




版權(quán)說(shuō)明:本文檔由用戶提供并上傳,收益歸屬內(nèi)容提供方,若內(nèi)容存在侵權(quán),請(qǐng)進(jìn)行舉報(bào)或認(rèn)領(lǐng)
文檔簡(jiǎn)介
1、7 Filter Design Techniques7.0 Introduction 7.1 Design of Discrete-Time IIR Filters from Continuous-Time Filters7.2 Design of FIR Filters by Windowing7.6 Comments on IIR and FIR Discrete-Time Filters7.7 Summary7.0 IntroductionFilters are a particularly important class of linear time-invariant systems
2、. The design of filters involves the following stages: (1) the specification of the desired properties of the system, (2) the approximation of the specifications using a causal discrete-time system, and (3) the realization of the system.4、 濾波器的設(shè)計(jì)步驟濾波器的設(shè)計(jì)步驟 按照實(shí)際任務(wù)要求, 確定濾波器的性能指標(biāo)。 用一個(gè)因果穩(wěn)定的離散線性時(shí)不變系統(tǒng)的系統(tǒng)函
3、數(shù)去逼近這一性能要求。根據(jù)不同要求可以用IIR系統(tǒng)函數(shù),也可以用FIR系統(tǒng)函數(shù)去逼近。 利用有限精度算法來(lái)實(shí)現(xiàn)這個(gè)系統(tǒng)函數(shù)。這里包括選擇運(yùn)算結(jié)構(gòu)(如第4章中的各種基本結(jié)構(gòu)),選擇合適的字長(zhǎng)(包括系數(shù)量化及輸入變量、中間變量和輸出變量的量化)以及有效數(shù)字的處理方法(舍入、截尾)等。 As shown in Section 4.4, if a linear time-invariant discrete-time system is used as in Figure 7.1, and if the input is bandlimited and the sampling frequency i
4、s high enough to avoid aliasing, then the overall system behaves as a linear time-invariant continuous-time system with frequency response In such cases,it is straightforward to convert from specifications on the effective continuous-time filter to specifications on the discrete-time filter through
5、the relation . That is is specified over one period by the equation,/0, ,/jTeffHeTHjT T,jeffH eHjTjH e axtTT x n y n aytjH e/C D/D CThis type of conversion is illustrated in Example 7.1Figure7.1 Basic system for discrete-time filtering of continuous-time signals.1 選頻濾波器的分類選頻濾波器的分類 數(shù)字濾波器是數(shù)字信號(hào)處理的重要基礎(chǔ)。
6、在對(duì)信號(hào)的過(guò)濾、檢測(cè)與參數(shù)的估計(jì)等處理中, 數(shù)字濾波器是使用最廣泛的線性系統(tǒng)。 數(shù)字濾波器是對(duì)數(shù)字信號(hào)實(shí)現(xiàn)濾波的線性時(shí)不變系統(tǒng)。它將輸入的數(shù)字序列通過(guò)特定運(yùn)算轉(zhuǎn)變?yōu)檩敵龅臄?shù)字序列。因此, 數(shù)字濾波器本質(zhì)上是一臺(tái)完成特定運(yùn)算的數(shù)字計(jì)算機(jī)。 由第1章1.3節(jié)已經(jīng)知道,一個(gè)輸入序列x(n),通過(guò)一個(gè)單位脈沖響應(yīng)為h(n)的線性時(shí)不變系統(tǒng)后,其輸出響應(yīng)y(n)為 nmnxmhnhnxny)()()()()(將上式兩邊經(jīng)過(guò)傅里葉變換,可得 )()()(jjjeHeXeY式中,Y(ej)、X(ej)分別為輸出序列和輸入序列的頻譜函數(shù), H(ej)是系統(tǒng)的頻率響應(yīng)函數(shù)。 可以看出,輸入序列的頻譜X(ej )
7、經(jīng)過(guò)濾波后,變?yōu)閄(ej)H(ej)。如果|H(ej)|的值在某些頻率上是比較小的,則輸入信號(hào)中的這些頻率分量在輸出信號(hào)中將被抑制掉。因此,只要按照輸入信號(hào)頻譜的特點(diǎn)和處理信號(hào)的目的,適當(dāng)選擇H(ej),使得濾波后的X(ej)H(ej)符合人們的要求,這就是數(shù)字濾波器的濾波原理。和模擬濾波器一樣,線性數(shù)字濾波器按照頻率響應(yīng)的通帶特性可劃分為低通、高通、帶通和帶阻幾種形式。它們的理想模式如圖5-1所示。(系統(tǒng)的頻率響應(yīng)H(ej)是以2為周期的。) 圖 5-1 數(shù)字濾波器的理想幅頻特性 )(ejHo22)(ejHo22)(ejHo22)(ejHo22(a)(b)(c)(d )低通高通帶通帶阻 滿足
