L 1 Adaptive Controller for a Missile Longitudinal Autopilot Design

Size: px
Start display at page:

Download "L 1 Adaptive Controller for a Missile Longitudinal Autopilot Design"

Transcription

1 AIAA Guidance, Navigation and Control Conference and Exhibit 8-2 August 28, Honolulu, Hawaii AIAA L Adaptive Controller for a Missile Longitudinal Autopilot Design Jiang Wang Virginia Tech, Blacksburg, VA 246, USA Chengyu Cao University of Connecticut, Storrs, CT, 6269, USA Naira Hovakimyan University of Illinois at Urbana-Champaign, Urbana, IL 68, USA Richard E. Hindman and D. Brett Ridgely Raytheon Missile Systems, Tucson, AZ 85734, USA This paper considers application of L adaptive output feedback controller to a missile longitudinal autopilot design. The proposed adaptive controller has satisfactory performance in the presence of parametric uncertainties and time-varying disturbances. Simulations demonstrate the benefits of the control method and compare the results to Linear Quadratic Regulator (LQR) and Linear Quadratic Gaussian (LQG) with Loop Transfer Recovery (LTR) design. I. Introduction This paper presents the application of an L adaptive output feedback controller to longitudinal autopilot design for a missile in the presence of uncertainties in system dynamics. The uncertainties include parametric variations in the transfer function and time-varying disturbances. The parametric variations of the system s transfer function are caused by changes in aerodynamic coefficients. The missile model, taken from Mracek and Ridgely [], is an unstable non-minimum phase system. The nominal optimal controller in Mracek and Ridgely [] uses both system outputs (pitch rate and normal acceleration) to compute the feedback control signal. In this paper we only use the acceleration. The proposed adaptive output feedback controller, taken from Cao and Hovakimyan [2, 3], allows tracking of reference systems that do not verify the SPR condition for their input-output transfer functions. Similar to the earlier Graduate Research Assistant, Student Member AIAA, Department of Aerospace and Ocean Engineering, 25 Randolph Hall; jwang5@vt.edu (Corresponding Author). Research Assistant Professor, Member AIAA,Department of Mechanical Engineering; ccao@engr.uconn.edu Professor, Associate Fellow AIAA, Department of Mechanical Science and Engineering; nhovakim@illinois.edu Principal Systems Engineer, Member AIAA, Rick.Hindman@raytheon.com Senior Department Manager, Senior Member AIAA, dbridgely@raytheon.com of 2 Copyright 28 by Jiang Wang. Published by the, Inc., with permission.

2 results of Cao and Hovakimyan [4], the L -norms of both input/output error signals between the closed-loop adaptive system and the reference system can be rendered arbitrarily small by reducing the step-size of integration. The key difference between [2,3] and [4] is the new piece-wise continuous adaptive law which enables tracking of the reference systems without imposing the SPR requirement on their input-output transfer function. The adaptive control is defined as the output of a low-pass filter, resulting in a continuous signal despite the discontinuity of the adaptive law. The L adaptive output feedback controller aims at achieving a guaranteed transient performance for the system s output, and regulating the frequency spectrum and the performance bound for the system s input signal as well, by rendering them arbitrarily close to the corresponding output/input signals of a bounded reference system. We consider the longitudinal dynamics of a missile in the presence of uncertainties in aerodynamics and timevarying disturbances. For comparison purposes, we first consider the nominal optimal controller from Mracek and Ridgely [], which is a classic three-loop topology autopilot designed by LQR methods. Then assuming only measured acceleration is available, we design a LQG controller with Loop Transfer Recovery (LTR) to recover the robustness of the LQR controller. This serves as the nominal output feedback controller in the absence of uncertainties. We augment the baseline LQG controller by L adaptive output feedback loop to compensate for the modeling uncertainties. The paper is organized as follows. Section II presents the problem formulation. Section III shows the nominal controller design. Section IV gives an overview of the key results from Cao and Hovakimyan [2, 3], and Section V discusses some design solutions to achieve the desired performance specifications. Section VI presents the simulations, while Section VII concludes the paper. II. Problem Formulation The missile s longitudinal dynamics can be described using the short period approximation of the longitudinal equations of motion []: ẋ p (t) = A p x p (t) + B p [δ p (t) + v(t, y(t))] () y p (t) = C p x p (t) + D p [δ p (t) + v(t, y(t))] (2) y(t) = A zm (t), (3) where δ p (t) is the elevator input, v(t, y(t)) is time-varying disturbance and depends on y(t), x p (t) and y p (t) are the state and the output vectors respectively, given by x p (t) = α(t) q(t), y p (t) = A z m (t) q m (t), (4) 2 of 2

3 while α(t) is angle of attack, q(t) is pitch rate, A zm (t) is normal acceleration and q m (t) is measured pitch rate. In () and (2) the system matrices are C p = A p = [ ] QSCzα V m m A X QSC zα mg QSdC mα I Y Y QSdC mα x gi Y Y, B p =, D p = QSC zδp mv m QSdC mδp QSC zδp mg I Y Y, QSdC mδp x gi Y Y. The numerical values of the simulation example in this paper are listed in Table, and are taken from Mracek and Ridgely []. In this paper we consider uncertainties in the aerodynamic coefficients C zα, C mα, C zδp and C mδp. Table. Numerical Values of Model Variable Value Units Description V m 335 ft/sec Total Missile Velocity m. slug Total Missile Mass I Y Y 37.8 slug ft 2 Pitch Moment of Inertia x.2 ft Distance from CG to IMU Positive Forward A X -6 ft/sec 2 Axial Acceleration Positive Forward C zα Pitch Force Coefficient due to Angle of Attack C mα Pitch Moment Coefficient due to Angle of Attack C zδp Pitch Force Coefficient due to fin Deflection C mδp Pitch Moment Coefficient due to fin Deflection Q 3332 lb/ft 2 Dynamic Pressure S.5454 ft 2 Reference Area d.8333 ft Reference Length g ft/sec 2 Gravity Constant These uncertainties do not satisfy the restrictive matching condition required by conventional adaptive feedback control schemes. Therefore we apply the adaptive output feedback method from Cao and Hovakimyan [2, 3]. We assume that the maximum possible variations of aerodynamic coefficients with respect to nominal values are known conservatively: C zα.5 C zα, C zδp.5 C zδp, C mα.5 C mα, C mδp.5 C mδp, while the exact values of these coefficients are unknown. In simulations, the actual values of aerodynamic coefficients 3 of 2

4 are selected to be C z α =.4 C zα, C m α =.4 C mα, C z δp =.7 C zδp, C m δp =.7 C mδp. (5) The control objective is to design adaptive output feedback controller to achieve satisfactory tracking performance for the output A zm (t), in the presence of parametric uncertainties and time-varying disturbances, using only A zm (t). III. Nominal Controller Design III.A. LQR solution We first develop the classical three-loop topology for the nominal controller design [], assuming that the required output signals in the three-loop topology are available. The system is augmented by considering z(t) = δ p (t) + v(t, y(t)) as an additional state: ẋ (t) = A x (t) + B (u(t) + d(t, y(t))) y (t) = C x (t), (6) where x (t) = α(t) q(t) z(t), u(t) = δ p (t), d(t, y(t)) = v(t, y(t)), y (t) = A zm (t) q m (t) q m (t), (7) and the transformed state space matrices are A = A p B p [], B = [], C = C p D p A p (2, :) B p (2, :) We also assume that the time-varying disturbance d(t, y(t)) satisfies the following assumption.. (8) Assumption There exists a constant L > such that the following inequality holds uniformly in t for all y, y : d(t, y) d(t, y ) L y y A block diagram of the plant and controller structure is shown in Fig.. When the aerodynamic uncertainties in (5) and the disturbance d(t, y(t)) are not present, the optimal Linear Quadratic Regulator (LQR) solution in [] is given by: u(t) = δ p (t) = K opt A zm (t) K ss r q m (t) q m (t), (9) 4 of 2

