The　amount　of　lubricating　oil　supply　of　ball　screw　mechanism　is　adjusted　only　by　artificial　experience　in　the　computer　numerical　control　(CNC)　feed　system,　which　lacks　scientific　basis　and　theoretical　guidance.　On　the　one　hand,　the　raceway　of　ball　screw　mechanism　is　worn　frequently,　on　the　other　hand,　lubricating　oil　is　wasted　and　environment　is　polluted.　To　meet　the　requirements　of　green　manufacturing,　realize　reasonable　and　accurate　lubrication,　and　improve　the　accuracy　retention　and　reliability　of　the　ball　screw　mechanism,　it　is　necessary　to　study　intelligent　lubrication　control　method　of　the　ball　screw　mechanism　based　on　the　actual　lubrication　state　online　perception.　Firstly,　the　calculation　methods　of　contact　characteristics,　oil　film　thickness　distribution,　and　friction　coefficient　with　the　vary　of　lubrication　degradation　process　were　established　based　on　the　theory　of　fractal　and　point-contact　elastohydrodynamic　lubrication　(EHL)　theory.　Subsequently,　according　to　the　quasi-static　force　balance　equation　under　the　constant　screw　speed　operating　condition,　the　calculation　method　of　friction　torque　of　the　ball　screw　mechanism　with　the　varies　of　lubrication　state　was　proposed,　and　the　law　of　friction　torque　with　lubrication　degradation　process　in　different　screw　speeds　was　determined.　Finally,　the　online　monitoring　experiment　on　the　lubrication　degradation　process　of　the　ball　screw　mechanism　was　implemented　based　on　the　linear　relationship　between　the　friction　torque　of　the　ball　screw　mechanism　and　the　current　of　the　driving　motor　under　the　no　external　load　and　constant　screw　speed　conditions,　and　the　experimental　results　verified　the　correctness　of　the　theoretical　analysis.　©　2021　IEEE.