IIR 滤波器设计外文翻译资料

 2023-01-12 10:01

IIR 滤波器设计

摘要:我们现在考虑各种数字滤波器的设计方法。FIR和IIR,这两种类型的滤波器有不同的设计方法,因此将被分开考虑。在本章中,我们将重点关注设计IIR或递归滤波器的技术。我们将从简单介绍滤波器的要领开始。

关键词:IIR; FIR;滤波器

4.1 章预览

我们现在考虑各种数字滤波器的设计方法。FIR和IIR,这两种类型的滤波器有不同的设计方法,因此将被分开考虑。在本章中,我们将重点关注设计IIR或递归滤波器的技术。我们将从简单介绍滤波器的要领开始。

4.2滤波器基本知识

在开始介绍数字滤波器设计之前,我们有必要对滤波器进行一次非常简短的回顾。

理想的滤波器将拥有至少统一的恒定增益通带和恒定零增益阻带。另外,增益应由阻带的零增至在单一频率的通带的高增益,即它应该有一个“砖墙”配置文件。理想的低通、高通、 带通和带阻滤波器的幅度响应如下(b)、(c)和(d)图所示。

要设计一个实用滤波器是不可能的,不管是模拟滤波器还是数字滤波器,它们将具有这些配置文件。例如,图4.2显示了一个实用的低通滤波器的幅频响应特性。通带和阻带幅频曲线不是十分平坦,这两个区域之间的“肩”是非常滚圆的,它们之间的过渡,“滚降”区域发生在很宽的频率范围内。越接近我们所要求的符合理想特性的滤波器,滤波的传递函数就越为复杂。

实际中的滤波增益不会在带通和阻通之间垂直下降,因此我们需要确定“截止”滤波频率即有效终止通带的一些方法。我们选择的点是“-3分贝”,这是在该增益下降3分贝,或至其最大值的频率(gain in )。

如果你对模拟滤波器的基本原则比较生疏,这将是一个很好的做一些背景阅读的机会。要查找的一些关键字:低通、高通、带阻、带通、截止频率、滚降、第一、第二(等等),以、无源和有源滤波器、波特图和分贝。 Howatson(1996)是丰富的将相关的电路原理和分析文本之一。

在设计模拟滤波器时,我们已经做了大量的工作,结果我们得出了非常高规格的可用于模拟滤波器的标准设计方程。然而,如前面已在这本书中所强调的一样,所有的模拟系统特性的改变是由温度变化和老化引起的,两个模拟系统执行结果不可能相同。而数字滤波器就没有这些缺陷。它们也比模拟滤波器更为灵活,因为它们是可程控的。

现在,我们将着眼于设计数字滤波器的各种方法。

4.3 FIR和IIR滤波器

在前面的章节中我们说到,数字滤波器大致可分为两种类型:有限脉冲响应(FIR)和无限脉冲响应(IIR)滤波器。如果一个单脉冲被用作输入的FIR滤波器的输出脉冲持续一段有限的时间,然而由IIR滤波器的输出,从理论上说,将会持续不断。一般地,FIR滤波器是非递归的,即不使用反馈,而IIR却要使用反馈。应当指出,在此阶段,我们有可能设计出一些递归的FIR滤波器。然而,为了方便起见,在这本书中, FIR滤波器将指非递归滤波器。

FIR滤波器传递函数的一般表达式T(z)为:

而IIR滤波器传递函数的一般表达式为:

尽管它们有需要更多的系数去实现类似的滤波器性能的缺点,但是FIR滤波器确实具有的优点是它们决不会不稳定,不会像IIR滤波器一样。FIR滤波器也具有线性相位反应的潜力,但将在后面详细介绍。主要的一点是,这两种类型的滤波器在它们的性能以及设计方面都有很大的不同。在本章中,我们只讨论IIR滤波器的设计;FIR滤波器将在下一章中加以考虑。

设计IIR滤波器有几种方法。一种普遍的方法是取一个标准的模拟滤波器,如巴特沃思、切比雪夫、贝塞尔等,并转换到这一点的离散等价滤波器。另一种方法是由z-平面p-z图开始,试图确定极点和零点,以产生所需的频率响应。这就是通常说的“直接”设计法。现在我们将仔细讨论这两种方法。

