基于IIR的语音信号滤波课程设计(论文).doc

PAGE 34 课程设计(论文) 题 目 名 称 基于IIR的语音信号滤波 目录 TOC \o 1-3 \h \u 幅频特性精度很高，不是线性相位的。在设计上可以借助成熟的模拟滤波器的成果，如巴特沃斯、契比雪夫和椭圆滤波器等，有现成的设计数据或图表可查，其设计工作量比较小，对计算工具的要求不高。由于对语音信号的延时效果没有限制，因此在相对FIR数字滤波器设计较为复杂，成本较高的基础上选择IIR数字低通滤波器是十分理想的。IIR数字滤波器是借助于模拟滤波器的设计方法进行的，为了避免设计的过程中由脉冲响应不变法带来的频谱混叠想象，可以通过采用双线性变换法进行设计。下面从先介绍IIR数字低通滤波器设计的基本流程开始，然后通过一个从取语音信号到最终语音信号的滤波结束的实例，在此过程中重点阐述通过分析语音信号的频谱，更具分析结果，制定指标，设计滤波器，最终达到由理论到实践，实践中得体会的学习目的。 关键词：MATLAB；频谱分析；IIR数字低通滤波器；双线性变换法；语音信 Abstract This paper introduces the use of bilinear transform design IIR digital low- pass filter with the basic idea of the steps , and the combination of theory and practice now , by the admission of a voice , using MATLAB software to analyze voice signal spectrum , identify indicators designed filters, write program design filter , filtering the speech signal processing , so as to eliminate high frequency noise , making speech more pure effect. Digital filtering of the speech signal is to use a fast Fourier transform (FFT) spectral analysis of the speech signal , a digital filter design using the index obtained in the analysis , the unnecessary portion by a digital filter designed to filter out in order to optimize voice purposes. There are a variety of filters , this design is a digital low-pass filter IIR , IIR digital filter frequency characteristics high precision, not linear phase . In the design of analog filters can help mature the fruits , such as Butterworth, Chebyshev and Elliptic filters , etc., have an existing design data or charts to be investigated , the design work is relatively small, no request for computational tools high . Because of the speech signal delay effect is not limited, so the relative FIR digital filter design is more complex, higher-cost choice based on IIR digital low -pass filter is ideal . IIR digital filter by means of an analog filter design methods , in order to avoid the process of designing the same response brought by the pulse method aliasing imagination, can be designe