Ilchmann, Achim: Non-identifier-based high-gain adaptive control

Size: px
Start display at page:

Download "Ilchmann, Achim: Non-identifier-based high-gain adaptive control"

Transcription

1 lchmann, Achim: Non-identifier-based high-gain adaptive control Zuerst erschienen bei: London : Springer, 1993 SBN (Lecture notes in control and information sciences ; 189). Zugl.: Hamburg, Univ., FB Mathematik, Habil.-Schr. : 1993

2 Lecture Notes in Control and nformation Sciences 189 Editors: M. Thoma and W. Wyner

3

4 r Achim lchmann Non-dentifier-Based High-Gain Adaptive Control Springer-Verlag -ondon Berlin Heidelberg New York Paris Tokyo Hong Kong Barcelona Budapest

5 Series Advisory Board L.D. Davisson' M.J. Grimble' H. Kwakernaak' A.G.J. MacFarlane J.L. Massey 'J. Stoer Y. Z. Tsypkin ' A.J. Viterbi Author Achim lchmann, PhD nstitut frr Angewandte Mathematik, Universität Hamburg, Bundesstrasse 55, Hamburg, Germany SBN 3-54G'19t45-8 Springer-Verlag Berlin Heidelberg New York SBN G Springer-Verlag New York Berlin Heidelberg Apart from any fair dealing for the purposes of research or private surdy, or criticism or review, as permined under the Copyright, Designs and Patents Act 19E8, this publication may only be rcproduced, storcd or transmitted, in any form or by any me8ns, with the prior permission in writing of the publishers, or in the case of reprographic reproduction in accordance with the terms of licences issued by the Copyright Licensing Agency. Enquiries concerning rcproduction outside those terms strould be sent to the publistrers. G Springer-Verlag London Limited 1993 Printed in Great Britain The publistrermakes no rcprcscnbtion, expreer or implicd, with regard to the accuracy of the information contained in thie book and cannot accept any legal responsibility or liability for any emors or omissions that may be made. Tlpesetting: Camen ready by author hinted and bound by Antony Rowe Ltrl., Chippentram, Wiltstrirp 69/383G' Printed on acid-free paper

6 'r\ P rrerace '[']rc sr:rninal work of Morse, Nussbaurn and Willems & l3yrrres irr 1983/4 has initiated the study of adaptive controllers for dyramical systems in which the adaptation strategy does not invoke any identification mechanism. Over the last clecaclc, this field of adaptive control has become a rnajor research t,opic. ''he present work gives a rather complete'state of the arrt'of the following more specific area: The system classes under consideration contain linear (possibly nonlinearly perturbed), finite dimensional, continuous tirne systems which are stabilizable by high-gain output feedback, therefore, in particular the system is rninimum phase. Simple aclaptive controllers involving a simple switching strategy in the feedback are designed. The switching strategy is mainly tuned by a one parameter controller based on output data alone. Control objectives consiclcrcd are stabilization, tracking, and servomechanisnr action. n addition, robustness with respect to nonlinear perturbations and perforrnance improvements are investigated. wrote the present text during a two years research visit to the Centre for Systerrrs :rnd Control Engineering at the University of Exeter, U.K., frorn October September The hospitality of the centre, with its stimulating environrnent, made a big contribution and thank especially Dave Owens and Stuart, 'lownley, of the centre, for numerous helpful discussions and suggestions. [ :rlso benefitted frorn many stimulating discussions rn'ith Gene Ryan of the Univcrsity of Bath and Hartmut Logemann of the University of Bremen. am indebted to leinrich Voß of the University of larnburg who read the mzrnuscript ancl rnade several critical and helpful comments. Finally, thanks are due to Dir:t,er Neuffer frorn the University of Stuttgart who spent considerable tirne an<l pat,iencc on introclucing me to SMULNK and MATLA. My visit was nr:rclr: possible by o two years research grzrnt from Deutsche Forschutrgsgt:rrte,ittschzrft (DFC;) ancl additional support carr]e frorn the University of Exeter zrnd thc iic SCFNC prograrrrne, whicir zrre hereby gratefrrlly acknowledged. arnl;urg, tebrrtary, 1993 Ar:hzrn ilr:htnann

7

8 Contents Nornenclature ix rrtroduction High-Gain Stabilizability 2.1 Minimurn phase systems 2.2 lligh-gain stabilizable systems Notes and References 28 Ahnost Strict Positive Realness 3.1 Almost strictly positive real systems 3.2 Minimum phase systems 3.3 Notes and References Jniversal Adaptive Stabilization Switching functions Stabilization via Nussbaum-type switching Stabilizatiorr via switching decision function FJxponential stabilization via exponential weighting factor Exponential stabilization via piecewise constant gain Notes and References 101 fjniversal Adaptive T\'acking 5.1 Asymptotic tracking 5.2 )-'lracking 5.3 Notes ancllteferences l1'2 t.)9 L.)

9 vltt CONTENt'S 6 R.obustness L Additive nonlinear state and input perturbations 6.2 Sector bounded input-output nonlinearities 6.3 )-Tracking controller r25 t34 r Notes and References Perforrnarlce mproved output and gain behaviour Arbitrary good transient and steady-state response Notes and References 183 Exponential Stability of the Terrninal Systern Root-loci of minimum phase systems 8.2 Topological aspects 8.3 Notes and References 186 r Referencers t97 ndex 203

10 NorrrenclatLlre l/ ltr+ (rr_ ) a+(a-) R[.r]'' "" R(")- " - G L,,(K) r].r (o),(a) /r"'i" (1)) /r-,r* (Ä) lal l'l l" ll'll LrQ) {1,..., i/} the set of non-negativc (non-posit,ive) real open right- (left-) half conrplex plarre the set of rn x rrl rnatrices ovcr the polynornials ring of real the set of rn x m matrices over the fielcl of rational real functions general linear group of invertible n x n rnatrices with entries in 1{ {* e l- lllrll < Ä} for rn N, ) > 0 the spectrurn of the matrix,4 g (ln x't rnininral singular value of the rnatrix A C,"^" maxinral singular value of thc rnatrix A e C,"*'n the determinant of the matrix A e C"^" numbers positive- \t< r,p'r > for r [t',p - pt' ]R'"' definite ll'llrtlre vector space of rnea,srrrable firnctions / : --- U1.", 1C R an interval, z being defirrecl by the context,, srrctr that l/( ) lt4,1(,):), wht:re

11 NOMENC-4TURE ll/( ) llr"1'; c, (1, R-) wr,* (1, R-) d.r( ), Dr(') - ( r rr/p l/ll/(")ll"a"l { Lr es sup llf (") ll lr s for p [l,oo) f"r p - oo the vector space of ptimes continuously differentiable functions/ : --* R*,p NU{oo} the Sobolev space of functions / :, -* R- which are absolutely continuous on compact intervals and /('), /(') g.["" (R) ' 'distance' functions defined in (5.12) respectively (5-18)-

12 Chapter 1 ntroduction A wide range of control theory deals with the design of a feedback controller for a. knoutn plant so tirat certain control objectives are achieved. The fundarnental clifference between this approach and that of adaptiue controlis that in adaptive corrtrol the plant is ttot knoun exactly, only structural information is available, e.g. n-rinirnality, rninimum phase, or known relative degree. The aim is therefore to clesign a single controller which achieves prespecified control objectives for every rnember of a given class. The controller has to learn from the output data and, b:rsed on this information, to adjust its parameters. The first, attempts in adaptive control go back to the late 1950's, but it was only in the 1970's that a breakthrough was made. Subsequently, during the 1980's the field of adaptive control has matured. For a survey see Aströln (1987) and Narendra (1991). Up to the end of the 1970's, adaptive controllers were a combination of identification or estimation mechanisms of the plant paranreters together with a feedback controller. An area of non-identzfier-based adaptive control was initiated by Mareels (1984), Märtensson (1985), Morse (1983), Nussbaum (1983), and Willems and Byrnes (1984). n their approach, the adaptive feedback strategy is not based on any identification or estirnation of the process to be controlled. This seminal work opened up an intensively studied specialised field within adaptive control, where the class of systems under consideration are either minirnum phase or, more generally, only stabilizable and detectable. See lchmann (1991) for a survey. n the present text, non-identifier-based adaptive controllers for minimum phase systems are studied, thus all controllers are designed according to the high-gazn properties of the system class. No assumptions are made on the upper bound of the order of the process, nor on the upper bound of the sign of the highfrecluency gain, no injection of probing signals is required, and the control strat,egy is rnore efficient than for non-minimum phase systems. 'he objective is to provide a single controller (consisting of a feedback law ancl a pararrreter zrdaptation l:rw) which can control each system belonging to a certain cla.ss of

