State Space Analysis of Control Systents - BE It (Mesh), MVSR Engg Col Sate variable Method of Analysis: Topics covered: What is a state model Comparison with TF model Methods of obtaining state-space models of dynamic systems Transfer function from SS model Solution of state equation a. State Transition Matrix b. Solution of Homogeneous (free) system equation €. Solution of non-homogeneous (forced) system equations d. Properties of State Transition Matrix 6. Controllability and Observability of a Control system oRene 1. State Model ( or Sate variable or State space modell) : State: The state of a dynamic system is the minimum set of variables called state variables, which if known at time to, along with the input, the system behaviour at t >to can be determined. A dynamic system represented by an nth order differential equations can be split into n first order equations, which can represent a state model. The model consists of the chosen state variables x1, x2, ...n, the input u, and an output y (for a single-input single output system) The state model representation of a system is X=AX+Bu y-CX+Du, where X is (x1, x9, ...xn) - a (nx1) matrix of state variables called State matrix, Ais a (nxn) matrix called Coefficient matrix, Bis a (nx1) matrix called Driving matrix, Cisa(1xn) matrix called Output matrix , and D is (nx1) matrix called Transmission matrix. U is input and y is the output. ‘The state variables are the minimum set of variables. The choice of the variables in the first order equations may be different, which leads to a different state model of the same system. For example: consider a transfer function Cls)/RIs) = k/(s? + as + b), +at+be=kr ... (1) ie., Let x: = cand x2 = i= Eqn. 1 becomes, %,= - ax2-b x1 + kr Above equations are written asState Space Analysis of Control Systems - BEN! (Mech), MVSR Engg College, or (i IEE ke -b -a} |x} [Lk y [1 0} [xr ("| K=AX+Bu y=CX +Du ~ DrPA Sastry is the state model of the given second order system. 2. Comparison of Classical (TF) method and State variable method Transfer function State variable (State Space) The classical Transfer function is a frequency domain approach Time-domain approach 2 | Transfer function is unique State model is not unique 3 | Only input, output and error Chosen state variables need signals are important and are not represent physical or physical variables and measurable _| measurable variables 4 | Design methods are trial and error _| Not trial and error. Systems methods and are approximate can be designed optimally. 5 | Can handle only linear and time _| Also handles non-linear and invariant systems time-varying systems 6 | Generally limited to single input-_| Can handle SISO and MIMO single output (SISO) systems. Multi | systems with equal ease. input —-Multi output systems (MIMO) are difficult 7 [Initials conditions (ICs) are Tes can be considered. neglected. @ | Internal variables cannot be fed Internal variables can be fed back back 9 | Requires Laplace transforms for Treats both similarly. continuous signals and Z- transforms for discrete signals. 