COMPENS TOR NET WORK= = _TIME-DOMAIN DESIGN USING COMPENSATOR NETWORKS Most of the design procedures of an automatic control system are based on trial and error approach. Each ‘control system has to mect some specifications, ic., transient response and steady state error, already discussed in Chapter 3.on Time Response Analysis. When system is unstable, a compensating network having poles and zeros could be used to stabilize it as well as to meet all the specification to the best possible extent, If the system is stable, a compensator could be designed to mect the set specifications to the maximum possible extent as well as keeping the stability status imact. ‘Compensator can be put in cascade or in series with the plant or put in feedback path or parallel, as shown in Fig 6.18 (a) and Fig 6.18 (6). Gels) (5) Compensator in feedback path. fa) Compensator in cascade Fig. 6.8 (Compensatar can be electronic network, clectrical, mechanical orhydraulic type. Lead and lag compensators are used quite extensively in control systems. Fora sinusoidal input if the sinusoidal output leads the input, it isknown as Lead Compensator and if it Ings the input, itis known as Lag Network. For Lag-Lead Network, for the small range of frequencies, the network behaves as Lag Network and for higher range of frequencies it acts as Lead Network. A lead compensator can increase the stability or speed of response of a system; a lag compensator can reduce (but not eliminate) the steady state error. Depending on the effect desired, ane or more lead and lag ‘compensators may be used in various combinations. Lead, lag, and lead-lag compensators are usually designed for a system in transfer function form, [n general the compensator network can have the simplest transfer function of the form(s+1/t) Gas)" eran" i aT tO A614) PO It may be noted that (i) The Compensator Edit Coption in the SISO Design window of MATLAB cam heused for designing ‘appropriate compensator. Gi). By right clicking on the rootlocrs curve we ean hue options af design and representation stich as — Add pole/zero; Delete pote’zero; Edit Compensator; Design Constraints; Grid; Zoom, etc, 6.4.1 Lead Compensator Note that ar 1, ensures that the pole is located 10 the left of the zero, such a compensator is referred as first= order phase lead compensatar. The transfer function of a phase lead compensator may be expressed as Gta) el <1 Pel . A615) This compensator tends to shift the root locus ard the left half s-plane and, therefore, results in improvement in the system’s stability and an increase in the response speed, ‘How is this accomplished? Let us recall—for finding the asymptotes of the root locus that lead to the zeros at infinity, the equation to determine intersection of the asymptotes (centroid) along the real axis is: La-ka 4° ‘When a lead compensator is added to a system, the value of this intersection will be a larger negative number than it was before, The net number of zeros and pales will be the same (ane zero and one pole are ‘sdded), bul the ackled pole is a larger negalive number than the added zero, Thus, the result of a bead conipensator is that the asymptotes’ intersection is moved further into the lefi-half plane, and the entire reat locus will be shifted to the left. This can inerease the region of stability as well as the response speed. In MATLAB a phase lead compensator in root lacus is implemented by using the transfer function in the fom umlead = ke*(1 2); deniead = [1 p):and using the conv () function to implement it with the numerator and denominator of the plant newnum © conv (num, numlead); newden = conv (den, denlead); Realisation of a Phase Lead Compensator A firstorder phase lead compensator can be realised using RC Network of Fig. 6.19 a8 at “are TTR RRC ct with r= RC and @= RAR, + R:)< 1 in Eg, 6.14, Fora given or desired value of rand a, we can choose appropriate (practical) values of &,, 8; and C. The first step in the design of Ry Lead Compensator, using Root locus Technique is to convert time domain specifi-cations in to dominant set of destred close loop pole ¥ aM Sy With real and imaginary part ~ a, and jaa, as shown in Fig. ‘6,20(a) and then check whether this desired pole lies on the rootlocus caf the unity feedback system with forward path transfer function Fig, 6.19 First-order lead Compensatorsinw = sinty + ¢) | Sn == “il es lane =O=y= a The angle y for the greatest cris obtained by solving dandy = 0 pe (12) (r= 6-0) But this value of edocs not guarantee the dominance of the desired closed loop poles in the compensated root locus and it must be checked. The steady state error is also checked and if it is not satisfactory, the fixation of pole zero location is repeated at some other location but keeping the angle fixed. Uf the campensator poles lies ow the positive real-axis, then a single lead network cannot help and nore than one lead network in cascade can be used for final compensation. Example 6.14 Design a lead compensation for the system with feed forward transfer function Gb) lL ase" to provide a closed-loop damping > 0.5 and natural frequency of oscillation «, > 7 rad/s. The general wansfer function of a lead compensator is given as {s (sit) GAN Garg) ear a=(l, th Let us design for the limiting condition of the damping ratio and natural frequency of oscillation. Therefore, lectus choose §= 0.5 and w, = 7 rad/s,Hence the closed loop pales of the system are given by, ~ fa, +ja, qi = , ordesired dominant closed loop complex ‘conjugate poles are s = -3,5 + j 6.062, Let us choose z = -2, The angle subtended by all the poles cand zeros of the feed forward transfer function, with the closed loo] 3.5 + 6.062 is - Op - 112.41-120 + 