基于DM642的窄带噪声主动控制系统的实现

发布时间：2018-02-04 00:09

本文关键词： 窄带主动噪声控制 FXLMS算法 DM642 多通道设计　出处：《哈尔滨工业大学》2016年硕士论文　论文类型：学位论文

【摘要】：噪声污染在当今已变得越来越严重,影响着人们的正常生活与生产。很多噪声是由旋转设备产生的,如管道风机、汽车引擎、螺旋桨等,而这些噪声主要是低频噪声,传统被动消噪技术难以解决这一问题。因此主动噪声控制技术(Active Noise Control,ANC)被提出,该技术主要针对窄带低频段噪声的消噪。窄带低频噪声的降噪技术研究需要从理论和实验入手,理论方面的研究已经相对成熟,近年来众多的ANC算法被提出并通过计算机仿真。然而,受制于电子硬件技术的发展,实际应用方面并不理想。本文以成熟的FXLMS自适应滤波算法作为ANC技术的主算法,研究了基于TMS320DM642的ANC系统硬件实现过程。主要研究工作有如下几个方面:首先,研究了主动噪声控制技术的理论知识,包括管道声场理论-分析管道声场的消噪的原理、自适应滤波理论、三种类型控制器的比较和自适应滤波算法。分析对比后选择出合适的控制器类型,最后对本课题采用的有源噪声控制算法-FXLMS算法做了详细介绍并分析其实现过程,且对算法做了计算机仿真。其次,给出ANC控制器硬件总体设计方案,详细具体地研究作为ANC控制器的TMS320DM642的硬件知识,对比分析其在作为ANC控制器上的优势,根据需要设计出ANC控制器最小系统板。除此之外,还研究其外围接口Mc ASP及音频芯片TLV320AIC23B的具体特性、通讯协议及使用方法,为了后期研究需要,完成多通道噪声采集电路板的设计。然后,完成基于TMS320DM642的ANC系统软件程序的编写。首先介绍了ANC软件的开发流程,给出其算法的程序流程图,分别对基于FXLMS算法的次级通道在线、离线辨识算法仿真对比,阐述其应用场景和优缺点。搭建管道硬件平台,编写上位机程序用于初级噪声产生和误差噪声监控。最后,合理选取初级噪声源,DSP只负责消噪,完成管道窄带噪声降噪实验,包括单频、多频降噪实验。对比分析实验结果,指出实验中遇到的问题与解决方案,对后续研究提出新要求。
[Abstract]:Noise pollution has become more and more serious nowadays, which affects people's normal life and production. A lot of noise is produced by rotating equipment, such as pipe blower, automobile engine, propeller and so on. But these noises are mainly low frequency noise, which is difficult to be solved by traditional passive noise elimination technology, so active Noise Control (ANC) is proposed. This technology is mainly aimed at the noise elimination of narrow band low frequency band noise. The research on the noise reduction technology of narrow band low frequency noise needs to start from theory and experiment, and the theoretical research has been relatively mature. In recent years, many ANC algorithms have been proposed and simulated by computer. However, they are restricted by the development of electronic hardware technology. The mature FXLMS adaptive filtering algorithm is used as the main algorithm of ANC technology. The hardware implementation process of ANC system based on TMS320DM642 is studied. The main research work is as follows: firstly, the theoretical knowledge of active noise control technology is studied. It includes the theory of pipeline sound field, the principle of analyzing the noise of pipeline sound field, the theory of adaptive filtering, the comparison of three types of controllers and the adaptive filtering algorithm. Finally, the active noise control algorithm-FXLMS is introduced in detail and its realization process is analyzed, and the algorithm is simulated by computer. Secondly. The hardware design scheme of ANC controller is given, the hardware knowledge of TMS320DM642 as ANC controller is studied in detail, and the advantages of TMS320DM642 as ANC controller are compared and analyzed. The minimum system board of ANC controller is designed according to the need. In addition, the specific characteristics of the peripheral interface MC ASP and audio chip TLV320AIC23B are also studied. Communication protocol and usage method, for later research needs, complete the design of multi-channel noise acquisition circuit board. Then. The ANC system software program based on TMS320DM642 is completed. Firstly, the development process of ANC software is introduced, and the program flow chart of its algorithm is given. The secondary channel online and off-line identification algorithms based on FXLMS algorithm are simulated and compared, and their application scenarios, advantages and disadvantages are described. The pipeline hardware platform is built. Finally, selecting the primary noise source DSP is only responsible for noise elimination, complete the pipeline narrow band noise noise reduction experiment, including single frequency. The experimental results of multi-frequency noise reduction are compared and analyzed. The problems and solutions encountered in the experiments are pointed out, and new requirements for further research are put forward.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TB535

【相似文献】