10 | TF models can be handled by ‘Computational ntensive and manual methods often needs use of computers.State Space Analysis of Control Systems - BE Iit (Mech), MVSR Eng: rata modde ean be obtained by Aiforout Loclntquos: Enanpe YO) is , a UG) ~ D3+7D4\4D+8 ve. o WS) 2 8 Ula parte rlese = we We MOY Theveiore nos tevwt ie (a numenaley Hemee WE berite ta” governing differeuhal sqpakow ac: » ye Tor mer sy -68u — tu = . 26) Leltwg z ape rny eaten 2 a . a, =3 2a, an <%2 _— 3) euler er eae ) M ao thew ™ mde m Eqns. (2), @) «Ge were es s|fel* Ele —© x -8 -y “YL 3 and ys % oo Pepper tone 8 le —— ) $0 * Eqns: (5) 4 @) repreaent dea state-weorralle medal of We givew syshoms 1 Reform Pe Sa NiRe Bid go Som where ke[f]. x= [iB]: Aq ae he lil ex[1 06] ey 7 C8 ly 7 & D0 -3-State Space Analysis Melkod 2: this ts le } s 1h odor terms and nemerato is a corstauk C@ wilouk storm) ao in ys) 2 8 Systems - BE II! (Mech), MVSR Eng DrPA Sastry applies hou UL denouinaky of TF ie Ues) ~ a1) cs+2ysr3> a: ee s+! Gr2) Gr4) We may uncte = oa _ 4. utsy+ 4/2 — us) ys) = oF US) “eS a) os) = %,CS) + Has) + C8) —a) in ohich X%CA= Sia urs) or RRB) + XiC= As acs) in time-domain , a) ee eee 5 a, +Zu — © an are) Ts ~£. uc, euulior ) Gusset? _ Oo X= 72 %27 i ale uts) and AGS? = “sag a) Oe te a -4%gr eu : a be onion Eqns: (2) (2) aud 2qn-(1) 8 x, al ONS % Pfs a B -2: == +) -4 |U i. o oe -¥ Le 4s OA Y [! ( Ife which veprosaut Ur so model fue given syshew . Ny} (P) \l 4tote space Analysis of Control Systems - BE Ill (Mech), MVSR. E499 ge, - DY.PA Sastry Floweror, ne of Mae devo ral fui TR to given on ey narreicr! ae in tat Qermple Mefied 4, iss 5 Gy geha difarend state modal, we) 8 : UG) ~ (SHC 8 iene > 0465+ nbsey =8u This tu homdled? as in Method 4, where Yat, Yortae%Urb- Hw aud U =-6yg-Ny-649 +84 = % leadug to 2X, o4o cay °o =) - (4 2]8)El | we -p 6] L%s} LF @) Pe an, stake spree wadole (P) @ @ ove UE models Ms), 2 4 ere Um ~Grhcs+2)Cra) Noles Ke can obtain Wa bramshon funckaw of a given, = stole made) ae ws) _@ csr-ANB RD — CR) aes) Ry sabstithing KE atperefiiake ruatrices tr form (P) or ©) info agn(R), We Ama TP obtained “ae Xe ka Sauer tram chen can be pros pi Gen a Be.state space Analysis of Contyol Systems - BE Il (Mesh), MVSR Engg College, - PHPA Sastry Malket 3. (ou nunercbe Pao £6y B)bame of oder D ul) Lebeig ‘miktal conditions (0) =0 (te 1D =% x, (20° ok.) Wade @ SIA OU + Dus) = yes) Cc CSI-A) 12+D is Ca branspy freon: Ul) Exantles Find TF. for lhe Broke wedel (B= Ta tA) Oe vat AG WO = ¢ CarayerD sehr sb 5]- Ts so etsy tle => Bal oh da Te so 1 Steeda Saws lee q i) - ae U8: 7 Fea 2s+d fence TF Zz = pass efor ie guew stle- space model, -9-State Space Analysis of Control Systents - BE Ml (Mesh), MVSR Engg College, - PrPA Sastry 5. Solution of state Equation Scion means fending Xt) ig CH, %CE).cke . givem a (te) ie(er K(0) hou te=s) fr Q given dale modet - 5a We know that for a diffoveuhal equation Cheworeaems) = ax, dha delation is xs Kot where K = 2x60), dothst x@) = Lacey. Similedy for the mabeix epabow, & = AX, Ne Sduhon ts ~aCL) atic Te i k, ¢ X(t) ig Ke wierod stale of x teks , Xt) = Ate oe KB XU) alee elene Seta xt) = ohttte) x) Heo, xa = xo ee NO) Su on me te we con fed Ue Alsho x(t) ak time t’, Knouong a ott, state at «seme ihe) Hme; aur msabbiplyeicy Or ve" foues ht ic called tote Tronsetion Malicix a0 ‘te transfers x10) to xt. Just as Pk t+ ata at a = r at n sett. ott a T+ ate Ate Loe ‘ ee Od [ae