To　improve　system　efficiency　and　reduce　the　risk　of　working　life　reduction　caused　by　the　adverse　operation　conditions　of　fuel　cell,　an　adaptive　Pontryagin＇s　minimum　principle　(PMP)　energy　management　strategy　with　the　fuel　cell　performance　degradation　consideration　was　proposed　for　the　fuel　cell/supercapacitor　hybrid　power　system　in　the　city　bus.　The　off-line　PMP　energy　distributions　of　the　fuel　cell/supercapacitor　hybrid　power　system　in　five　different　driving　conditions　were　analyzed,　and　the　relation　between　the　initial　value　of　the　costate　variable　and　the　difference　of　state　variable　of　the　energy　management　system,　i.e.,　the　state-of-charge　difference　of　the　supercapacitor,　between　the　beginning　and　the　end,　was　obtained.　The　corresponding　relation　between　the　initial　value　of　the　costate　for　online　PMP　and　state-of-charge　could　be　interpolated.　Integrating　with　the　normal　equation　of　the　PMP,　the　instant　costate　variable　was　obtained,　and　thus　the　adaptively　updating　costate　variable　in　the　online　PMP　leading　to　keep　up　state-of-charge　could　be　formulated.　Choosing　the　fuel　cell　performance　degradation,　the　power　change　rate,　number　of　start-stop,　and　maximal　power　of　the　fuel　cell　as　the　constraints　of　the　energy　management　system,　the　fuel　cell　power　change　rate　was　further　formulated　as　a　penalty　term　in　the　cost　function　of　the　adaptive　PMP　to　meet　the　energy　management　constraints　as　well　as　achieve　better　fuel　economy.　The　controller　hardware-in-the-loop　(HIL)　simulation　test　was　carried　out　to　validate　the　practical　efficacy　of　the　proposed　energy　management　strategy.Research　results　show　that,　under　the　bus　condition　SC03　and　New　York　City　cycle　condition　(NYCC)　that　are　different　from　the　costate　variable　formulation　used　typical　operation　conditions,　the　adaptive　PMP　energy　management　can　make　the　terminal　state-of-charge　approach　the　target　value,　and　compared　with　the　off-line　PMP,　the　losses　of　fuel　economy　are　only　1.27%　and　0.94%,　respectively.　Under　the　China　bus　driving　cycle　(CBDC)　comprehensive　test　condition　that　is　different　from　the　operation　conditions　used　for　the　costate　determination　and　cost　function　weighting　adjustment,　the　proposed　adaptive　optimal　energy　management　strategy　can　implement　the　energy　distributions　of　fuel　cell　and　supercapacitor　under　the　provided　constraints,　and　the　fuel　economy　can　keep　up　90.76%　compared　with　the　off-line　optimization.　Under　the　CBDC　and　NurembergR36　test　conditions,　the　average　errors　between　HIL　and　numerical　simulations　are　less　than　5%.　Consequently,　the　proposed　adaptive　optimal　energy　management　strategy　considering　the　performance　degradation　of　fuel　cell　can　implement　the　high　operation　of　the　hybrid　power　system　with　the　potential　of　a　long　working　life.　©　2022,　Editorial　Department　of　Journal　of　Traffic　and　Transportation　Engineering.　All　right　reserved.