光学薄膜BRDF五参数模型的粒子群优化  

Five Parameters Optimization of BRDF Model of Optical Film by Particle Swarm Algorithms

作　　者：王党社[1] 张建科[2] 吴振森[3] 

机构地区：[1]西安工业大学数理系,西安710032 [2]西安邮电学院应用数理系,西安710121 [3]西安电子科技大学理学院,西安710071

出　　处：《西安工业大学学报》2009年第3期214-218,共5页Journal of Xi’an Technological University

基　　金：国家自然科学基金资助项目(60371020);陕西省薄膜技术与光学检测重点实验室资助项目(ZSKJ200702)

摘　　要：为得到光学薄膜表面双向反射分布函数的统计模型,测量了不同薄膜材料表面的激光(0.808μm)双向反射分布函数.采用多参数优化的惯性权重模型粒子群算法,权重因子随迭代代数增加线性减小,从1.0线性递减至0.4,优化建立了材料BRDF五参数模型,即材料表面双向反射分布函数由kb,kd,kr,a,b五个参数确定,其中kd决定漫反射分量的大小,kb决定镜向反射分量的大小,kr为小面元分布函数的参数,决定表面的斜率分布,a,b为待定参数,与介质的折射率有关.To obtain the BRDF model of optical film,the LBRDF（0. 808μm） of film is measured. Based on measured data of LBRDF, the particle swarm optimization of multigle parameter model with inertia weights is utilized to create five parameter model of BRDF of above mentioned material. BRDF of material is determined namely by kb ,ka ,kr ,a ,b. kd decides diffuse reflection component, kb decides mirror reflection component, kr is a parameter of small surface distribution function,deciding the slope of the surface, a,b are two parameters to he determined. They are relevant to refractive index of material. The inertia weight factor reduces linearly from 1.0 to 0. 4 with the increasing of iterative time. Compared to the result of genetic algorithm, the particle swarm optimization show better efficiency and precision in computing.

关 键 词：双向反射分布函数 粒子群算法 遗传算法 优化 

分 类 号：O484.41[理学—固体物理]