As one of the modern high technologies, robotics has been developed and widely used in various industries, which greatly promotes the progress of productivity. However, when the task is complex which single robot can not achieve it, multi-robot cooperation is needed to complete the work. The significance of multi-robot cooperation is not the function superposition of each one. The solving task of multi-robot system is seeking one or more robots that can perform tasks coefficient in the system. And more, the robot group is a combination of the lowest price of the task. Depending on the type of task, the multi-robot cooperation was divided into three different forms: sequential cooperation, synchronous cooperation and free cooperation. Researches of the three forms of cooperative planning and path optimization were achieved through the optimization algorithm of design. The feasibility and correctness of the method are verified by simulation.