5 A zc D p K ~ ss d K opt u δ p + v s ( p p A, B ) x p C p y p Figure. Block diagram of system. where K ss is chosen to ensure zero steady-state error for step commands, while r is the steady-state value of the reference command r(t). Based on the nominal numerical values given in Table, the optimal controller gains are: K opt = [ ], K ss =.855. In the current setup, q m (t) is not measurable, and the above computed optimal controller is actually the derivative of δ p (t). Since we have a linear system, we can integrate both sides of (9) to determine δ p (t) (assuming constant gains) as: t (A t z m (τ) K ss r ) dτ (A z m (τ) K ss r ) dτ δ p (t) = K opt t q m(τ)dτ = K t opt q m(τ)dτ. () t q m(τ)dτ q m (t) If the available feedback signal is only A zm, the above optimal controller cannot be implemented. The proposed adaptive control method can have satisfactory tracking performance in the presence of parametric uncertainties and disturbance, using only A zm as feedback signal. For comparison purposes, we design an observer that can recover the original LQR control performance. At the same time we need to recover the inherent robustness of baseline LQR controller as much as possible, because we are looking into the controller s robustness to parametric uncertainties and disturbance. III.B. Output feedback solution: LQG/LTR Based on the LQR solution of the above nominal optimal control, we can design a Linear Quadratic Gaussian (LQG) with Loop Transfer Recovery (LTR) controller using only the output A zm. Since the LQR controller is ready, we only need to design the Kalman filter which can help recover the robustness of LQR controller. We note that due to the non-minimum phase property of the system, the robustness recovery is limited. In Mracek and Ridgely [], the LQR design is based on the transformed system with state transformation x 2 = C x, as shown below ẋ 2 (t) = A 2 x 2 (t) + B 2 u(t) y 2 (t) = x 2 (t) () 5 of 2

6 where x 2 (t) = A zm (t) q m (t) q m (t), A 2 = C A C, B 2 = C B. We design a Kalman filter based on this LQR solution. Plant noise and measurement noise are introduced to produce ẋ 2 (t) = A 2 x 2 (t) + B 2 u(t) + B 2 ω(t) y k (t) = A zm (t) = [ ]x 2 (t) + v(t) (2) where the plant noise ω(t) and the measurement noise v(t) are white noises with the spectral densities S ω and S v respectively, and they are uncorrelated and orthogonal. Furthermore, the plant noise and the initial states of the system (2) are assumed to be uncorrelated and orthogonal; so are the measurement noise and the states. The Kalman filter equation is ˆx 2 (t) = A 2ˆx 2 (t) + B 2 u(t) + G (y k (t) [ ]ˆx 2 (t)) (3) where G is the Kalman gain. The Loop Transfer Recovery design is done by increasing the spectral density S ω of the plant noise ω(t). We choose different values of S ω to design the Kalman filter, and compare the results to the LQR results. The spectral density of the measurement noise is set as S v =.. The Kalman filter gain is obtained by MATLAB command kalman. Next we show the system response of LQG/LTR control. First, the unit step responses of the LQR and LQG/LTR control are shown. In Fig. 2, the LQG controller is designed with S ω = and S v =.. It can be seen that the time responses of these two controllers are identical. However, the robustness of these two controllers are different, due to the small value of spectral density S ω, which means that the loop transfer recovery is not enough. This can be seen in Fig. 3. In this figure, the parametric uncertainties in (5) are present. The LQR controller has certain inherent robustness to parametric uncertainties, hence its performance is acceptable. However, the LQG controller has degraded performance due to lack of robustness. We need to increase the spectral density S ω to recover as much as possible the robustness of the LQR controller. In Fig. 4, the spectral density is increased to, and we can see that the performance is improved in the presence of parametric uncertainties. This robustness recovery is limited by the system s non-minimum phase property. Hence, increasing S ω cannot recover the robustness completely, as shown in Fig. 5. With very large value of S ω, the system s performance is no better than that in the case of smaller S ω. We now introduce a disturbance d(t) =. sin(.5πt) into the system. In Fig. 6, the system output under the LQR controller drifts from the desired steady state position, and the performance of LQG/LTR is also unacceptable. This is expected because of the limitation of the loop transfer recovery applied to a non-minimum phase system. 6 of 2

7 .2.8 Acceleration (g) LQR LQG time (sec) Figure 2. Comparison between LQR and LQG - no uncertainty, no disturbance 2.5 Acceleraton (g).5 LQR LQG time (sec) Figure 3. Comparison between LQR and LQG - with uncertainty, no disturbance 2.5 Acceleration (g).5 LQR LQG S ω = LQG/LTR S ω = time (sec) Figure 4. Comparison between LQR, LQG and LQG/LTR - with uncertainty, no disturbance 7 of 2

8 LQG/LTR S ω = LQG/LTR S ω = LQG/LTR S ω = Acceleration (g) time (sec) Figure 5. Different values of S ω for LQG/LTR designs - with uncertainties, no disturbance Acceleraton (g).5.5 LQR LQG S ω = LQG/LTR S ω = time (sec) Figure 6. Comparison between LQR, LQG and LQG/LTR - with uncertainties and disturbance 8 of 2

9 IV. L Adaptive Output Feedback Control In system (6), if we let y(t) = C (, :)x (t) = c x (t) = A zm (t), the longitudinal dynamics of the missile can be presented in the following form: y(s) = A(s) [u(s) + d(s)], y() =, (4) where u(t) = δ p (t) R is the input, y(t) = A zm (t) R is the system output, A(s) = c (si A ) B is the unknown transfer function of the system, d(s) is the Laplace transform of the time-varying disturbances in (6). Notice that d(t, y) depends on the system output y, and the upper bound of the growth rate of d(t, y) with respect to y is L, as stated in Assumption. Substituting the numerical values from Table into the system in (6), we get the nominal system of A(s) A (s) = A n = 3.5s s A d s s 2. (5) 29.3s To achieve the control objective, we need to design an adaptive output feedback controller u(t) such that in the presence of uncertainties the system output y(t) tracks the reference input r(t) with satisfactory performance. This can be done by selecting a minimum-phase, strictly proper and stable transfer function M(s), and designing an adaptive control law to achieve y(s) M(s)r(s). The selection of M(s) needs to satisfy the sufficient conditions for stability and performance, and we postpone the discussion on selection of M(s) to Section V. The system in (4) can be rewritten as: y(s) = M(s) (u(s) + σ(s)) (6) ( ) σ(s) = (A(s) M(s))u(s) + A(s)d(s) /M(s). (7) Let (A m R N N, b m R N, c m R N ) be the minimal realization of M(s). Hence, (A m, b m, c m ) is controllable and observable, with A m being Hurwitz. Thus, the system in (6) can be rewritten as: ẋ(t) = A m x(t) + b m (u(t) + σ(t)) (8) y(t) = c mx(t), x() = x =. Next we introduce the closed-loop reference system that defines an achievable control objective for the L adaptive controller. Closed-loop reference system: The reference system is given by y ref (s) = M(s)(u ref (s) + σ ref (s)) (9) σ ref (s) = ( (A(s) M(s))u ref (s) + A(s)d ref (s) )/ M(s) u ref (s) = C(s)(r(s) σ ref (s)) where C(s) is a low pass filter with DC gain C() =. 9 of 2

