基于粒子群优化算法的自适应IIR滤波器设计

Particle swarm optimization algorithm for adaptive IIR digital filters design

针对自适应无限冲激响应(infinite impulse responseI,IR)数字滤波器的设计实质上是一个多参数优化问题,提出了一种用粒子群优化算法(particleswarmoptimization,PSO)设计IIR数字滤波器的方法。将滤波器的设计问题转化为滤波器参数的优化问题,利用粒子群优化算法对整个参数空间进行高效并行搜索以获得参数的最优化,基于多个典型系统的随机数值仿真以及与最小二乘方法的比较研究,验证了该方法的有效性、全局性和对初值的鲁棒性。

With optimization design of adaptive IIR digital filters being multiple-parameter optimization problems, a satisfactory optimization method based on particle swarm optimization （PSO） is proposed to design adaptive IIR digital filters. The design of adaptive IIR digital filters is converted into the optimization of the parameters of IIR digital filters. PSO is used to search the whole parameters space effectively and globally in order to optimize parameters. Random simulation based on some typical systems as well as comparison with least-mean-square method demonstrates the effectiveness, global optimization ability and robustness on initial values of this method.