13 2 systems. The control objectives are stabilization, tracking or servornechanism action, partly under performance requirements and in the presence of nonlinear perturbations. We illustrate the idea by the simplest example we can think of:''he system to be st,abilized belongs to the class of scalar systems describecl by b(t) y(t) ar(t) + öu(l) cx(t), z(0) - xs) (1 t) where e,b,c,ro R are unknown and the only structural assumption is cö > 0, i.e. the sign of the high-frequency gain is known to be positive. lf we apply the feedback law u(t) - -ky(t) to (1.1), then the closed-loop system has the form i(t) -la- kcblr(t),r(0) - x:s (12) Clearly, if aflcbl < lßl and sfgn(ß) : sign(cb), then (1 2) isexponentiallyst,able. {owever, a,b,c are not known and thus the problem is to find adaptively an appropriate k so that the motion of the feedback system tends to zero. Now a time-uarying feedback is built into the feedback law u(t) - -k(t)y(t), (i.3) where,t(t) has to be adjusted so that it gets large enough to ensure stability but also remains bounded. This can be achieved by the adaptation rule Ä1ry - y(t)2, e(0) R (1 4) The nonlinear closed-loop system (1.1), (1.3),(1.4), i.e. J'(t) - la - k(t)cbfu(t), k(t) - "' l: "(")'ds * e(0), (*(0),'(0)) R2 (1 5) has at least a solution on a small interval [0,r), and the non-trivial solution r(t) _ "flt"-ft( s )crjas r19; is rnonotonically increasing as long as c - k(t)cb ) 0. lence k(l) 2 l(cr(o))' + ft(o) increases as well. Therefore, there exists at* ) 0 such that o -k(t. )cä - g and (1.5) yields o - ft(f)cö < 0 for all t > t*. lence the solution r(t) decays exponentially and liml*"" e(r) - ßoo R exists. This is a special example for the following concept of universal adaptive control. Suppose E denotes a certain class of linear, finite dimensional, tinrc-invariant systems of the form i(t) vu) Ar(t) + Bu(t), z(0) - C r(t) + Du(t) ' g \ t (1 6)

14 3 wlrere: (tl, B,C,D) [R'xn x R'',x'2 x R,'-"" x R'r'x') are unknown, nt is usually fixecl, t,he state dirnension n is an arbitrary and unknown nurnber. 'fhe aim is to design a single adaptive output feedback mechanisrn,rr(l) - f (y( )lto,rl) which is a universal stabilizer for the class. i.e. if "(l) - f (y( )llo'l) is applied to any systenr (1.6) belonging to, then the output 37(l) of the closed-loop system tends to zero as tends to infinity and the intcrtral variables are bounded. n the present t,ext, most of the adaptive stabilizers are of the following sinrple forrn (cf. F igtrre 1.1): A 'tuning' parameter A(l), generated by an aclaptation law k(t) - g(y(l)),,t(0)-,lo, (r 7 t whcrc g :R'' -* R is continuous and locally Lipschitz, is irnplementecl into thc fc,'rlback law via u(t) - F(,L(t))y(t), whert: l/ :R -r trr"*- is piecewise continuous ancl locally Lipschitz. (t 8) r t Adaptive Controller Adaptation Law Process i:arbu a:cr*du L Figure 1.1: Universal adaptive stabilizer Definitiorr A controller, consisting of the aclaptation law (1 7) ancl the feedback rule (1.8), is c.af lr:rl a ttniue.rsul adaptiue stabilizerfor the class of systems X, if for arbitrary init,iäl condition ro l" and any system (1.6) belonging to, tiie closc.rl-loop systenr (1 6)-(1.8) has a solution the properties (i) t,lrere cxists zr unique solution (r( ),Ä'()):[U, x) - R'.'+1, (ii) e,( ), y(.), rr( ), t'(.) are boundccl,

15 (iii) lirnl*oo y(t) - 0, (i") liml*oo k(r) - ßoo R exists. The concept of adaptive tracking is similar. Suppose a class./rer of reference signals is given. t is desired that the error between the output y(t) of (1.6) and the reference signal y."r(l) "(t) :- y(t) - y..r(t) is forced, via a simple adaptive feedback mechanism, either to zero or towards a ball around zero of arbitrary small prespecified radius Ä > 0. The latter is called )-tracking. To achieve asymptotic tracking, an internal model itr) u(t) - A- e (t) * B* u(t), (0) : 40 :- c- e (t) * D* u(t) ( 1.e) wlrere (A*, B*, C*, D") e [t"'" n' y Tf7.n'Xtnx R-"'' x R-t-, i, implemented in series interconnection with an universal The precompensator resp. internal model reference signals. An internal model is not adaptive stabilizer, cf. Figure 1.2. (1.9) contains the dynamics of the necessary if A-tracking is desired. Adaptive Tracking Controller nternal Model Process 9."f : A' * B'u u:c'4* D' i:ar*bu g-cr*du Figure 1.2: Universal Adaptive Tlacking with nternal Model Definition A controller, consisting of an adaptation lau,t(1.7), a feedback lau (1 8), and an internal model (1 9) is called a uniuersal adaptiue tracking controller for the class of systems! and reference signals.)y'..r, if for every U."r(.).)."r. ts R', o R"', and every system (1.6) belonging to X, the closed-loop systern (1 6)-(1.9) satisfies

16 5 (i) t,here exists a unique solution (r( ), ( ),f(.))'[0,m) * R72+n'+i (ii) the variables r(/),y(t), r(r), (l) 'blow up' no faster than 9.er(l), (iii) limt*o,[y(t) - 9..r(l)] - 0, (ir') liml*"o fr(l) - k.'. R exists. F'or prespecifiecl Ä > 0, (1 7), (1,8) is called a un,zaer-sal adapt.iue \-stabzlizer resp. ),-tracking controllerif (i),(ii), and (iv) hold true, but instead of (iii) the weaker condition [v(t) - e.er(l)] - E^(o) as l-*oo is satisfied, i.e. the error e(l) approacires the closed as tcnds to oo. f zr universal adaptive controller is applied to ( 1.6) then we call the systern L the termznol systern, provided it is well defined. ball of radius ) around zero for some (ro, Ao) R." x R, '(,t-,))-tcl r(r) J Many results presented in this text fit into the framework described above. We will also consider linear systerns subjected to nonlinear perturbations in the state, input and output, corrupted input and output, noise, and nonclillerentiable gain adaptation. Due to the nonlinearities, the solution of the closed-loop systern of t,he norninal system and ttre aclaptive feedback rnechanisrn is no longer unique, but all solution will meet the desired control objectives. Some feedback strategies contain non-differential gain adaptation, other have multiparameter gain adaptation, i.e. g(.) i" (1 7) is a mapping from Rto R-' for some n{ > 1. Many results can be extended, by no mearrs trivially, to classes of infinite-dimensional systems. This will not be treated here, but the zrva.ilable literature is discussed in the sections'notes and References'ending each chapter. ''tre outline of the chapters is as follows. n Chapter 2 ancl 3 we do not deal with any adaptation rnechanisrns, inst,ead the systern classes under consi<ieration are analysecl ancl results to be used later are prr:parcd. n (-'Lrapter 2, we study irr detail the piropi:rtics of nrult,ivzrri:rblc rnininlunr plrzrse sysl,errrs, give corrvenit:nt, slat,e spac.e forrns, prove the so called high-gain lcmrnat,a for relative clegree ancl 2 systr:rns, and clr:rive irnl>ortant, incclualitit:s relaling past inputs ;rncl out,1>uts with t]re present, output. T'lris k:ads t,o a ral,- her cornplete understanding of high-gain stzrbilizable systerns and toc-,ls which

17 will be used throughout the remaining chapters. n Chapter 3, the class of strictly positive real ancl alrnost strictly positivc real systems is investigated. Although this class is more restrictive than minimunr phase systems it turns out that, by simplc input and output transformations, every minimum phase system is equivalent to an almost strictly positive real system. The result will become important for the stability proofs of adaptive stabilizers given in Chapter 4. Recently published results on multivariable strictly positive real systems, in particular its relationsphip to the Lur'e equations, are used to understand the effect of high-gain control, and simplify the proofs for adaptive stabilizers. Moreover, it allows consideration not only of strictly proper, but of proper linear systems. n the first section of Chapter 4, the concept of switching functions is studied. Different switching functions are introduced and it is shown how they relate to the Nussbaum conditions. These results are needed for the remaining sections of the chapter. n Section 4.2, various universal adaptive stabilizers based on Nussbaum-type switching are derived for systems (A, B,C,D) under milcl assumptions on the high-frequency gain matrix. A universal adaptive stabilizer for the class of all single-input, single-output, positive or negative high-gain stabilizable systems is given. Throughout the section the feedback strategy is continuous respectively piecewise continuous, i.e. f'(.) in (1.8) is continuous respectively piecewise constant. n Section 4.3, an alternative switching strategy based on a switching decision function is introduced. Apart from robustness properties shown in Chapter 6, the advantage of this different approach is tliat there is no need to implement a scaling-invariant Nussbaum function whicli is behaving very rapidly. n Sections 4.4 and 4.5, we deal with the problem of how to obtain exponential decay of the state r(l) of the closed-loop system. n Section 4.4, this is achieved by introducing an exponential weighting factor in the gain adaptation, whereas in Section 4.5 a different approach uses piecewise constant gain implementation in the feedback strategy. n Chapter 5, the problem of how to track signals belonging to a class of reference signals is investigated. One solution to the problem of asymptotic tracking is presented in Section 5.1. We use series interconnection between arr internal model, representing the dynamics of the reference signals, and universal adaptive stabilizers studied in Chapter 4. This approach is restricted to sinusoid reference signals, whereas in Section 5.2, at the expense of A-tracking, a controller using dead-zones is introduced which does not invoke an internal model and works for a much larger class of reference signals. Well posedness and robustness properties of the universal adaptive stabilizers of Chapter 4 and the asymptotic and )-tracking controllers of Chapter 5 are investigated in Chapter 6. n Section 6.1, it is proved that the problern of universal adaptive stabilization is well posed with respecr to nonlinearities in the state equation of the nominal system. Robustness with respect to other nonlinear perturbations is proved. n section 6.2, rve show that rnanl-universal adaptive stabilizers of single-input, single-output systems tolerate sector bouncled input-output nonlinearities, for multi-input, multi-output almost strictly positive real systems ever multivariable sector bouncled input-output nonlinea-

