Underwater Object Search and Power Recharging Using Information Network and Multiple Underwater Robots

摘要

Consider a bounded workspace where the information network is already deployed to cover a cluttered underwater environment where Global Positioning System (GPS) is not available. Later, underwater robots are deployed in the workspace where the network has already existed. Our purpose is to search for all objects, such as mines or enemy submarines, in the workspace in a provably complete manner. As a strategy of finding all objects in the workspace, this paper proposes using both multiple underwater robots and the information network. This paper considers an underwater robot with object detection sensors, such as side scan sonar (SSS), for detecting an object close to the robot. Our strategy is to make the robots search for all objects, assisted by the information network. In the information network, one or more nodes are set as power recharging nodes, and every robot needs to visit a recharging node for recharging its power. In the proposed search approach, a robot keeps visiting an information node, while searching for objects that are positioned close to the node. Whenever a robot is lack of sufficient power, it visits the nearest recharging node for recharging its power. The proposed multi-robot search based on the information network is unique in addressing a provably complete search strategy for finding all objects. The proposed search and recharging strategy is unique, since one does not need to access the global position of an information node or an underwater robot. In practice, drifting nodes or broken nodes can generate coverage holes in the network. We thus present how to cover sensing holes that can be generated while information nodes operate in changing underwater environments. MATLAB simulations are used for demonstrating the outperformance of the proposed environment search strategy.