基于DSP的自适应消噪器研究与设计

发布时间：2018-03-30 08:11

本文选题：自适应算法　切入点：自适应消噪器　出处：《西南交通大学》2014年硕士论文

【摘要】：自适应消噪器是基于自适应滤波技术的一种运用,它能在时变、非平稳噪声背景环境中有效消除噪声干扰,在自适应信号处理领域有着广泛运用。随着大规模数字集成电路和计算机技术的发展,特别是高性能数字信号处理器的出现,为自适应消噪器的构建提供硬件支持。本文在研究自适应滤波技术及相关算法的基础上,致力于构建一个基于DSP的自适应消噪硬件平台,并运用于实际噪声环境中语音信号消噪处理,为自适应消噪技术的硬件实现与运用做出有益尝试。本文采用理论分析、软件仿真与硬件设计实现相结合的研究方法。在研究自适应算法的基础上,用软件仿真验证相关参数对自适应算法性能的影响,并通过建模仿真将自适应算法运用于自适应消噪实验中,通过消噪结果对比算法的性能。最后基于DSP构建自适应消噪硬件平台,完成自适应消噪器的硬件实现。主要完成的工作有： 1.对自适应算法进行研究,重点研究了LMS算法和RLS算法的推导过程,并对影响LMS算法性能指标的因素进行分析。针对LMS算法在收敛速度、稳态误差对步长因子取值要求的矛盾性,研究了几种变步长算法,在此基础上提出一种改进的变步长算法。 2.通过在MATLAB环境中仿真,验证了改进变步长算法在自适应噪声抵消的运用中,具有良好的性能,能有效提高消噪系统的消噪性能。 3.采用SIMULINK建模仿真,建立了基于LMS算法和RLS算法的自适应消噪实时仿真模型,通过实时仿真结果确定了基于所建模型的最优算法参数选取。 4.采用TI公司2000系列DSP芯片TMS320F28335为核心,构建了一个自适应消噪最小硬件系统,完成软硬件电路的设计与调试,并通过一个实际噪声环境中的语音信号去噪测试实验,验证了所设计的系统能实现自适应消噪功能。
[Abstract]:Adaptive noise canceller is a kind of application based on adaptive filtering technology. It can effectively eliminate noise interference in time-varying, non-stationary noise background environment. With the development of large scale digital integrated circuit and computer technology, especially the emergence of high performance digital signal processor, Based on the research of adaptive filtering technology and related algorithms, this paper is devoted to constructing a hardware platform of adaptive denoising based on DSP. And it is applied to the speech signal de-noising processing in the actual noise environment, which makes a beneficial attempt for the hardware realization and application of the adaptive de-noising technology. Based on the research of adaptive algorithm, the influence of relevant parameters on the performance of adaptive algorithm is verified by software simulation. Through modeling and simulation, the adaptive algorithm is applied to the experiment of adaptive denoising, and the performance of the algorithm is compared with the results of de-noising. Finally, the hardware platform of adaptive de-noising is constructed based on DSP. The hardware implementation of the adaptive de-noising device is completed. The main work accomplished is as follows:. 1. The adaptive algorithm is studied, the derivation process of LMS algorithm and RLS algorithm is studied, and the factors influencing the performance of LMS algorithm are analyzed. Based on the contradiction of steady-state error to the requirement of step size factor, several variable step size algorithms are studied, and an improved variable step size algorithm is proposed. 2. Through the simulation in MATLAB environment, it is proved that the improved variable step size algorithm has good performance in the application of adaptive noise cancellation, and it can effectively improve the denoising performance of the de-noising system. 3. The real-time simulation model of adaptive de-noising based on LMS algorithm and RLS algorithm is established by using SIMULINK modeling and simulation. The parameters of the optimal algorithm based on the established model are determined by the real-time simulation results. 4. Using TI 2000 series DSP chip TMS320F28335 as the core, a minimum hardware system for adaptive de-noising is constructed. The design and debugging of the hardware and software circuit are completed, and the test experiment of speech signal de-noising in a real noise environment is carried out. It is verified that the designed system can realize adaptive de-noising function.
【学位授予单位】：西南交通大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TB535.2

【参考文献】