窄带主动噪声控制系统实时仿真及硬件实现

发布时间：2018-12-07 21:22

【摘要】：噪声是由物体振动产生的对人造成干扰的声音。窄带主动噪声控制(Active Noise Control-ANC)系统针对能量集中在有限个频率的窄带噪声进行主动控制降噪。当采用非声学传感器获取参考信号时,窄带ANC系统中往往存在频率不匹配(Frequency Mismatch-FM),极大削弱系统的降噪性能。此外,初级噪声信号频率发生跳变,即非平稳也是存在于系统中的另一问题。本文分别从理论仿真和硬件实现对窄带ANC进行研究。首先,在理论仿真部分,利用MATLAB/Simulink的实时仿真功能,构建了针对前馈窄带ANC系统的实时仿真平台。对传统窄带ANC控制系统进行噪声抑制仿真分析,包括系统存在FM时系统的性能评价。针对FM问题对传统FM补偿方法以及一种改进的FM补偿方法进行了实时仿真并进行了性能比较。进一步,在FM补偿系统中引入了非平稳的情况,对两种FM补偿方法进行了比较,对其性能进行评估。其次,对窄带ANC系统进行硬件实现设计。在实时仿真的基础上,设计基于高速浮点型C6000系列DSP芯片的窄带主动噪声控制器,以一维管道为实验平台,实现对初级和次级噪声信号的采集转换、次级通道辨识、相关ANC算法的处理,使次级噪声信号与初级噪声信号相消干涉,实现降噪目的。分别对传统窄带前馈ANC系统及基于自回归(Auto Regressive-AR)模型的频率补偿系统进行实验,验证了两种系统的有效降噪性。
[Abstract]:Noise is the noise produced by the vibration of an object that interferes with human beings. The narrow band active noise control (Active Noise Control-ANC) system is used to reduce the noise of narrow band noise which is concentrated on a limited frequency. When the reference signal is obtained by the non-acoustic sensor, the frequency mismatch (Frequency Mismatch-FM) often exists in the narrowband ANC system, which greatly weakens the noise reduction performance of the system. In addition, the frequency jump of the primary noise signal, that is, non-stationary, is another problem in the system. In this paper, narrowband ANC is studied by theoretical simulation and hardware implementation. Firstly, in the theoretical simulation part, a real-time simulation platform for feedforward narrowband ANC system is constructed by using the real-time simulation function of MATLAB/Simulink. The noise suppression simulation of the traditional narrow band ANC control system is carried out, including the performance evaluation of the system with FM. Aiming at the FM problem, the traditional FM compensation method and an improved FM compensation method are simulated in real time and the performance is compared. Furthermore, the non-stationary condition is introduced into the FM compensation system. The performance of the two FM compensation methods is compared and evaluated. Secondly, the hardware implementation of narrowband ANC system is designed. On the basis of real-time simulation, a narrow band active noise controller based on high-speed floating-point C6000 series DSP chips is designed. The acquisition and conversion of primary and secondary noise signals and secondary channel identification are realized by using one-dimensional pipeline as experimental platform. The processing of correlation ANC algorithm eliminates the interference between the secondary noise signal and the primary noise signal and realizes the purpose of noise reduction. Experiments on the traditional narrowband feedforward ANC system and the frequency compensation system based on autoregressive (Auto Regressive-AR) model are carried out to verify the effective noise reduction of the two systems.
【学位授予单位】：南京航空航天大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TB535

【参考文献】