementioned　multi-objective　problem.　A　modification　of　the　IEEE　39　bus　system　with　a　large　number　of　inductive　loads　is　proposed　to　verify　the　proposed　approach.　©　2023　IEEE.