Few　studies　have　considered　the　information　in　the　frequency　domain　to　detect　the　structural　breaks　in　financial　markets.　This　paper　provides　a　mixed　model　that　integrates　BEMD,　AFD　and　Chow　test　to　study　the　fluctuation　characteristics　of　the　soybean　futures　price.　Due　to　the　application　of　AFD,　the　mixed　model　can　detect　the　structural　breaks　of　the　price　fluctuation　through　obtaining　the　high-resolution　information　in　the　frequency　domain.　According　to　our　results,　in　general,　the　soybean　futures　price　in　China　is　mainly　determined　by　IMF11,　followed　by　IMF9,　IMF8　and　IMF7,　and　there　exist　many　structural　breaks　of　different　IMFs.　Interestingly,　national　macro-controls　have　opposite　effects　at　IMF8　and　IMF11.　The　results　show　the　effectiveness　of　the　proposed　mixed　model　for　detecting　structural　breaks　of　financial　markets　in　frequency-domain　by　revealing　the　impact　of　external　events　on　the　soybean　futures　price　at　multi-scales.