Signal Processing and Linear Systems-B.P.Lathi copy

# e k yt ytt jr

This preview shows page 1. Sign up to view the full content.

This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: e, • E xample 1 1.10 F ind t he o utput y (t) for all t in Example 11.9. In Example 11.9, we found t he response y[kJ only a t t he sampling instants. To find t he o utput values between sampling instants, we use t he modified z-transform. The procedure is t he same as before, except t hat we use modified z-transform corresponding t o continuous-time systems a nd signals. For t he system G (s) = 1 / s + 4 w ith T = 0.5, t he modified z-transform [Eq. (11.50b) w ith>' = - 4, a nd T = 0.5J is Z3 Y[zJ = T[zJF[zJ = 7(z---I"'-)('-z---0-=.3-94-'-)"-(z---0-.1-74""') Moreover t o find the modified z-transform corresponding t o f (t) = u (t) [>' = 0 in Eq. (11.50a)], we have F[z, J.iJ = z /(z - 1). S ubstitution of these expressions in those found in Example 11.9, we o btain a nd z (z - 1)(z - 0.394)(z - 0.174) 1 0.583 = z - 1 - z - 0.394 + 0.083 z - 0.174 Y[ J- z, J.i - e -2" [Z 0.583z z _ 1 - z _ 0.394 0.083z ] 0.174 +z- 7 04 11 D iscrete-Time Systems Analysis Using t he Z -Transform 1 From Eqs. (11.50), we obtain the inverse (modified) z-transform of thi~ equation as y[(k + JL)TJ = e - 2" [1 - 0.583(0.394)k t - y k u [ -(k+ I )] + 0.083( - 0.174)k] -7 T he complete response is also shown in Fig. U .ll. 705 11. 7 T he B ilateral Z - Transform -5 -3 -\ k- • D esign o f S ampled-Data S ystems As with continuous-time control systems, s ampled-data systems are designed t o m eet c ertain transient (PO, t r , ts, e tc.) and s teady-state specifications. T he design procedure follows along the lines similar to those used for continuous-time systems. W e begin with a general second-order system. t he r elationship between closedloop pole locations a nd t he corresponding transient p arameters PO, t r , t s ," . are determined. Hence, for a given t ransient specifications, a n a cceptable region in the z -plane where the dominant poles of t he closed-loop transfer function T[z] should lie is d etermined. Next, we s ketch t he root locus for t he system. T he rules for s ketching t he r oot locus are t he s ame as those for continuous-time systems. I f t he r oot locus passes through t he a cceptable region, t he t ransient specifications can be m et b y simple adjustment of t he gain K . I f not, we m ust use a compensator, which will s teer t he r oot locus in t he a cceptable region. Fig. 1 1.12 _ ·/u[-(k + I}J and the region of convergence of its z-transform. =- = 1 1. 7 T he Bilateral Z- Transform 00 L + ...J +( 1- [1 + ~ + (~r + (~r + ...J I~I < 1 'Y z Izl < hi Therefore f [k]z-k Izl < k =-oo T he inverse z-transform is given by f [k] = r ~r 1 =1--1- ~ S ituations involving noncausal signals or systems c annot b e handled b y t he (unilateral) z-transform discussed so far. Such cases c an b e analyzed by t he b ilateral (or two-sided) z - transform defined by F[z] == [~ + ( ~ ~fF[Z]zk-ldZ 21rJ T hese e quations define t he b ilateral z-transform. T he u nilateral z-transform discussed so far is a special case, where t he i nput signals a re r estricted t o b e causal. R estricting signals in this way results in considerable simplification in t he region of convergen...
View Full Document

## This note was uploaded on 04/14/2013 for the course ENG 350 taught by Professor Bayliss during the Spring '13 term at Northwestern.

Ask a homework question - tutors are online