信息处理课群课设解决方案.docVIP

课程设计任务书 学生姓名： 专业班级： 指导教师： 李政颖 工作单位： 信息工程学院 题 目: 基于LMS 算法的多麦克风降噪LMS ；麦克风；降噪；MATLAB Abstract Noise reduction of speech is mainly about how to use signal processing techniques to eliminate the noise signal,so as to improve the ratio to extract useful signal output signal-to-noise. Usually the method to eliminate the noise pollution in the signal is that signal polluted through a noise suppression and signal relative invariant filter, which has the input noise from the signal can not be detected.then through the filter, the original noise is counteracted, so as to achieve the improvement of signal to noise ratio. The least mean square (LMS) adaptive algorithm is known to the error between the desired filter response and the output signal of the mean square value of minimum standard, based on the input signal to estimate the gradient vector in the iterative process,and the adaptive iterative algorithm and update the weight in order to achieve the optimal. LMS algorithm is a gradient descent method, whose characteristics and advantages are its simplicity, this algorithm needs to calculate the correlation function of the corresponding, also does not need matrix operation. In order to complete the noise reduction of speech based on LMS, I have conducted a deep study and research about the modulation and demodulation principle of LMS .by using the simulation software MATLAB,I design and emulate the modulation and demodulation of the modulation and demodulation principle of LMS, and the simulation results are analyzed. Key word: LMS；microphone；Noise reduction of speech；MATLAB 目 录 摘 要 I Abstract II 1. 前言 1 2. 基本原理及数学模型 2 2.1 语音与噪声的基本知识 2 2.1.1 语音与噪声的特性 2 2.1.2 带噪语音的模型 3 2.2 自适应噪声抵消原理 4 2.2.1 滤波器结构 4 2.2.2 最佳滤波准则 5 2.3 LMS自适应算法 6 2.3.1 权向量的收敛 9 2.3.2 性能指标 9 2.3.3 性能分析 10 2.4 LMS降噪系统设计 11 3. 仿真设计与实现方法 14 3.1 仿真环境介绍 14 3.2 MATLAB中的算法实现 16 3.3 MATLAB中的仿真实现 17 3.3.1 读取语音信号 17 3.3.2 播放语音信号 17 3.3.3 语音信号波形实现 17 3.3.4 语音信号频谱实