基于模拟退火算法的高阶数字微分器设计

通过简单有效的搜索操作和参数,提出了设计高阶数字微分器的一种模拟退火算法.在有效频段最小化二次型误差的意义下,算法能够快速而简单地得到最优滤波器参数.通过与著名的McClellan-Parks算法的比较表明,所提算法不仅能够实现误差函数的全局最优,而且在绝大多数频段能够取得更好的性能,同时具有较强的初值鲁棒性.

Simulated Annealing for Designing Higher-Order Digital Differentiator

A simulated annealing approach with effective search operator and parameters is proposed for designing higher-order digital differentiator, which can rapidly and simply obtain the optimal filter coefficients in the sense of minimizing the quadratic measure of the error in the frequency band. Comparisons to the well-known McClellan-Parks algorithm show that the proposed simulated annealing is not only optimal in sense of error function, but also can achieve much better performance in most of the frequency band, as well as is of much strong robustness on initial conditions.