When　an　emergency　occurs　in　the　city,　a　large　area　of　road　congestion　usually　occurs.　Therefore,　it　is　particularly　important　to　provide　an　effective　emergency　path　planning　strategy　for　vehicles.　However,　existing　emergency　path　planning　methods　do　not　take　into　account　the　connectivity　characteristics　of　the　road　network　and　commuting　capacity.　For　this　purpose,　a　Grid　Map　Emergency　Path　Planning　(GMEPP)　framework　based　on　a　novel　model　Grid　Road　Network　(GRN)　is　designed　in　this　paper.　First,　the　road　network　data　is　divided　into　grids　under　equal　spacing　bands,　and　the　roads　data　divided　into　different　grids　and　use　the　commuting　capacity　of　each　road　as　the　weight　of　each　edge　in　the　grid.　Then　a　Grid　PageRank　(GPR)　algorithm　will　be　introduced,　the　output　value　of　this　methodology　is　calculated　based　on　the　capacity　and　number　of　connected　edges　of　all　vertices　pointed　by　the　external　grid　in　each　grid.　The　higher　value　of　the　grid　will　be　recommended　to　users　first　when　the　path　is　planning.　According　to　the　GRN　model,　an　improved　Bidirectional　Dijkstra　will　be　applied　to　query　the　shortest　path　between　two　points,　which　is　called　Gird　Bidirectional　Dijkstra　(GBD).　At　last,　GMEPP　uses　Reverse　Contraction　Hierarchies　(RCH)　and　Multiple　Reverse　Contraction　Hierarchies　(MRCH)　originality　methodologies　based　on　intersection　type　to　speed　up　the　query　algorithm　GBD.　To　compare　the　efficiency　of　the　proposed　method,　this　paper　conducted　extensive　experiments　to　verify.　The　results　of　the　test　showed　that　the　Gird　Bidirectional　Dijkstra　Multiple　Reverse　Contraction　Hierarchies　(GBD-MRCH)　is　better　than　other　methods　in　different　grid　distributions.