The Fuzzy-Based Visual Intelligent Guidance System of an Autonomous Underwater Vehicle: Realization of Identifying and Tracking Underwater Target Objects

This study aims to develop a fuzzy-based visual intelligent guidance system (VIGS) that executes missions involving the identification and tracking of underwater target objects for an autonomous underwater vehicle (AUV). To demonstrate the VIGS functions, a series of tests were conducted in the stability tank and towing tank at National Cheng Kung University. The characteristic of the VIGS is to immediately capture continuous real-time images from the bow to calculate and identify visual information concerning the AUV's surroundings. By mapping the target's information in the AUV visual coordinate system (two-dimensional) onto the earth-fixed coordinates (three-dimensional), the relationships of corresponding distance and azimuth between the target and the AUV were defined. Eventually, this information was used to calculate the control parameters of the AUV's moving speed and heading angle, completing the visual motion control framework. In the dynamic guidance experiments, the moving speed of the target object is proved to be a predominant factor leading to different trends of AUV's steering performances for pitch and yaw controls.