Research On Multi-AUV Coordination Method With Limited Communication

Multiple robots coordination is one of the most popular and difficult research orientations. Formatin means that multiple robots march with special geometric shape in unknown environment. It is widely used in space search,ocean explore,military application and so on.Communication and coordination are the most important parts of the formation research. Communication is considerred as the base of multiple robots Cooperation.Muti-AUV real time coordination control is extremely difficult because underwater communication is restricted to low transmit rate,high data error probability and large transmit delay.First of all, the paper introduces the related rearch about Multiple UUV/AUV. After that it describes an underwater communication simulation method for Multiple-AUVS formation. It bases on the statistics principle to simulate underwater communication channel which can reduce the difficulty. It is able use point to point,broadcast or random communication mode and can meet the requirement of centralized and decentralized control.Compared to general method, it uses Virtual Leader AUV to concentrate the communitication coordination for one part of the divided channel which can avoid the dissipation of contending for channel. For the other part of channel, it use the random feature of S-ALOHA single channel sharing communication to deal with Burst event which is used for exchange data that is less imminent and important. It may cause retransmission but is effective for decreasing Handshaking.Then, it introduce the formation coordination control method. Artifical Potential method is used to control the form of the AUVs swarm. The Potential function is defined which make the formatiom has the capacity of di-direction coordination. Experiment shows that AP method is difficult to transform AUVs form and it is seriously affected by the communication "disable" while avoiding obstacles. In the case of that,behavior-based mothod is used to improve the flexibility and quick respond. In order to deal with the data delay and loss, least squares is used to motion forecast. When AUV is loss or "disable", an replacement AUV is arranged to fill up the vacant position.Following that simulation system and experiment are introduced. AmigoBot robots is used to simulate AUVs formation control Algorithm and communication is simulated viaing Wireless Local Area Networks. Last, communication form is described and experiment result is analysised. Simulation result shows that the methods are effective for AUVs formation coordination.