You are on page 1of 5
4 Analysis of Discrete-Time Systems Chap. 3 Output PLU UU Input S 0 5 tT Time Figure 3.20 The step response and the control signal of the double inte- grator when the controller of (3.30) is used. 3.6 Problems at Determine if the following equations have all their roots inside the unit dise: (a) 2-152 +09 =0 (b) 2 27 422-05 = 0 (0) 2-227 + 2-05 =0 (a) 2% + 52" ~ 0252 1.25 = 0 {e) A= 112 +172 -0.7=0 Consider the system in Fig. 3.5 and let 12)( ‘Determine the values of K for which the closed-loop syatem ia stable. Consider the system in Fig. $.21. Assume that the sampling is done periodically ‘with the period h and that the D-A converter holds the control signal constant over Figure 3.21 Closed-loop system for Problem 3.3, Sec. 3.6 Problems: 118 the sampling interval, The control algorithm is assumed to be u(hh) = (uth 7} = 9th = )) where K > 0 and r is the computation time. The transfer function of the process is (a) How large are the values of the regulator gain, K, for which the closed-loop system is stable if 7 = O and r =A? (©) Compare this system with the corresponding continuous-time systems, that is, when there is n continuous-time proportional controller and a time delay in the process, %4 Determine the Nyquist curve for the system 8.5 From the system x{k+ t= (; Hi a+ (4) fh) vk) = (0 1) ath ‘the following values are obtained yaj=0 u}=1 yQ)=1 u(2} Determine the value of the state at A = 3. 86 Is the following system (a) cheervable, (b) reachable? xk+ 1 = (° asl x(t) + (i u(t) yh) (2 ~4) x) 3.7 Is the following system reachable? x(k) = (0 os} aw(} al uff) Assume that a scalar input u'(k) such that w= (2) aw is introduced. Is the system reachable from u'{k)? 116 Analysis of Discrete-Time Systems Chap. 3 38. Given the system o12 x(t+tj=]0 0 3 000 0 x(k) + | | uk) 0 (0) Determine a control sequence such that the system is taken from the initial stata 7(0) = ( 1 1) to the origin, (©) Which is the minimum number of steps that solve the problem in (a)? (©) Explain why itis not possible to find « sequence of control signals such that the state (1 1 1)" is reached from the origin. 89° Vorify the formula for W," given in Example 3.9 for an nth-order systam, 3.10 The system (b+ 1) = @x{b) + Tul) has been obtained from the system a{k+ 1) = Fe(h) + Gulh) by a linear transformation 2=Te (a) Use the result in Sec. 3.4 to derive a formula for 7 when dim(u) = 1 and dim(u) =r (b) Use the result to solve Problem 27. 3.11 Determine the stability and the stationary value of the output for the system de- scribed by Fig. 3.21 with 1 M0) = 39-05 when 1, is a step function and (a) H.(q) = K (proportional controller), K > 0; and (b} H.(q) = Kq/(q — 1) (integral controller), K > 0. ‘9.42 Consider the system in Problem 3.11. Determine the steady-stata error hetween ‘the command signal, u,, and the output when u, is a unit ramp, that is, u¢(k} = b. Assume that H. is (a} a proportional controller and (b) an integral controller. 813 Sample the system set O68) Sy oastt and determine the sampling intervals for which the response of the system will have hidden oscillations. Verify by simulations, Sec.3.6 —Probiems WwW ‘3.14 Consider the tank system with the pulse-transier operator given in Problem 2.10(b}, that is, when the system is sampled with 4 = 12. {a} Introduce a controller as in Fig. 8.21. Let the command input be a step and determine the steady-state error when using a proportional controller K and fan integral controller K/(1 ~ q74) (b) Simulate the system using the controllers in (a). Investigate the influence of the controller gain X. Determine K such that the poles of the closed-loop system correspond to a damping of f = 017 3.15 Consider the system in Fig. 3.5, Derive a formula for the velocity error coefficient. Thal is an expression for the steady-state error when the reference signal u, is a unit ramp. 8.16 Determine the values af X > 0 for which the system act +o ata) J) = BT oie is stable under simple feedback, 8.17 Determine a coordinate transformation = Tx that transfers the system afk 1) = (i | afk) + (") a(h) wfth= (8 6} x) to controllable canonical form and to observable canoniea! form, 3.18 Assume that the continuous-time system (CT) dx Onde i +Bu yas is sampled and gives the discrete-time system (IY?) (bh +h} = Ox(Ah) + Tu(hh) tle) = Cx(Bh) Consider the following statements: (a) CT stable = DT stable (b) CT unsteble = DT unstable (e) CT stable inverse = DT stable inverse (a) CT unsteble inverse => DT unstable inverse {e) CT contrallable = DT controllable (8 CT observable =» DT observable (g) CT pole excess r = DT pole excess r ne Analysis of Discrete-Time Systems Chap. 3 Which statements are true for the following eases: (i) All sampling intervals & > 0. (ii) All A > except for isolated values. (iii) Neither (f) nor (ii). 3:19 Consider the system 0-3 2 0 xeene |3 -12 7 vel ath) 6 -21 12 2 ‘Determine whether (a) the system is reachable, {b) the system is controllable, 3.20 Given the system (q? + 0.4q)y(&) = w(k) {a} For which values of X in the proportional controller uh) = K (ue (R) ~ (8) is the closed-loop system stable? {tb} Determine the stationary error u, ~y when u, is a step and when K = 0.5 in the controller in (a). B21 Assume that the system lk) - L2y(k - 2) + OSy(e- 2) = Odute ~ 1) + O8u(k - 2) is controlled by uk) = —Ky(h) (a) Determine for which vaiucs of K the closed-loop system is stable. {b} Assume that there is a computational delay in the controller, that is, uk) = ~Ky(e - 1) For which values of K is the closed-loop system now stable? 3.7 Notes and References Original papers on tests for checking if a polynomial hae all its poles inside the unit circle are Schur (1918) and Cohn (1922). Jury's test is a simplification of the Schur-Cohn test and is found in Jury and Blanchard (1961). A simple

You might also like