18 rities are allowed. n Section 6.3, we prove t,hat the Ä-stabilization respectively )-tracking controller is capable of toierating a nruch larger class of nonlinear perturbations in the input ancl state as well as sect,or bounded input-output nonlinearities, even input and output corruptecl noise is allou'ecl. 'fhe purpose of Section 7.1. is to illustrate the qualitative dynarnical behaviour of nrany universal adaptivc stabilizers and to introcluce modifications of the previous universal adaptive stabilizers and )-tracking controllers which lead to an improvement of the transient behaviour. n Section 7.2, a universal adaptive stabilizer is designed whic.h, at the expense of derivative feedback, achieves (prespecified) arbitrarily srnall overshoot, of the outprut, and, rrloreover, guarantees that the output is less thern an zrrbitrarily small, prespecified constant in an arbitrarily small, prespecified period of time. 'he results on root-loci of single-input, single-outjrut nrininrum phase systems rlerived in Section 8.1 are used in Section 8.2 to show that that the piecewise constant stabilizer introduced in Section 4.5'almost always'(w.r.t. tlie sequence of thresholds) yields an exponentially stable terminal system. Each chapter is finalized with a section on the literature and related problenrs, in particular, extendecl results for infinite-dinrensional systems are quoted.

19 8

20 Chapter 2 High- Gain Stab ilizability n this chapter, we derive several properties of high-gain stabilizable and/or minimum phase systems which are essential for getting a deeper insight into these system classes, and wtrich will be used throughout the remaining chapters. 2.L Minirnurn phase systems We shall show that all non-trivial systems which are stabilizable by high-gain output feedback are necessarily minimum phase. Therefore, it is important to study the class of minimum phase systems. We will also give simple and convenient state space forms for relative degree or 2 systems and prove a crucial integral inequality relating past input and output data to the present output. Definition Let G(.) e m(s;rrlxln be a rational matrix with Smilh-McMillan form r.. (s) diag {( "r(s)'',ä3,0',0) - U(r)-tC(")tz(")-' (2.r) where U(.),V(.) e R["]-x- are unimodular, r,bp1")g(.) : r, e;(.),,lrro R[s] are monic and coprime and satisfy et1.)le;+r(-), rlrr+t(.)lrlrr(.) for i -,...,r. Set + + e (s):- ll e;(s), ü(") '- ll ü;(') i=l.s6 is a (transntzssion) zero of G(.), if e(so) : 0, and a pole of G(-), if y'(.so) : 0. f G(-) - g( ) tr[s], then d"gy'(.) - dege( ) is called Lhe relatzue degree of s (-) i-

21 10 CHAP']:T:R 2. HGH-GÄ.N STABLZABL'Y {;(-) is pt-oper- resp. strtctly propet' if deg r/,( ) > dege(.) resp. d"g,/,(.) > dege(.). 'he system i(t) = Ar(t)*Ru(t), v(t) - cr(t)+du(t) with (A,B.,C,D) R"' x m,'"^ x fli. n'*' x l.-*'', i" called a minimal realization of G() R(s)-*-, if (A, B) is controllable and (,4, C) is observable ancl G(s) - C(s^ - A)-rB + D. G(-) is said to be minimum phase, if e(s) 0 for all.s e O1. A state space system (A,B,C,D) R'x'r x TR'z'^ x R*"' x R',*- is called minimum phase, if it is stabilizable and detectable and G(") ha^s no zeros in c+. A characterization of the minimum pha-se condition for state space system is given in the following proposition. Proposition 2..2 (A, B, C, D) R'x' det xil RN i' K TfT all s O1 if, a.d only if, the (f following,win. three f l conditio.s lol ns are af( :e satisfied (i) rklsln-a,bl- :n. n for all s 6i e ( C1, i.e.(a,b) isstabilizablebystate feedback,.["t-_ Al - (tt) rk : t for all s C+, i.e. (,4,C) is detectable, L C J (iii) G(") has no zeros in O=,.. r, l"i L s,. C a) A rh xr ffl. B D X J T. J R 0 ftx' fi7 fo)r satisfies Proof: We trse the notation of Definition Coppel (1974), 'fheoren-r 10, has proved that, if (Ä, B,C) is detectable ancl stabilizable, then so O+ is a zero of r/,( ) (including multiplicity) if and only if it is a zero of det(.1, - A). Using this result together with Schur's forrnula, see e.g. Gantmacher (195g), the proposition follows from s[,, - '- A -B ', _ t^t Al r-r /1/^r,rr-1n, ls1r. -Ä D

22 2.1. M{TMUN{ PHASE S}'S?EMS 1l ''he following lenrrna provides a useful state space fortn into which every systenr u (l) v(t) Ar(t) * Bu(t), C r(t), r(0)-ro R' (2 2) witlr (,4,8,C) R'"'x.lRrlx- x R-"'with det(cb) f 0 can be converted. 'f5r-- state space transformation is representing the direct sum of the range of B alcl t,he kernel of C. t makes possible the separation of the inputs and or-rtputs fronr the rest of the system states. Leurrna 2-L.3 (-lorrsicler the system (2.2) with det(c13) a basis rnatrix of ker C. t follows that g-,-- lc', N')t, where l/ :- (V'V) state space transformation (y',,')' - (2.2) into 0 and let,/ JR'x('.--l clenote.9 :- lb(cb)-t, Y] has t,he inverse -t v' lt. - B(CB)-1cr]. lerrce the S- 1c - ((C *)t., (Nr)t)' converts v(t) t (t) Ary(t) * Azz(t) + CBU(t) Asy(t) + A4z(t) ( v(0) '(0) J n-1 L) & t). (23) lere At R-*^,A, R-x(n-rn),At e 6(n-zn)*^,,An R("-'n)x(n-"n), so that Av A2 A3 Aa - S-r AS. f (,.4,8,C) is minimum phase, then,4a in (2 3) is asymptotically stable. Proof: The proof of the transformation is straightforward ancl therefore onrittecl. Stability of Aa is a consequence of det(c B) + 0, the minirnum phase assumption, and of the following equation which holds for all s e C1 sln-a C B 0 slrn - A1 -As rn -42 sn-rn - Aq 0 CB lc B l.l"1, -rn - Aol * 0.! ior strictly proper, scalar, nrinimurn pliase systerrrs of relative dcgree 2, a slightly rnore complicated brrt still very useful stat,e space clescription is also avail:rblr:.

23 2 CHAP'ET 2. HGH.GÄN STAB,ZABLTY d('y,).\ 1 o1/ f: nlü(t) - 1", #f a[ llüio l+l,e \ z(t)l Lo, 'öo ;")\",t>) L-;"1 where dz R, 04, as R'-', Au R('" -2)x(n-2). lf (2.2) is minimum phase, then a(aa) C C-. u(t,) (2.4) Proof: Choose V R'x('-2) of full rank so thar k",l:") - V trt"-2 t is easily verified, that the inverse of t si' : lcabv. V* :- (V'V)-tV', and hence e :- cs1 and eäb LoAyields A t- (SrSz)-taSrSz Moreover, ö52 - e, S;tb -6, and cazb wrth S:.9rS2. Using (2.4), we have,91 :: : c o :: - eä2b - [Ab, b,v](cab)-1 ca _ n#" [r" - (ou [1,0,...,0], ä :- 5, 1ö: J-.:r., oj - cab. Applying elementrary row and o o si,,-[t 1 t L: ;,:_,] L Lernrna 2.L.4 f the system (2.2) is single-input, single-output, (A, B,C) - (A., b, c), and of relative degree 2,i.e. cb - 0, cab f 0, then there exists a coordinate transforinto mation S e GL,(R), such that S-rr : (y,ü,t')' converts (2.2) 4..- columm operation r o o * 1 * l 0 0 n_z J 1 0 a3 "[ bca(cab) s1.4s, - 0 A e a3cab, l and is givcnby -'ll S tt t, * 1 * + * + * * * 1 of the fornt hence (2 4) holds )1,, -A b c 0 ) )-os a[ 0 Ä,.*z - Ae cab 0 0 -cab.l)/,-z -,4ol

24 2. j. MNNUM PASE SYSTEMS 1.) Lr) ff (2.2) is nrinimum Phase, then arrd hetrce Ae is asymptotically it follorvs that län-z - Aal + 0 for all ) stable. This completes the proof. O+, D Anot her itr-rportant consequenceof strictly proper minimurn phase systerns rvitir linr,*.',, s(i(s) Gf'.(R) (for single-input, single-output this simply means they are of relative degree 1) is, that a sirnple input-output description of the svsteln is Possible. Lernrna 2-L-5 f t,he syst,em (2.2) is minin-rum phase r,r'ith det(cf)) f 0, then there exists a bounded and causal operator L : Lo(O, oo) * Lr(0, m), for all P [], rc], so that the input-output behaviour of (2.2) is described by y(t) - A1y(r) + L(a()) (t) * CBu(t) + u(t), y(0) - Cro (2 5) wittr,4r R-"-, and -(')'[0,-)---* R- all exponentially decaying analytic function taking into account the initial condition of a part of the internal state. Figure 2.1: nput-output Description Proof: Without lossof generality we nray in the form (2 3). Definethe causal opcrator assume that ttre systern (2.2) is L(a( ))tt) - Az.l o t ea,(t-.') A*lG)dr.

