Selamat datang di Scribd!
Akademik Dokumen
Profesional Dokumen
Budaya Dokumen
Hobi & Kerajinan Dokumen
Pertumbuhan Pribadi Dokumen
0-84, +B, +128 + 34, +4B, = 48 y"(5) =0 > 0= A, +5B, + 6(25)+6—-R=> A, +5B,— R= —-156 y"(5) =0 > 0=B, +12(5)+12-R > B,-R=-72 which solve to give Aj = 18, By = —25.5, R= 46.5 Thus 5 — 4.2523 +927, 0 F(t—n) n=0 u(t) = F(t) + fE—m) + Fe The graph may be plotted by computer and should take the form $0) 1 : +1)? taking f(t) = and g(t) =te™t in the convolution theorem = 23(a) Since L{t} = , L{te~*} Lo'F(s)G(s)] = f = g(t) © © Pearson Education Limited 2004148 Glyn James: Advanced Modern Engineering Mathematics, Third edition gives : = ff fe natrver = [ree = [-(t-r)ret — (t 27 1 242% + tet (s +2) vee 23(b) y(t) =#+2 fp ww) cos(t — u)du Taking f(t) = y(t), g(t) = cost > F(s) = Y(s),G(s) = ea transforms (s? + 1—28)¥(s) etl 2 2 2 ere Ras 2 ste and y(t) = £-'{¥(s)} =2+4—2ef + 2te’. ‘Taking transforms (s°¥(s) — sy(0) — ¥'(0))(s¥ (s) — y(0)) = ¥(8) (°¥(s) — wi)(s¥(s)) = yo giving Y(s) = 0 or Y(s) = 4+ = which on inversion gives lo y(t) = 0 or y(t) = a + by In the second of these solutions the condition on y’(0) is arbitrary. © Pearson Education Limited 2004Glyn James: Advanced Modern Enginecring Mathematics, Third edition = 24 y Be Equation for displacement is dy EI, = -Wo(x-6 det @-4) with y(0) =0, (30 =0, y’(0) =y/(2 with y’(0) = A, y'"(0) = B then taking Laplace transforms gives giving y(x) = Wao He—H+4 6EI 2Wwe pope oa 3 “9 ET * si ET" © © Pearson Education Limited 2004 149150 Glyn James: Advanced Modern Engineering Mathematics, Third edition With the added uniform load the differential equation governing the deflection is pit = —W6(x — 0) — wlH(x) — H(w — 2) = 25(a) Taking Laplace transforms (s?— 3s +.3)X(s) = few giving a(t) = £7 X(s)} = [! ae He —") (cos Bu —a)— vain — «| H(t—a) 1 6 25(b) fs) =O ‘ _ w X(s) = G(s)C{sinwt} = G(s) w =o (8 wr aye 7a) Since the system is stable all the poles of G(s) have negative real part. Expanding in partial fractions and inverting gives Fue x(t) = 2R, oe e/*) + terms from G(s) with negative exponentials Thus as t + co the added terms tend to zero and a(t) + 2.(t) with [teow ta(t) = Re 7 © Pearson Education Limited 2004Glyn James: Advanced Modern Engineering Mathematics, Third edition 151 = 26(a) In the absence of feedback the system has poles at s=—3ands=1 and is therefore unstable. cy G8) 1 = 1 26(b) Cil8) = TRG = GoDEFHN LK ~ PEt KS) 26(c) Poles G(s) given by s=-1+ V-K. These may be plotted in the s-plane for different values of K, Plot should be as in the figure TAG) Res) 26(d) Clearly from the plot in (c) all the poles are in the left half plane when K > 3. Thus system stable for K > 3. a a ay 26 , - 1s? + 28 + (K-3)=0 Routh-Hurwitz determinants are A, =2>0 alee cose oe . A 0 ao 0 K_3| = 2K —-3) >On K >3 thus confirming the result in (a) = 27(a) Closed loop transfer function is © © Pearson Education Limited 2004152 Glyn James: Advanced Modern Engineering Mathematics, Third edition 2 Thus oN oyaspe) = A(t) = 2c sint ie, CY{—__2 ____} pe Ftsin (5 — 2 t= 20 sine (s+ $P+6-%) giving a=4 27(b) Closed loop transfer function is _ 10 ~ x? 4+ (10K —1)s +10 Poles of the system are given by s?4 (10K —1)s+10=0 which are both in the negative half plane of the s-plane provided (10K — 1) > 0; that is, K > 4. Thus the critical value of KC for stability of the closed loop system isk = 3. 