This　study　aims　to　analyze　the　cascading　failure　process　of　traffic　network　in　tourist　attraction　under　unexpected　emergency　events,　and　discuss　the　measures　to　improve　traffic　network　invulnerability.　For　that　purpose,　taking　tourist　attraction　as　the　research　object,　the　topology　model　of　traffic　network　in　tourist　attraction　is　established　based　on　Space　L　method.　After　different　types　of　emergency　events　are　simulated　as　different　attack　strategies,　we　discuss　the　spatial　and　temporal　distribution　characteristics　and　evolution　process　of　tourism　emergency　events.　The　cascading　failure　model　of　traffic　network　based　on　load-capacity　is　constructed,　then　the　main　factors　affecting　the　scale　of　dynamic　cascading　failures　are　given　and　their　sensitivity　are　analyzed.　Taking　the　Summer　Palace　in　Beijing　as　an　example,　the　invulnerability　analysis　of　the　traffic　network　is　carried　out　and　the　node　protection　strategies　with　different　network　load　are　proposed.　The　study　reveals　that　(a)　the　traffic　network　of　the　Summer　Palace　has　a　typical　network　characteristics　of　small　world;　(b)　network　load,　node　capacity,　node　attack　strategy,　adjacent　nodes　relationship,　load　distribution　rule　significantly　affect　the　scale　of　cascading　failure　of　the　traffic　network　in　tourist　attraction,　(c)　when　the　network　load　coefficient　$\delta　＜　0.7$　,　the　cascading　failure　rate　(CFR)　can　be　effectively　controlled　within　0.2　to　avoid　large-scale　cascading　failure;　(d)　the　scale　of　cascading　failure　can　be　effectively　reduced　by　22.14,　40.91,　and　63.66,　after　increasing　the　capacity　of　the　nodes　which　are　the　top　5,　10,　and　20　in　the　ranking　of　CFR,　respectively.