8、奈奎斯特采樣定理時(shí),信號(hào)的頻率特性只能限帶于|的范圍。由圖5-1可知,理想低通濾波器選擇出輸入信號(hào)中的低頻分量,而把輸入信號(hào)頻率在cN時(shí),H(z)可看成是一個(gè)N階IIR子系統(tǒng)與一個(gè)(M-N)階的FIR子系統(tǒng)的級(jí)聯(lián)。 以下討論都假定MN。 IIR濾波器的系統(tǒng)函數(shù)的設(shè)計(jì)就是確定各系數(shù)ak, bk或零極點(diǎn)ck,dk和A,以使濾波器滿足給定的性能要求。利用模擬濾波器的理論來(lái)設(shè)計(jì)數(shù)字濾波器 首先,設(shè)計(jì)一個(gè)合適的模擬濾波器;然后,變換成滿足預(yù)定指標(biāo)的數(shù)字濾波器。這種方法很方便,因?yàn)槟M濾波器已經(jīng)具有很多簡(jiǎn)單而又現(xiàn)成的設(shè)計(jì)公式,并且設(shè)計(jì)參數(shù)已經(jīng)表格化了,設(shè)計(jì)起來(lái)既方便又準(zhǔn)確。 利用模擬濾波器來(lái)設(shè)計(jì)數(shù)字濾波器
9、,就是從已知的模擬濾波器傳遞函數(shù)Ha(s)設(shè)計(jì)數(shù)字濾波器的系統(tǒng)函數(shù)H(z)。因此,它歸根結(jié)底是一個(gè)由S平面映射到Z平面的變換,這個(gè)變換通常是復(fù)變函數(shù)的映射變換,這個(gè)映射變換必須滿足以下兩條基本要求: (1)H(z)的頻率響應(yīng)要能模仿Ha(z)的頻率響應(yīng),也即S平面虛軸j必須映射到Z平面的單位圓ej上。(2) 因果穩(wěn)定的Ha(s)應(yīng)能映射成因果穩(wěn)定的H(z),也即S平面的左半平面Res0必須映射到Z平面單位圓的內(nèi)部|z|1。 下面首先分別討論由模擬濾波器設(shè)計(jì)IIR數(shù)字濾波器的兩種常用的變換方法:脈沖響應(yīng)不變法和雙線性變換法,然后介紹一下常用模擬低通濾波器的特性。FIR數(shù)字濾波器的設(shè)計(jì)方法與IIR
10、數(shù)字濾波器設(shè)計(jì)方法明顯不同,這將在下一章中介紹。 5.4 Design of Continuous-Time Filters 常用的模擬原型濾波器有巴特沃思(Butterworth)濾波器、切比雪夫(Chebyshev)濾波器、橢圓(Ellipse)濾波器、貝塞爾(Bessel)濾波器等。這些濾波器都有嚴(yán)格的設(shè)計(jì)公式,現(xiàn)成的曲線和圖表供設(shè)計(jì)人員使用。這些典型的濾波器各有特點(diǎn):巴特沃思濾波器具有單調(diào)下降的幅頻特性;切比雪夫?yàn)V波器的幅頻特性在通帶或者在阻帶有波動(dòng),可以提高選擇性;貝塞爾濾波器通帶內(nèi)有較好的線性相位特性;橢圓濾波器的選擇性相對(duì)前三種是最好的, 但在通帶和阻帶內(nèi)均為等波紋幅頻特性。這樣
11、根據(jù)具體要求可以選用不同類型的濾波器。 圖 5-3 各種理想模擬濾波器的幅頻特性 o低通o帶通o帶阻o高通)j (aH)j(aH)j (aH)j (aH5.5 The Transform from Continuous-Time Low-Pass Filter to Discrete-Time Filter 首先,把數(shù)字濾波器的性能要求轉(zhuǎn)換為與之相應(yīng)的作為“樣本”的模擬濾波器的性能要求,根據(jù)此性能要求設(shè)計(jì)模擬濾波器, 這可以用查表的辦法, 也可以用解析的方法。然后,通過(guò)脈沖響應(yīng)不變法或雙線性變換法,將此“樣本”模擬低通濾波器數(shù)字化為所需的數(shù)字濾波器H(z)。我們討論采用雙線性變換法和脈沖響應(yīng)不
12、變法來(lái)設(shè)計(jì)低通濾波器的過(guò)程。 例例 5-6 用脈沖響應(yīng)不變法設(shè)計(jì)一個(gè)三階巴特沃思數(shù)字低通濾波器,采樣頻率為fs=4 kHz(即采樣周期為T=250s),其3 dB截止頻率為fc=1 kHz。 解解 查表可得歸一化三階模擬巴特沃思低通濾波器的傳遞函數(shù) 322211)(ssssHNa然后, 以s/c代替其歸一化頻率,則可得三階模擬巴特沃思低通濾波器的傳遞函數(shù)為 32)/()/(2)/(211)(cccasssssH式中,c=2fc。上式也可由巴特沃思濾波器的幅度平方函數(shù)求得。 為了進(jìn)行脈沖響應(yīng)不變法變換,將上式進(jìn)行因式分解并表示成如下的部分分式形式: 2/ )31 ()3/(2/ )31 ()3/
13、()(6/6/jsejsessHcjccjccca將此部分分式系數(shù)代入(5-40)式就得到 12/ )31(6/12/ )31(6/11)3/(1)3/(1)(zeezeezezHjjcjjccccc式中,c=cT=2fcT=0.5是數(shù)字濾波器數(shù)字頻域的截止頻率。 將上式兩項(xiàng)共軛復(fù)根合并,得 22/12/1123cos21623cos26cos231)(zeezezzezHcccccccc 從這個(gè)結(jié)果我們看到,H(z)只與數(shù)字頻域參數(shù)c有關(guān),也即只與臨界頻率fc與采樣頻率fs的相對(duì)值有關(guān),而與它們的絕對(duì)大小無(wú)關(guān)。 例如fs=4kHz,fc=1 kHz與fs=40 kHz,fc=10kHz的數(shù)字
14、濾波器將具有同一個(gè)系統(tǒng)函數(shù)。這個(gè)結(jié)論適合于所有的數(shù)字濾波器設(shè)計(jì)。 將c=cT=2fcT=0.5代入上式,得 21112079. 01905. 01551. 0571. 12079. 01571. 1)(zzzzzH 這個(gè)形式正好適合用一個(gè)一階節(jié)及一個(gè)二階節(jié)并聯(lián)起來(lái)實(shí)現(xiàn)。 脈沖響應(yīng)不變法由于需要通過(guò)部分分式來(lái)實(shí)現(xiàn)變換,因而對(duì)采用并聯(lián)型的運(yùn)算結(jié)構(gòu)來(lái)說(shuō)是比較方便的。 圖5-18給出了脈沖響應(yīng)不變法得到的三階巴特沃思數(shù)字低通濾波器的頻響幅度特性,同時(shí)給出例5-5雙線性變換法設(shè)計(jì)的結(jié)果。由圖可看出,脈沖響應(yīng)不變法存在微小的混淆現(xiàn)象,因而選擇性將受到一定損失,并且沒(méi)有傳輸零點(diǎn)。 圖 5-18 三階巴特沃思