25 T4 CAPTER 2. {G-GÄ{ STABLZAB.TY Since Aq is asymptotically stable, there exist Mr, ) 0, so that ll"o"'ll 1 M1e-'t for all t > 0. This yields, see e.g. Vidyasagar (1978) pp , that lll(y)( )llr p(o,t) L 'fherefore, f : Lo(0, oo) * Lr(0, oo) is well defined and bounded. By applying Variations of Constants to the second equatiorr in (2.3), and inserting z(l) into the first equation, we obtain l " ü0 = As(t)*Azl"o^'z(0) + *CBu(t). J"on(t-')ery1";a"l L U J Setting u(t) - AzeAn'r(0), the proof of the lemma is complete. n For strictly proper minimum phase systems (A, B, C) with det(c B) 0, it is possible to relate the present output to past output and input data via the following inequality, where no information of the state variables is required. This will be an another important tool for the stability proofs of universal adaptive stabilizers in Chapter 4. Lerrrrna 2.L.6 Suppose the system (2.2) is rninimurn ptraseand satisfies del(cb) t' 0. Let P() : R^ * R-, v,* lfu) - (., ) ffic if a+0 l. o if v:0. Then for every positive-definite matrix P - Pr R-"'' (depending only on A, B,C and P) such tl-rat there exists ' > A r ^ ' r i = tlvt,)llä 2""' "' ' J" J, or, more general, for arbitrary p > 1, it holds that, i i _llrtt)llp"< tvlllrsllp+,lr / ttots)llpds+ ltrt')llot'(p(v(")),p('btt(s))d P t r t, (2 T) for arbitrary initial condition r(0) R', for arbitrary piecervise continuous u(') : -* [0,c.,') R'', where cr (0,.c], a1d for all [0,.)

26 2.1. MNTMUM PHASE SYSTEMS 15 Proof: (.), We first consider the case p:2 (2 5) yields, for all s [0,c.,'), r d,,,,.rr1. t,ß(llv(")llf') - (v("), PAty(s) + l'u,(s) + PL(y) (") + PCBu(s)) ö o for + (v("), PCBU(s)) (2 8) Mz:-llPAtll + llpll Applying ölder's inequality and using (26) gives r / r \ r l r /', \'/', \' J ttrt")il ill(yt(s)llds s l,lv(")il'd"l l,tctü(s)il'. \l / \i ),rs 0 t llv(")ll'd' (2 e) and, since tr(s) - A2eAn'z(0), S e t. rl r llly(s)ll lltr(s)llds o ( (, t r \ 6 ; \ ' tlrt'lll'a" 6 / M3llz(0)ll llv(")ll2ds + u3ll,(o)ll' (2 10) for Ms:- sup r>0 (t,azea",'r") 2 0 ntegration of (2.8) over [0,1] and [0,c,,,), inserting (2.9) and (2.10) yields, for all )no...lt', 1 2 llv(o)ll? +M., - llz )-t z + t + MrMSll:10)ll2 + (v("), P C Bu(s))ds.l 0.l 0 t lly(") ll'd". ) 7),i- ''his proves the first inequalit,lrt itv. (b), 'he map ** lly(r)l lp isnot differentiable but, since V(.) is differentiable, it is absolutely continuous. Tl T] fherefore, the set,11 :- {t' = f0, [0, ). t^t ) llv(t )llp is nol cliflr:renl,iable]

27 16 CAPTER 2. GH.GÄN SA'ABLZABLTY is of measure zero. LeL,, Jz :- "(') Now a routine calculation gives, for all s Thus it follows frorn : ; *(ttv(")ttä) where An application,(-) Lp(0,t), {t e [0, c,.,) fttrt"ltto = f (v(") { llv (2.5) that, for all s llv(")llä-' ( ri is not continuous at t ) R+ \ (/r g ir),. Pü(s)) / \ ffi,v(')*o 0,y(s) - 0. R+ \ (/t g Jz), ), P A1y(s) M4:- llpll"+ illp1'll + llplll of Hölder's inequality gives, for q - -FPu(s) + PL(y) (") + PCBU(s)l [llv(")l lo + lll(v) (")llllv(")llo-' +l ly(") lloi' PCBU(s)), f,,o li J ttrtrllr-' llu(s)llas L'. J L"t J f-, +lly(")llo-' ll-(") lll 1 r 1 - ' q ' p 1 and every.f',0 11) Since Jru Jz is of measure zero, integration of *llv(")llpo over [0,r] \ (./r u yields, by using (2.11), for all t [0,r,r), r - r - i s + ;ltatt)llä ittvtolllä lla{")llp-'(p(y(s)), PCBU(s))ds Jz) t f *Ms 0 llv(")llo + llv(")llo-' illc(v)(")ll + ll-(")lll d" jttrroltt'" t +.l 0 lly(") llr-' U3(y(")), PC Bu(s))ds +,v/s llv( )ill," (o,r) + llv(')tl?lä,'t llc(v) ( )llro(0,r) +llv( )ll?lä,,rll-ollr"1o,,.l],

CONTROLLABILITY. Chapter 2. 2.1 Reachable Set and Controllability. Suppose we have a linear system described by the state equation

CONTROLLABILITY. Chapter 2. 2.1 Reachable Set and Controllability. Suppose we have a linear system described by the state equation Chapter 2 CONTROLLABILITY 2 Reachable Set and Controllability Suppose we have a linear system described by the state equation ẋ Ax + Bu (2) x() x Consider the following problem For a given vector x in

More information

B y R us se ll E ri c Wr ig ht, DV M. M as te r of S ci en ce I n V et er in ar y Me di ca l Sc ie nc es. A pp ro ve d:

B y R us se ll E ri c Wr ig ht, DV M. M as te r of S ci en ce I n V et er in ar y Me di ca l Sc ie nc es. A pp ro ve d: E ff ec ts o f El ec tr ic al ly -S ti mu la te d Si lv er -C oa te d Im pl an ts a nd B ac te ri al C on ta mi na ti on i n a Ca ni ne R ad iu s Fr ac tu re G ap M od el B y R us se ll E ri c Wr ig ht,

More information

Example 4.1 (nonlinear pendulum dynamics with friction) Figure 4.1: Pendulum. asin. k, a, and b. We study stability of the origin x

Example 4.1 (nonlinear pendulum dynamics with friction) Figure 4.1: Pendulum. asin. k, a, and b. We study stability of the origin x Lecture 4. LaSalle s Invariance Principle We begin with a motivating eample. Eample 4.1 (nonlinear pendulum dynamics with friction) Figure 4.1: Pendulum Dynamics of a pendulum with friction can be written

More information

Victims Compensation Claim Status of All Pending Claims and Claims Decided Within the Last Three Years

Victims Compensation Claim Status of All Pending Claims and Claims Decided Within the Last Three Years Claim#:021914-174 Initials: J.T. Last4SSN: 6996 DOB: 5/3/1970 Crime Date: 4/30/2013 Status: Claim is currently under review. Decision expected within 7 days Claim#:041715-334 Initials: M.S. Last4SSN: 2957

More information

Control Systems Design

Control Systems Design ELEC4410 Control Systems Design Lecture 17: State Feedback Julio H Braslavsky julio@eenewcastleeduau School of Electrical Engineering and Computer Science Lecture 17: State Feedback p1/23 Outline Overview

More information

Control Systems 2. Lecture 7: System theory: controllability, observability, stability, poles and zeros. Roy Smith

Control Systems 2. Lecture 7: System theory: controllability, observability, stability, poles and zeros. Roy Smith Control Systems 2 Lecture 7: System theory: controllability, observability, stability, poles and zeros Roy Smith 216-4-12 7.1 State-space representations Idea: Transfer function is a ratio of polynomials

More information

MATH 304 Linear Algebra Lecture 9: Subspaces of vector spaces (continued). Span. Spanning set.

