We shall derive several direct sampling methods for the recovery of wave-penetrable inhomogeneous scatterers in a three-dimensional stratified ocean waveguide. The proposed methods are simple, straightforward and computationally efficient since no matrix inversions, optimizations or solutions of ill-posed linear systems are involved. Moreover, they are highly tolerant to noise and are applicable with a few scattered fields corresponding to only two or four incidences. In addition, a priori knowledge of either the number of disconnected components or the physical features of the unknown obstacles is not required by the direct sampling methods. Several extensive numerical experiments are shown to evaluate the effectiveness and feasibility of the newly proposed methods.

刘可伋, 上海财经大学讲席副教授, 毕业于香港中文大学, 主要从事反散射问题的算法与理论研究。曾获得上海市“浦江人才”计划及“晨光学者”计划支助, 目前主持在研国家自然科学基金一项, 主持完成上海市重大科技攻关子项目一项, 在SIAM J. Sci. Comp.，SIAM J. Imag. Sci., Inverse Problem等国际SCI期刊上发表多篇论文。