28(a) Overall closed loop transfer funetion is ek s(1+ KK\)+K G(s) “T+ (+ Kis) K ses) 28(b) Assuming zero initial conditions step response x(t) is given by . _ K X(s) = G(s)L{LA(t)} = Pas = RR) =F] Un ~ |g? + 2€w,8 + we] 1 $+ 2, 8 9+ wns + we 1 (s+ wn) + Sn s- le + Ett)? + [wi — ral (% + &wn) + & © L (8+ Ew)? + w’ giving x(t) = £-1{X(s)} = eS" Teos wat + © Pearson Education Limited 2004Glyn James: Advanced Modern Engin ering Mathematics, Third edition 1 d: # 28(c) The peak time ty is given by the solution of SF],_,, =0 dx swat [fe,, Swa Lun te = eee! (gu, — = ) 008 wat (+ wa) sin wat sin wat fA-e ‘Thus t, given by the solution of Wn . en bnty sin watp ie, sinwaty = 0 Since the peak time corresponds to the first peak overshoot t a watp =F oF ty =~ wa ‘The maximum overshoot M, occurs at the peak time tp. Thus wig fe . ws [cos@ + —— sine [own + Spans] = set =e lV F ge We wish M, to be 0.2 and t, to be 1s, thus eS/VI-@ — 0.2 giving € = 0.456 and <1 giving wg = 3.14 a Then it follows that w, = = 3.53 from which we deduce that K =w? = 12.5 and Ky = 2&4 _ ozs, © Pearson Education Limited 2004154 Glyn James: Advanced Modern Engineering Mathematics, Third edition 28(d) The rise time t¢, is given by the solution of a(t,) = 1L= 1— eS [eos waty + Since e~Stntr é 008 Wate + = vie giving tan wath = so the curves 1+ are the envelope curves of the transient response to a unit step input and have a time constant T = ~—. ‘The settling time ¢, may be measured in terms of T. Using the 2% criterion t, is approximately 4 times the time constant and for the 5% criterion it is approximately 3 times the time constant. Thus 29 criterion + t, 47 = = 2.485 fw 3 5% criterion 1 t, = 37 = —~ = 1.86s GW Footnote ‘This is intended to be an extended exercise with students being encouraged to carry out simulation studies in order to develop a better understanding of how the transient response characteristics can be used in system design. 29 As for Exercise 28 this is intended to be an extended problem supported by simulation studies. The following is simply an outline of a possible solution. © Pearson Education Limited 2004Glyn James: Advanced Modern Engineering Mathematics, Third edition 155 Figure 2.63(a) is simply a mass-spring damper system represented by the differential equation és de Assuming that it is initially in a quiescent state taking Laplace transforms M, + BE + Kye =sinwt X(s)- +.” X)= RPE Ret Ry P40? The steady state response will be due to the forcing term and determined by the as+ 3 3? + w? response will be of the form Asin(wé +); that is, a sinusoid having the same term in the partial fractions expansion of X(s). Thus, the steady state frequency as the forcing term but with a phase shift d and amplitude scaling A. In the situation of Figure 2.63(b) the equations of motion are Px - dx . Mae =—-Kir— BE + Ko(y— 2x) +sinwt ay Maye = —Koly - 2) Assuming an initial quiescent state taking Laplace transforms gives (Mis? + Bs + (Ki + Kx)|X(s) — K2Y(s) = w/(s? + w?) —K2X(s) + (s?Mz + K2)¥(s) = 0 Eliminating Y(s) gives ' w(s?Mp + Ky XS) = Cary pls) where p(s) = (Mis? + Bs + K, + K3)(s?M> + Ka). Because of the term (s? + w?) in the denominator 2(t) will contain terms in sinwt and coswt. However, if (s?M2+K2) exactly cancels (s? + w?) this will be avoided. Thus choose Ky = Mzw?. This does make practical sense for if the natural frequency of the secondary system is equal to the frequency of the applied force then it may resonate and therefore damp out the steady state vibration of M. It is also required to show that the polynomial p(s) does not give rise to any undamped oscillations. That is, it is necessary to show that p(s) does not possess purely imaginary roots of the form j@,@ real, and that it has no roots with a positive real part. This can be checked using the Routh-Hurwitz criterion. © Pearson Education Limited 2004156 Glyn James: Advanced Modern Engineering Mathematics, Third edition ‘To examine the motion of the secondary mass Mp solve for ¥(s) giving Kou NO)" Sw) Clearly due to the term (s? + w?) in the denominator the mass Mp possesses an undamped oscillation. Thus, in some sense the secondary system has absorbed the energy produced by the applied sinusoidal force sin wt. 30 Again this is intended to be an extended problem requiring wider exploration by the students. The following is an outline of the solution. 30(a) Students should be encouraged to plot the Bode plots using the steps used in example 2.62 of the text and using a software package. Sketches of the magnitude and phase Bode plots are given in the figures below. 30(b) With unity feedback the amplifier is unstable, Since the —180° crossover gain is greater than dB (from the plot it is +924B). 30(c) Due to the assumption that the amplifier is ideal it follows that for marginal stability the value of 4 must be 924B (that is, the plot is effectively lowered by 92dB). Thus 1 20log— = 92 O85 30(d) From the amplitude plot the effective 0dB axis is now drawn through the 100dB point. Comparing this to the line drawn through the 92dB point, corresponding to marginal stability, it follows that Gain margin = —8dB and Phase margin = 24°. © Pearson Education Limited 2004gy Glyn James: Advanced Modern Engineering Mathematics, Third edition 1) 30(e) K (1+ s71)(1 — 872)(1 + 873) Given low frequency gain K = 120dB so G(s) = 2Wlog K = 120 = K = 10° 1 T, == where fj is the oscillating frequency in cycles per second of the pole. Since IMHz = 10 cycles per second 1 1 71> 7 = jg since fi = IMHa n= “Tar since fy = 10MHz = = 5 = since fy = 5 n=, — since fy = 25MHz Thus 5 Gls) i "U4 BOC+ Toi)E + sie) _ 250.10°4 ~ (s+ 108)(s + 107)(s + §.107) ‘The closed loop transfer function G(s) is Gs) G9) =F + BG(s) 30(f) The characteristic equation for the closed loop system is (s+ 109)(s + 10")(s + §.107) + 625.10 =0 or 8° + 36(10°)s? + (285)101?s + 101°(25 + 25g10°) = 4 4 4 Al Ag A3 By Routh-Hurwitz criterion system stable provided A, > 0 and AyAg > Ag. If =1 then A,Ap < Ag and the system is unstable as determined in (b). For marginal stability A;Ap = Ag giving 8 = 1.40~* (compared with = 2.5.10-® using the Bode plot). © Pearson Education Limited 2004158 Glyn James: Advanced Modern Engineering Mathematics, Third edition Magnitude vs Frequency Ptot 120 T I Data margin CELL CECH Hay ULE fl 8 63] FEL 4005 on - Hep bt - aol 60} @ 40 8 & 3 20 FH