15、數(shù)字低通濾波器的頻響 1.00.500.5)(ejH01.02.0f / kHz1:脈沖響應(yīng)不變法2:雙線性變換法12 下面我們總結(jié)利用模擬濾波器設(shè)計(jì)IIR數(shù)字低通濾波器的步驟。 (1)確定數(shù)字低通濾波器的技術(shù)指標(biāo):通帶截止頻率c、通帶衰減c、阻帶截止頻率s、阻帶衰減s。 (2)將數(shù)字低通濾波器的技術(shù)指標(biāo)轉(zhuǎn)換成模擬低通濾波器的技術(shù)指標(biāo)。 21tan()2TT 如果采用雙線性變換法,邊界頻率的轉(zhuǎn)換關(guān)系為 (3)按照模擬低通濾波器的技術(shù)指標(biāo)設(shè)計(jì)模擬低通濾波器。 (4)將模擬濾波器Ha(s),從s平面轉(zhuǎn)換到z平面,得到數(shù)字低通濾波器系統(tǒng)函數(shù)H(z)。 一、脈沖相應(yīng)不變法設(shè)計(jì)數(shù)字巴特沃思低通濾波器 例
16、例 5-7 設(shè)計(jì)一個(gè)巴特沃思低通數(shù)字濾波器,給定抽樣頻率fs=10KHz,要求頻率小于 1KHz的通帶內(nèi),幅度特性下降小于1dB;在頻率大于 1.5KHz的阻帶內(nèi),衰減大于15dB 解解 (1) 求對(duì)應(yīng)的各數(shù)字域頻率: 數(shù)字低通的技術(shù)指標(biāo)為 c=0.2rad,c=1dB; s=0.3rad,s=15dB 模擬低通的技術(shù)指標(biāo)為 3333221 1020.210 10221.5 1020.310 10cccsstststsff Tfff Tf 44142/1/10/0.210 ,1/0.310 ,15sccststTTTfTdBTdB 設(shè)計(jì)巴特沃斯低通濾波器。先計(jì)算階數(shù)N及3dB截止頻率c。 22
17、2101033221010320.11 ()1 () 20log()10log 1 ()210310 10log 1 ()110log 1 ()15210 1 ()101 (aNcNacNNccNcHjHj 即:; 先用等號(hào)來(lái)滿足指標(biāo),可得; 321.533310)105.88586,7.032 10 6,7.032 10NcccNNNNN解此兩方程,可得:。因?yàn)?是濾波器的階數(shù),必須取整數(shù),為了滿足指標(biāo),故應(yīng)選比所求出 大一點(diǎn)的整數(shù),取代入通帶條件,得,即(4)用查表法 根據(jù)階數(shù)N=6,查表5.2.1,得到歸一化傳輸函數(shù)為234561( )1 3.86377.46419.14167.4641
18、3.8637aHsssssss 為去歸一化,將s=s/c代入Ha(s)中,得到實(shí)際的傳輸函數(shù)Ha(s), 66524334256( )3.86377.46419.14167.46413.8637caccccccHsssssss1212630 132,334,5( ), 0,1,.,5-1.820015536.79239041) 10-4.972374894.97237489) 10-6.792390411.82001553) 10( )akjkcaHsseksjsjsjHs ,另一種方法是求的極點(diǎn)為: 其中左半平面的三對(duì)極點(diǎn)為: ( ( ( 傳輸函數(shù)20 ( )()NcakkHsss為:即可得
19、傳輸函數(shù)。(5)用脈沖響應(yīng)不變法將Ha(s)轉(zhuǎn)換成H(z)。首先將Ha(s)進(jìn)行部分分式,并按照(5-36)式,可得到:1112121120.28710.44662.14281.1454( )10.12970.69491 1.06910.36991.85580.630410.99720.2570zzH zzzzzzzz,jwze當(dāng)即得到數(shù)字濾波器的頻率響應(yīng)。圖5.4.7 例5.4.2圖用脈沖響應(yīng)不變法設(shè)計(jì)的數(shù)字低通濾波器的幅度特性Example 7.1 Determining Specifications for a Discrete-Time FilterWe want the overal
20、l system of that figure to have the following properties when the sampling rate is samples/s ( )1. The gain should be within0.01 of unity in the frequency band 2. the gain should be no greater than 0.001 in the frequency band Such a set of lowpass specifications on can be depicted as in Figure7.2(a)
21、 , where the limits of tolerable approximation error are indicated by the shaded horizontal lines. For this specific example, the parameters would be 410Ts02 (2000)2 (3000)effHj120 .0 1 ,0 .0 0 1 ,2( 2 0 0 0 )2( 3 0 0 0 )ps410effHjSince the sampling rate is samples/s, the gain of the overall system
22、is identically zero above , due to the ideal discrete-to-continuous( D/C ) converter in Figure7.1The tolerance scheme for the discrete-time filter is shown in Figure7.2(b).From Eq.(7.1b), it follows that in the passband the magnitude of the frequency response must approximate unity within an error o