4.4 IIR 滤波器的直接设计

用这种方法,极点和零点被放在z平面上,以试图达到所要求的频率响应。当以这种方式设计数字滤波器时,会涉及到“测试和出错”的一个要素。通常这很小,而只是用来“tweak”所计算的零极点位置,从而改善了反应,但在其他时候,它很大——通常是难以准确计算最佳零极点位置的时候。越来越被广泛使用的CAD软件包,如MATLAB,已明显地令对测试和出错的依赖性更加可行。这些软件包允许我们进行仿真,并且能非常迅速地改善我们的设计。

合适的p-z图一旦建立,该系统传递函数可以得到。使用数字滤波器的众多乐趣之一是,传递函数转换成实际的滤波器,也就是将其转换成相应的DSP软件是非常容易实现的。这点远非模拟滤波器能比的。在这里,一旦我们到达适当的p-z图,则相应的传递函数就需要被转换成合适的硬件,即电气元件。这是一个更加困难的任务!

这种IIR滤波器的“直接” 设计法最好是通过实施例的方式来描述,下面将举两个例子。

例4.1

要求一个低通滤波器,有1直流增益和为采样频率0.25的截止频率,该滤波器具有如下形式的传递函数.

解决方案

我们的首要任务是找到一个零点和极点。为了完成设计,我们要计算一个合适的k值。由于我们需要一个低通滤波器,因此在初始通带后,增益必须随着频率的增加而下降,也就是随着我们绕z-平面的单位圆以逆时针方向运动时,在z =1(d.c.)开始。

记住公式

因此,令在Nyquist频率的零点距离为零的做法是明智的。换句话说,我们需要把滤波器的零点放在z=-1处。

.. a = 1

常识表明,为了达到所要求的反应,单极可能需要被放在正实轴的某处,这是为了确保低频率的高增益对和高频率的低增益。在图4.3中,零点为z =-1和极点为z= d。点A对应于0 赫兹和点B对应于截止频率,它等于Nyquist频率的二分之一(四分之一采样频率)。

使用公式

在0 赫兹 和 0.5 的增益由一下公式给出:

但 B点的值为-3 分贝,因此

因此极点必须放在原点。

应用公式

A点处:

把a、b、k的值代入传递公式,我们得到:

图4.4 表示 -3 分贝点的值为0.5 ,并且直流增益为1(0 分贝), 所以我们的设计是符合滤波器规范的。

我们需要的数字陷波滤波器有20 赫兹 的频率,不超过4赫兹的带宽和陷波频率至少为40 分贝的衰减。在通带内的增益为1,采样频率为160 赫兹。

陷波滤波器是一种带宽很窄的带阻滤波器——随着频率的增大,它的增益会先急剧下降而后急剧上升,因此幅度响应具有切口的形状。

如公式

接下来我们必须令陷波频率的“零点距离”非常小。正如说明书中规定的,衰减必须至少40分贝,然后做最容易的事,是把一个零点在单位圆上对应于20 赫兹陷波频率的位置。理论上,这将导致增益下降到零。当然,这个z-平面零点是两种共轭复数零之一。

因为采样频率为160赫兹,Nyquist频率是80赫兹,因此20赫兹的陷波频率对应于z-平面图上的一个角。图4.5示出了零,zd,以及它的共轭复数零,放置在单位圆对应于该频率的位置上。现在,我们需要放置复共轭极点,以获得不超过4赫兹的正确带宽。换言之,增益必须下降到-3 分贝或约至其通带增益,频率为18赫兹和22赫兹。

由频率响应的对称性,把极点和零点半径设为相等的似乎是合理。两极将必须充分接近相应的零点,以确保到“缺口”的增益在所有的频率范围内大致统一。因为它们是如此接近,则要求k = 1,以得到1的通带增益。他们的位置也必须如此,以使得带宽为4赫兹。我们可能会发现,本说明书中是不可能实现的。