Corresponds to 180" phase iag| plot it " To phase O° —90° — 180° = 270° | 1 10 25 In freq. MHZ © Pearson Education Limited 20043 The z Transform Exercises 3.2.3 = 1a) F(s) = 1(b) 1c) ws oye Fe) = Y - s|s2 mo 1(a) oO he Fe) = OP = & 1(e) from (3.6) whence zn, ABR} = 87 py itl 2 [> 1 whence © Pearson Education Limited 2004160 Glyn James: Advanced Modern Engineering Mathematics, Third edition Exercises 3.3.6 a3 a log loo 2 {sink} = 5 oor ~ yy poet zsinwT = 2 coswT $1 ad so Proceeding directly 2{m} = kas 7 = 5(a) z {-3} = 5(b) {eos kr} = {(-1)"} 2 {eos ke} = a jz|>1 =6 By (3.5) © Pearson Education Limited 2004Glyn James: Advanced Modern Engineering Mathematics, Third edition 161 thus etm 2 {(ka")} = Gop whence k . 1\"(_ 2{h(3) \- oe = 7a) 1 1 a ok oy ke sinh ka = 5(e*)* — 5(e~°y so oY fot. L esinho {sinh ko} -3( - ~ Becosha #1 7b) 1 1 L — yk —aeyk cosh ka = 5(e")" + 5(e*) then proceed as above. = 8(a) a(b) zsinT —2zcosT +1 8(c) 1 oo, _ . uy = = (MT + eT) then proceed as above. = 9 Initial value theorem: obvious from definition. © Pearson Education Limited 2004162 Glyn James: Advanced Modern Engineering Mathematics, Third edition = 9 Final value theorem Gar )x@= m1 —% , w—2 Tt tet = 0+ As 2—+landif |" a, exists, then lim (1 — 271)X(2) = lim 2, zal rato = 10 Multiplication property (3.19): Let 2 {a} = Dy SE = X(2) then = X(2/a) akxy ae 2 {aka} y 10 Multiplication property (3.20) d —X(2) = ‘The general result follows by induction, Exercises 3.4.2 = 11(a) 43 ftom tables uy = 1 11(b) - a = : from tables w, = (—1)* 11(c) . _ A Soya from tables ws = (1/2) © Pearson Education Limited 2004Glyn James: Advanced Modern Engineering Mathematics, Third edition 163 11(a) = _ Fe +173 1/3) 11(e) a} from tables w, = (9 )* 11(f) /a\k poy OI? py avy y 11(g) k=0 k>0 using first shift property. 11(h) = 12(a) so 12(b) oe} (g- sip) Homan 12(c) 1 1 1 + erpip at ey © Pearson Education Limited 2004164 Glyn James: Advanced Modern Engineering Mathematics, Third edition Be) . 2 2 22 2k 2, vk YO= se yp ar ar gt) — 2a yoy k oa yet = Fats 2- 12(e) ' 1 YO= a (e57 z—( =) a z ~ oy ( {0,0,1,1} + {(-1/3)*} 13(e) -2-1/2)1, bh © Pearson Education Limited 2004166 Glyn James: Advanced Modern Engineering Mathematics, Third edition 13(f) Ok 1-%k—-Y)+2"k21 _ fo k=0 ~ 3 -2h 424, 13(g) Exercises 3.5.3 14(a) If the signal going into the left D-block is we and that going into the right D-block is v%, we have 1 Vet = We = Tk ~ 50k Uk = so 1 Ueto = Vet = Tk — GUE 1 1 = bE 5 Uk = Tk SUREL ie. 1 Uke + Uk = @e 14(b) Using the same notation 1 ot Yeti = Ves Yeti = Wk = Re Fe — Ue © Pearson Education Limited 2004Glyn James: Advanced Modern Engineering Mathematics, Third edition 167 Then 1 1 Ykp2 = ke — TYR — 5 Uk 4 5 or toast oti ty =a, Ueto quett Bue he 15(a) PY (2) — yo — 2m — 2(2Y (2) — eyo) + ¥(2) = 9 with yo =0, y= 1 ¥(2)= 50 yp — kk > 0. 15(b) Transforming and substituting for yo and y, thus 15(c) Transforming and substituting for yp and y: . : YO) = ane re) a z * © ay (22072 2 2e-377 thus 1 ok (cake /2 _ p-ake/2 pk-1 es > == 7/2 g-akr/2) 9 2, k wa 7? (e eri?) sin kr/2, k= 0 © Pearson Education Limited 2004168 Glyn James: Advanced Modern Engineering Mathematics, Third edition 15(d) Transforming, substituting for yo and yi, and rearranging - 62-11 YOlP= Bre 3) so YO=2 5p t ya thus ue = (1/2) +38, k= 0 = 16(a) Synto + Yeti — Ye = 3, yoo yi =O ‘Transforming with yo = y: = 0, (627 + 2-DY(2) = . 