Abstarct: Target tracking is one of the important parts in systems including supervision systems, tracking systems and those applied to avoid collisions in water and space. Underwater tracking is a difficult task due to low propagation velocity acoustical waves and also the unwanted noise signals as a result of ships’ noises and environmental problems. In this thesis, using passive sonars signals emitted from the target with different angles are received by two sensors. The measurement vector received by the sensors, consists of the input angle between the acoustical signals and the horizontal axis (Bearing angle) that it has a non-linear relation with state vector. Particle filter is utilized since this model is non-linear and the data are noisy which has been put into practice according to its efficiency for non-linear equations. But in this filter we are faced with the loss of particles phenomenon that practically after some stages, only a number of particles are utilized to estimate the target position. In other words, diversity of particles is removed. In order to overcome this problem, fuzzy logic has been applied. When this phenomenon occurs, particles with appropriate scattering coefficient are propagated in search space. This scattering coefficient is selected at every instance using particle swarm optimization. This optimal value of scattering coefficient leads to a better location of particles. The simulation results show that the optimized fuzzy particle filter is reduced the problem of divergence and convergence time so as a result has been improved accuracy of estimation. Also according to accuracy and the time estimate, the proposed particle filter has better performance compared to particle filter local linearization and is more efficient for real-time tracking.