The　rapid　development　of　rail　transit　in　big　cities　caues　the　increase　of　passenger　flow　year　by　year,　and　hub　stations　have　become　the　main　distributing　center　of　massive　passenger　flow,　leading　to　potential　safety　hazard　of　passenger　flow　congestion.　The　purpose　of　the　paper　is　to　study　on　the　dynamic　identification　method　of　passenger　flow　congestion　risk　in　rail　transit　hub　station.　Based　on　the　real-time　data　of　AFC　(Automatic　Fare　Collection)　system,　the　passenger　flow　of　rail　transit　hub　station　was　extracted,　and　the　passenger　flow　was　divided　into　three　types:　arrival　passenger　flow,　departure　passenger　flow　and　transfer　passenger　flow　based　on　the　analysis　of　time-varying　characteristics,　choosing　passenger　flow　of　three　types　as　evaluation　indexes.　Then　the　operation　time　(5:00　to　23:00)　of　rail　transit　was　divided　into　72　periods　of　time　with　the　smallest　unit　in　15　minutes,　and　the　evaluation　model　of　passenger　flow　congestion　risk　in　rail　transit　hub　station　was　established　based　on　grey　clustering,　realizing　dynamic　identification　to　passenger　flow　congestion　risk　level　of　rail　transit　hub　station　in　different　periods　of　time.　Finally,　the　method　was　applied　to　evaluate　the　passenger　flow　congestion　risk　of　Dongzhimen　hub　station,　verifying　the　validity　of　the　method.　©　2017　IEEE.