基于压缩感知理论的数据采集系统研究与设计

发布时间：2018-06-27 23:30

本文选题：数据采集 + 压缩感知　；参考：《西南石油大学》2015年硕士论文

【摘要】：随着时代的发展,现代数据采集系统需要处理的信息量越来越大。随着信号频率的增大,带宽越来越宽和采样点数的增加,采样频率将会越来越高,最后采集到的数据量将会越来越大,不利于后续的信号处理及存储,导致在某些应用场合产生局限性。模拟数字转换器(ADC)作为数据采集系统的核心,受到技术、工艺等方面的制约,其采样频率并不能做到很高,且采样频率越高的ADC芯片价格越贵,限制了采样系统对宽带高频信号的采集。近年来兴起的压缩感知理论为解决数据量的剧增与硬件信号处理的限制提供了新的思路。本文针对压缩感知理论及其硬件实现展开研究,具有十分重要的实际应用价值。本文将压缩感知理论引入到高频信号的数据采集系统中,主要完成了三个方面的工作。首先,论文分析了传统的奈奎斯特采样定理在某些应用场合的局限性,由此引出了压缩感知理论并对其基本原理进行了详细介绍,分析了该理论的三大关键要素,通过对离散信号的MATLAB仿真,验证了压缩感知理论的可行性。其次,论文介绍了基于压缩感知理论的模拟信息转换技术,详细分析了随机解调转换器模型和调制宽带转换器模型的基本原理。给出了系统评价指标,并根据系统要求对随机解调转换器模型进行了MATLAB仿真研究。最后,结合理论研究,通过对比分析上述两种模型,设计了以随机解调转换器模型框架,以FPGA为控制核心的数据采集硬件系统。该系统主要包括模拟信号处理部分(混频电路、滤波电路、m伪随机序列发生器)和数字信号处理部分(FPGA主控单元、AD转换电路、存储电路),通过对硬件系统的调试,验证了压缩感知理论在对信号的低速采样的应用方面的可行性。与传统的数据采集流程不同,压缩感知理论将信号的压缩与采集合二为一,在采样之前就对信号进行了压缩,让我们可以以低于原信号的奈奎斯特采样频率进行采样,从而降低了信号的采样频率,减轻了模数转换的压力及硬件成本。本文实现了以FPGA为主控器件基于压缩感知理论的信号采集系统的开发,通过MATLAB仿真验证以及硬件调试实现了对信号的低速采样。
[Abstract]:With the development of the times, the modern data acquisition system needs to deal with more and more information. With the increase of signal frequency, the bandwidth and the number of sampling points increase, the sampling frequency will become higher and higher, and the amount of data collected will become larger and larger, which is not conducive to the subsequent signal processing and storage. Lead to limitations in some applications. As the core of data acquisition system, analog digital converter (ADC) is restricted by technology, technology and so on. The sampling frequency of ADC is not high, and the higher the sampling frequency is, the more expensive the ADC chip is. The acquisition of wideband high frequency signals by sampling system is limited. The theory of compression sensing developed in recent years provides a new way to solve the problem of increasing data volume and limiting hardware signal processing. This paper focuses on the theory of compression sensing and its hardware implementation, which has very important practical application value. In this paper, the theory of compression sensing is introduced into the data acquisition system of high frequency signal. Firstly, the limitation of the traditional Nyquist sampling theorem in some applications is analyzed, and the theory of compression perception is introduced in detail, and the three key elements of the theory are analyzed. The feasibility of compression sensing theory is verified by MATLAB simulation of discrete signals. Secondly, the paper introduces the technology of analog information conversion based on compression sensing theory, and analyzes the basic principle of random demodulation converter model and modulation broadband converter model in detail. The evaluation index of the system is given, and the random demodulation converter model is simulated by MATLAB according to the system requirements. Finally, by comparing and analyzing the above two models, a data acquisition hardware system based on random demodulation converter and FPGA is designed. The system mainly includes analog signal processing part (mixing circuit, filter circuit) and digital signal processing part (FPGA main control unit, AD conversion circuit, storage circuit). The feasibility of the compression sensing theory in the application of low-speed signal sampling is verified. Different from the traditional data acquisition process, the compression sensing theory combines the compression and acquisition of the signal and compresses the signal before sampling, so that we can sample at the Nyquist sampling frequency lower than the original signal. Thus, the sampling frequency of signal is reduced, the pressure of A / D conversion and the cost of hardware are reduced. In this paper, a signal acquisition system based on compression sensing theory based on FPGA is developed, and the low speed signal sampling is realized by MATLAB simulation and hardware debugging.
【学位授予单位】：西南石油大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TP274.2;TN911.7

【相似文献】