23、f Where and radians . The other approximation band is the stopband, in which the magnitude response must approximate zero with an error less than In this example , and radians.41025000 1, . .,i e1111,jpH e10.012, . .,ie423000 /100.6s20 .0 0 1422000 /100.4p2,jsHeFigure 7.2 (a) Speciffications for eff
24、ective frequency response of overall system in Figure7.1 for the case of a lowpass filter. (b) Corresponding specifications for the discrete-time system in Fihure7.1111-1)(ejHPassbandTransitionstophand2o1111-1)(ejHPassbandTransitionstophand2o1 (a)(b)7.1 DESIGN OF DISCRETE-TIME IIR FILTERS FROM CONTI
25、NUOUS-TIME FILTERSThe traditional approach to the design of discrete-time IIR filters involves the transformation of a continuous-time filter into a discrete filter meeting prescribed specifications . This is a reasonable approach for several reasons:The art of continuous-time IIR filter design is h
26、ighly advanced , and since useful results can be achieved, it is advantageous to use the design procedures already developed for continuous-time filters.Many useful continuous-time IIR design methods have relatively simple closed-form design formulas. Therefore, discrete-time IIR filter design metho
27、ds based on such standard continuous-time design formulas are rather simple to carry out.The stand approximation methods that work well for continuous-time IIR filters do not lead to simple closed-form design formulas when these methods are applied directly to the discrete-time IIR case.7.1.1 Filter
28、 Design by Impulse InvarianceImpulse invariance provides a direct means of computing samples of the output of a bandlimited continuous-time system for bandlimited input signals . Alternatively , in the context of filter design, we can think of impulse invariance as a method for obtaining a discrete-
29、time system.In the impulse invariance design procedure for transforming continuous-time filters into discrete-time filters , the impulse response of the discrete-time filter is chosen proportional to equally spaced samples of the impulse response of the continuous-time filter; i.e., where represents
30、 a sampling interval. dcdh nT hnTdTWhen impulse invariance is used as a means for designing a discrete-time filter with a specified frequency response , we are especially interested in the relationship between the frequency response of the discrete-time and continuous-time filters.If the continuous-
31、time filter is bandlimited, so that then i.e., the discrete-time and continuous-time frequency response are related by a linear scaling of the frequency axis , namely ,2,(7.5)jckddH eHjjkTT0,/,(7.6)cdHjT ,;(7.7)jcdHeHjTdT forUnfortunately, any practical continuous-time filter cannot be exactly bandl
