In underwater target search path planning, the accuracy of sonar models directly dictates the accurate assessment of search coverage. In contrast to physics-informed sonar models, traditional geometric sonar models fail to accurately characterize the complex influence of marine environments. To overcome these challenges, we propose an acoustic physics-informed intelligent path planning framework for underwater target search, integrating three core modules: The acoustic-physical modeling module adopts 3D ray-tracing theory and the active sonar equation to construct a physics-driven sonar detection model, explicitly accounting for environmental factors that influence sonar performance across heterogeneous spaces. The hybrid parallel computing module adopts a message passing interface (MPI)/open multi-processing (OpenMP) hybrid strategy for large-scale acoustic simulations, combining computational domain decomposition and physics-intensive task acceleration. The search path optimization module adopts the covariance matrix adaptation evolution algorithm to solve continuous optimization problems of heading angles, which ensures maximum search coverage for targets. Large-scale experiments conducted in the Pacific and Atlantic Oceans demonstrate the framework's effectiveness: (1) Precise capture of sonar detection range variations from 5.45 km to 50 km in heterogeneous marine environments. (2) Significant speedup of 453.43 × for acoustic physics modeling through hybrid parallelization. (3) Notable improvements of 7.23% in detection coverage and 15.86% reduction in optimization time compared to the optimal baseline method. The framework provides a robust solution for underwater search missions in complex marine environments.
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号