【作者】 张博； 黄翔东；

【Author】 ZHANG Bo;HUANG Xiangdong;School of Electronics and Information Engineering,Tianjin University;

【机构】 天津大学电子信息工程学院；

【摘要】 提出基于遗传算法的边界频率精确可控的两阶段FIR滤波器设计法:第1阶段采用基于偶对称频率采样的单窗全相位滤波器设计,可获得边界频率落在期望频点上的待补偿滤波器;第2阶段引入遗传算法在窄频带优化配置补偿滤波器的Kaiser窗参数,结合奇对称频率采样的单窗全相位滤波器设计,即可实现期望的幅频特性补偿效果,且可获得通带波纹小于1%的幅频曲线。另外,还总结出滤波器的参数配置与设计流程。数值仿真表明,该方法可设计出既能精确控制边界频率又具备优良传输特性的FIR滤波器。更多还原

【Abstract】 This paper proposes one novel two-stage genetic algorithm based scheme for designing FIR filter with controllable critical frequency.In stage 1,a closed-form analytical formula related to single-window all-phase filter design employing the even symmetric frequency sampling mode is derived,which yields an uncompensated filter with a critical frequency located at the expected frequency.In stage 2,genetic algorithm combined with the odd symmetric frequency sampling mode single-window all-phase filter design is implemented in narrow band to optimize a Kaiser window parameter that is closely related to the final filter’s transfer characteristics.Thus the expected compensation performance of amplitude transfer characteristic can be realized,which can restrict the pass-band ripples to the range of 1%.In addition,this paper also sums up the parameter configuration and filter design flow.Numerical results show that the proposed scheme can not only accurately locate the critical frequency but also can possess excellent transfer characteristic.更多还原