32、imited, and consequently , interference between successive terms in Eq.(7.5) occurs , causing aliasing, as illustrated in Figure 7.3. 3 2)j(aHoo23 T)(ejHT2TTT2While the impulse invariance transform from continuous time to discrete time is defined in terms of time-domain sampling , it is easy to carr
33、y out as a transformation on the system functions.The corresponding impulse response isThe impulse response of the discrete-time filter obtained by sampling is 1,(7.9)NkckkAHsss 1,0,(7.10)0,0kNs tkkcA eth tt 11(),(7.11)kdk dNNns nTs Tdcddkdkkkh nT h nTT Aeu nT A eu n( )dcT h tThe system function of
34、the discrete-time filter is therefore given byIn comparing Eqs.(7.9) and (7.12) , we observe that a pole at in the s-plane transforms to a pole at in the z-plane and the coefficients in the partial fraction expansions of and are equal , except for the scaling multiplier Td. Hz cHsk ds Tzekss 11,(7.1
35、2)1kdNdks TkT AHzezExample 7.2 Impulse Invariance with a Butterworth FilterLet us consider the design of a lowpass discrete-time filter by applying impulse invariance to an appropriate Butterworth continuous-time filter. The specifications for the discrete-time filter areSince the parameter Td cance
36、ls in the impulse invariance procedure , we can choose Td=1, so that .0.891251,00.20.17783,0.3jjH eH eBecause of the preceding considerations , we want to design a continuous-time Butterworth filter with magnitude function for which Since the magnitude response of an analog Butterworth filter is a m
37、onotonic functions of frequency.0.891251,00.2 ,(7.14 )0.17783,0.3,(7.14 )ccHjaHjb Eqs.(7.14a) and (7.14b) will be satisfied if and 0.20.89125,(7.15 )cHja0.30.17783,(7.15 )cHjbSpecifically , the magnitude-squared function of a Butterworth filter is of the formSo that the filter design process consist
38、s of determining the parameters N and to meet the desired specifications. Using Eq.(7.16) in Eqs.(7.15) with equality leads to the equations and The solution of these two equations is N=5.8858 and 22/1,(7.16)1ccNHjc220.211,(7.17 )0.89125Nca220.311,(7.17 )0.17783Ncb0.70474c With and with N=6, the 12
39、poles of the magnitude-squared function are uniformly distributed in angle on a circle of radius as indicated in Figure 7.4. consequently , the poles of are the three pole pairs in the left half of the s-plane with the following coordinates: Pole pair 1:-0.182j(0.679) Pole pair 2:-0.497j(0.497) Pole
40、 pair 3:-0.679j(0.182)Figure 7.4 s-plane locations for poles of for sixth-order Butterworth filter in Example 7.2.0.7032c 21/ 1/NcccHs Hssj0.7032c cHs ccHs HsTherefore ,If we express as a partial fraction expansion , perform the transformation of Eq.(7.12), and then combine complex-conjugate terms,
