In　the　merge　areas　of　highway,　vehicles　often　misjudge　the　traffic　situations　of　themselves　and　their　adjacent　lanes,　so　it　is　easy　to　cause　problems　such　as　sudden　stops　or　collisions,　which　can　lead　to　traffic　jams.　Advances　in　Connected　and　Automated　Vehicles　allow　us　to　build　a　computational　framework　for　real-time　control　and　coordination　of　vehicles.　This　paper　establishes　a　decentralized　optimal　control　framework　to　solve　the　congestion　in　two　adjacent　merge　areas,　so　as　to　reduce　the　travel　time　in　merge　areas　and　fuel　consumption.　Firstly,　we　determine　the　priority　of　vehicles　on　different　roads　and　coordinate　the　order　and　time　of　vehicles　passing　through　the　merge　areas.　Then　we　get　the　analytic　solution　of　the　optimal　control　without　collision　by　adding　an　interior　point　constraint.　The　simulation　results　show　the　effectiveness　of　the　proposed　method　by　comparing　with　baseline　scenario.