10 According to [2, Lemma ] the selection of C(s) and M(s) must ensure that ( ) H(s) = A(s)M(s)/ C(s)A(s) + ( C(s))M(s) (2) is stable and that the L -gain of the cascaded system is upper bounded as follows: H(s)( C(s)) L L < (2) Then the reference system in (9) is stable. The elements of the L adaptive controller are introduced next. State predictor (passive identifier): Let (A m R n n, b m R n, c m R n ) be the minimal realization of M(s). Hence, (A m, b m, c m ) is controllable and observable with A m being Hurwitz. Then the system in (4) can be rewritten as ẋ(t) = A m x(t) + b m (u(t) + σ(t)) (22) y(t) = c mx(t) The state predictor is given by: ˆx(t) = A mˆx(t) + b m u(t) + ˆσ(t) (23) ŷ(t) = c mˆx(t) where ˆσ(t) R n is the vector of adaptive parameters. Notice that in the state predictor equations ˆσ(t) is not in the span of b m, while in the equation (22) σ(t) is in the span of b m. Further, let ỹ(t) = ŷ(t) y(t). Adaptation law: Let P be the solution of the following algebraic Lyapunov equation: A mp + P A m = Q where Q >. From the properties of P it follows that there always exists a nonsingular P such that P = P P. Given the vector c m( P ), let D be the (n ) n-dimensional nullspace of c m( P ), i.e. D(c m( P ) ) = (24) and let Λ = c m D P R n n (25) The update law for ˆσ(t) is defined via the sampling time T > a : a T defines the sampling rate of the available CPU. ˆσ(iT ) = Φ (T )µ(it ), i =, 2,, (26) of 2

11 where and Φ(T ) = T e ΛA mλ (T τ) Λdτ (27) µ(it ) = e ΛAmΛ T ỹ(it ), i =, 2, (28) Here denotes the basis vector in the space R n with its first element equal to and other elements being zero. Control law: The control law is defined via the output of the low-pass filter: u(s) = C(s)r(s) C(s) M(s) c m(si A m ) ˆσ(s). (29) The complete L adaptive controller consists of the state predictor in (23), the adaptation law in (26), and the control law in (29), subject to the L -gain upper bound in (2). The performance bounds of the L adaptive output feedback controller are given by the following theorem. Theorem ( ) lim ỹ L T ( ) lim y yref L T ( ) u uref L lim T = = = The result in this theorem follows immediately from [2, Theorem ] and [2, Lemma 3]. V. Design Issues of L Adaptive Output Feedback Control V.A. Stability The first step of designing an L adaptive output feedback controller is to guarantee stability of the closed-loop system. From Theorem it can be seen that the output of the closed-loop system tracks that of the closed-loop reference system arbitrarily closely for all t >. Hence the goal of the first step in the design is to find C(s) and M(s) to satisfy the sufficient conditions given in (2) and (2). These two conditions can guarantee the stability of closed-loop reference system. We first discuss the classes of systems that can satisfy (2) via the choice of M(s) and C(s). We demonstrate that stability of H(s) is equivalent to stabilization of A(s) by C(s) M(s)( C(s)). (3) Consider the closed-loop system, comprised of the system A(s) and negative feedback of (3). The closed-loop transfer function is: + A(s) A(s) C(s) M(s)( C(s)). (3) of 2

12 Letting it follows from (2) that where A(s) = A n(s) A d (s), C(s) = C n(s) C d (s), M(s) = M n(s) M d (s), (32) H(s) = C d(s)m n (s)a n (s) H d (s), (33) H d (s) = C n (s)a n (s)m d (s) + M n (s)a d (s)(c d (s) C n (s)). (34) Incorporating (32), one can verify that the denominator of the system in (3) is exactly H d (s). Hence, stability of H(s) is equivalent to the stability of the closed-loop system in (3). The selection of M(s) and C(s) can be restricted due to the properties of the plant A(s). Thus, it is not a trivial task. However, it can be done using linear systems theory. The essential objective in this step is to design, based on the nominal system A (s), a feedback controller that can be decomposed into C(s) and M(s) according to the equation (3), while achieving stability of H(s) in (2) and verifying the condition in (2) based on conservative knowledge of parametric variations in A(s). In the following subsection we describe one method towards the selection of C(s) and M(s). V.A.. Design via pole placement We use a pole placement method (see examples in Ioannou and Sun [5]) to design a dynamic compensator for A (s). The block diagram in Fig. 7 shows the structure of the closed-loop system, where the dynamic controller P (s)/l(s) is determined by the solution of the following equation A n (s)p (s) + A d (s)l(s) = A cl (s). (35) All terms in (35) are polynomials of s. The Hurwitz polynomial A cl (s) defines the desired pole locations of the closedloop system. The coefficients of polynomials P (s) and L(s) may be obtained by solving the algebraic equation β l = S l α l containing the Sylvester matrix S l of A n and A d, while β l is a vector containing coefficients of both P (s) and L(s), and α l is a vector containing coefficients of A cl (s) (defined next). Definition Given two polynomials a(s) = a n s n + a n s n + + a, b(s) = b n s n + b n s n + + b, the 2 of 2

13 Sylvester Matrix S l is defined to be the following 2n 2n matrix: a n b n a n a n b n b n. a n a.... n. b n b... n a S l = a n b b n a a a n b b b n a b a a... b b a b. Definition 2 If L(s) = s n + l n 2 s n l s + l, P (s) = p n s n + p n 2 s n p s + p, then β l = [, l n 2, l n 3,, l, l, p n, p n 2,, p, p ]. If A cl (s) = s 2n + a 2n 2s 2n a s + a, then α l = [, a 2n 2, a 2n 3,, a, a ]. The polynomials A n (s) and A d (s) should be coprime to ensure the existence and uniqueness of solutions of P (s) and L(s), and non-singularity of S l. A A n d ( s) ( s) P( s) L( s) Figure 7. Block diagram of pole placement method. Incorporating (32) we have C(s) M(s) ( C(s)) = C n (s)m d (s) M n (s) (C d (s) C n (s)). (36) 3 of 2

14 Since the low pass filter C(s) has DC gain of, the polynomial C d (s) C n (s) has no constant term, which means that the transfer function (36) has at least one pole at the origin. To obtain suitable dynamic compensators which can be decomposed into M(s) and C(s), we need to design a dynamic compensator for the system A (s). Observing Fig. s 8, it can be seen that the two closed-loop systems have the same characteristic equation. Hence we can find C(s) and M(s) from the transfer function P (s). Upon selection of the structure of C(s) and M(s), we can write explicitly s L(s) the transfer function in (36) and obtain the coefficients of s by equating it to P (s) s L(s). A ( s) s A(s) P( s) L( s) s P( s) L( s) (a) (b) Figure 8. Block Diagrams C(s) Upon obtaining the dynamic compensator, one can apply Nyquist criterion or Root Locus methods to tune the gains of this compensator by changing the poles in A cl (s). If the low-pass filter C(s) and the desired M(s) ( C(s)) system M(s) do not lead to satisfactory performance, we need to re-select the desired locations of the closed-loop poles. Notice that the above method is an intuitive one, and other methods from linear systems theory can be equivalently explored for determining a structure for C(s) and M(s). We show how to select M(s) and C(s) for our missile model. If we choose the desired pole locations as 2 ± 2j, 2 ± 2j, 2, 2 and 2, then A cl (s) = s s E5s E7s 4 +.Es Es E2s + 5.E3. The Sylvester Matrix is determined by the coefficients of A n (s) and sa d (s). The vectors containing the coefficients of A n (s) and sa d (s), respectively, are shown below [ ], [ ]. 4 of 2