41、the resulting system function of the discrete-time filter is 2220.12093,(7.18)0.3400.49450.99450.49451.35850.4945cssssssHs 1111212120.2871 0.44662.1428 1.14551.8557 0.6303,(7.19)1 0.99720.25701 1.06910.36991 0.99720.2570azzH zzzzzzz cHsAs is evident from Eq.(7.19), the system function resulting from
42、 the impulse invariance design procedure may be realized directly in parallel form.The basis for impulse invariance is to choose an impulse response for the discrete-time filter that is similar in some sense to the impulse response of the continuous-time filter.Figure 7.5 Frequency response of sixth
43、-order Butterworth filter transform by impulse invariance. (a) Log magnitude in dB. (b) magnitude . (c ) Group delay5.2 Filter Design by Impulse Invariance 一、一、 變換原理變換原理 利用模擬濾波器來(lái)設(shè)計(jì)數(shù)字濾波器,也就是使數(shù)字濾波器能模仿模擬濾波器的特性,這種模仿可以從不同的角度出發(fā)。 脈沖響應(yīng)不變法是從濾波器的脈沖響應(yīng)出發(fā),使數(shù)字濾波器的單位脈沖響應(yīng)序列h(n)模仿模擬濾波器的沖激響應(yīng)ha(t),即將ha(t)進(jìn)行等間隔采樣,使h(n)
44、正好等于ha(t)的采樣值,滿足 h(n)=ha(nT) (5-31)式中, T是采樣周期。 如果令Ha(s)是ha(t)的拉普拉斯變換,H(z)為h(n)的Z變換,利用第2章2.5節(jié)采樣序列的Z變換與模擬信號(hào)的拉普拉斯變換的關(guān)系,即利用式(2-53)(P71),得 kTjsXTjksXTzXkaskaezsT21)(1)(5-32) 則可看出,脈沖響應(yīng)不變法將模擬濾波器的S平面變換成數(shù)字濾波器的Z平面,這個(gè)從s到z的變換z=esT正是第2章2.5節(jié)中從S平面變換到Z平面的標(biāo)準(zhǔn)變換關(guān)系式(2-51)。 圖 5-9 脈沖響應(yīng)不變法的映射關(guān)系 j3/ T/ T3/ T/ Too11jImzRezZ
45、平面S平面二、二、 混疊失真混疊失真 由式(5-32)知,數(shù)字濾波器的頻率響應(yīng)和模擬濾波器的頻率響應(yīng)間的關(guān)系為 TkjHTeHkaj21)(5-33) 這就是說(shuō),數(shù)字濾波器的頻率響應(yīng)是模擬濾波器頻率響應(yīng)的周期延拓。正如第1章1.4節(jié)采樣定理所討論的,只有當(dāng)模擬濾波器的頻率響應(yīng)是限帶的,且?guī)抻谡郫B頻率以內(nèi)時(shí),即 0)(jHa2|sT(5-34) 才能使數(shù)字濾波器的頻率響應(yīng)在折疊頻率以內(nèi)重現(xiàn)模擬濾波器的頻率響應(yīng),而不產(chǎn)生混疊失真,即 TjHTeHaj1)(|(5-35) 但是,任何一個(gè)實(shí)際的模擬濾波器頻率響應(yīng)都不是嚴(yán)格限帶的, 變換后就會(huì)產(chǎn)生周期延拓分量的頻譜交疊,即產(chǎn)生頻率響應(yīng)的混疊失真,如圖
46、5-10所示。這時(shí)數(shù)字濾波器的頻響就不同于原模擬濾波器的頻響,而帶有一定的失真。當(dāng)模擬濾波器的頻率響應(yīng)在折疊頻率以上處衰減越大、越快時(shí),變換后頻率響應(yīng)混疊失真就越小。這時(shí),采用脈沖響應(yīng)不變法設(shè)計(jì)的數(shù)字濾波器才能得到良好的效果。 圖 5-10 脈沖響應(yīng)不變法中的頻響混疊現(xiàn)象 32)j(aHoo23 T)(ejHT2TTT2 對(duì)某一模擬濾波器的單位沖激響應(yīng)ha(t)進(jìn)行采樣,采樣頻率為fs,若使fs增加,即令采樣時(shí)間間隔(T=1/fs)減小,則系統(tǒng)頻率響應(yīng)各周期延拓分量之間相距更遠(yuǎn),因而可減小頻率響應(yīng)的混疊效應(yīng)。 三、三、 模擬濾波器的數(shù)字化方法模擬濾波器的數(shù)字化方法 由于脈沖響應(yīng)不變法要由模擬系
47、統(tǒng)函數(shù)Ha(s)求拉普拉斯反變換得到模擬的沖激響應(yīng)ha(t),然后采樣后得到h(n)=ha(nT),再取Z變換得H(z),過(guò)程較復(fù)雜。下面我們討論如何由脈沖響應(yīng)不變法的變換原理將Ha(s)直接轉(zhuǎn)換為數(shù)字濾波器H(z)。 設(shè)模擬濾波器的系統(tǒng)函數(shù)Ha(s)只有單階極點(diǎn),且假定分母的階次大于分子的階次(一般都滿足這一要求,因?yàn)橹挥羞@樣才相當(dāng)于一個(gè)因果穩(wěn)定的模擬系統(tǒng)),因此可將 NkkkassAsH1)( (5-36) 其相應(yīng)的沖激響應(yīng)ha(t)是Ha(s)的拉普拉斯反變換,即 11( )( )( )kNs taakkh tLHsA e u t式中, u(t)是單位階躍函數(shù)。 在脈沖響應(yīng)不變法中,要求
48、數(shù)字濾波器的單位脈沖響應(yīng)等于對(duì)ha(t)的采樣,即 NknTskNknTskanueAnueAnThnhkk11)()()()()( (5-37) NkTskNknTsnkNknTsknnnzeAzeAzeAznhzhkkk111101101)()()()(對(duì)h(n)求Z變換,即得數(shù)字濾波器的系統(tǒng)函數(shù) (5-38) 將式(5-36)的Ha(s)和式(5-38)的H(z)加以比較,可以看出: (1)S平面的每一個(gè)單極點(diǎn)s=sk變換到Z平面上z=eskT處的單極點(diǎn)。 (2) Ha(s)與H(z)的部分分式的系數(shù)是相同的,都是Ak。 (3)如果模擬濾波器是因果穩(wěn)定的,則所有極點(diǎn)sk位于S平面的左半平
