论文部分内容阅读
为了表征复杂涂层表面的光学散射特性,在微面元理论的基础上,建立了典型涂层样片的偏振双向反射分布函数模型。由于实验数据与模型参数之间存在复杂的非线性关系,采用了遗传算法对模型参数进行反演。针对遗传算法收敛速度慢及易限于局部极小的特点,在传统遗传算法参数反演的基础上,在适应度计算中引入了模拟退火算法对偏振双向反射分布函数模型进行优化建模。实验结果表明:模型的计算结果与实验结果吻合较好。从误差收敛曲线来看,这种混合遗传算法优化方法不仅可以有效避免目标函数陷入局部极小,而且可以有效缩短目标函数的收敛时间。这可以为后续的目标特征提取与识别工作提供参考。
In order to characterize the optical scattering properties of the complex coating surface, the polarization birefringence distribution function model of a typical coating sample is established based on the theory of the micro-plane element. Due to the complex nonlinear relationship between the experimental data and the model parameters, the genetic algorithm is used to invert the model parameters. Aiming at the slow convergence speed and the limitation of local minimization of genetic algorithm, a simulated annealing algorithm was introduced to optimize the polarization bi-directional reflection distribution function model based on the inversion of traditional genetic algorithm parameters. The experimental results show that the calculated results agree well with the experimental results. From the error convergence curve, the hybrid genetic algorithm optimization method can not only effectively avoid the objective function from falling into a local minimum, but also effectively shorten the convergence time of the objective function. This can provide reference for the subsequent extraction and identification of target features.