论文部分内容阅读
将一种广泛用于求解复杂系统优化问题的技术———模拟退火法———用来求解椭偏反演方程。首先假设一个薄膜模型 ,计算出其相应的椭偏参数 (Ψ ,Δ)的值 ,在这个计算值的基础上加入不同标准偏差的高斯噪声 ;然后将加入噪声后的值 (Ψm ,Δm)作为模拟的测量数据 ,采用模拟退火算法进行求解 ,验证得知这种方法求得的薄膜参数很接近于假设的薄膜模型参数的真值 ,与其他文献的报道结果一致 ,而且在扩大搜寻范围时 ,仍然可以得到准确解 ,从而证明了该方法的可行性以及有效性。
A widely used technique for solving complex system optimization problems - simulated annealing - is used to solve the ellipsometric inversion equation. First, suppose a thin film model is used to calculate the corresponding values of ellipsometric parameters (Ψ, Δ). Based on the calculated values, Gaussian noise with different standard deviations is added. Then, the added noise values (Ψm, Δm) Simulated measurement data were simulated by simulated annealing algorithm. The results show that the thin film parameters obtained by this method are close to the true values of the assumed thin film model parameters, which are consistent with those reported in other literatures. And when the search range is expanded, The exact solution can still be obtained, which proves the feasibility and effectiveness of the method.