基于非均匀采样的激光海洋湍流传输数值仿真研究

Numerical Simulation of Laser Propagation through Oceanic Turbulence Based on Non-uniform Mesh Sampling

摘要:

针对海洋湍流环境下光束传播建模中精度与效率难以兼顾的问题,提出了一种基于非均匀网格采样的数值仿真方法。该方法利用Fibonacci螺旋与Delaunay三角剖分生成初始网格,并通过局部动态调整实现复杂区域的自适应加密,在边界与扰动密集区域细化网格,而在平稳区域保持稀疏划分,从而兼顾数值精度与计算效率。数值结果表明,在1km海洋湍流传输条件下,网格点数仅增加约10%,即可实现误差幅度降低50%、平均误差下降30%-40%,同时保持网格几何稳定,瘦长三角形比例不足2%。与传统规则网格方法相比,该方法在确保计算稳定性的同时显著提升了空间分辨率与建模效率,有效突破了规则网格与傅里叶谱方法在复杂湍流建模中的局限性,为水下激光通信与探测系统的设计优化提供了重要参考。

Abstract:

To address the trade-off between accuracy and efficiency in beam propagation modeling under oceanic turbulence, a numerical simulation method based on non-uniform mesh sampling is proposed. The method employs Fibonacci spiral and Delaunay triangulation to generate the initial mesh, and introduces local dynamic refinement: meshes are densified in boundary and perturbation-sensitive regions while sparse allocation is maintained in stable regions, thereby balancing computational precision and efficiency. Numerical results show that, under 1 km oceanic turbulence transmission, the number of mesh points increases by only ~10%, while the error magnitude is reduced by 50% and the mean error decreases by 30–40%. Meanwhile, the mesh geometry remains stable, with slender triangles accounting for less than 2%. Compared with conventional regular-mesh methods, the proposed approach ensures numerical stability while significantly enhancing spatial resolution and simulation efficiency. It effectively overcomes the limitations of regular meshes and Fourier spectral methods in complex turbulence modeling, providing valuable references for the design and optimization of underwater laser communication and detection systems.