In　this　paper,　a　novel　human　decision-making　behavior　model　is　proposed　for　human-robot　interaction　(HRI)　for　the　control　of　multi-robot　systems　(MRS).　The　proposed　human　drift　diffusion　model　(HDDM)　combines　the　traditional　drift　diffusion　model　(DDM)　and　the　null-space-based　behavioral　control　(NSBC)　method　by　introducing　a　data-processing　station　and　a　human　cognitive　system.　In　the　HDDM,　the　evolution　of　human-decision　information　is　computed.　By　using　a　threshold　of　such　information　to　trigger　human　interaction,　accurate　human　decision-making　timing　can　be　obtained.　In　addition,　a　cooperative　controller　is　designed　for　robots　to　follow　human　instructions.　Simulations　under　various　scenarios　show　that　by　using　the　proposed　HDDM　and　the　controller,　robots　can　complete　human　instructions　more　accurately　comparing　to　traditional　methods.　An　experiment　using　a　group　of　quadrotors　subject　to　external　wind　disturbances　also　demonstrate　the　effectiveness　of　the　proposed　HDDM　in　real-world　uncertain　environments.