The　performance　of　predictive　energy　management　(PEM)　strategy　depends　greatly　on　the　precise　prediction.　Inevitably,　the　imprecise　prediction　occurs　under　changeable　driving　conditions.　This　study　proposes　an　online　correction　predictive　energy　management　strategy　to　address　the　above　issue.　The　first　is　development　of　a　PEM　using　the　sequential　quadratic　programming　(SQP)　combined　with　back　propagation　neural　network-based　velocity　prediction.　Second,　a　correction　algorithm　based　Fuzzy　Neural　Network　(FNN)　is　designed　to　evaluate　the　correction　factor.　The　Q-learning　based　Swarm　Optimization　(QSO)　algorithm　is　implemented　to　optimize　the　parameters　of　the　FNN　and　evaluate　accurately　the　correction　factor.　Finally,　in　combination　with　the　above　efforts,　a　novel　correction　algorithm　based　on　the　SQP　backup　control　strategy　and　integrated　the　QSO　algorithm　with　the　FNN　has　been　further　established　accordingly.　The　numerical　validation　results　demonstrate　that　the　superior　performance　of　the　proposed　strategy　in　economic　compared　with　the　benchmark　strategies,　and　the　effectiveness　of　the　proposed　strategy　is　validated　by　Hardware-in-the-loop　(HIL)　experiments.　Both　the　numerical　validation　and　HIL　results　indicate　that　the　combination　of　QSO　and　FNN　made　it　possible　to　develop　the　online　correction　algorithm　capable　of　significantly　further　improving　the　hydrogen　consumption　of　the　fuel　cell　electric　vehicle.　(c)　2021　Elsevier　Ltd.　All　rights　reserved.