基于粒子群优化策略的海洋环境参数反演研究OA
Research on the ocean environmental parameter inversion based on improved particle swarm optimization algorithm
本文将三维海面电磁散射建模与机器学习算法相结合,提出了一种基于粒子群优化策略的海洋环境参数反演模型.该模型有效考虑了海面大小尺度波浪对电磁波的散射效应,采用半确定面元法构建了三维海面散射模型.在此框架上,对传统粒子群算法进行了结构性增强,并利用改进后的粒子群优化策略对支持向量机方法中的参数进行自适应搜索,从而构建了用于反演海面上方风速、海水介电常数等海洋环境参数的有效模型.同时,为了评估所提反演模型在噪声环境下的稳健性,本文在输入样本中叠加高斯白噪声并开展对比实验.实验结果显示,本文方法具有良好的反演效果,在海洋遥感领域应用前景广泛.
This study integrates three-dimensional(3D)sea surface electromagnetic scattering modeling with machine learning algorithms to propose an ocean environmental parameter inversion model based on a particle swarm optimization(PSO)strategy.In this model,the electromagnetic scattering effects of multi-scale sea waves are considered by employing the semi-deterministic facet method to establish a 3D sea surface electromagnetic scattering model.Within this framework,structural enhancements are applied to the conventional particle swarm algorithm,and the improved particle swarm optimization strategy is employed to adaptively search the key parameters of the support vector machine(SVM).This enables the construction of an effective inversion model to retrieve ocean environmental parameters,such as the near-surface wind speed and seawater dielectric constant.Meanwhile,to assess the robustness of the proposed inversion model under noisy conditions,Gaussian white noise is superimposed onto the input samples,and a series of comparative experiments are conducted.The experimental results demonstrate that the proposed method achieves reliable inversion performance and exhibits considerable potential for applications in marine remote sensing.
黄猛;王玉菊;窦松然;黎佳艺;董春雷
中国人民解放军 91977 部队,北京 102249中国人民解放军 91977 部队,北京 102249船舶信息研究中心,北京 100012中国人民解放军 91977 部队,北京 102249西安电子科技大学,西安 710071
航空航天
海面电磁散射建模机器学习粒子群优化策略支持向量机方法海洋环境反演模型
sea surface electromagnetic scattering modelingmachine learningparticle swarm optimization strategySVM methodocean environmental parameter inversion model
《空间电子技术》 2026 (4)
179-188,10
评论