Research on Unmanned Surface Vehicle Detection Strategy Based on Game Theory

Abstract

Owing to the evasive action of small underwater targets, they are lost and difficult to detect, which leads to a high false alarm rate of the security system. Based on stationary sonar detection of the target and continuously obtaining relevant data, this study considers the scene of an unmanned surface vehicle (USV) that carries the image sonar to verify the target at a short distance. First, the trajectory data of the target are obtained using stationary sonar, and then the particle filter method is used to predict the trajectory data. Then, a game theory model between our USV and the enemy target is established. According to the action of the enemy target at every moment and the payment function in the model, our USV chooses the most favorable decision to form the confrontation process of both sides. Finally, the target points and detection strategies of the USV are obtained through a numerical simulation. The accuracy of the target point was verified by the test data, which is convenient for the USV to successfully detect the target. The results of this study provide a theoretical basis for unmanned systems to detect small targets.

Keywords

• unmanned surface vehicle
• particle filter
• game theory
• image sonar
• detection strategy