Research On Infrared Sensing Based Flocking Behavior Of Swarm Robotics

Flocking behavior of swarm robotics means the swarm should remain ordered in space during the movement. Flocking is a typical multi-robotics coordination problem and the foundation of more complex tasks. The difficulty of flocking is maintaining the formation dynamically. Most existing algorithms use communication to coordinate different robotic nodes, however the communication mechanism has its bottle neck: When the scale of the swarm is too large, the communication quantity between the nodes would soar drastically, and this communication problem will confine the swarm scale being enlarged further. So a new flocking model with limited infrared sensing and no communication is presented.The infrared sensing model can avoid the communication bottle neck. But exact infrared sensing is difficult to get in the multi-robotics background; furthermore, no communication means any information can hardly be shared between the nodes. Both these make the task even harder to be finished.With all the concerns above, this paper will include these several parts:1) The infrared sensing module is designed. Infrared is used very often as a sensing mechanism in the robotics research field. But the infrared sensing can't be exact with interference from the environment. Here we will analyze the performance characteristic of infrared and design one infrared sensing module which is reliable, including the circuit and the algorithm.2) Flocking model is designed using the infrared module above. The flocking model would be realized on the hardware platform in our lab. There are two models we will discuss: inactive infrared flocking and active infrared flocking.3) Analyze the flocking algorithms we designed and check the merits and demerits with others in this research field, showing the direction for the further work.With the infrared sensing module and the flocking algorithm designed in this paper, we finished the two robots beeline flocking and curve flocking, and the four robots beeline flocking. It is showed that the infrared sensing mechanism is viable for the flocking control. And some advice is also offered for the hardware development in our lab.