To　address　the　problem　of　unpredictable　trajectory　caused　by　sensing　limitation　of　UGV,　in　this　paper,　a　heterogeneous　multi-agent　behavioral　control　architecture　with　predictive　function　is　proposed,　in　which　UAV　intervenes　in　the　decision-making　of　UGV.　The　basic　task　functions　of　UGV　and　UAV　are　defined　respectively　and　the　intervention　signal　of　UAV　is　connected　to　the　task　function　of　UGV　in　null-space-based　control　framework　to　solve　the　problem　of　multi-task　conflict.　The　UAV　platform　adopts　the　nonlinear　model　predictive　control　to　predict　the　trajectory　according　to　the　environment　data　collected　in　advance　and　the　current　state　of　the　UGV.　The　prediction　results　are　input　into　the　decision　model　to　produce　the　final　intervention　signal　and　sent　to　the　UGV　platform.　The　UGV　adjusts　the　task　strategy　in　time　to　ensure　the　selection　of　the　optimal　trajectory　for　the　UGV.　Finally,　the　simulation　results　show　the　effectiveness　of　the　proposed　cooperative　control　method.　©　2021　IEEE