As　an　effective　means　to　improve　the　efficient　use　of　energy,　multi-energy　coupled　systems　have　indi-cated　the　important　direction　for　the　development　of　Energy　Internet　in　the　foreseeable　future.　However,　the　uncertainties　that　exist　in　various　energy　networks　present　severe　challenges　to　the　implementation　of　this　innovative　system.　In　this　regard,　this　paper　proposed　a　multi-stage　robust　optimization　model　for　the　coordinated　operation　of　an　electricity-gas-transportation　coupled　system,　which　simultaneously　considered　the　uncertainties　of　traffic　demands,　wind　power,　and　gas　fuel　consumption　by　gas-fired　units.　To　conform　to　the　non-anticipative　sequential　decision-making　process　for　power　system　sched-uling,　the　multi-stage　robust　unit　commitment　(UC)　problem　was　established　and　considered　as　the　master　problem,　which　was　solved　by　introducing　a　linear　affine　rule　and　the　robust　counterpart　method.　The　gas　network　operation　feasibility　checks　in　each　period　were　then　addressed　as　the　subproblems,　and　the　generalized　benders　decomposition-based　decoupling-coordination　strategy　was　developed　to　solve　the　above　master-subproblems　iteratively.　Simulation　results　demonstrate　that　the　proposed　multi　-stage　robust　UC　model　can　achieve　a　scheduling　solution　with　superior　economic　performance　and　reduced　conservativeness　compared　to　the　two-stage　robust　UC　model.　Furthermore,　the　effectiveness　of　the　decoupling-coordination　solving　strategy　was　verified　for　the　multi-energy　coupled　system.　(c)　2021　Elsevier　Ltd.　All　rights　reserved.