15 The Sylvester Matrix S l is S l = We solve the following algebraic equation β l = S l α l to get the vector β l : β l = [ E6 2.4E8 3.E5.2E7.8E8.E9], where α l is the vector of the coefficients of A cl (s). Hence, The first four elements of β l are the coefficients of L(s), and the rest of the elements are the coefficients of P (s). L(s) = s s E6s + 2.4E8, P (s) = 3.E5s 3 +.2E7s 2 +.8E8s +.E9. If we select C(s) to be a second order, relative degree 2 transfer function, and M(s) be third order, relative degree transfer function, we can write explicitly the transfer function in (36) and obtain the coefficients of M(s) and C(s) by equating (36) to P (s). The transfer functions for C(s) and M(s) take the form: s L(s) M(s) = C(s) = s s + 4.5E6 s s s , (37) 3.E5 s s + 3.E5. (38) This selection of M(s) and C(s) generate satisfactory performance according to simulation results shown in Section VI. V.A.2. Stability check We notice that the design of the dynamic compensator is based on our knowledge of the nominal plant A (s). We know the bounds for the variation of the system parameters, but not the exact values of these parameters. The stability 5 of 2

16 of the transfer function (3), or equivalently stability of the condition (2) can be checked by Kharitonov s Theorem. Towards that end, consider the set I(s) of real polynomials of degree n of the form δ(s) = δ + δ s + δ 2 s 2 + δ 3 s δ n s n, where the coefficients lie within the given ranges: δ [X, Y ], δ [X, Y ],, δ n [X n, Y n ]. Let c δ = [δ, δ,, δ n ] and consider the polynomial δ(s) with its coefficient vector c δ. Introduce the hyper-rectangle box of the coefficients B := { c δ : c δ R n+, X i δ i Y i, i =,,, n }. We assume that the degree remains invariant over the family, so that / [X n, Y n ]. Kharitonov s Theorem provides a (conservative) necessary and sufficient condition for Hurwitz stability of the entire family. Theorem 2 (Kharitonov Theorem) [6] Every polynomial in the family I(s) is Hurwitz if and only if the following four extreme polynomials are Hurwitz: K (s) = X + X s + Y 2 s 2 + Y 3 s 3 + X 4 s 4 + X 5 s 5 + Y 6 s 6 +, K 2 (s) = X + Y s + Y 2 s 2 + X 3 s 3 + X 4 s 4 + Y 5 s 5 + Y 6 s 6 +, K 3 (s) = Y + X s + X 2 s 2 + Y 3 s 3 + Y 4 s 4 + X 5 s 5 + X 6 s 6 +, K 4 (s) = Y + Y s + X 2 s 2 + X 3 s 3 + Y 4 s 4 + Y 5 s 5 + X 6 s 6 +. Proof of this theorem is omitted. To show the relationship between the box B and the vertices corresponding to extreme polynomials, a plot is shown for a family of second order polynomials [7], Fig. 9. δ 2 2 K Β K 4 K 3 K δ δ Figure 9. The box B and four Kharitonov vertices If we consider the bounds of parametric variations in A(s) in (34) upon designing C(s) and M(s), we can obtain the variation range [X i, Y i ], i =,,, n of each coefficient of the polynomial H d (s). Substituting X i and Y i into 6 of 2

17 the four extreme polynomials, we only need to check the stability of these four polynomials. If these four polynomials are stable, the designed C(s) and M(s) are acceptable for verification of the condition in (2). Upon the design of transfer functions C(s) and M(s), the condition (2) needs to be checked. This follows from the same procedure of the results in Cao and Hovakimyan [8, 9]. V.B. Performance The second step is to ensure satisfactory performance. Upon determining the structure of M(s) and C(s), which can satisfy the sufficient conditions for stability, we can tune the parameters of M(s) and C(s) within the acceptable parameter space to achieve satisfactory performance. The tuning of parameters for M(s) follows from conventional linear systems theory, which we omit here. The guideline for tuning the low-pass filter C(s) follows the same lines of Cao and Hovakimyan [8, 9]. The trade-off between the time-delay margin and the performance of the L adaptive controller depends solely upon C(s). Increasing the bandwidth of C(s) leads to improved performance at the price of reduced time-delay margin. In [], we consider constrained optimization of the performance and/or the robustness of L adaptive controller by resorting to appropriate Linear Matrix Inequality (LMI) type conditions. If the corresponding LMI has a solution, then arbitrary desired performance bound can be achieved, while retaining a prespecified lowerbound on the time-delay margin. In summary, to gain more freedom in design, it is important for designers to find the largest possible acceptable parameter space in the first step discussed in section V.A. VI. Simulation Example The nominal transfer function of the unstable non-minimum phase missile plant in (5) is repeated below: A(s) = 3.5s s s s 2, (39) 29.3s and the desired system M(s) and the low-pass filter C(s) are taken from (37) and (38): M(s) = C(s) = s s + 4.5E6 s s s , (4) 3.E5 s s + 3.E5. (4) We select T =.. The L output feedback adaptive control approach is applied to this system. Fig. shows the system outputs with L controller, in the absence and in the presence of parametric uncertainties. We can see that the system output tracks the step command satisfactorily. Although this response is different from that of the baseline LQR controller, we demonstrate later that in different unknown scenarios, the L controlled system still has a uniform response close to the one shown in Fig., independent of the nature of the uncertainty. This verifies the theoretical claim on uniform approximation of the corresponding signals of a bounded reference system. In Fig. the control signal is shown, which is guaranteed to stay in low frequency range. 7 of 2

18 The disturbance is then introduced, as shown in Fig. 2. Since d(t) does not depend on the system output y(t), the condition in (2) is satisfied automatically. We see that the output response is slightly different than that of the nominal L case, but is still satisfactory. To explain this, we look into the closed loop reference system (9). It can be shown that y ref (s) = M(s) [C(s)r(s) + ( C(s))σ ref (s)] [ ] C(s)( C(s))(A(s) M(s)) = M(s) C(s)r(s) + M(s) + (A(s) M(s))C(s) r(s) + ( C(s))A(s) M(s) + (A(s) M(s))C(s) d(s). The transfer function from d(s) to y ref (s) can be expressed as (C d (s) C n (s))a n (s)m n (s) (C d (s) C n (s))a d (s)m n (s) + A n (s)m d (s)c n (s). (42) The magnitude curve of the Bode diagram is given in Fig. 3, which shows disturbance attenuation at low and high frequencies. This behavior of the reference system can be improved by manipulating the bandwidths of M(s) and C(s). For our non-minimum phase, unstable system A(s), the possible selections are not many..2.8 Acceleration (g) L no uncertainty no disturbance L with uncertainty, no disturbance time (sec) Figure. Closed loop response of L controller - with/without uncertainties, no disturbance Finally the parametric uncertainties are changed due to a change in aerodynamic coefficients given below: C z α =.2 C zα, C m α =.5 C mα, C z δp =.8 C zδ p, C m δp =.7 C mδ p. (43) The system output is shown in Fig. 4. We can see that the L output feedback adaptive control still has uniform performance. 8 of 2

19 .5 L control signal with uncertainty, no disturbance..5 δ p (rad) time (sec) Figure. Control signal of L controller.2.8 Acceleration (g) L no uncertainty no disturbance L with uncertainty and disturbance time (sec) Figure 2. Closed loop response of L controller - with uncertainties and disturbance Bode Diagram 2 Magnitude (db) Frequency (rad/sec) Figure 3. Frequency response of transfer function from d to y ref 9 of 2

