蚂蚁算法在FIR数字滤波器优化设计中的参数

Parameters in FIR Digital Filters Optimal Design Based on Ant Algorithm

蚂蚁算法中参数的准确分析和合理配置直接影响着算法的性能。在已完成的蚂蚁算法应用于有限冲激响应(Finite impulse response,FIR)数字滤波器优化设计研究基础上,分析了各个参数的不同配置对算法性能的影响,推导了参数配置的基本公式,提出了参数之间的一般配置原则。不失一般性,在最小最大优化准则下进行的仿真实验结果表明,文中提出的算法参数配置原则对于提高FIR数字滤波器的优化设计性能较为有效,同时还验证了蚂蚁算法在其他应用领域中的参数设置也满足参数配置原则,进一步表明本文的参数选取原则的可行性,有利于蚂蚁算法在优化问题中的推广和应用。

Accurate analysis and reasonable configuration of the parameters affect the perform- ance of ant algorithm （AA）. Based on the well-studied AA for FIR digital filters optimal de- sign, the influence of the parameters in the algorithm is analyzed. Then the basic parameters configuration formula is derived and a general configuration principle is proposed. Without loss of generality, simulation results show that the proposed parameters configuration principle is effective in improving optimal design performance of FIR digital filters under mini-max （MM） optimality criterion. Meanwhile, they also show that AA for other application fields is adapta- ble to the parameters configuration principle, and the parameters configuration is feasible. It is believed that the results are beneficial to the development and application of AA in optimization problems.