In　high-speed　milling　process,　chatter　is　often　encountered　due　to　the　self-excitation　mechanism,　which　not　only　limits　the　productivity,　but　also　results　in　inferior　work　piece　quality,　shorter　tool　life,　cutting　noise　and　so　on.　In　order　to　control　chatter,　an　active　chatter　control　method　based　on　model　predictive　control　is　proposed　in　this　work.　Firstly,　an　active　control　structure　is　designed　by　integrating　the　specific　tool　holder,　piezoelectric　actuators　and　sensors　into　the　spindle.　Then,　a　time-varying　model　integrating　the　chatter　dynamics　is　developed　comprehensively　considering　the　designed　structure　and　the　milling　process.　The　model　is　then　transformed　into　a　linear　time　invariant　(LTI)　approximate　model　through　zero　order　approximation　(ZOA)　and　Pade　approximation.　Next,　on　the　base　of　the　model,　a　model　predictive　controller　is　designed　to　control　only　the　chatter　dynamics　caused　by　regenerative　effect.　Finally,　both　simulation　analysis　and　milling　experiments　are　carried　out　to　verify　the　effectiveness　of　the　active　chatter　control　method.　The　results　show　that　the　chatter　is　effectively　controlled　in　the　milling　process　and　the　milling　stability　is　improved　using　the　proposed　active　control　method.　(C)　2019　Elsevier　Ltd.　All　rights　reserved.