20 .2.8 Acceleration (g) L no uncertainty no disturbance L with uncertainty case 2, and disturbance time (sec) Figure 4. Closed loop response of L controller with different uncertainties VII. Conclusion Longitudinal autopilot design for a missile model is performed using L adaptive output feedback controller, appropriate for non-spr reference system dynamics. The new piece-wise constant adaptive law along with the lowpass filtered control signal ensures uniform performance bounds for system s both input/output signals as compared to the corresponding signals of a non-spr reference system. The simulation responses of the proposed controller are compared to those of baseline LQR and LQG/LTR design, and the benefits of the L adaptive controller are clearly demonstrated. VIII. Acknowledgment This material is based upon work supported by the AFOSR under Contract FA Any opinions, findings and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of United States Air Force. References C. P. Mracek and D. B. Ridgely, Missile Longitudinal Autopilots: Connections Between Optimal Control and Classical Topologies, Proceedings of AIAA Guidance, Navigation, and Control Conference and Exhibit, San Francisco, CA, C. Cao and N. Hovakimyan, L Adaptive Output Feedback Controller for Systems of Unknown Relative Degree, Submitted to American Control Conference, C. Cao and N. Hovakimyan, L Adaptive Output Feedback Controller for Non Strictly Positive Real Reference Systems, Submitted to American Control Conference, C. Cao and N. Hovakimyan. L Adaptive Output Feedback Controller for Systems of Unknown Dimension. IEEE Transactions on Automatic Control, 53:85 82, April P. A. Ioannou and J. Sun, Robust Adaptive Control. NJ, Prentice Hall, of 2

21 6 V. L. Kharitonov, Asymptotic Stability of an Equilibrium Position of a Family of Systems of Linear Differential Equations, Differential Uravnen, 4: , S. P. Bhattacharyya, H. Chapellat and L. H. Keel, Robust Control: The Parametric Approach. NJ, Prentice Hall, C. Cao and N. Hovakimyan, Design and Analysis of a Novel L Adaptive Control Architecture, Part I: Control Signal and Asymptotic Stability, American Control Conference, pages , C. Cao and N. Hovakimyan, Design and Analysis of a Novel L Adaptive Control Architecture, Part II: Guaranteed Transient Performance, American Control Conference, pages , 26. D. Li, N. Hovakimyan, C. Cao and K. Wise, Filter Design for Feedback-loop Trade-off of L Adaptive Controller: A Linear Matrix Inequality Approach, Proceedings of AIAA Guidance, navigation and Control Conference, Honolulu, HI, August of 2

L 1 Adaptive Output Feedback Controller to Systems of Unknown

L 1 Adaptive Output Feedback Controller to Systems of Unknown Proceedings of the 27 American Control Conference Marriott Marquis Hotel at Times Square New York City, USA, July 11-13, 27 WeB1.1 L 1 Adaptive Output Feedback Controller to Systems of Unknown Dimension

More information

L 1 Adaptive Output Feedback Controller for Non Strictly Positive Real Reference Systems with Applications to Aerospace Examples

L 1 Adaptive Output Feedback Controller for Non Strictly Positive Real Reference Systems with Applications to Aerospace Examples AIAA Guidance, Navigation and Control Conference and Exhibit 18-21 August 28, Honolulu, Hawaii AIAA 28-7288 L 1 Adaptive Output Feedback Controller for Non Strictly Positive Real Reference Systems with

More information

L 1 Adaptive Controller for a Class of Systems with Unknown

L 1 Adaptive Controller for a Class of Systems with Unknown 28 American Control Conference Westin Seattle Hotel, Seattle, Washington, USA June -3, 28 FrA4.2 L Adaptive Controller for a Class of Systems with Unknown Nonlinearities: Part I Chengyu Cao and Naira Hovakimyan

More information

L 1 Adaptive Controller for Multi Input Multi Output Systems in the Presence of Unmatched Disturbances

L 1 Adaptive Controller for Multi Input Multi Output Systems in the Presence of Unmatched Disturbances 28 American Control Conference Westin Seattle Hotel, Seattle, Washington, USA June -3, 28 FrA4.4 L Adaptive Controller for Multi Input Multi Output Systems in the Presence of Unmatched Disturbances Chengyu

More information

Filter Design for Feedback-loop Trade-off of L 1 Adaptive Controller: A Linear Matrix Inequality Approach

Filter Design for Feedback-loop Trade-off of L 1 Adaptive Controller: A Linear Matrix Inequality Approach AIAA Guidance, Navigation and Control Conference and Exhibit 18-21 August 2008, Honolulu, Hawaii AIAA 2008-6280 Filter Design for Feedback-loop Trade-off of L 1 Adaptive Controller: A Linear Matrix Inequality

More information

Design and Analysis of a Novel L 1 Adaptive Controller, Part I: Control Signal and Asymptotic Stability

Design and Analysis of a Novel L 1 Adaptive Controller, Part I: Control Signal and Asymptotic Stability Proceedings of the 26 American Control Conference Minneapolis, Minnesota, USA, June 4-6, 26 ThB7.5 Design and Analysis of a Novel L Adaptive Controller, Part I: Control Signal and Asymptotic Stability

More information

Linear State Feedback Controller Design

Linear State Feedback Controller Design Assignment For EE5101 - Linear Systems Sem I AY2010/2011 Linear State Feedback Controller Design Phang Swee King A0033585A Email: king@nus.edu.sg NGS/ECE Dept. Faculty of Engineering National University

More information

Introduction. 1.1 Historical Overview. Chapter 1

Introduction. 1.1 Historical Overview. Chapter 1 Chapter 1 Introduction 1.1 Historical Overview Research in adaptive control was motivated by the design of autopilots for highly agile aircraft that need to operate at a wide range of speeds and altitudes,

More information

Übersetzungshilfe / Translation aid (English) To be returned at the end of the exam!

Übersetzungshilfe / Translation aid (English) To be returned at the end of the exam! Prüfung Regelungstechnik I (Control Systems I) Prof. Dr. Lino Guzzella 3.. 24 Übersetzungshilfe / Translation aid (English) To be returned at the end of the exam! Do not mark up this translation aid -

More information

Lecture 7 : Generalized Plant and LFT form Dr.-Ing. Sudchai Boonto Assistant Professor

Lecture 7 : Generalized Plant and LFT form Dr.-Ing. Sudchai Boonto Assistant Professor Dr.-Ing. Sudchai Boonto Assistant Professor Department of Control System and Instrumentation Engineering King Mongkuts Unniversity of Technology Thonburi Thailand Linear Quadratic Gaussian The state space

More information

Design Methods for Control Systems

Design Methods for Control Systems Design Methods for Control Systems Maarten Steinbuch TU/e Gjerrit Meinsma UT Dutch Institute of Systems and Control Winter term 2002-2003 Schedule November 25 MSt December 2 MSt Homework # 1 December 9

More information

Return Difference Function and Closed-Loop Roots Single-Input/Single-Output Control Systems

Return Difference Function and Closed-Loop Roots Single-Input/Single-Output Control Systems Spectral Properties of Linear- Quadratic Regulators Robert Stengel Optimal Control and Estimation MAE 546 Princeton University, 2018! Stability margins of single-input/singleoutput (SISO) systems! Characterizations

More information

Output Adaptive Model Reference Control of Linear Continuous State-Delay Plant

Output Adaptive Model Reference Control of Linear Continuous State-Delay Plant Output Adaptive Model Reference Control of Linear Continuous State-Delay Plant Boris M. Mirkin and Per-Olof Gutman Faculty of Agricultural Engineering Technion Israel Institute of Technology Haifa 3, Israel

More information

Controls Problems for Qualifying Exam - Spring 2014

Controls Problems for Qualifying Exam - Spring 2014 Controls Problems for Qualifying Exam - Spring 2014 Problem 1 Consider the system block diagram given in Figure 1. Find the overall transfer function T(s) = C(s)/R(s). Note that this transfer function

More information

CDS 101/110a: Lecture 8-1 Frequency Domain Design

CDS 101/110a: Lecture 8-1 Frequency Domain Design CDS 11/11a: Lecture 8-1 Frequency Domain Design Richard M. Murray 17 November 28 Goals: Describe canonical control design problem and standard performance measures Show how to use loop shaping to achieve

