《所谓数字信号处理就是指能够在计算机或DSP芯片上处理信....doc》由会员分享，可在线阅读，更多相关《所谓数字信号处理就是指能够在计算机或DSP芯片上处理信....doc（3页珍藏版）》请在taowenge.com淘文阁网|工程机械CAD图纸|机械工程制图|CAD装配图下载|SolidWorks_CaTia_CAD_UG_PROE_设计图分享下载上搜索。
1、2022年-2023年建筑工程管理行业文档 齐鲁斌创作1.1介绍所谓数字信号处理就是指能够在计算机或DSP芯片上处理信号使用代码的任何东西。为了减少在信号幅度上一定的正弦信号部分的频率，数字滤波器应运而生了。人们总想获得完整的信号。如果信号来自于汽车的速度计，完整的信号就能给出汽车的位置。如果信号是噪声，那么为了减小噪声频率幅度而过滤信号就能改善信号的质量。例如，噪声也许发生在户外音乐表演时的风或者雨。滤除非音乐自己产生的那些频率的信号的正弦部分能够使记录的音乐中几乎没有风和雨的噪声。有时信号并不是被噪声干扰而是被其它频率的无效信号干扰。如果信号是对使用从外部的贴近头部的探针取得的脑电波的电子
2、测量，其它电信号也被探针取出，但是物理学家也许只对某个特殊频率的信号感兴趣。通过使用数字滤波器，物理学家感兴趣的信号才能被表达。1.2历史的视角最初信号处理只是在模拟信号处理中处理模拟或时间连续信号。直到19世纪50年代末，数字计算机还没有在商业中使用。当它们确实在商业中使用时，它们又笨重又昂贵，它们被用于模拟信号处理的演示仿真，以判断信号的有效性。然而，这些仿真使仿真或演示的数字处理器编码与模拟系统在信号样本中所做的任务几乎一样。一段时间后，人们意识到模拟系统的仿真编码其实是一个作用于离散时间间隔的输入和输出样本的数字信号处理系统。但是进行数字信号处理而不使用模拟系统仍然是一个问题。第一个问
3、题是一个输入的模拟信号必须被重新表达为信号取样的序列，然后这些序列被转换成计算机的数字表达式。相同的处理也必须被应用到数字化信号的输出。第二个问题是：因为处理的过程是在又笨重又缓慢而且昂贵的计算机上进行的，所以信号样本之间实际的即时处理是不可能的。最后，正如我们在第9章中将看到的，即使为了充分地表达输入信号，在输入样本进行足够快的数字处理，高抽样率也比慢的数字处理要求更精确的比特。更快、更便宜和更小的输入信号取样器和数字数据与模拟数据的输出转换器的发展开始成功地进行即时处理的数字信号处理实践。同时，处理器正变得更小、更快、更便宜而且能处理更多比特。即时处理替代模拟系统也许能相当小、便宜而且精确
4、，能够以一定的足够的抽样速率处理模拟信号。然而，对数字信号处理系统编码的测试和更改使没有模拟信号处理等价物的数字信号处理系统有时比发达到替代模拟系统的数字信号编码更好地演示信号处理过程。对于数字滤波，这些处理方法将在第10章和第11章中讨论。1.1 IntroductionDigital signal processing (DSP) refers to anything that can be done to a signal using code on a computer or DSP chip. To reduce certainsinusoidal frequency compone
5、nts in a signal in amplitude, digital filtering is done. One may want to obtain the integral of a signal. If the signal comes from a tachometer, the integral gives the position. If the signal is noisy, then filtering the signal to reduce the amplitudes of the noise frequencies improves signal qualit
6、y. For example, noise may occur from wind or rain at an outdoor music presentation. Filtering out sinusoidal components of the signal that occur at frequencies that cannot be produced by the music itself results in recording the music with little wind and rain noise. Sometimes the signal is corrupte
7、d not by noise, but by other signal frequencies that are of no present interest. If the signal is an electronic measurement of a brain wave obtained by using probes applied externally to the head, other electronic signals are picked up by the probes, but the physician may be interested only in signa
8、ls occurring at a particular frequency. By using digital filtering, the signals of interest only can be presented to the physician.1.2 Historical PerspectiveOriginally signal processing was done only on analog or continuous timesignals using analog signal processing (ASP). Until the late 1950s digit
9、alcomputers were not commercially available. When they did become commercially available they were large and expensive, and they were usedto simulate the performance of analog signal processing to judge itseffectiveness. These simulations, however, led to digital processor codethat simulated or perf
10、ormed nearly the same task on samples of thesignals that the analog systems did on the signal. After a while it wasrealized that the simulation coding of the analog system was actually aDSP system that worked on samples of the input and output at discretetime intervals.But to implement signal proces
11、sing digitally instead of using analogsystems was still out of the question. The first problem was that an analoginput signal had to be represented as a sequence of samples of the signal,which were then converted to the computers numerical representation.The same process would have to be applied in
12、reverse to the output ofthe digitally processed signal. The second problem was that because theprocessing was done on very large, slow, and expensive computers,practical real-time processing between samples of the signal was impossible. Finally, as we will see in Chapter 9, even if digital processin
13、g could be done quickly enough between input samples in order to adequately represent the input signal, high sample rates require more bits of precision than slower ones.The development of faster, cheaper, and smaller input signal samplers(ADCs) and output converters from digital data to analog data
14、 (DACs)began to make real-time DSP practical. Also, the processors werebecoming smaller, faster, and cheaper and used more bits. Real-timereplacements for analog systems may be just as small, cheap, and accurateand be able to process at a sample rate adequate for many analog signals.However, testing
15、 and modification of the coding for DSP systems led toDSP systems that have no analog signal processing equivalents, yetsometimes perform the signal processing better than the DSP codingdeveloped to replace analog systems. For digital filtering, these processing methods are discussed in Chapters 10 and 11.