中文摘要：

提出了一种新的基于布谷鸟搜索算法的片上电感模型参数优化算法。布谷鸟搜索算法对复杂非线性非凸函数的优化效果显著,这种函数的特点与半导体器件紧凑型模型参数和优化目标之间的非线性特性相近。该算法开发了模型参数交叉操作,将两个解空间的对应参数交换位置,避免敏感度高的参数对敏感度低的参数的过度影响。引入了自适应参数,使算法的搜索步长能自适应调整,既不会因为步长太大跳过最优解,也不会因为步长太小导致收敛速度慢。采用集成电路工艺片上螺旋电感的实测数据对该算法进行验证,得到较好的拟合度。提出的模型优化算法可适用于集成电路器件模型的自动优化。

英文摘要：

A new on-chip inductor parameters extraction optimization algorithm based on cuckoo search algorithm was presented,which was good at the optimization of complex nonlinear non-convex function. These nonlinear characteristics were similar to the characteristics of semiconductor devices compact model. The crossover operation that exchanged the positions of two solutions corresponding parameter space was developed,in order to avoid the low sensitive parameters being influenced extremely by the high sensitive parameters. Further,adaptive parameters were also introduced to adjust the search step of the algorithm automatically,by which neither the search step was too large to skip the optimal solution,nor the search step was too small to slow the convergence rate. The measured data of on-chip spiral inductor was used to verify the proposed algorithm. The better fitting results showed that the new algorithm could be used to optimize the models of integrated circuits automatically.