More information

Topic # Feedback Control Systems

Topic # Feedback Control Systems Topic #17 16.31 Feedback Control Systems Deterministic LQR Optimal control and the Riccati equation Weight Selection Fall 2007 16.31 17 1 Linear Quadratic Regulator (LQR) Have seen the solutions to the

More information

OPTIMAL CONTROL AND ESTIMATION

OPTIMAL CONTROL AND ESTIMATION OPTIMAL CONTROL AND ESTIMATION Robert F. Stengel Department of Mechanical and Aerospace Engineering Princeton University, Princeton, New Jersey DOVER PUBLICATIONS, INC. New York CONTENTS 1. INTRODUCTION

More information

EECE 460 : Control System Design

EECE 460 : Control System Design EECE 460 : Control System Design SISO Pole Placement Guy A. Dumont UBC EECE January 2011 Guy A. Dumont (UBC EECE) EECE 460: Pole Placement January 2011 1 / 29 Contents 1 Preview 2 Polynomial Pole Placement

More information

Chapter 2. Classical Control System Design. Dutch Institute of Systems and Control

Chapter 2. Classical Control System Design. Dutch Institute of Systems and Control Chapter 2 Classical Control System Design Overview Ch. 2. 2. Classical control system design Introduction Introduction Steady-state Steady-state errors errors Type Type k k systems systems Integral Integral

More information

I. D. Landau, A. Karimi: A Course on Adaptive Control Adaptive Control. Part 9: Adaptive Control with Multiple Models and Switching

I. D. Landau, A. Karimi: A Course on Adaptive Control Adaptive Control. Part 9: Adaptive Control with Multiple Models and Switching I. D. Landau, A. Karimi: A Course on Adaptive Control - 5 1 Adaptive Control Part 9: Adaptive Control with Multiple Models and Switching I. D. Landau, A. Karimi: A Course on Adaptive Control - 5 2 Outline

More information

Control System Design

Control System Design ELEC ENG 4CL4: Control System Design Notes for Lecture #14 Wednesday, February 5, 2003 Dr. Ian C. Bruce Room: CRL-229 Phone ext.: 26984 Email: ibruce@mail.ece.mcmaster.ca Chapter 7 Synthesis of SISO Controllers

More information

Here represents the impulse (or delta) function. is an diagonal matrix of intensities, and is an diagonal matrix of intensities.

Here represents the impulse (or delta) function. is an diagonal matrix of intensities, and is an diagonal matrix of intensities. 19 KALMAN FILTER 19.1 Introduction In the previous section, we derived the linear quadratic regulator as an optimal solution for the fullstate feedback control problem. The inherent assumption was that

More information

Chapter 3. LQ, LQG and Control System Design. Dutch Institute of Systems and Control

Chapter 3. LQ, LQG and Control System Design. Dutch Institute of Systems and Control Chapter 3 LQ, LQG and Control System H 2 Design Overview LQ optimization state feedback LQG optimization output feedback H 2 optimization non-stochastic version of LQG Application to feedback system design

More information

Supplementary Material for Load Capacity. Improvements in Nucleic Acid Based Systems. Using Partially Open Feedback Control

Supplementary Material for Load Capacity. Improvements in Nucleic Acid Based Systems. Using Partially Open Feedback Control Supplementary Material for Load Capacity Improvements in Nucleic Acid Based Systems Using Partially Open Feedback Control Vishwesh Kulkarni,, Evgeny Kharisov, Naira Hovakimyan, and Jongmin Kim Institute

More information

Advanced Adaptive Control for Unintended System Behavior

Advanced Adaptive Control for Unintended System Behavior Advanced Adaptive Control for Unintended System Behavior Dr. Chengyu Cao Mechanical Engineering University of Connecticut ccao@engr.uconn.edu jtang@engr.uconn.edu Outline Part I: Challenges: Unintended

More information

Analysis of SISO Control Loops

Analysis of SISO Control Loops Chapter 5 Analysis of SISO Control Loops Topics to be covered For a given controller and plant connected in feedback we ask and answer the following questions: Is the loop stable? What are the sensitivities

More information

LQG/LTR CONTROLLER DESIGN FOR AN AIRCRAFT MODEL

LQG/LTR CONTROLLER DESIGN FOR AN AIRCRAFT MODEL PERIODICA POLYTECHNICA SER. TRANSP. ENG. VOL. 8, NO., PP. 3 4 () LQG/LTR CONTROLLER DESIGN FOR AN AIRCRAFT MODEL Balázs KULCSÁR Department of Control and Transport Automation Budapest University of Technology

More information

D(s) G(s) A control system design definition

D(s) G(s) A control system design definition R E Compensation D(s) U Plant G(s) Y Figure 7. A control system design definition x x x 2 x 2 U 2 s s 7 2 Y Figure 7.2 A block diagram representing Eq. (7.) in control form z U 2 s z Y 4 z 2 s z 2 3 Figure

More information

CONTROL DESIGN FOR SET POINT TRACKING

CONTROL DESIGN FOR SET POINT TRACKING Chapter 5 CONTROL DESIGN FOR SET POINT TRACKING In this chapter, we extend the pole placement, observer-based output feedback design to solve tracking problems. By tracking we mean that the output is commanded

More information

Control Systems Design

Control Systems Design ELEC4410 Control Systems Design Lecture 18: State Feedback Tracking and State Estimation Julio H. Braslavsky julio@ee.newcastle.edu.au School of Electrical Engineering and Computer Science Lecture 18:

More information

Control System Design

Control System Design ELEC ENG 4CL4: Control System Design Notes for Lecture #36 Dr. Ian C. Bruce Room: CRL-229 Phone ext.: 26984 Email: ibruce@mail.ece.mcmaster.ca Friday, April 4, 2003 3. Cascade Control Next we turn to an

More information

Synthesis via State Space Methods

Synthesis via State Space Methods Chapter 18 Synthesis via State Space Methods Here, we will give a state space interpretation to many of the results described earlier. In a sense, this will duplicate the earlier work. Our reason for doing

More information

Automatic Control 2. Loop shaping. Prof. Alberto Bemporad. University of Trento. Academic year

Automatic Control 2. Loop shaping. Prof. Alberto Bemporad. University of Trento. Academic year Automatic Control 2 Loop shaping Prof. Alberto Bemporad University of Trento Academic year 21-211 Prof. Alberto Bemporad (University of Trento) Automatic Control 2 Academic year 21-211 1 / 39 Feedback

More information

Subject: Optimal Control Assignment-1 (Related to Lecture notes 1-10)

Subject: Optimal Control Assignment-1 (Related to Lecture notes 1-10) Subject: Optimal Control Assignment- (Related to Lecture notes -). Design a oil mug, shown in fig., to hold as much oil possible. The height and radius of the mug should not be more than 6cm. The mug must

More information

Linear Quadratic Gausssian Control Design with Loop Transfer Recovery

Linear Quadratic Gausssian Control Design with Loop Transfer Recovery Linear Quadratic Gausssian Control Design with Loop Transfer Recovery Leonid Freidovich Department of Mathematics Michigan State University MI 48824, USA e-mail:leonid@math.msu.edu http://www.math.msu.edu/

More information

Lecture 13: Internal Model Principle and Repetitive Control

Lecture 13: Internal Model Principle and Repetitive Control ME 233, UC Berkeley, Spring 2014 Xu Chen Lecture 13: Internal Model Principle and Repetitive Control Big picture review of integral control in PID design example: 0 Es) C s) Ds) + + P s) Y s) where P s)

More information

Übersetzungshilfe / Translation aid (English) To be returned at the end of the exam!