如图4.5中所示,两极到两个零点的距离为d。B和A这两个- 3分贝点,频率为18赫兹和22赫兹(不按比例)。

ACB角对应于4赫兹,因此,=-80赫兹,rad

使用公式,其中s是一个圆的圆弧,为半径,弧所对应的角为,然后,当时,弧长AB也必为0.16,所以ZA=ZB=0.08 。

由于点A非常接近点Z,极点、零点和点A(或点B)构成一个近似的直角三角形,。因此,极点到点A或点B 的距离必定非常接近。

如公示

k=1,

因此,这些极点到原点的距离必为0.92。

接下来,我们必须设置零点为,即,极点为或。

所以

N.B. 另一极点和零点的影响在计算中不予考虑。这是因为他们是如此接近,往往相互抵消,所以对增益的影响应该不会很明显。

该设计已经完成;所有剩下的工作就是检查响应——MATLAB图形如Fig.4.6所示。由图可知,该幅度响应满足本说明书的要求,有0.25(20赫兹)的陷波频率,陷波的衰减是至少40分贝,带宽约0.05(4赫兹)。

如果我们发现,该反应不那么令人满意,在MATLAB的辅助下则可以很容易地 “tweaked” 该磁极位置,以提高性能。

问题

如果只用零点设计IIR滤波器,这有可能吗?

答案

绝对不可能。

这种方法有一个致命的缺陷,并且注定要失败的。从理论上讲,虽然把零的p-z图上以满足幅度响应是有可能的,但实际上,离散系统的传递函数的零点要比极点多,这是不可能的。这样的系统会在接收输入信号之前产生输出信号!如果你不相信,请看第三章的结尾处的问题4。

N.B.实际上,如果能更多地借助MATLAB确定极点和零点的位置,那我们在这些例子中所做的计算可能过多了。例如,一旦我们算出了“缺口”问题的零位,我们可以通过反复的试误法找到合适的位置,这将省掉大量的工作。然而,了解设计过程中的基本原则是重要的,因此额外的计算是值得的,这样想你会不会感觉好多了?

introduction to digital signal processing,2000,pages 71-78.

4 The design of IIR filters

4.1 CHAPTER PREVIEW

We are now in a position to consider the various approaches to the design of digital filters. The two types of filter, FIR and IIR, have very different design methods and so will be considered separately. In this chapter we will focus on the techniques used to design IIR or recursive filters. We begin with a very brief resume of filter essentials.

4.2 FILTER BASICS

Before we move on to the design of digital filters, it is probably worth having a very brief recap on filters.

An ideal filter will have a constant gain of at least unity in the passband and a constant gain of zero in the stopband.Also, the gain should increase from the zero of the stopband to the higher gain of the passband at a single frequency, i.e. it should have a brick wall profile. The magnitude responses of ideal lowpass, highpass, bandpass and bandstop filters are as shown in Fig. 4.1(a), (b), (c) and (d).

It is impossible to design a practical filter, either analogue or digital, that will have these profiles. Figure 4.2, for example, shows the magnitude response for a practical lowpass filter. The passband and stopband are not perfectly flat, the shoulder between these two regions is very rounded and the transition between them, the roll-off region, takes place over a wide frequency range. The closer we require our filter to agree with the idea

剩余内容已隐藏,支付完成后下载完整资料


4 The design of IIR filters

4.1 CHAPTER PREVIEW

We are now in a position to consider the various approaches to the design of digital filters. The two types of filter, FIR and IIR, have very different design methods and so will be considered separately. In this chapter we will focus on the techniques used to design IIR or recursive filters. We begin with a very brief resume of filter essentials.

4.2 FILTER BASICS

Before we move on to the design of digital filters, it is probably worth having a very brief recap on filters.

An ideal filter will have a constant gain of at least unity in the passband and a constant gain of zero in the stopband.Also, the gain should increase from the zero of the stopband to the higher gain of the passband at a single frequency, i.e. it should have a brick wall profile. The magnitude responses of ideal lowpass, highpass, bandpass and bandstop filters are as shown in Fig. 4.1(a), (b), (c) and (d).