MATH 304 Linear Algebra Lecture 9: Subspaces of vector spaces (continued). Span. Spanning set. MATH 304 Linear Algebra Lecture 9: Subspaces of vector spaces (continued). Span. Spanning set. Vector space A vector space is a set V equipped with two operations, addition V V (x,y) x + y V and scalar

More information

1.- L a m e j o r o p c ió n e s c l o na r e l d i s co ( s e e x p li c a r á d es p u é s ).

1.- L a m e j o r o p c ió n e s c l o na r e l d i s co ( s e e x p li c a r á d es p u é s ). PROCEDIMIENTO DE RECUPERACION Y COPIAS DE SEGURIDAD DEL CORTAFUEGOS LINUX P ar a p od e r re c u p e ra r nu e s t r o c o rt a f u e go s an t e un d es a s t r e ( r ot u r a d e l di s c o o d e l a

More information

OPTIMAl PREMIUM CONTROl IN A NON-liFE INSURANCE BUSINESS

OPTIMAl PREMIUM CONTROl IN A NON-liFE INSURANCE BUSINESS ONDERZOEKSRAPPORT NR 8904 OPTIMAl PREMIUM CONTROl IN A NON-liFE INSURANCE BUSINESS BY M. VANDEBROEK & J. DHAENE D/1989/2376/5 1 IN A OPTIMAl PREMIUM CONTROl NON-liFE INSURANCE BUSINESS By Martina Vandebroek

More information

Lecture 13 Linear quadratic Lyapunov theory

Lecture 13 Linear quadratic Lyapunov theory EE363 Winter 28-9 Lecture 13 Linear quadratic Lyapunov theory the Lyapunov equation Lyapunov stability conditions the Lyapunov operator and integral evaluating quadratic integrals analysis of ARE discrete-time

More information

G(s) = Y (s)/u(s) In this representation, the output is always the Transfer function times the input. Y (s) = G(s)U(s).

G(s) = Y (s)/u(s) In this representation, the output is always the Transfer function times the input. Y (s) = G(s)U(s). Transfer Functions The transfer function of a linear system is the ratio of the Laplace Transform of the output to the Laplace Transform of the input, i.e., Y (s)/u(s). Denoting this ratio by G(s), i.e.,

More information

Linear Maps. Isaiah Lankham, Bruno Nachtergaele, Anne Schilling (February 5, 2007)

Linear Maps. Isaiah Lankham, Bruno Nachtergaele, Anne Schilling (February 5, 2007) MAT067 University of California, Davis Winter 2007 Linear Maps Isaiah Lankham, Bruno Nachtergaele, Anne Schilling (February 5, 2007) As we have discussed in the lecture on What is Linear Algebra? one of

More information

B I N G O B I N G O. Hf Cd Na Nb Lr. I Fl Fr Mo Si. Ho Bi Ce Eu Ac. Md Co P Pa Tc. Uut Rh K N. Sb At Md H. Bh Cm H Bi Es. Mo Uus Lu P F.

B I N G O B I N G O. Hf Cd Na Nb Lr. I Fl Fr Mo Si. Ho Bi Ce Eu Ac. Md Co P Pa Tc. Uut Rh K N. Sb At Md H. Bh Cm H Bi Es. Mo Uus Lu P F. Hf Cd Na Nb Lr Ho Bi Ce u Ac I Fl Fr Mo i Md Co P Pa Tc Uut Rh K N Dy Cl N Am b At Md H Y Bh Cm H Bi s Mo Uus Lu P F Cu Ar Ag Mg K Thomas Jefferson National Accelerator Facility - Office of cience ducation

More information

Quadratic Equations in Finite Fields of Characteristic 2

Quadratic Equations in Finite Fields of Characteristic 2 Quadratic Equations in Finite Fields of Characteristic 2 Klaus Pommerening May 2000 english version February 2012 Quadratic equations over fields of characteristic 2 are solved by the well known quadratic

More information

Factorization Theorems

Factorization Theorems Chapter 7 Factorization Theorems This chapter highlights a few of the many factorization theorems for matrices While some factorization results are relatively direct, others are iterative While some factorization

More information

Fuzzy Differential Systems and the New Concept of Stability

Fuzzy Differential Systems and the New Concept of Stability Nonlinear Dynamics and Systems Theory, 1(2) (2001) 111 119 Fuzzy Differential Systems and the New Concept of Stability V. Lakshmikantham 1 and S. Leela 2 1 Department of Mathematical Sciences, Florida

More information

3.7 Non-autonomous linear systems of ODE. General theory

3.7 Non-autonomous linear systems of ODE. General theory 3.7 Non-autonomous linear systems of ODE. General theory Now I will study the ODE in the form ẋ = A(t)x + g(t), x(t) R k, A, g C(I), (3.1) where now the matrix A is time dependent and continuous on some

More information

Matrix Representations of Linear Transformations and Changes of Coordinates

Matrix Representations of Linear Transformations and Changes of Coordinates Matrix Representations of Linear Transformations and Changes of Coordinates 01 Subspaces and Bases 011 Definitions A subspace V of R n is a subset of R n that contains the zero element and is closed under

More information

NOTES ON LINEAR TRANSFORMATIONS

NOTES ON LINEAR TRANSFORMATIONS NOTES ON LINEAR TRANSFORMATIONS Definition 1. Let V and W be vector spaces. A function T : V W is a linear transformation from V to W if the following two properties hold. i T v + v = T v + T v for all

More information

Continued Fractions and the Euclidean Algorithm

Continued Fractions and the Euclidean Algorithm Continued Fractions and the Euclidean Algorithm Lecture notes prepared for MATH 326, Spring 997 Department of Mathematics and Statistics University at Albany William F Hammond Table of Contents Introduction

More information

Are We There Yet? IPv6 as Related to GDP per Capita Alain Durand, October 26 th 2016

Are We There Yet? IPv6 as Related to GDP per Capita Alain Durand, October 26 th 2016 Are We There Yet? as Related to Alain Durand, October 26 th 2016 Quesons for this Study: I. Where are we across the globe with adoption? a. Is deployed uniformly? b. Is there a rich country/poor country

More information

Lecture 7: Finding Lyapunov Functions 1

Lecture 7: Finding Lyapunov Functions 1 Massachusetts Institute of Technology Department of Electrical Engineering and Computer Science 6.243j (Fall 2003): DYNAMICS OF NONLINEAR SYSTEMS by A. Megretski Lecture 7: Finding Lyapunov Functions 1

More information

Chapter 12 Modal Decomposition of State-Space Models 12.1 Introduction The solutions obtained in previous chapters, whether in time domain or transfor

Chapter 12 Modal Decomposition of State-Space Models 12.1 Introduction The solutions obtained in previous chapters, whether in time domain or transfor Lectures on Dynamic Systems and Control Mohammed Dahleh Munther A. Dahleh George Verghese Department of Electrical Engineering and Computer Science Massachuasetts Institute of Technology 1 1 c Chapter

More information

FedEx Ground Rates. Effective January 7, 2013

FedEx Ground Rates. Effective January 7, 2013 FedEx Ground Rates Effective January 7, 2013 Introduction FedEx Ground shipping services provide you with dependable, cost-effective, day-definite delivery for packages that don t require the speed of

More information

A matrix over a field F is a rectangular array of elements from F. The symbol

A matrix over a field F is a rectangular array of elements from F. The symbol Chapter MATRICES Matrix arithmetic A matrix over a field F is a rectangular array of elements from F The symbol M m n (F) denotes the collection of all m n matrices over F Matrices will usually be denoted

More information

Matrix Polynomials in the Theory of Linear Control Systems

Matrix Polynomials in the Theory of Linear Control Systems Matrix Polynomials in the Theory of Linear Control Systems Ion Zaballa Departamento de Matemática Aplicada y EIO Matrix Polynomials in Linear Control Matrix Polynomials in Linear Control Coefficient Matrices

More information

Week 5: Binary Relations

Week 5: Binary Relations 1 Binary Relations Week 5: Binary Relations The concept of relation is common in daily life and seems intuitively clear. For instance, let X be the set of all living human females and Y the set of all

More information

NON SINGULAR MATRICES. DEFINITION. (Non singular matrix) An n n A is called non singular or invertible if there exists an n n matrix B such that

NON SINGULAR MATRICES. DEFINITION. (Non singular matrix) An n n A is called non singular or invertible if there exists an n n matrix B such that NON SINGULAR MATRICES DEFINITION. (Non singular matrix) An n n A is called non singular or invertible if there exists an n n matrix B such that AB = I n = BA. Any matrix B with the above property is called

More information

THREE DIMENSIONAL GEOMETRY

THREE DIMENSIONAL GEOMETRY Chapter 8 THREE DIMENSIONAL GEOMETRY 8.1 Introduction In this chapter we present a vector algebra approach to three dimensional geometry. The aim is to present standard properties of lines and planes,

More information

Numerical Analysis Lecture Notes

Numerical Analysis Lecture Notes Numerical Analysis Lecture Notes Peter J. Olver 5. Inner Products and Norms The norm of a vector is a measure of its size. Besides the familiar Euclidean norm based on the dot product, there are a number

More information

THE FUNDAMENTAL THEOREM OF ALGEBRA VIA PROPER MAPS

THE FUNDAMENTAL THEOREM OF ALGEBRA VIA PROPER MAPS THE FUNDAMENTAL THEOREM OF ALGEBRA VIA PROPER MAPS KEITH CONRAD 1. Introduction The Fundamental Theorem of Algebra says every nonconstant polynomial with complex coefficients can be factored into linear

More information

Matrix Methods for Linear Systems of Differential Equations

Matrix Methods for Linear Systems of Differential Equations Matrix Methods for Linear Systems of Differential Equations We now present an application of matrix methods to linear systems of differential equations. We shall follow the development given in Chapter

More information

PUTNAM TRAINING POLYNOMIALS. Exercises 1. Find a polynomial with integral coefficients whose zeros include 2 + 5.

