There　is　weak　source-load　matching　and　inadequate　energy　storage　application　in　a　microgrid.　It　needs　to　coordinate　and　optimize　the　source-load　and　energy　storage　considering　the　demand-side　management　and　carbon　emission.　In　this　paper,　the　optimization　of　a　grid-connected　microgrid　is　divided　into　two　stages　including　demand-side　management　and　dispatching.　First,　in　the　first　stage,　a　demand-side　management　model　is　introduced.　The　energy　storage　battery　is　considered　and　a　crisscross　optimization　algorithm　is　used　to　minimize　the　net　load　of　the　microgrid　and　maximize　the　self-supply　rate.　Then,　in　the　second　stage,　from　the　economy　and　environmental　protection　perspective,　a　wind,　photovoltaic,　micro-gas　turbine,　fuel　cell　and　storage　energy　microgrid　day-ahead　optimal　dispatching　model　is　established　to　minimize　total　costs　and　carbon　emissions.　It　relies　on　the　information　after　the　demand-side　management　in　the　first　stage.　The　above　model　is　analyzed　using　the　improved　multi-objective　gray　wolf　optimization　algorithm.　Finally,　taking　an　actual　microgrid　in　Fujian　as　an　example,　simulation　shows　that　the　introduction　of　demand-side　management　can　effectively　use　energy　storage　batteries　to　improve　the　matching　degree　of　microgrid　source-load　and　fully　tap　the　demand　response　potential.　Compared　with　single-stage　optimization,　two-stage　optimization　is　more　conducive　to　improving　the　penetration　rate　of　renewable　energy,　reducing　the　daily　operating　costs　and　carbon　emissions　of　microgrids,　and　realizing　their　low-carbon　economic　operation.　©　2022　Power　System　Protection　and　Control　Press.　All　rights　reserved.