Rail　transit　network　(RTN)　has　become　the　core　of　urban　public　transport　system.　Because　RTN＇s　topological　structure　has　become　complicated　with　its　scale,　the　phenomenon　of　overload　operation　frequently　appears　at　stations　or　lines,　reducing　the　stability　of　RTN　in　operation.　There　is　a　need　to　identify　the　key　stations　and　to　analyze　the　vulnerability　of　RTN　in　depth.　This　paper　quantitatively　analyzes　statistical　topology　parameters　of　Beijing　rail　transit　network　(BRTN)　based　on　complex　network　theory　(CNT).　The　key　stations　are　evaluated　by　different　evaluation　indicators,　such　as　node　degree,　betweenness,　and　strength.　Then,　a　model　is　proposed　to　analyze　the　cascading　failures　of　weighted　BRTN　considering　loading　and　redistribution　of　multi-static　passenger　flow　based　on　coupled　map　lattices　(CML).　Results　show　that　when　the　external　perturbation　is　larger,　the　time　of　all　stations　failure　is　earlier　once　the　failure　is　triggered.　Moreover,　the　perturbation　threshold　for　the　cascading　failure　of　weighted　BRTN　is　determined.　Transfer　regularities　of　the　risk　stations　in　BRTN　are　researched　under　different　combinations　of　the　topological　coupling　coefficients　and　the　flow　coupling　coefficients.　Under　attack,　the　threshold　of　loop　line　damaged　is　smaller　than　a　straight　line.　And,　while　attacking　the　loop　line,　the　discrete　degree　of　peak　time　and　all　station　failure　time　are　higher　under　different　perturbations,　meaning　that　failure　is　hard　to　control.　This　study　provides　recommendations　on　developing　strategies　for　RTN　design　and　performance　evaluation.　Future　studies　will　focus　on　the　coordination　of　weighted　complex　public　transit　networks　other　than　passenger　flow　redistribution　within　RTN.