PUTNAM TRAINING POLYNOMIALS. Exercises 1. Find a polynomial with integral coefficients whose zeros include 2 + 5. PUTNAM TRAINING POLYNOMIALS (Last updated: November 17, 2015) Remark. This is a list of exercises on polynomials. Miguel A. Lerma Exercises 1. Find a polynomial with integral coefficients whose zeros include

More information

Control Systems Design

Control Systems Design ELEC4410 Control Systems Design Lecture 20: Scaling and MIMO State Feedback Design Julio H. Braslavsky julio@ee.newcastle.edu.au School of Electrical Engineering and Computer Science Lecture 20: MIMO State

More information

IRREDUCIBLE OPERATOR SEMIGROUPS SUCH THAT AB AND BA ARE PROPORTIONAL. 1. Introduction

IRREDUCIBLE OPERATOR SEMIGROUPS SUCH THAT AB AND BA ARE PROPORTIONAL. 1. Introduction IRREDUCIBLE OPERATOR SEMIGROUPS SUCH THAT AB AND BA ARE PROPORTIONAL R. DRNOVŠEK, T. KOŠIR Dedicated to Prof. Heydar Radjavi on the occasion of his seventieth birthday. Abstract. Let S be an irreducible

More information

Mathematics Course 111: Algebra I Part IV: Vector Spaces

Mathematics Course 111: Algebra I Part IV: Vector Spaces Mathematics Course 111: Algebra I Part IV: Vector Spaces D. R. Wilkins Academic Year 1996-7 9 Vector Spaces A vector space over some field K is an algebraic structure consisting of a set V on which are

More information

STEEL PIPE NIPPLE BLACK AND GALVANIZED

STEEL PIPE NIPPLE BLACK AND GALVANIZED Price Sheet CWN-616 Effective June 06, 2016 Supersedes CWN-414 A Member of The Phoenix Forge Group CapProducts LTD. Phone: 519-482-5000 Fax: 519-482-7728 Toll Free: 800-265-5586 www.capproducts.com www.capitolcamco.com

More information

SCO TT G LEA SO N D EM O Z G EB R E-

SCO TT G LEA SO N D EM O Z G EB R E- SCO TT G LEA SO N D EM O Z G EB R E- EG Z IA B H ER e d it o r s N ) LICA TIO N S A N D M ETH O D S t DVD N CLUDED C o n t e n Ls Pr e fa c e x v G l o b a l N a v i g a t i o n Sa t e llit e S y s t e

More information

BOUNDED, ASYMPTOTICALLY STABLE, AND L 1 SOLUTIONS OF CAPUTO FRACTIONAL DIFFERENTIAL EQUATIONS. Muhammad N. Islam

BOUNDED, ASYMPTOTICALLY STABLE, AND L 1 SOLUTIONS OF CAPUTO FRACTIONAL DIFFERENTIAL EQUATIONS. Muhammad N. Islam Opuscula Math. 35, no. 2 (215), 181 19 http://dx.doi.org/1.7494/opmath.215.35.2.181 Opuscula Mathematica BOUNDED, ASYMPTOTICALLY STABLE, AND L 1 SOLUTIONS OF CAPUTO FRACTIONAL DIFFERENTIAL EQUATIONS Muhammad

More information

Form: Parental Consent for Blood Donation

Form: Parental Consent for Blood Donation A R C Wt, C 20006 Ptl Ct f B i Ifi T f t y t ll f i y tl t q y t l A R C ly. Pl ll 1-800-RE-CROSS (1-800-733-2767) v. if y v q r t t i I iv t f yr,, t, y v t t: 1. Y y t t l i ly, 2. Y y t t t l i ( k

More information

1 Quiz on Linear Equations with Answers. This is Theorem II.3.1. There are two statements of this theorem in the text.

1 Quiz on Linear Equations with Answers. This is Theorem II.3.1. There are two statements of this theorem in the text. 1 Quiz on Linear Equations with Answers (a) State the defining equations for linear transformations. (i) L(u + v) = L(u) + L(v), vectors u and v. (ii) L(Av) = AL(v), vectors v and numbers A. or combine

More information

Summary of week 8 (Lectures 22, 23 and 24)

Summary of week 8 (Lectures 22, 23 and 24) WEEK 8 Summary of week 8 (Lectures 22, 23 and 24) This week we completed our discussion of Chapter 5 of [VST] Recall that if V and W are inner product spaces then a linear map T : V W is called an isometry

More information

Linear Control Systems Lecture # 16 Observers and Output Feedback Control

Linear Control Systems Lecture # 16 Observers and Output Feedback Control Linear Control Systems Lecture # 16 Observers and Output Feedback Control p. 1/2 p. 2/2 Observers: Consider the system ẋ = Ax + Bu y = Cx + Du where the initial state x(0) is unknown, but we can measure

More information

T ( a i x i ) = a i T (x i ).

T ( a i x i ) = a i T (x i ). Chapter 2 Defn 1. (p. 65) Let V and W be vector spaces (over F ). We call a function T : V W a linear transformation form V to W if, for all x, y V and c F, we have (a) T (x + y) = T (x) + T (y) and (b)

More information

Decision-making with the AHP: Why is the principal eigenvector necessary

Decision-making with the AHP: Why is the principal eigenvector necessary European Journal of Operational Research 145 (2003) 85 91 Decision Aiding Decision-making with the AHP: Why is the principal eigenvector necessary Thomas L. Saaty * University of Pittsburgh, Pittsburgh,

More information

1 if 1 x 0 1 if 0 x 1

1 if 1 x 0 1 if 0 x 1 Chapter 3 Continuity In this chapter we begin by defining the fundamental notion of continuity for real valued functions of a single real variable. When trying to decide whether a given function is or

More information

MATH 304 Linear Algebra Lecture 4: Matrix multiplication. Diagonal matrices. Inverse matrix.

MATH 304 Linear Algebra Lecture 4: Matrix multiplication. Diagonal matrices. Inverse matrix. MATH 304 Linear Algebra Lecture 4: Matrix multiplication. Diagonal matrices. Inverse matrix. Matrices Definition. An m-by-n matrix is a rectangular array of numbers that has m rows and n columns: a 11

More information

Robust output feedback stabilization of the angular velocity of a rigid body

Robust output feedback stabilization of the angular velocity of a rigid body Systems & Control Letters 39 (2000) 203 20 www.elsevier.com/locate/sysconle Robust output feedback stabilization of the angular velocity of a rigid body F. Mazenc, A. Astol Centre for Process Systems Engineering,

More information

a 11 x 1 + a 12 x 2 + + a 1n x n = b 1 a 21 x 1 + a 22 x 2 + + a 2n x n = b 2.

a 11 x 1 + a 12 x 2 + + a 1n x n = b 1 a 21 x 1 + a 22 x 2 + + a 2n x n = b 2. Chapter 1 LINEAR EQUATIONS 1.1 Introduction to linear equations A linear equation in n unknowns x 1, x,, x n is an equation of the form a 1 x 1 + a x + + a n x n = b, where a 1, a,..., a n, b are given

More information

C o a t i a n P u b l i c D e b tm a n a g e m e n t a n d C h a l l e n g e s o f M a k e t D e v e l o p m e n t Z a g e bo 8 t h A p i l 2 0 1 1 h t t pdd w w wp i j fp h D p u b l i c2 d e b td S t

More information

Chapter 8. Matrices II: inverses. 8.1 What is an inverse?

