高级检索

光散射模型参量反演中的遗传模拟退火算法

Genetic simulated annealing algorithin in the parameter retrieval of light scattering model

  • 摘要: 为了研究遗传模拟退火算法在光散射模型参量反演中的迭代搜索性能问题,分别采用遗传模拟退火算法和单一遗传算法迭代搜索了几种介质的双向反射分布函数模型的相关参量。将两种算法的反演结果与在特定激光波长下的双向反射分布函数实验数据进行了对比,通过理论分析和实验验证,取得了两种算法所得到的拟合值,两种拟合值都与实验数据吻合得较好;同时比较了遗传模拟退火算法和单一遗传算法在迭代次数、计算时间和均方误差等之间的差异。结果表明,两种算法在不同介质表面双向反射分布函数模型参量反演时都可以得到满意的结果,且前者优化效果更优。这一结果对研究不同算法的迭代搜索性能是有帮助的。

     

    Abstract: In order to study iteration searching properties of genetic simulated annealing algorithm (GSAA) parameters in bidirectional reflectance distribution, function (BRDF) model, GSAA and genetic algorithm(GA) were employed, respectively, to search iteratively parameters of BRDF model for several media.The retrieved results of the two algorithms were compared with the BRDF experimental data at a certain laser wavelength.By theoretically analyzing and experimentally demonstrating, fitting values for both the algorithms were obtained, which were consistent with experimental values very well.Meanwhile, the difference of iteration number,computation time and mean square error in both the algorithms was compared.The results show that, the BRDF experimental data can be retrieved satisfactorily using both GSAA and GA, and the former is better than the latter.The aforementioned results are of help to the study of iteration searching properties of algorithms.

     

/

返回文章
返回