摘要
传统的基于声传播时间的声速剖面反演需要多次搜索不同声速剖面条件下的特征声线,降低了声速剖面反演的速度.通过一个简单的近似,微扰法将实际声传播时间表示为背景声速下的声传播时间与扰动项之和,从而使得声速剖面反演由原来的非线性优化问题转化为线性方程组的求解形式.实验数据处理结果表明,与传统方法相比,虽然微扰法反演的精度下降了近50%,但是计算时间由10h减少到了3s,从而实现声速剖面海上实时监测.
The conventional sound speed profile (SSP) inversion algorithms based on sound transmission time influence the inversion speed because they have to trace the eigen-ray of every SSP. The actual transmission time was expressed as the sum of that for referenced SSP and perturbation terms, and SSP inversion was changed from non-linear optimization to the solution of linear equations by the perturbation method. The experimental inversion results indicate that although the root mean square (RMS) error is doubled in comparison with that obtained by conventional algorithms, the calculating time can be reduced from 10 hours to 3 seconds by the perturbation method and therefore satisfies real-time monitoring of SSP.
出处
《上海交通大学学报》
EI
CAS
CSCD
北大核心
2013年第8期1287-1291,共5页
Journal of Shanghai Jiaotong University
基金
国家自然科学基金资助项目(11104045)
国家安全重大基础研究项目(613110010202)
关键词
微扰法
声速剖面反演
传播时间
特征声线
perturbation method
sound speed profile (SSP)inversion
transmission time
eigen-ray