Chapter 8. Matrices II: inverses. 8.1 What is an inverse? Chapter 8 Matrices II: inverses We have learnt how to add subtract and multiply matrices but we have not defined division. The reason is that in general it cannot always be defined. In this chapter, we

More information

OPTIMAL CONTROL OF A COMMERCIAL LOAN REPAYMENT PLAN. E.V. Grigorieva. E.N. Khailov

OPTIMAL CONTROL OF A COMMERCIAL LOAN REPAYMENT PLAN. E.V. Grigorieva. E.N. Khailov DISCRETE AND CONTINUOUS Website: http://aimsciences.org DYNAMICAL SYSTEMS Supplement Volume 2005 pp. 345 354 OPTIMAL CONTROL OF A COMMERCIAL LOAN REPAYMENT PLAN E.V. Grigorieva Department of Mathematics

More information

Chem 115 POGIL Worksheet - Week 4 Moles & Stoichiometry

Chem 115 POGIL Worksheet - Week 4 Moles & Stoichiometry Chem 115 POGIL Worksheet - Week 4 Moles & Stoichiometry Why? Chemists are concerned with mass relationships in chemical reactions, usually run on a macroscopic scale (grams, kilograms, etc.). To deal with

More information

2.3 Convex Constrained Optimization Problems

2.3 Convex Constrained Optimization Problems 42 CHAPTER 2. FUNDAMENTAL CONCEPTS IN CONVEX OPTIMIZATION Theorem 15 Let f : R n R and h : R R. Consider g(x) = h(f(x)) for all x R n. The function g is convex if either of the following two conditions

More information

7 - Linear Transformations

7 - Linear Transformations 7 - Linear Transformations Mathematics has as its objects of study sets with various structures. These sets include sets of numbers (such as the integers, rationals, reals, and complexes) whose structure

More information

Lecture 12 Basic Lyapunov theory

Lecture 12 Basic Lyapunov theory EE363 Winter 2008-09 Lecture 12 Basic Lyapunov theory stability positive definite functions global Lyapunov stability theorems Lasalle s theorem converse Lyapunov theorems finding Lyapunov functions 12

More information

LOOP TRANSFER RECOVERY FOR SAMPLED-DATA SYSTEMS 1

LOOP TRANSFER RECOVERY FOR SAMPLED-DATA SYSTEMS 1 LOOP TRANSFER RECOVERY FOR SAMPLED-DATA SYSTEMS 1 Henrik Niemann Jakob Stoustrup Mike Lind Rank Bahram Shafai Dept. of Automation, Technical University of Denmark, Building 326, DK-2800 Lyngby, Denmark

More information

State-Space Representation of LTI Systems

State-Space Representation of LTI Systems 214 Analysis and Design of Feedback Control Systems State-Space Representation of LTI Systems 1 Introduction Derek Rowell October 2002 The classical control theory and methods (such as root locus) that

More information

Linear Control Systems

Linear Control Systems Chapter 3 Linear Control Systems Topics : 1. Controllability 2. Observability 3. Linear Feedback 4. Realization Theory Copyright c Claudiu C. Remsing, 26. All rights reserved. 7 C.C. Remsing 71 Intuitively,

More information

max cx s.t. Ax c where the matrix A, cost vector c and right hand side b are given and x is a vector of variables. For this example we have x

max cx s.t. Ax c where the matrix A, cost vector c and right hand side b are given and x is a vector of variables. For this example we have x Linear Programming Linear programming refers to problems stated as maximization or minimization of a linear function subject to constraints that are linear equalities and inequalities. Although the study

More information

Course 221: Analysis Academic year , First Semester

Course 221: Analysis Academic year , First Semester Course 221: Analysis Academic year 2007-08, First Semester David R. Wilkins Copyright c David R. Wilkins 1989 2007 Contents 1 Basic Theorems of Real Analysis 1 1.1 The Least Upper Bound Principle................

More information

MATRIX ALGEBRA AND SYSTEMS OF EQUATIONS

MATRIX ALGEBRA AND SYSTEMS OF EQUATIONS MATRIX ALGEBRA AND SYSTEMS OF EQUATIONS Systems of Equations and Matrices Representation of a linear system The general system of m equations in n unknowns can be written a x + a 2 x 2 + + a n x n b a

More information

H ig h L e v e l O v e r v iew. S te p h a n M a rt in. S e n io r S y s te m A rc h i te ct

H ig h L e v e l O v e r v iew. S te p h a n M a rt in. S e n io r S y s te m A rc h i te ct H ig h L e v e l O v e r v iew S te p h a n M a rt in S e n io r S y s te m A rc h i te ct OPEN XCHANGE Architecture Overview A ge nda D es ig n G o als A rc h i te ct u re O ve rv i ew S c a l a b ili

More information

Vilnius University. Faculty of Mathematics and Informatics. Gintautas Bareikis

Vilnius University. Faculty of Mathematics and Informatics. Gintautas Bareikis Vilnius University Faculty of Mathematics and Informatics Gintautas Bareikis CONTENT Chapter 1. SIMPLE AND COMPOUND INTEREST 1.1 Simple interest......................................................................

More information

MATH10212 Linear Algebra. Systems of Linear Equations. Definition. An n-dimensional vector is a row or a column of n numbers (or letters): a 1.

MATH10212 Linear Algebra. Systems of Linear Equations. Definition. An n-dimensional vector is a row or a column of n numbers (or letters): a 1. MATH10212 Linear Algebra Textbook: D. Poole, Linear Algebra: A Modern Introduction. Thompson, 2006. ISBN 0-534-40596-7. Systems of Linear Equations Definition. An n-dimensional vector is a row or a column

More information

Some stability results of parameter identification in a jump diffusion model

Some stability results of parameter identification in a jump diffusion model Some stability results of parameter identification in a jump diffusion model D. Düvelmeyer Technische Universität Chemnitz, Fakultät für Mathematik, 09107 Chemnitz, Germany Abstract In this paper we discuss

More information

GENERALIZED PYTHAGOREAN TRIPLES AND PYTHAGOREAN TRIPLE PRESERVING MATRICES. Mohan Tikoo and Haohao Wang

GENERALIZED PYTHAGOREAN TRIPLES AND PYTHAGOREAN TRIPLE PRESERVING MATRICES. Mohan Tikoo and Haohao Wang VOLUME 1, NUMBER 1, 009 3 GENERALIZED PYTHAGOREAN TRIPLES AND PYTHAGOREAN TRIPLE PRESERVING MATRICES Mohan Tikoo and Haohao Wang Abstract. Traditionally, Pythagorean triples (PT) consist of three positive

More information

ISOMETRIES OF R n KEITH CONRAD

ISOMETRIES OF R n KEITH CONRAD ISOMETRIES OF R n KEITH CONRAD 1. Introduction An isometry of R n is a function h: R n R n that preserves the distance between vectors: h(v) h(w) = v w for all v and w in R n, where (x 1,..., x n ) = x

More information

LINEAR ALGEBRA W W L CHEN

LINEAR ALGEBRA W W L CHEN LINEAR ALGEBRA W W L CHEN c W W L Chen, 1997, 2008 This chapter is available free to all individuals, on understanding that it is not to be used for financial gain, and may be downloaded and/or photocopied,

More information

The Ideal Class Group

The Ideal Class Group Chapter 5 The Ideal Class Group We will use Minkowski theory, which belongs to the general area of geometry of numbers, to gain insight into the ideal class group of a number field. We have already mentioned

More information

Diagonals of rational functions

Diagonals of rational functions Diagonals of rational functions Main Conference of Chaire J Morlet Artin approximation and infinite dimensional geometry 27 mars 2015 Pierre Lairez TU Berlin Diagonals: definitions and properties Binomial

More information

6. Metric spaces. In this section we review the basic facts about metric spaces. d : X X [0, )

6. Metric spaces. In this section we review the basic facts about metric spaces. d : X X [0, ) 6. Metric spaces In this section we review the basic facts about metric spaces. Definitions. A metric on a non-empty set X is a map with the following properties: d : X X [0, ) (i) If x, y X are points

More information

Inner products on R n, and more

Inner products on R n, and more Inner products on R n, and more Peyam Ryan Tabrizian Friday, April 12th, 2013 1 Introduction You might be wondering: Are there inner products on R n that are not the usual dot product x y = x 1 y 1 + +

More information

OPTIMAL SELECTION BASED ON RELATIVE RANK* (the "Secretary Problem")

OPTIMAL SELECTION BASED ON RELATIVE RANK* (the Secretary Problem) OPTIMAL SELECTION BASED ON RELATIVE RANK* (the "Secretary Problem") BY Y. S. CHOW, S. MORIGUTI, H. ROBBINS AND S. M. SAMUELS ABSTRACT n rankable persons appear sequentially in random order. At the ith

More information

General Theory of Differential Equations Sections 2.8, 3.1-3.2, 4.1

General Theory of Differential Equations Sections 2.8, 3.1-3.2, 4.1 A B I L E N E C H R I S T I A N U N I V E R S I T Y Department of Mathematics General Theory of Differential Equations Sections 2.8, 3.1-3.2, 4.1 Dr. John Ehrke Department of Mathematics Fall 2012 Questions

More information

Transient Voltage Suppressor SMBJ5.0 - SMBJ440CA

Transient Voltage Suppressor SMBJ5.0 - SMBJ440CA Features: Glass passivated junction Low incremental surge resistance, excellent clamping capability 600W peak pulse power capability with a 10/1,000μs waveform, repetition rate (duty cycle): 0.01% Very

More information

Section 8.8. 1. The given line has equations. x = 3 + t(13 3) = 3 + 10t, y = 2 + t(3 + 2) = 2 + 5t, z = 7 + t( 8 7) = 7 15t.

