中文摘要：

光照是树木生长需要的重要资源,是树木生长仿真计算中必不可少的因素.但在森林演化的计算机模拟中,由于光照模拟的复杂性,使得光照模型的计算量十分巨大.本文采用了光照指数（Gap Light Index,GLI）为因子的光照模型,并针对该模型开展了快速计算研究.由于该模型计算中存在着大量限制计算效率的几何求交运算,本文根据光照指数与植物暴露面积所具有的共同特点,提出采用基于暴露面积的计算方法来近似拟合GLI的值,并在并行计算架构CUDA上实现了该算法.最后通过不同实验比较,验证了本文提出的方法针对较大的树木规模,在保证较小误差率的前提下,获得了比GPU并行求交方法快数十倍以上的加速比.

英文摘要：

Light is the important resource for the growth of tree and an indispensable factor in ecological model.However,in the simulation of forest evolution the computational load of light is extremely large because of high complexity of the light simulation.We apply the light model which defines the Gap Light Index（GLI） as a factor and research on the fast computing method based on this model.There are a large number of geometric intersections in the model,which limit the computational efficiency.For solving this problem,this paper proposed an approach to calculate GLI of a tree by approximately fitting its value with the exposing area of the tree based on their common attributes.Furthermore,we implement the algorithm by computing the exposed area of trees on parallel computing architecture CUDA.In the last,the comparison of different experimental results shows that our approach obtains more than dozens of times speed-up than the naive GPU light intersection approach with reasonable error in the case of large scale of trees.