3 YO = Ce - Des) and 1 9 2 Y¥@)=574- 2-18 > Betlp Inverting 1 9 2 k w= 37 (lls + e-1/2) 16(b) Transforming with yo = 0, ys: = 1, —52+6)¥(2)= 245 whence © Pearson Education Limited 2004Glyn James: Advanced Modern Engineering Mathematics, Third edition 169 viny - 4 2 ¥@) 1b2— 1/2 whence 4 > > = = (1/2) — =(2)* + 213)" vn = (0/2) — 32)" + 208) 16(d) Transforming with yo = 1, m1 = 0, (2-32 +3)¥(2) 16(e) Transforming with yo = 1, m1 922 — 3x —2)¥(2) = 22 2 (28 Bs —2)V(2) = 2 be tO © Pearson Education Limited 2004170 Glyn James: Advanced Modern Engineering Mathematics, Third edition 80 Y@)= and 1 1 Un =l-n— 32)" — gy" 17 Write the transformed equations in the form 2-3/2 1 e2)\ _ (2Co 021 2-1/2) \el2)/ \2Bo (03) = s=aeroa Corse) Ge) Solve for (2) as Then 2-12 + 4800-—G3 e(2) = 1200 and Ox = 1200(1.2)* + 4800(0.8)* ‘This shows the 20% growth in C;, in the long term as required. Then Ey, = 1.50, — Cry = 1800(1.2)* + 7200(0.8)* — 1200(1.2)**! — 4800(0.8)*** Differentiate wrt k and set to zero giving 0.6 log( 1.2) + 5.62 log(0.8) = 0 where x = (0.8/1.2)* Solving, ¢ = 0.0875 and so log 0.0875 ee _ 6.007 iog(0.8/1.2) ~ °° ‘The nearest integer is k — 6, corresponding to the seventh year in view of the labelling, and Cg = 4841 approx. 18 Transforming and rearranging 2774 22-3) © Pearson Education Limited 2004Glyn James: Advanced Modern Engineering Mathematics, Third edition 171 so Y@) thus =19 Ty = Ce t+ Pr+ Ge = aly_1 + (Cy -— Ci) + Ge = aly-1 + Bayt — Iya) + Ge 80 Teyg — (1 + b)gy 1 + abl, = Gayo ‘Thus substituting Tita — Tega + she =G Using lower case for the 2 transform we obtain 2+ Si) = (22 +2)G4+6 whence Thus 201 / p =a)o kf gpkn/A _ g-ake/a Ty apres) {e e i k =2G [ + (+) sinks] v2 © Pearson Education Limited 2004172 Glyn James: Advanced Modern Engineering Mathematics, Third edition = 20. Elementary rearrangement leads to ing2 —2eosha ing tin = 9 with cosha — 1+ R,/2R. Transforming and solving for I(z)/z gives zig + (i1 — 2io cosha) Lajfe= (2 = eal 8) _ 1 fine + (iy = Big cosh a) — ige® + (#1 — 2% cosh) ~ 2sinha zee z- Thus 4, — (ine? + (i = Bio cosh a))e"* ~ (ine + (i = Bip cosha))em" .- 2sinha sinhna ~ fg sinh(n — 1a} Exercises 3.6.5 = 21 Transforming in the quiescent state and writing as Y(2) = H(2)U(z) then 21(a) H(2) 21(b) H(2) 21(c) = 22 For the first system, transforming from a quiescent state, we have (2? 40.52 + 0.25)¥ (2) = Ul2) ‘The diagram for this is the standard one for a second order system and is shown in Figure 3.1 and where ¥(2) = P(2), that is yx — pe. © Pearson Education Limited 2004Glyn James: Advanced Modern Engineering Mathematics, Third edition 173 Ua) zP(z) P(z) Figure 3.1: The block diagram for the basic system of Exercise 22. Transforming the second system in the quiescent state we obtain (2? + 0.52 + 0.25)¥(z) = (1 — 0.6)U(z) Clearly (2? + 0.52 + 0.25)(1 — 0.6z)P(z) = (1 — 0.6z)U(z) indicating that we should now set ¥(2) = P(2) — 0.6zP(z) and this is shown in Figure 3.2 O=O- + ui) 2P(z) P(z) 7 ie +| 1/z . + Figure 3.2: The block diagram for the second system of Exercise 22 © Pearson Education Limited 2004174 Glyn James: Advanced Modern