中文摘要：

利用物理光学方法预估高频雷达散射截面（RCS）时，需要判别目标表面的阴影区域。本文提出了一种基于树型结构的多重阴影判别算法，将判断面元是否被面元遮挡转变为判断面元是否被分组遮挡，极大地降低了计算复杂度。与此同时，引入并行技术，将基于树型结构的多重阴影遮挡算法并行化，进一步节省计算时间。数值算例表明，基于树型结构的多重阴影判别算法与传统的阴影判别方法相比，在保证计算精度一致的前提下，计算效率更加高效。

英文摘要：

Using physical optics method to evaluate high frequency radar cross section （RCS）, the shadow recognition is required and time consuming. A tree structure based method is proposed for multiple shadow recognition. Judging whether a facet is obstructed by another facet is replaced by judging whether a facet is obstructed by another sub-group. The computational complexity is able to be reduced greatly by this method. Furthermore, the parallel technique is introduced. The computation time can be saved by the paralleled tree-base multiple shadow recognition algorithms. Numerical results demonstrate that the proposed method is more efficient than traditional method.