Electric　energy　is　closely　related　to　people＇s　life,　in　recent　years,　the　construction　of　smart　grid　has　already　been　proposed.　Short-term　load　forecasting　is　a　research　hotspot　in　the　process　of　smart　grid.　In　this　paper,　we　proposed　a　combined　forecasting　method　based　on　random　forest　and　artificial　neural　network,　the　final　result　is　the　weighted　sum　of　the　two　single　models,　and　the　weight　of　each　single　model　is　obtained　by　the　least　square　method.　The　data　of　experiment　is　the　load　data　of　a　power　plant　in　Hunan　province　from　2012　to　2017,　and　the　corresponding　weather　information,　the　sampling　granularity　of　the　data　is　15　minutes.　The　combined　model　we　proposed　can　combine　the　advantages　of　random　forest　and　artificial　neural　network,　and　the　result　of　experiment　shows　that　the　combined　model　improves　the　accuracy　of　short　term　load　forecasting.