Übersetzungshilfe / Translation aid (English) To be returned at the end of the exam! Prüfung Regelungstechnik I (Control Systems I) Prof. Dr. Lino Guzzella 3. 8. 24 Übersetzungshilfe / Translation aid (English) To be returned at the end of the exam! Do not mark up this translation aid

More information

QUANTITATIVE L P STABILITY ANALYSIS OF A CLASS OF LINEAR TIME-VARYING FEEDBACK SYSTEMS

QUANTITATIVE L P STABILITY ANALYSIS OF A CLASS OF LINEAR TIME-VARYING FEEDBACK SYSTEMS Int. J. Appl. Math. Comput. Sci., 2003, Vol. 13, No. 2, 179 184 QUANTITATIVE L P STABILITY ANALYSIS OF A CLASS OF LINEAR TIME-VARYING FEEDBACK SYSTEMS PINI GURFIL Department of Mechanical and Aerospace

More information

The Rationale for Second Level Adaptation

The Rationale for Second Level Adaptation The Rationale for Second Level Adaptation Kumpati S. Narendra, Yu Wang and Wei Chen Center for Systems Science, Yale University arxiv:1510.04989v1 [cs.sy] 16 Oct 2015 Abstract Recently, a new approach

More information

Automatic Control II Computer exercise 3. LQG Design

Automatic Control II Computer exercise 3. LQG Design Uppsala University Information Technology Systems and Control HN,FS,KN 2000-10 Last revised by HR August 16, 2017 Automatic Control II Computer exercise 3 LQG Design Preparations: Read Chapters 5 and 9

More information

Extension of L1 Adaptive Control with Applications

Extension of L1 Adaptive Control with Applications University of Connecticut DigitalCommons@UConn Doctoral Dissertations University of Connecticut Graduate School 12-15-2016 Extension of L1 Adaptive Control with Applications jiaxing che University of Connecticut,

More information

Übersetzungshilfe / Translation aid (English) To be returned at the end of the exam!

Übersetzungshilfe / Translation aid (English) To be returned at the end of the exam! Prüfung Regelungstechnik I (Control Systems I) Prof. Dr. Lino Guzzella 5. 2. 2 Übersetzungshilfe / Translation aid (English) To be returned at the end of the exam! Do not mark up this translation aid -

More information

Let the plant and controller be described as:-

Let the plant and controller be described as:- Summary of Fundamental Limitations in Feedback Design (LTI SISO Systems) From Chapter 6 of A FIRST GRADUATE COURSE IN FEEDBACK CONTROL By J. S. Freudenberg (Winter 2008) Prepared by: Hammad Munawar (Institute

More information

Dr Ian R. Manchester

Dr Ian R. Manchester Week Content Notes 1 Introduction 2 Frequency Domain Modelling 3 Transient Performance and the s-plane 4 Block Diagrams 5 Feedback System Characteristics Assign 1 Due 6 Root Locus 7 Root Locus 2 Assign

More information

Lecture 9. Introduction to Kalman Filtering. Linear Quadratic Gaussian Control (LQG) G. Hovland 2004

Lecture 9. Introduction to Kalman Filtering. Linear Quadratic Gaussian Control (LQG) G. Hovland 2004 MER42 Advanced Control Lecture 9 Introduction to Kalman Filtering Linear Quadratic Gaussian Control (LQG) G. Hovland 24 Announcement No tutorials on hursday mornings 8-9am I will be present in all practical

More information

Analysis and Synthesis of Single-Input Single-Output Control Systems

Analysis and Synthesis of Single-Input Single-Output Control Systems Lino Guzzella Analysis and Synthesis of Single-Input Single-Output Control Systems l+kja» \Uja>)W2(ja»\ um Contents 1 Definitions and Problem Formulations 1 1.1 Introduction 1 1.2 Definitions 1 1.2.1 Systems

More information

Optimal Polynomial Control for Discrete-Time Systems

Optimal Polynomial Control for Discrete-Time Systems 1 Optimal Polynomial Control for Discrete-Time Systems Prof Guy Beale Electrical and Computer Engineering Department George Mason University Fairfax, Virginia Correspondence concerning this paper should

More information

Aircraft Stability & Control

Aircraft Stability & Control Aircraft Stability & Control Textbook Automatic control of Aircraft and missiles 2 nd Edition by John H Blakelock References Aircraft Dynamics and Automatic Control - McRuler & Ashkenas Aerodynamics, Aeronautics

More information

5. Observer-based Controller Design

5. Observer-based Controller Design EE635 - Control System Theory 5. Observer-based Controller Design Jitkomut Songsiri state feedback pole-placement design regulation and tracking state observer feedback observer design LQR and LQG 5-1

More information

EEE 184: Introduction to feedback systems

EEE 184: Introduction to feedback systems EEE 84: Introduction to feedback systems Summary 6 8 8 x 7 7 6 Level() 6 5 4 4 5 5 time(s) 4 6 8 Time (seconds) Fig.. Illustration of BIBO stability: stable system (the input is a unit step) Fig.. step)

More information

LQR, Kalman Filter, and LQG. Postgraduate Course, M.Sc. Electrical Engineering Department College of Engineering University of Salahaddin

LQR, Kalman Filter, and LQG. Postgraduate Course, M.Sc. Electrical Engineering Department College of Engineering University of Salahaddin LQR, Kalman Filter, and LQG Postgraduate Course, M.Sc. Electrical Engineering Department College of Engineering University of Salahaddin May 2015 Linear Quadratic Regulator (LQR) Consider a linear system

More information

Every real system has uncertainties, which include system parametric uncertainties, unmodeled dynamics

Every real system has uncertainties, which include system parametric uncertainties, unmodeled dynamics Sensitivity Analysis of Disturbance Accommodating Control with Kalman Filter Estimation Jemin George and John L. Crassidis University at Buffalo, State University of New York, Amherst, NY, 14-44 The design

More information

6 OUTPUT FEEDBACK DESIGN

6 OUTPUT FEEDBACK DESIGN 6 OUTPUT FEEDBACK DESIGN When the whole sate vector is not available for feedback, i.e, we can measure only y = Cx. 6.1 Review of observer design Recall from the first class in linear systems that a simple

More information

Lyapunov Stability of Linear Predictor Feedback for Distributed Input Delays

Lyapunov Stability of Linear Predictor Feedback for Distributed Input Delays IEEE TRANSACTIONS ON AUTOMATIC CONTROL VOL. 56 NO. 3 MARCH 2011 655 Lyapunov Stability of Linear Predictor Feedback for Distributed Input Delays Nikolaos Bekiaris-Liberis Miroslav Krstic In this case system

More information

Stochastic Optimal Control!

