IIR数字滤波器的序列约束Chebyshev设计方法.docVIP

IIR数字滤波器的序列约束Chebyshev设计方法 侯秀竹，赖晓平，杨吉祥，罗家浩 （学院，） 摘要无限冲击响应数字滤波器的无约束Chebyshev设计问题，得到的滤波器幅值响应在过渡带过冲，转化为一系列具有相同的目标函数滤波器在过渡带的频率响应与某个具有单调幅值的过渡响应函数之间的误差函数之比雪夫模我们称该方法为序列约束切比雪夫方法既可以使频率响应误差小又可以有效减小以及边附近大的群延迟误差关键词数字滤波器Chebyshev设计 中图分类号： 文献标识码： A Sequential Constrained Chebyshev Design Method of IIR Digital Filters HOU Xiuzhu, LAI Xiaoping,YANG Jixiang, LUO Jiahao (College of Automation, Hangzhou Dianzi University, 310018, Hangzhou, China) ABSTRACT：The magnitude response of an unconstrianed Chebyshev infinite impulse response digital filter usually exists an obvious overshoot in the transition band and large group-delay error near the passband edges.This paper presents a sequential constrained Chebyshev design method that formulates the design problem as a series of constrained Chebyshev problems which take the error amplitude of the impulse response in the transition bands and some preset monotonically increasing and/or decreasing transient response as the same objective function. The method can not only minimizes the maximum error of the magnitude responses in the transition bands and some preset monotonically increasing and/or decreasing transient response functions, but also effectively decreases the maximum error of frequency response on the passband and stopband and the group-delay error near the passband edges. Simulation results demonstrate the effectiveness of the proposed method. KEYWORDS：IIR digital filter; Chebyshev design; overshoot; sequential constrained minimax method 1 引言 (IIR)滤波器和有限冲击响应(FIR)滤波器。相比较FIR滤波器，IIR滤波器具有更高的选频特性、更小的通带群延迟，因此研究如何有效地设计IIR滤波器具有重要的现实意义。 IIR滤波器没有精确的线性相位和内禀稳定性，其设计问题形成的数学模型通常是非线性约束非凸规划模型。针对这一问题，提出了Levy-Sanathanan-Koerner(LSK)[1,2]策略和高斯牛顿(GN)[3] 策略进行IIR滤波器的优化设计。 从性能指标来看，IIR滤波器的优化设计大致可以分为最小二乘设计和Chebyshev设计两大类。在几十年的研究中，人们提出了许多有意义的设计方法[4-13]。IIR数字滤波器原始最小二乘设计存在收敛性问题，其初始点选取很关键。Chebyshev设计要求最大逼近误差幅值达到最小，并实现误差在整个感兴趣的频带上均匀分布，备受工程技术研究人员的青睐。IIR滤波器Chebyshev设计问题可以采用线性规划(LP)[4,5,6]方法实现，但LP方法需要较大的储存空间和计算时间。随后提出用迭代重加权最小二乘实现IIR数字滤波器的Chebyshev设计，将问题转化成