The invention relates to a multi-command processing method and device for robot system interaction. The method comprises the steps that a command set determined according to user semantics is received, wherein the command set comprises multiple service instructions corresponding to at least one preset function scene; the command set is analyzed, and whether the command set supports command concurrence output or not is judged; if the command set supports command concurrence output, the command set is converted according to the function scene, and a corresponding set of at least one single command is obtained; service commands corresponding to the set of the single command are called, and corresponding execution results are output. The multiple commands corresponding to multiple scenes included in one voice message of a user can be output to achieve simultaneous trigger and output of the commands, the cooperative mode of the commands is regulated, and therefore the command processing efficiency and stability of a robot system are improved.