While　bidirectional　automatic　fare　gates　(BAFGs)　serve　as　important　tools　for　mitigating　the　congestion　of　passengers　in　subway　stations,　most　of　BAFGs　function　merely　as　single-directional　automatic　fare　gates　(AFGs)　in　most　cities　of　China　and　other　countries.　This　paper　proposes　a　pioneering　work　on　the　dynamic　regulation　problem　of　BAFGs.　We　build　a　macro-to-micro　model　which　simulates　the　movement　of　passengers　in　stations　from　both　the　macroscopic　and　microscopic　perspectives.　The　system　dynamics　based　macromodel　simulates　the　macroscopic　movement　of　passengers　in　stations,　while　the　social　force　based　micromodel　simulates　the　microscopic　movement　of　passengers.　Benefitting　from　the　negative　feedback　regulation　mechanism　of　the　system　dynamics,　we　adjust　different　number　of　BAFGs　according　to　the　congestion　extent　of　the　passenger　flow　on　the　platform　in　the　macromodel,　and　change　the　pedestrian　flowline　in　the　micromodel　to　ensure　that　there　is　no　new　conflict　produced　by　the　regulation　scheme.　We　conduct　case　studies　on　Dawanglu　station　and　Jiangtai　station,　Beijing,　China,　and　propose　optimized　configurations　and　dynamic　regulation　scheme　of　BAFGs　accordingly.　The　simulation　results　show　that　the　macro-to-micro　model　not　only　significantly　improves　both　the　BAFGs＇　efficiency　and　the　passengers＇　efficiency,　but　also　exhibits　a　high　generalization　ability　towards　different　types　of　subway　stations.　Our　study　benefits　the　subway　operation　and　management,　and　is　conducive　to　sustainable　urban　development.