Section 8.8. 1. The given line has equations. x = 3 + t(13 3) = 3 + 10t, y = 2 + t(3 + 2) = 2 + 5t, z = 7 + t( 8 7) = 7 15t. . The given line has equations Section 8.8 x + t( ) + 0t, y + t( + ) + t, z 7 + t( 8 7) 7 t. The line meets the plane y 0 in the point (x, 0, z), where 0 + t, or t /. The corresponding values for x and

More information

MATRIX ALGEBRA AND SYSTEMS OF EQUATIONS. + + x 2. x n. a 11 a 12 a 1n b 1 a 21 a 22 a 2n b 2 a 31 a 32 a 3n b 3. a m1 a m2 a mn b m

MATRIX ALGEBRA AND SYSTEMS OF EQUATIONS. + + x 2. x n. a 11 a 12 a 1n b 1 a 21 a 22 a 2n b 2 a 31 a 32 a 3n b 3. a m1 a m2 a mn b m MATRIX ALGEBRA AND SYSTEMS OF EQUATIONS 1. SYSTEMS OF EQUATIONS AND MATRICES 1.1. Representation of a linear system. The general system of m equations in n unknowns can be written a 11 x 1 + a 12 x 2 +

More information

MATH 304 Linear Algebra Lecture 8: Inverse matrix (continued). Elementary matrices. Transpose of a matrix.

MATH 304 Linear Algebra Lecture 8: Inverse matrix (continued). Elementary matrices. Transpose of a matrix. MATH 304 Linear Algebra Lecture 8: Inverse matrix (continued). Elementary matrices. Transpose of a matrix. Inverse matrix Definition. Let A be an n n matrix. The inverse of A is an n n matrix, denoted

More information

Design Methods for Control Systems

Design Methods for Control Systems Design Methods for Control Systems Okko H. Bosgra Delft University of Technology Delft, The Netherlands Huibert Kwakernaak Emeritus Professor University of Twente Enschede, The Netherlands Gjerrit Meinsma

More information

Sign pattern matrices that admit M, N, P or inverse M matrices

Sign pattern matrices that admit M, N, P or inverse M matrices Sign pattern matrices that admit M, N, P or inverse M matrices C Mendes Araújo, Juan R Torregrosa CMAT - Centro de Matemática / Dpto de Matemática Aplicada Universidade do Minho / Universidad Politécnica

More information

Adaptive Online Gradient Descent

Adaptive Online Gradient Descent Adaptive Online Gradient Descent Peter L Bartlett Division of Computer Science Department of Statistics UC Berkeley Berkeley, CA 94709 bartlett@csberkeleyedu Elad Hazan IBM Almaden Research Center 650

More information

EE 580 Linear Control Systems VI. State Transition Matrix

EE 580 Linear Control Systems VI. State Transition Matrix EE 580 Linear Control Systems VI. State Transition Matrix Department of Electrical Engineering Pennsylvania State University Fall 2010 6.1 Introduction Typical signal spaces are (infinite-dimensional vector

More information

Assignment 3. Solutions. Problems. February 22.

Assignment 3. Solutions. Problems. February 22. Assignment. Solutions. Problems. February.. Find a vector of magnitude in the direction opposite to the direction of v = i j k. The vector we are looking for is v v. We have Therefore, v = 4 + 4 + 4 =.

More information

Classification of Cartan matrices

Classification of Cartan matrices Chapter 7 Classification of Cartan matrices In this chapter we describe a classification of generalised Cartan matrices This classification can be compared as the rough classification of varieties in terms

More information

EIGENVALUES AND EIGENVECTORS

EIGENVALUES AND EIGENVECTORS Chapter 6 EIGENVALUES AND EIGENVECTORS 61 Motivation We motivate the chapter on eigenvalues b discussing the equation ax + hx + b = c, where not all of a, h, b are zero The expression ax + hx + b is called

More information

ANALYTICAL MATHEMATICS FOR APPLICATIONS 2016 LECTURE NOTES Series

ANALYTICAL MATHEMATICS FOR APPLICATIONS 2016 LECTURE NOTES Series ANALYTICAL MATHEMATICS FOR APPLICATIONS 206 LECTURE NOTES 8 ISSUED 24 APRIL 206 A series is a formal sum. Series a + a 2 + a 3 + + + where { } is a sequence of real numbers. Here formal means that we don

More information

Chem 115 POGIL Worksheet - Week 4 Moles & Stoichiometry Answers

Chem 115 POGIL Worksheet - Week 4 Moles & Stoichiometry Answers Key Questions & Exercises Chem 115 POGIL Worksheet - Week 4 Moles & Stoichiometry Answers 1. The atomic weight of carbon is 12.0107 u, so a mole of carbon has a mass of 12.0107 g. Why doesn t a mole of

More information

An example of a computable

An example of a computable An example of a computable absolutely normal number Verónica Becher Santiago Figueira Abstract The first example of an absolutely normal number was given by Sierpinski in 96, twenty years before the concept

More information

U.C. Berkeley CS276: Cryptography Handout 0.1 Luca Trevisan January, 2009. Notes on Algebra

U.C. Berkeley CS276: Cryptography Handout 0.1 Luca Trevisan January, 2009. Notes on Algebra U.C. Berkeley CS276: Cryptography Handout 0.1 Luca Trevisan January, 2009 Notes on Algebra These notes contain as little theory as possible, and most results are stated without proof. Any introductory

More information

PERIODIC TABLES: I. Directory, II. Traditional, III. Vertical, IV. Toxicity (LD 50 values), V. Native (elemental form)

PERIODIC TABLES: I. Directory, II. Traditional, III. Vertical, IV. Toxicity (LD 50 values), V. Native (elemental form) PERIODIC TABLES: I. Directory, II. Traditional, III. Vertical, IV. Toxicity (LD 50 values), V. Native (elemental form) Site developed by Steve Murov, Professor Emeritus of Chemistry, Modesto Junior College,

More information

Chapter 2: Binomial Methods and the Black-Scholes Formula

Chapter 2: Binomial Methods and the Black-Scholes Formula Chapter 2: Binomial Methods and the Black-Scholes Formula 2.1 Binomial Trees We consider a financial market consisting of a bond B t = B(t), a stock S t = S(t), and a call-option C t = C(t), where the

More information

WHICH LINEAR-FRACTIONAL TRANSFORMATIONS INDUCE ROTATIONS OF THE SPHERE?

WHICH LINEAR-FRACTIONAL TRANSFORMATIONS INDUCE ROTATIONS OF THE SPHERE? WHICH LINEAR-FRACTIONAL TRANSFORMATIONS INDUCE ROTATIONS OF THE SPHERE? JOEL H. SHAPIRO Abstract. These notes supplement the discussion of linear fractional mappings presented in a beginning graduate course

More information

CONTROL SYSTEMS, ROBOTICS, AND AUTOMATION Vol. III Design of State Space Controllers (Pole Placement) for SISO Systems - Lohmann, Boris

CONTROL SYSTEMS, ROBOTICS, AND AUTOMATION Vol. III Design of State Space Controllers (Pole Placement) for SISO Systems - Lohmann, Boris DESIGN OF STATE SPACE CONTROLLERS (POLE PLACEMENT) FOR SISO SYSTEMS Lohmann, Boris Institut für Automatisierungstechnik, Universität Bremen, Germany Keywords: State space controller, state feedback, output

More information

Module 3F2: Systems and Control EXAMPLES PAPER 1 - STATE-SPACE MODELS

Module 3F2: Systems and Control EXAMPLES PAPER 1 - STATE-SPACE MODELS Cambridge University Engineering Dept. Third year Module 3F2: Systems and Control EXAMPLES PAPER - STATE-SPACE MODELS. A feedback arrangement for control of the angular position of an inertial load is

More information

Row Ideals and Fibers of Morphisms

Row Ideals and Fibers of Morphisms Michigan Math. J. 57 (2008) Row Ideals and Fibers of Morphisms David Eisenbud & Bernd Ulrich Affectionately dedicated to Mel Hochster, who has been an inspiration to us for many years, on the occasion

More information

BANACH AND HILBERT SPACE REVIEW

BANACH AND HILBERT SPACE REVIEW BANACH AND HILBET SPACE EVIEW CHISTOPHE HEIL These notes will briefly review some basic concepts related to the theory of Banach and Hilbert spaces. We are not trying to give a complete development, but

More information

3 Signals and Systems: Part II

3 Signals and Systems: Part II 3 Signals and Systems: Part II Recommended Problems P3.1 Sketch each of the following signals. (a) x[n] = b[n] + 3[n - 3] (b) x[n] = u[n] - u[n - 5] (c) x[n] = 6[n] + 1n + (i)2 [n - 2] + (i)ag[n - 3] (d)

More information

7 Gaussian Elimination and LU Factorization

7 Gaussian Elimination and LU Factorization 7 Gaussian Elimination and LU Factorization In this final section on matrix factorization methods for solving Ax = b we want to take a closer look at Gaussian elimination (probably the best known method

More information

Solutions to old Exam 1 problems

Solutions to old Exam 1 problems Solutions to old Exam 1 problems Hi students! I am putting this old version of my review for the first midterm review, place and time to be announced. Check for updates on the web site as to which sections

More information