Engineering Mathematics, Third edition = 28(a) Co _ 1 Yolo? = Dae tt) *° 1 1 Yo) ~ 25544 22s 1A Leary hapa 23(b) YoM2 = a 3r43 whence © = (v3) e sinkn/6+ Feoskn/6 = 2(/3)* sin(k + 1)x/6 = 23(c) <0 then 23(a) 0 and ye = (2) + (aE © Pearson Education Limited 2004Glyn James: Advanced Modern Engineering Mathematics, Third edition 175 = 24(a) 24(b) = 25 Examining the poles of the systems, we find 25(a) Poles at > =—1/3 and > = —2/3, both inside | stable, 2 |= 1 so the system is 25(b) Poles at » = —1/3 and > = 2/3, both inside | 2 |= 1 so the system is stable, 25(c) Poles at x = 1/2+ 1/29, | x |= 1/./2, so both inside | system is stable. 25(d) Poles at 2 = 3/4 17/4, one of which is outside | 2 |= 1 and so the system is unstable. 25(e) Poles at 2 =—1/4 and 2 =1 thus one pole is on | > |= 1 and the other is inside and the system is marginally stable, © Pearson Education Limited 2004176 Glyn James: Advanced Modern Engineering Mathematics, Third edition = 26 To use the convolution result, calculate the impulse response as ys,x—(1/2)*. ‘Then the step response is k ve = SO 1 OP = 1/221 y= yay OP j=0 j=0 = 0/2)F AT 1) = 2 (1/2)* Direetly, , : YOM? = Ca pjye a) so ye = 2—(1/2)* = 27 Substituting Yt — In + KYn1 = K/2" or Ynia~ Yosit Kym = 1/2 ‘Taking 2 transforms from the quiescent state, the characteristic equation is e—2z+K=0 with roots Vvl—4AK wads 1 Vi Wands = For stability, both roots must be inside | 2 wie wie |= 1 so if K < 1/4 then Vl-4K <1=+K>0 and If K > 1/4 then 1 1 |g tagV4K-1P<1SK <1 ‘The system is then stable for 0< K <1. When k = 2/9 we have 2 1 Ynt2 — Unt + Fn = = Yoo? — Unt + 5m = 9 © Pearson Education Limited 2004Glyn James: Advanced Modern Engineering Mathematics, Third edition 177 ‘Transforming with a quiescent initial state 2 1 @-2+ YW) = 5 which inverts to Un = 2(1/3)" + 2(2/3)" — 4(1/2)” = 28 P+ 2242=(2-(-1+3))(2-(-1+3)) establishing the pole locations. Then So since (—14+ 9) = VQet#*/4 ete., un — (V2) sin Bh /4 Exercises 3.9.6 229 : NS) = ay ae2 Replace s with = 2— to giv eplace 8 with < “—— to give ; M241) HG) = Tey GAG? — 1) 4 2224 IE Ae +1? ~ (446A + 2A?}2? + (4A? — 8)2 + (4— 6A + 2A?) © Pearson Education Limited 2004178 Glyn James: Advanced Modern Engineering Mathematics, Third edition ‘This corresponds to the difference equation (Ag? + Bq + C)yx = A?(q? + 2g + Lux where A=4+6A+2A? B=4A?-8 C=4—-6A+ 2A? Now put g=1+A6 to get (AA28? + 2QA+ B)AS+A+B+C)yn = AP?(A75? + 4A0 + d)uy With f = 0.01 in the q form the system poles are at 2 = 0.9048 and 7 = 0.8182, inside | z |= 1. When t = 0.01 these move to 2 = 0.9900 and 2 = 0.9802, closer to the stability boundary. Using the 6 form with f = 0.1, the poles are at v = ~18182 and v = —0.9522, inside the circle centre (—10,0) in the v-plane with radius 10. When t = 0.01 these move to y = —1.9802 and vy = —0.9950, within the circle centre (—100,0) with radius 100, and the closest pole to the boundary has moved slightly further from it. 30 The transfer function is 242841 22-1 To discretise using the bi-linear form use 6 —+ Povi to give 5 (2 +1) fits) = 4 (2+) 23+ B22 +Cz+D and thus the discrete-time form (Ag + BY + Cq+ Dy, = Tq? + 3q? + 3q + Lux where A=T* 447? 48748, B=3T* 447? sP—3, C=3P -4T?-sT4+3, D=P>—4T?+8P-1 © Pearson Education Limited 2004