This　article　proposes　a　reference　optimization-based　compliant　control　framework　for　aerial　pipeline　inspection　using　a　hexacopter　with　a　robotic　contact　device.　The　system　dynamics　is　modeled,　and　the　aerial　interaction　performance　is　quantified　by　a　cost　function,　which　is　regulated　by　the　trajectory　tracking　error　and　the　interaction　force.　Then,　the　iteration-based　reference　learning　scheme　is　introduced　to　minimize　the　cost　function　and　realize　the　optimal　interaction　performance　without　the　knowledge　of　the　contact　environment.　Next,　a　command　filter-based　adaptive　backstepping　trajectory　tracking　controller　is　designed　to　avoid　the　complexity　explosion　problem　as　well　as　implement　stable　and　accurate　tracking　performance　under　external　wrenches.　Simulation　studies　and　real-world　experiments　illustrate　the　feasibility　of　the　proposed　framework　and　provide　promising　results.