It is impossible to design a practical filter, either analogue or digital, that will have these profiles. Figure 4.2, for example, shows the magnitude response for a practical lowpass filter. The passband and stopband are not perfectly flat, the shoulder between these two regions is very rounded and the transition between them, the roll-off region, takes place over a wide frequency range. The closer we require our filter to agree with the ideal characteristics, the more complicated is the filter transfer function.

As the gain of a real filter does not drop vertically between the passband and stopband, we need some way of defining the cut-off frequencies of filters, i.e. the effective end of the passband. The point chosen is the -3 dB point. This is the frequency at which the gain has fallen by 3 dB, or to of its maximum value (gain in ).

If you are rusty on the basic principles of analogue filters, this would be a good time to do some background reading. Some keywords to look for are: lowpass, highpass, bandstop, bandpass, cut-off frequency, roll-off, first, second (etc.) order, passive and active filters, Bode plots and dB. Howatson (1996) is just one of an abundance of circuit theory and analysis texts which will be relevant.

Much work has been carried out into the design of analogue filters and, as a result, standard design equations for analogue filters with very high specifications are available. However, as has been stressed earlier in this book, the characteristics of all analogue systems alter due to temperature changes and ageing. It is also impossible for two analogue systems to perform identically. Digital filters do not have these defects. They are also much more versatile than analogue filters in that they are programmable.

We will now look at various methods of designing digital filters.

4.3 FIR AND IIR FILTERS

You will remember from earlier chapters that digital filters can be divided broadly into two types-finite impulse response(FIR) and infinite impulse response (IIR) filters. If a single pulse is used as the input for an FIR filter the output pulses last for a finite time, while the output from an IIR filter will, theoretically, continue for ever. Generally, FIR filters are non-recursive, i.e. do not use feedback, while IIR do. It should be pointed out at this stage that it is possible to design some FIR filters which are recursive. However, for convenience, the term FIR filter will be taken to mean a non-recursive filter in this book.

The general expression for the transfer function, , of an FIR filter is:

while that for the IIR filter is:

Although they have the disadvantage of requiring more coefficients to achieve a similar filter performance, FIR filters do have the advantage that they will never be unstable, unlike IIR filters. FIR filters also have the potential to feature a linear phase response-but more on this later. The main point is that the two types of filter are very different in their performance and also in their design. In this chapter we will deal with the design of IIR filters only; FIR filters are considered in the next chapter.

There are several approaches that can be taken to the design of IIR filters. One common method is to take a standard analogue filter, e.g. Butterworth, Chebyshev, Bessell, etc., and convert this into its discrete equivalent. An alternative procedure is to start with the z-plane p-z diagram and attempt to place poles and zeros so as to produce the desired frequency response. This is often called the direct design method. We will now look at these two approaches in detail.

4.4 THE DIRECT DESIGN OF IIR FILTERS

With this method, poles and zeros are placed on the z-plane in an attempt to achieve the required frequency response. When designing digital filters in this way there is sometimes an element of trial and error involved. Often this is small and just serves to tweak the calculated pole-zero positions, so as to improve the response, but at other times it is significant- this is usually when it proves very difficult to calculate the optimum pole-zero positions accurately. The reliance on trial and error has obviously been made much more viable with the increased availability of CAD packages such as MATLAB. These packages allow us to simulate and then modify our designs very quickly.

Once a suitable p-z diagram has been established, the system transfer function can be derived. One of the many pleasures of using digital filters is that it is very easy to convert the transfer function into the actual filter, i.e. to convert it into the corresponding DSP software. This is far from true with analogue filters. Here, once we arrive at a suitable p-z diagram, the corresponding transfer function then needs to be converted into suitable hardware- i.e. electrical components. This is a much more difficult task!

This direct method of

剩余内容已隐藏,支付完成后下载完整资料


资料编号:[287053],资料为PDF文档或Word文档,PDF文档可免费转换为Word

原文和译文剩余内容已隐藏,您需要先支付 30元 才能查看原文和译文全部内容!立即支付

以上是毕业论文外文翻译,课题毕业论文、任务书、文献综述、开题报告、程序设计、图纸设计等资料可联系客服协助查找。