103.893", For this angle to be equal to = 180°, the angle subtended by the compensator pole with the closed loop pole must be 51.488°. ‘We can obtain graphically the pole as follows: With tan (51.488) @ 6,062/x, x © 4.824 and hence p, 1 + 3.5 = 8.324, ‘The parameter k of the compensated network is calculated ‘by satisfying the magnitude criterion at dominant pole location, 5" -3.5+/ 6.062. The parameter k comes out to 60, The root locus of the compensated system with lead network is obtained, as shown in Fig, 6.21 by using the following MATLABO script: a=[0 0 1 d=[l 1 0) g=tfin,d) SISO tool (g) ‘The transfer function of the lead compensator is obtained as sl 0 a f 4 2 0 eal Aci Fig. 6. Root locus of compensated system with lead network (Example 6.14)is+2) Gels} Gils) Gia) = 60 9) “The block diagram of the compensated system is shown in Fig. 6.22, 6.4.2 Lag Compensator A firs order phase lag compensator is: Fig. 6.22 Block diagram of the compensated system (+z) _ +i/o z ® 2 = at] AGAT, G0" Coen) teary p,” 6.1) ‘Since the magnitude of z, is greater than the magnitude of p,, a phase-lag compensator tends 10 shift the foot locus to the right, which is undesirable. For this reason, the pole and zero of a lag compensator miust be placed clase together (usually near the origin) sa they do not appreciably change the transient response ot stability characteristics of the system, How does the lag controller shift the rootlocus to the right? We recall—for finding the asymptotes of the Toot locus that lead to the zeros at infinity, the equation to determine the intersection of the asymptotes along the real-axis is: La-De 7S nm =a2 ‘When a lag compensator is added to a system, the value of this interscetion will be a smaller negative rnuraber than it was before. The net number of zeros and poles will again be the same (one zere and one pale are added), but the added pale is a smaller negative number than the added zero of compensator zero, Thus,‘the result ofa lag compensator is that the asymptotes’ intersection is mowed closer to the right-half plane, and ‘the entire root locus will be shifted to the right. It was previously stated that that lag controller should only minimally change the transient response because of its negative effect. If the phase-lag compensator is not supposed to change the transient response noticeably, ‘what is it good for? The answer is that aphase-lag eompensitor can improve the system's steady-state response, I works in the following manner, At high frequencies, the lag controller will have unity gain, Atlow frequencies, the gain willbe.s,/p,, whichis greater than 1, This factor, p, will multiply the position, velocity, or acceleration constant (K,, K,, or.X,), and the steady state error will thus decrease by the factor = /p,.. In MATLAB a phase fag compensator in root locus is implemented by using the transfer function in the form numlag = dentag = and using the conv ( } function to implement it with the numerator and denominator of the plant ‘newnum = conv (num, numlag); cnewden = conv (den, denlag); Realisation of a Phase Lag Compensator A first-order lag compensator can be realised using & C Network ofthe Fig. 6.23 as [a Ves) Rats (s+ /R,C) G48)" V5) B+ Ry + YC ee G.18) STR RVR IRC with t= RyCand f= (Ry + RyRy > Vin Eq. (6.17).with «t= AyCand B= (Ry + RyRy > Vin Eg. (6.17). Ry hi ¥ Ve {| t# | . (a) Log network: (0) Location of pove-zera Fig. 6.23 First-order lag compensator Appropriate (practical) values of &,, R, and C can be chosen for given or desired value of tand a For time-domain design of Lag Compensator using Root locus Technique the steps are: 1. Plot the root locus of the uncompensated system. 2. Change the time domain specification in vo dominant set of desired close loop poles 3, (real and imaginary part with E and co,) and then check whether this desired pole Tie on the root locus af G(s) of not. If the system response satisfies the transient specification, the dominant roots of the closed loop system will lie on or be close to the root locus of the system. . Calculate the gain k of the uncompensated system at the dominant root s, and calculate the corresponding error constant. 4. Determine the factor by which the error constant of the uncompensated network should be increased to meet the specified value. Select the a parameter of the lag compensator to be somewhat greater than this factor.5. Choose a zero of the compensator sufficiently close to the origin. For simplicity, construct a line making an angle of 10° with the desired ¢ line from 5. The intersection of this line with real-axis Bives location of compensator zero, 6. Compensator pole can then be located at =p, = =z/at In order to ensure that transient specifications age still met, the pole-zero pair should contribute amie wy sinall nearly $° or less at s, 50 that the _ Lag-Lead Compensator We recall that lead compensator is suitable for systems having unsatisfactory transient response but it provides only a limited improvement in steady state response. Whereas a lag compensator is a good choice if the transient response is satisfactory but the steady state response is unsatisfactory. ‘When both the transient and steady state responses are unsatisfactory we use combination of lag and lead compensators. In such a case we first design a lead compensator to meet the transient response specifications, If the error constant requirements are also met satisfactorily, the design is complete and we do not use lag ‘compensator. If the error constant is much higher than that obtained with lead compensator, then we design a lag network to compensate for the error constant, Usually acceptable design may be obtained afler several trials,