aay, of epaeala rel na ott: + Theve are vovieus mélhods af eyaluahig 2 one f Tam is Laplace “Trans ren aproaeh: senstate space Analysis of Cont Bib Soluctan pr ttomogeneous es We have Me LT form of . x= AX as &XA)— XC0) = A x8) From ufhich (I-A) xcs) = *t0) x4) = (BE-AY' x) = ——— (0) ow xtye LO [st- eA (0) As Re = eft x00) ' 4 We have 1c Stoke Feuciton Mie = £1 (55-4) Nso -egn(i) obove may be written ao el coe pea) XC) whow BCS)= as OnS x(t) = p(t) xCo) rere He [GH ect) coh ig he STM. The Laplace fore of of pt) & Hts) v8 in called Policy Mebux Houce cue wattiod of ooluakig STM eAt_ {Csi eo BI O= Example + Fad x0) given Xt) = (& ' ty] xen-b25) d Peer [tot alee (ois: t meeps aeG)= 6I- ee iL “le -[& Wf [s a) 1 jo To o des | fet aaah) cm 0° TEE ol ay KE) = pk) xO) = i [dark av agen 3\\ s | ystems ~ BE It (Mech), MVSR Engg College, - Dr.PA Sastry aide“state Space Analysis of Control Systents - BE Itt (Mech), MVSR Engg College, - DY-PA Sastry she aie sae S43 “s+ _ [2 e gee _ eal v4 a zdton, eater weal © Sd sauhon of Stake Equohe uation valhe forcing furchon Cu) Non - homogencows équshon) Kk) = Ax) + Bua) Pat een hide, on mualilying lee, enn ere 2u re a (é*%e) = eeu Inregaig | ae ote t, AT, 2 ub) — x(0) = Bu(c) at Meatteplysag, tRvranghest bay ett © ey ath wee) w oa of x0) + [eh Bucode oo x ~ Xe) = gee) xto) + PH) Jaer) eucode. The forcing function effect ate add fra faced et. Bt) x(0) hich ie ie uhm use Exauble + pad xit), for ME systen eUliei+Tde. a o -2 Tuital Condihion 0) =f aud uo a ewiek step mbState Space Analysis of Control Systems - BE Ill (Mech), MVSR Engg College, ~ Dr-PA Sastry x 2 We howe XA) = At x(0) 4 | ht ae neous } RD Gra = Ts -1 - | oan oo sta) = sey |e - 5 a0 ok 2 (GE-AY = fe ase) x: ° xem ih seiees oS) s = | lace i o <2 r oo [SIE “Td Ths co hE ABhen Sham inp =O & % roa “lac (eee) bee ° (2 B ute) dT = ie ere E oO 3 (eo one | Peete 1 = o : 2 ya\ cs, . Et Ee l x gerne Bi i ae) At t ACT *U8) 2 2 lo) + 2 guceddt wos ~ pele US US | vt Ose aa if ¥ aa aide 4 (2% > ov i co Wa Bduhow ce tens nedince 24) cows We pele Nowtables 240) aad 2GCk) , Whom We ier 2 aon step - oe ee —13-State Space Analysis of Control. Systems - BE Ill (Mech); MVSR Engg College, - DV.PA Sastry Anctian method of Evaluation of At (Aageralicahin of.) Su ha stale medel , ik Aw liagersl , 20 Az Bs = ; then Ue canbe vend teu dat Z _at ee Gray [é se] Hower, 4 A a rel os » bub has Aafueh eigen valaoe, (404 MouaRs), hau we com obtain ETM aa oft M eta at Bs bhere g=[E Re oud ‘ Mis tx model wabsix of aigenvertos ¢ Beeaiiplen lak = [oul * 6 Sam: 2 bRe 3 = Jar-al 20> c alt => Chex: wil RK, 2-4. oS UB O% Eigen vector aemard= (NF a d, e-) = AAICAE-AD = Bil? +E dye-e MIATA = Ty S| > = [Us| Heuce M=E% % = [3] pis eel (ka eqewalves ‘ ost . Le a at eo ~ wo Shee a} at -t =p se NobState Space Analysis of Control Systems ~BE Ill (Mech), MVSR Engg College, - DLPA Sastry Propertios af state Tramsthan Mahix , et) = a Properties of STM Can be veagily obtained Prepenliy Expansion. 1. oye t = be Dr.PA Sastry ao Conbvalae [eal i; Es | is oe fence Me cytlem is pot obseweble + cheek dhat Ike system () *- [Rs e JL Je cxitale 2 ke [eis]. a -[}}eet 212] is cedvotlable but not observable. EAS e). : oh s AE 2° [| ,e=[7 3 J is not Cotrollable 2 nak obsewable . ® madi