Unplanned　subway　interval　disruptions　caused　by　equipment　failures　frequently　occur　in　major　cities　around　the　world.　It　is　crucial　to　ensure　that　passengers　can　complete　their　subsequent　travel　with　less　delay.　Using　bridging　buses　to　evacuate　passengers　is　currently　the　most　common　response　strategy　used　by　management,　but　in　actual　operation　there　are　problems　with　unclear　passenger　demand　and　inefficient　connections.　To　address　this　gap,　we　developed　a　passenger　travel　behavior　decision-making　model　based　on　cumulative　prospect　theory　and　an　emergency　bus　bridging　model.　The　models　serve　two　primary　purposes:　(1)　to　identify　the　origin-destination　of　affected　passenger　flow　within　the　bridging　zone　during　disruptions;　and　(2)　to　propose　an　integrated　emergency　bus　bridging　strategy　that　takes　passenger　demand　into　account.　The　solution　efficiency　of　the　proposed　strategy　is　enhanced　using　a　simulated　annealing-improved　genetic　algorithm.　We　validate　the　model　through　a　case　study　of　Fuzhou　Subway　Line　1　in　China,　followed　by　a　sensitivity　analysis　examining　variables　such　as　the　number　of　emergency　buses,　passenger　demand,　subway　disruption　recovery　time,　and　total　passenger　travel　time.　The　results　indicate　that　the　proposed　strategy　significantly　improves　evacuation　efficiency,　allowing　passengers　to　incur　lower　travel　costs　compared　to　baseline　strategies.