In　this　paper,　we　present　a　new　stochastic　model　on　the　basis　of　the　switching　problem　of　two　modes　in　the　ARZ　traffic　flow　model　-　Markov　jump　traffic　flow　model.　The　model　is　a　continuous　time　model,　it　can　accurately　capture　the　state　switching　of　the　freeway　segment　at　different　times　and　describes　the　evolution　of　traffic　flow　on　the　freeway.　Compared　with　the　deterministic　ARZ　model,　the　model　proposes　the　switching　of　two　modes　satisfy　Markov　stochastic　processe　and　uses　Markov　state　transition　rate　matrix　to　describe　this　processe,　which　improve　the　accuracy　of　the　model.　In　the　last　part　of　the　paper,　some　NGSIM　data　are　substituted　into　the　Markov　jump　traffic　flow　model　for　numerical　evolution,　the　reliability　and　accuracy　of　the　model　are　verified　by　comparing　with　the　actual　data.