Cooperative Localization of UUVs with Decentralized Method Based on Underwater Weak Communication

With the increasing difficulty and complexity of underwater tasks, it becomes an inevitable trend for cooperative localization of Underwater Unmanned Vehicles (UUVs) in the form of clusters, which can extend the spatial datum of Global Navigation Satellite System (GNSS) to seafloor and further expand the perception range of single UUV. Due to the weak underwater communication conditions and the complexity of meteorological environment, it has some problems for UUV sensors, e.g., observation limitation and time delay. In the former work, the huge communication of traditional decentralized method brings extremely difficult to operate in actual underwater scene, and the centralized method has the vulnerability and low fault tolerance. Therefore, a decentralized cooperative localization method of UUVs under the condition of underwater weak communication is proposed. Based on the rigorous mathematical theory, the decentralized framework based on extend information filter is constructed. The single UUV establishes its own state chain according to local sensor, and the UUVs cooperate to complete the recursive correction of the information matrix. Observation data in Nankai trough (Japan) are selected as the underwater sound velocity field, and the UUVs on the surface and underwater are simulated. Experimental results show that the proposed method can realize the decentralization of UUVs cooperative localization on the basis of ensuring the accuracy and consistency with the centralized method. Furthermore, data traffic is significantly reduced and the application field of UUV is further expanded.