摘要
针对传统弹跳射线方法(SBR)在处理电大尺寸目标电磁散射问题时存在效率不高的问题,提出了改进的电大目标电磁散射弹跳射线算法。该算法在采用层次包围的二叉树加速数据结构的基础上,提出了在图形处理器(GPU)端实现射线管的分裂和追踪更新的计算统一设备构架(CUDA)并行计算SBR算法。通过实验算例验证表明,该方法可有效提高SBR方法处理电大尺寸目标电磁散射问题的的计算效率。
In order to counteract the low efficiency of traditional shooting and bouncing rays（SBR）method in calculation of electrically large target multiple reflections,This paper put forward an improved shooting and bouncing ray algorithm for EM scattering by electrically large target.On the ground of bounding volume hierarchy（BVH）binary tree acceleration data structures and GPU-based Compute Unified Device Architecture（CUDA）,the SBR calculation methods were proposed based on splitting of ray tube at the end of GPU and tracing renewed CUDA.Experiment proved that this method could effectively increase the calculation efficiency of SBR in dealing EM scattering by electrically large target.
出处
《探测与控制学报》
CSCD
北大核心
2015年第3期91-95,共5页
Journal of Detection & Control
基金
国家自然科学基金(11264033)项目资助
甘肃省高校基本科研业务费(2050205)项目资助
关键词
弹跳射线法
目标电磁散射
层次包围
计算统一设备构架
shooting and bouncing rays
EM scattering by target
bounding volume hierarchy
compute unified device architecture
