Study of obstacle information processing for unmanned underwater vehicle in unknown ocean environment

In ocean environment, to ensure the safety of navigation for unmanned underwater vehicle (UUV), it is the most important to obtain the position information and moving information of obstacles. In order to distinguish between static and dynamic obstacles, the detection of moving property is used to detect property of obstacles. Then, to meet the requirement of obstacle configurations, k-means clustering algorithm is put forward for clustering. Finally, least squares prediction is put forward for detecting position information of UUV, this method predicts the future position information of obstacles and the moving parameter, and solves the problem of lack of useful information for UUV. Simulations show that the proposed method could get the accurate position information and moving information of obstacles.