Abstarct: Swarm robotic is a new field of research that uses the robots with specific abilities to achieve goals. The concept of swarm is inspired by the robotic structure of the collective life of animals and insects living socially. The aim is to bring the robot group behavior closer social behavior of insects. So swam robotic control has a heuristic or exploratory nature and can’t be formal. The formal ways have analytic skill. Unlike that, modeling with the Petri nets have formal state and it can convert to the executive code. So if we can handle a swarm's control of a formal structure and its modeling capability with Petri's network, we will achieve better results. We will begin by analysis and comparing. Then, using the swarm robotic mechanism and forming them with Petri networks, we try to change heuristic natures to formal ones and simplification them. In this way, with the clustering of robots and Division tasks, in addition to optimizing the division of tasks, control task are greatly optimized. Also the complexity of the problem is reduced considerably. Finally, we can modeling the swarm robotic with the Petri networks.