Research on Submarine Position Prediction and Search and Rescue Strategy Based on Kalman Filter and Particle Swarm Optimization

In this study, we propose a hybrid model combining dynamic modeling, Bayesian networks, and particle swarm optimization (PSO) for accurate submarine position prediction and efficient search and rescue operations. The dynamic model, grounded in Newtonian mechanics, provides a foundational framework for simulating submarine movements. The Bayesian network is employed to assess the probability of the submarine's presence at various locations, integrating uncertainty in environmental factors. The Kalman filter further enhances the predictive accuracy by managing noise and refining trajectory estimates. Finally, PSO is utilized to optimize the search process, demonstrating rapid convergence towards the most probable locations. This integrated approach offers a robust solution for underwater operations, as validated through extensive simulations.