In　this　paper,　the　problem　of　a　fully　actuated　hexarotor　performing　a　physical　interaction　with　the　environment　through　a　rigidly　attached　tool　is　considered.　A　nonlinear　model　predictive　impedance　control　(NMPIC)　method　is　proposed　to　achieve　the　goal　in　which　the　controller　is　able　to　simultaneously　handle　the　constraints　and　maintain　the　compliant　behavior.　The　design　of　NMPIC　is　the　combination　of　a　nonlinear　model　predictive　control　and　impedance　control　based　on　the　dynamics　of　the　system.　A　disturbance　observer　is　exploited　to　estimate　the　external　wrench　and　then　provide　compensation　for　the　model　which　was　employed　in　the　controller.　Moreover,　a　weight　adaptive　strategy　is　proposed　to　perform　the　online　tuning　of　the　weighting　matrix　of　the　cost　function　within　the　optimal　problem　of　NMPIC　to　improve　the　performance　and　stability.　The　effectiveness　and　advantages　of　the　proposed　method　are　validated　by　several　simulations　in　different　scenarios　compared　with　the　general　impedance　controller.　The　results　also　indicate　that　the　proposed　method　opens　a　novel　way　for　interaction　force　regulation.