49、面,即Resk0, 則變換后的數(shù)字濾波器的全部極點(diǎn)在單位圓內(nèi),即|eskT|=eReskT1, 因此數(shù)字濾波器也是因果穩(wěn)定的。 (4)雖然脈沖響應(yīng)不變法能保證S平面極點(diǎn)與Z平面極點(diǎn)有這種代數(shù)對(duì)應(yīng)關(guān)系,但是并不等于整個(gè)S平面與Z平面有這種代數(shù)對(duì)應(yīng)關(guān)系,特別是數(shù)字濾波器的零點(diǎn)位置就與模擬濾波器零點(diǎn)位置沒(méi)有這種代數(shù)對(duì)應(yīng)關(guān)系,而是隨Ha(s)的極點(diǎn)sk以及系數(shù)Ak兩者而變化。 從式(5-35)看出,數(shù)字濾波器頻率響應(yīng)幅度還與采樣間隔T成反比: TjHTeHaj1)(| 如果采樣頻率很高,即T很小,數(shù)字濾波器可能具有太高的增益,這是不希望的。為了使數(shù)字濾波器增益不隨采樣頻率而變化,可以作以下簡(jiǎn)單的修正,
50、令 h(n)=Tha(nT) (5-39) 則有: NkTskzeTAzHk111)(TjHkTjTjHeHakaj2)(5-40) (5-41) 例例 5-3 設(shè)模擬濾波器的系統(tǒng)函數(shù)為 3111342)(2sssssHa試?yán)妹}沖響應(yīng)不變法將Ha(s)轉(zhuǎn)換成IIR數(shù)字濾波器的系統(tǒng)函數(shù)H(z)。 解解 直接利用式(5-40)可得到數(shù)字濾波器的系統(tǒng)函數(shù)為 TTTTTTTezeezeeTzezTezTzH423131311)(1)(11)(設(shè)T=1,則有 21101831. 04177. 013181. 0)(zzzzH 模擬濾波器的頻率響應(yīng)Ha(j)以及數(shù)字濾波器的頻率響應(yīng)H(ej)分別為: 2
51、201831. 04177. 013181. 0)(432)(jjjjaeeeeHjjH)(把|Ha(j)|和|H(ej)|畫在圖5-11上。由該圖可看出,由于Ha(j)不是充分限帶的,所以H(ej)產(chǎn)生了嚴(yán)重的頻譜混疊失真。 圖 5-11 例5-3的幅頻特性 / T2/ T2)j (aH)(ejHoo5.4.4 優(yōu)缺點(diǎn)優(yōu)缺點(diǎn) 從以上討論可以看出,脈沖響應(yīng)不變法使得數(shù)字濾波器的單位脈沖響應(yīng)完全模仿模擬濾波器的單位沖激響應(yīng),也就是時(shí)域逼近良好,而且模擬頻率和數(shù)字頻率之間呈線性關(guān)系=T。 因而,一個(gè)線性相位的模擬濾波器(例如貝塞爾濾波器)通過(guò)脈沖響應(yīng)不變法得到的仍然是一個(gè)線性相位的數(shù)字濾波器。 脈
52、沖響應(yīng)不變法的最大缺點(diǎn)是有頻率響應(yīng)的混疊效應(yīng)。 所以, 脈沖響應(yīng)不變法只適用于限帶的模擬濾波器(例如, 衰減特性很好的低通或帶通濾波器),而且高頻衰減越快,混疊效應(yīng)越小。 至于高通和帶阻濾波器,由于它們?cè)诟哳l部分不衰減, 因此將完全混淆在低頻響應(yīng)中。如果要對(duì)高通和帶阻濾波器采用脈沖響應(yīng)不變法, 就必須先對(duì)高通和帶阻濾波器加一保護(hù)濾波器,濾掉高于折疊頻率以上的頻率,然后再使用脈沖響應(yīng)不變法轉(zhuǎn)換為數(shù)字濾波器。 當(dāng)然這樣會(huì)進(jìn)一步增加設(shè)計(jì)復(fù)雜性和濾波器的階數(shù)。 7.1.2 Bilinear TransformationThe technique discussed in this subsection
53、 avoids the problem of aliasing by using the bilinear transformation , an algebraic transformation between the variables s and z that maps the entire -axis in the s-plane to one revolution of the unit circle in the z-plane.With denoting the continuous-time system function and H (z) the discrete-time
54、 system function, the bilinear transformation corresponds to replacing s by that is, cHs1121, (7.20)1dzsTz 1121,(7.21)1cdzH zHTzTo develop the properties of the algebraic transformation specified in Eq.(7.20), we solve for z to obtain1/ 2,(7.22)1/ 2ddTszTsand , substituting into Eq.(7.22) ,we obtain
55、1/ 2/ 2,(7.23)1/ 2/ 2ddddTjTzTjTsj If then , from Eq.(7.23), it follows that z 1 for all .Next , to show that the -axis of the s-plane maps onto the unit circle, we substitute into Eq.(7.22), obtaining From Eq.(7.24) , it is clear that z =1 for all values of s on the -axis . That is , the -axis maps
