The　inlet　and　outlet　temperature　of　the　mill　is　an　important　index　in　the　process　of　slag　powder　production.　The　specific　technological　process　of　the　production　of　slag　powder　is　introduced,　and　the　factors　that　influence　the　inlet　and　outlet　temperature　of　the　mill　are　analyzed.　The　actual　data　of　production　process　is　preprocessed　by　the　Pauta　criterion.　The　forecasting　model　of　inlet　and　outlet　temperature　of　mill　is　established　to　predict　the　temperature　in　an　hour　by　using　the　basic　BP　neural　network　and　BP　neural　network　optimized　by　mind　evolutionary　algorithm.　Through　the　simulation　of　Matlab　software,　the　prediction　effect　diagram,　prediction　error　chart,　prediction　mean　square　error,　mean　absolute　error,　mean　absolute　percentage　error,　and　coefficient　of　decision　of　the　two　algorithms　are　compared.　The　results　show　that　the　BP　neural　network　optimized　by　mind　evolutionary　algorithm　is　better　than　the　original　BP　neural　network　in　the　prediction　accuracy.　©　2018　IEEE.