This　paper　is　intended　to　describe　a　new　algorithm,　which　makes　entropy　function　method　with　Lagrange　function　and　Taylor　formula　for　solving　inseparable　variables　of　quadratic　programming.　Quadratic　programming　problem　are　an　important　in　the　fields　of　nonlinear　programming　problem.　Entropy　function　also　called　KS　function.　The　nature　and　related　certificate　of　KS　function　and　its　convergence　have　already　been　proved　at　home　and　abroad.　The　application　of　KS　function　nature　for　solving　quadratic　programming　is　a　very　good　method,　and　it　is　one　of　the　advantages　of　making　more　constraint　programming　problem　become　a　single　constraint　programming　problem,　and　the　original　problems　are　simplified.　Electing　three　examples　of　separated　variables　for　quadratic　programming　problems,　that　is　cross　terms　of　zero,　and　then　contrasted　with　the　new　method.　The　algorithm　resolves　implementation　of　separated　variables　of　quadratic　programming.　Our　numerical　experiments　show　the　proposed　algorithm　is　feasible.