56、 onto the unit circle , so Eq.(7.24) takes the formTo derive a relationship between the variables and , it is useful to return to Eq.(7.20) and substitute . We obtain 00jsj 1/ 2, (7.24)1/ 2ddjTzjTjj1/ 2,(7.25)1/ 2jddjTejT jze21,(7.26)1jjdesTe or, equivalently,Equating real and imaginary parts on bot
57、h sides of Eq.(7.27) leads to the relations and or/2/22sin/222tan/2 ,7.27)2cos/2jjddejjsjTeT 2tan/2 ,(7.28)dT2arctan/2 ,(7.29)dT0these properties of the bilinear transformation as a mapping from the s-plane to the z-plane are summarized in Figures 7.6 and 7.7. from Eq.(7.29) and Figure 7.7, we see t
58、hat the range of frequencies maps to while the range maps to .The bilinear transformation avoids the problem of aliasing encountered with the use of impulse invariance, because it maps the entire.0 00 0imaginary axis of the s-plane onto the unit circle in the z-plane. The price paid for this, howeve
59、r, is the nonlinear compression of the frequency axis depicted in Figure 7.7Consequently, the design of discrete-time filters using the bilinear transformation is useful only when this compression can be tolerated or compensated for, as in the case of filters that approximate ideal piecewise-constan
60、t magnitude-response characteristics. This is illustrated in Figure 7.8. where we show how a continuous-time frequency response and tolerance scheme maps to a corresponding discrete-time frequency response and tolerance scheme through the frequency warping of Eqs.(7.28) and (7.29).Typical frequency-
溫馨提示
- 1. 本站所有資源如無(wú)特殊說(shuō)明,都需要本地電腦安裝OFFICE2007和PDF閱讀器。圖紙軟件為CAD,CAXA,PROE,UG,SolidWorks等.壓縮文件請(qǐng)下載最新的WinRAR軟件解壓。
- 2. 本站的文檔不包含任何第三方提供的附件圖紙等,如果需要附件,請(qǐng)聯(lián)系上傳者。文件的所有權(quán)益歸上傳用戶所有。
- 3. 本站RAR壓縮包中若帶圖紙,網(wǎng)頁(yè)內(nèi)容里面會(huì)有圖紙預(yù)覽,若沒(méi)有圖紙預(yù)覽就沒(méi)有圖紙。
- 4. 未經(jīng)權(quán)益所有人同意不得將文件中的內(nèi)容挪作商業(yè)或盈利用途。
- 5. 人人文庫(kù)網(wǎng)僅提供信息存儲(chǔ)空間,僅對(duì)用戶上傳內(nèi)容的表現(xiàn)方式做保護(hù)處理,對(duì)用戶上傳分享的文檔內(nèi)容本身不做任何修改或編輯,并不能對(duì)任何下載內(nèi)容負(fù)責(zé)。
- 6. 下載文件中如有侵權(quán)或不適當(dāng)內(nèi)容,請(qǐng)與我們聯(lián)系,我們立即糾正。
- 7. 本站不保證下載資源的準(zhǔn)確性、安全性和完整性, 同時(shí)也不承擔(dān)用戶因使用這些下載資源對(duì)自己和他人造成任何形式的傷害或損失。
最新文檔
- 三級(jí)人力資源管理師-三級(jí)人力資源管理師考試《理論知識(shí)》押題密卷1
- 物理選修3-5全套講義第三章原子核與放射性章末檢測(cè)3
- 2025年江蘇省蘇州市中考模擬英語(yǔ)試題(二)(原卷版+解析版)
- DB11T-蒸壓加氣混凝土墻板系統(tǒng)應(yīng)用技術(shù)規(guī)程編制說(shuō)明
- 專業(yè)裝修工程合同范例
- 修理制式合同范例
- 個(gè)人吊車轉(zhuǎn)賣合同范例
- 低壓借款合同范例
- 企業(yè)贊助活動(dòng)合同范例
- 加工承攬貨架合同范例
- 2023年古文中的化學(xué)知識(shí)歸納及相關(guān)練習(xí)題(含答案)
- 《基礎(chǔ)寫作》試卷及答案
- 2025年高考數(shù)學(xué)復(fù)習(xí)大題題型歸納:解三角形(原卷)
- 醫(yī)院軟式內(nèi)鏡清洗消毒技術(shù)規(guī)范
- 2024年中央空調(diào)市場(chǎng)占有率分析:中央空調(diào)國(guó)產(chǎn)品牌市場(chǎng)占有率上升至52.57%
- 2024年電力交易員(中級(jí)工)職業(yè)鑒定理論考試題庫(kù)-下(多選、判斷題)
- 2024年江蘇廣播電視局事業(yè)單位筆試真題
- 輪胎英語(yǔ)詞匯
- 按摩技師簽訂勞動(dòng)合同注意事項(xiàng)
- 項(xiàng)目保證金協(xié)議書范本
- 2022-8口腔質(zhì)控督查表(培訓(xùn)用)
評(píng)論
0/150
提交評(píng)論