Abstarct: It is Importance of object detection and tracking under water in the commercial, leisure and defense field. The need to improve the accuracy, stability and risk avoidance, with the advent of new applications and the growing acceptance of undersea robots spreads in military and civilian spreads. One of the main factors in the operation of underwater operations, long-term ability to automatically and without supervision by a ship on the surface, which is the subject of research is necessary to highlight underwater navigation. In this thesis aims to signals emitted from different angles by several sensors are received. Vector measured by the sensors, including arrival angle between the horizontal axis is the acoustical signal that has a linear relationship with the state vector. Due to the nonlinearity of the model and the data noise, we use the Uncented Kalman filter. In this thesis used a type of Kalman filter called augumented uncented Kalman filter and compared with the uncented Kalman filter and extended Kalman filter. Due to the use of multiple sensors, the sensors need to combine data is essential. Covariance method used to combine data and compare our method with this method is that the efficiency of the proposed method is characterized in results. The proposed method combines data by combining both sensor, advantages over the conventional methods, which can be referred to a higher accuracy and reduce the tracking error. Finally, check the placement of sensors, three to four sensor placement suggested that the issue of Bayesian model with Visibility is better than the other models.