Stochastic Optimal Control! Stochastic Control! Robert Stengel! Robotics and Intelligent Systems, MAE 345, Princeton University, 2015 Learning Objectives Overview of the Linear-Quadratic-Gaussian (LQG) Regulator Introduction to Stochastic

More information

State Regulator. Advanced Control. design of controllers using pole placement and LQ design rules

State Regulator. Advanced Control. design of controllers using pole placement and LQ design rules Advanced Control State Regulator Scope design of controllers using pole placement and LQ design rules Keywords pole placement, optimal control, LQ regulator, weighting matrixes Prerequisites Contact state

More information

Time Response Analysis (Part II)

Time Response Analysis (Part II) Time Response Analysis (Part II). A critically damped, continuous-time, second order system, when sampled, will have (in Z domain) (a) A simple pole (b) Double pole on real axis (c) Double pole on imaginary

More information

A DESIGN METHOD FOR SIMPLE REPETITIVE CONTROLLERS WITH SPECIFIED INPUT-OUTPUT CHARACTERISTIC

A DESIGN METHOD FOR SIMPLE REPETITIVE CONTROLLERS WITH SPECIFIED INPUT-OUTPUT CHARACTERISTIC International Journal of Innovative Computing, Information Control ICIC International c 202 ISSN 349-498 Volume 8, Number 7(A), July 202 pp. 4883 4899 A DESIGN METHOD FOR SIMPLE REPETITIVE CONTROLLERS

More information

EL2520 Control Theory and Practice

EL2520 Control Theory and Practice EL2520 Control Theory and Practice Lecture 8: Linear quadratic control Mikael Johansson School of Electrical Engineering KTH, Stockholm, Sweden Linear quadratic control Allows to compute the controller

More information

Lecture 4 Stabilization

Lecture 4 Stabilization Lecture 4 Stabilization This lecture follows Chapter 5 of Doyle-Francis-Tannenbaum, with proofs and Section 5.3 omitted 17013 IOC-UPC, Lecture 4, November 2nd 2005 p. 1/23 Stable plants (I) We assume that

More information

Topic # Feedback Control. State-Space Systems Closed-loop control using estimators and regulators. Dynamics output feedback

Topic # Feedback Control. State-Space Systems Closed-loop control using estimators and regulators. Dynamics output feedback Topic #17 16.31 Feedback Control State-Space Systems Closed-loop control using estimators and regulators. Dynamics output feedback Back to reality Copyright 21 by Jonathan How. All Rights reserved 1 Fall

More information

Raktim Bhattacharya. . AERO 422: Active Controls for Aerospace Vehicles. Basic Feedback Analysis & Design

Raktim Bhattacharya. . AERO 422: Active Controls for Aerospace Vehicles. Basic Feedback Analysis & Design AERO 422: Active Controls for Aerospace Vehicles Basic Feedback Analysis & Design Raktim Bhattacharya Laboratory For Uncertainty Quantification Aerospace Engineering, Texas A&M University Routh s Stability

More information

ECEN 605 LINEAR SYSTEMS. Lecture 20 Characteristics of Feedback Control Systems II Feedback and Stability 1/27

ECEN 605 LINEAR SYSTEMS. Lecture 20 Characteristics of Feedback Control Systems II Feedback and Stability 1/27 1/27 ECEN 605 LINEAR SYSTEMS Lecture 20 Characteristics of Feedback Control Systems II Feedback and Stability Feedback System Consider the feedback system u + G ol (s) y Figure 1: A unity feedback system

More information

Introduction to Feedback Control

Introduction to Feedback Control Introduction to Feedback Control Control System Design Why Control? Open-Loop vs Closed-Loop (Feedback) Why Use Feedback Control? Closed-Loop Control System Structure Elements of a Feedback Control System

More information

Autonomous Mobile Robot Design

Autonomous Mobile Robot Design Autonomous Mobile Robot Design Topic: Guidance and Control Introduction and PID Loops Dr. Kostas Alexis (CSE) Autonomous Robot Challenges How do I control where to go? Autonomous Mobile Robot Design Topic:

More information

Control Systems. Root Locus & Pole Assignment. L. Lanari

Control Systems. Root Locus & Pole Assignment. L. Lanari Control Systems Root Locus & Pole Assignment L. Lanari Outline root-locus definition main rules for hand plotting root locus as a design tool other use of the root locus pole assignment Lanari: CS - Root

More information

The Tuning of Robust Controllers for Stable Systems in the CD-Algebra: The Case of Sinusoidal and Polynomial Signals

The Tuning of Robust Controllers for Stable Systems in the CD-Algebra: The Case of Sinusoidal and Polynomial Signals The Tuning of Robust Controllers for Stable Systems in the CD-Algebra: The Case of Sinusoidal and Polynomial Signals Timo Hämäläinen Seppo Pohjolainen Tampere University of Technology Department of Mathematics

More information

L 1 Adaptive Control of a UAV for Aerobiological Sampling

L 1 Adaptive Control of a UAV for Aerobiological Sampling Proceedings of the 27 American Control Conference Marriott Marquis Hotel at Times Square New York City, USA, July 11-13, 27 FrA14.1 L 1 Adaptive Control of a UAV for Aerobiological Sampling Jiang Wang

More information

Adaptive Output Feedback Based on Closed-Loop. Reference Models for Hypersonic Vehicles

Adaptive Output Feedback Based on Closed-Loop. Reference Models for Hypersonic Vehicles Adaptive Output Feedback Based on Closed-Loop Reference Models for Hypersonic Vehicles Daniel P. Wiese 1 and Anuradha M. Annaswamy 2 Massachusetts Institute of Technology, Cambridge, MA 02139 Jonathan

More information

TRACKING AND DISTURBANCE REJECTION

TRACKING AND DISTURBANCE REJECTION TRACKING AND DISTURBANCE REJECTION Sadegh Bolouki Lecture slides for ECE 515 University of Illinois, Urbana-Champaign Fall 2016 S. Bolouki (UIUC) 1 / 13 General objective: The output to track a reference

More information

Richiami di Controlli Automatici

Richiami di Controlli Automatici Richiami di Controlli Automatici Gianmaria De Tommasi 1 1 Università degli Studi di Napoli Federico II detommas@unina.it Ottobre 2012 Corsi AnsaldoBreda G. De Tommasi (UNINA) Richiami di Controlli Automatici

More information

Dr Ian R. Manchester Dr Ian R. Manchester AMME 3500 : Review

Dr Ian R. Manchester Dr Ian R. Manchester AMME 3500 : Review Week Date Content Notes 1 6 Mar Introduction 2 13 Mar Frequency Domain Modelling 3 20 Mar Transient Performance and the s-plane 4 27 Mar Block Diagrams Assign 1 Due 5 3 Apr Feedback System Characteristics

More information

Verifiable Adaptive Control Solutions for Flight Control Applications

Verifiable Adaptive Control Solutions for Flight Control Applications Verifiable Adaptive Control Solutions for Flight Control Applications Jiang Wang Dissertation submitted to the Faculty of the Virginia Polytechnic Institute and State University in partial fulfillment

More information

Digital Control Engineering Analysis and Design

Digital Control Engineering Analysis and Design Digital Control Engineering Analysis and Design M. Sami Fadali Antonio Visioli AMSTERDAM BOSTON HEIDELBERG LONDON NEW YORK OXFORD PARIS SAN DIEGO SAN FRANCISCO SINGAPORE SYDNEY TOKYO Academic Press is

More information

Introduction. Performance and Robustness (Chapter 1) Advanced Control Systems Spring / 31

Introduction. Performance and Robustness (Chapter 1) Advanced Control Systems Spring / 31 Introduction Classical Control Robust Control u(t) y(t) G u(t) G + y(t) G : nominal model G = G + : plant uncertainty Uncertainty sources : Structured : parametric uncertainty, multimodel uncertainty Unstructured

More information

Department of Electronics and Instrumentation Engineering M. E- CONTROL AND INSTRUMENTATION ENGINEERING CL7101 CONTROL SYSTEM DESIGN Unit I- BASICS AND ROOT-LOCUS DESIGN PART-A (2 marks) 1. What are the

More information

MRAGPC Control of MIMO Processes with Input Constraints and Disturbance

MRAGPC Control of MIMO Processes with Input Constraints and Disturbance Proceedings of the World Congress on Engineering and Computer Science 9 Vol II WCECS 9, October -, 9, San Francisco, USA MRAGPC Control of MIMO Processes with Input Constraints and Disturbance A. S. Osunleke,

More information

MAS107 Control Theory Exam Solutions 2008

MAS107 Control Theory Exam Solutions 2008 MAS07 CONTROL THEORY. HOVLAND: EXAM SOLUTION 2008 MAS07 Control Theory Exam Solutions 2008 Geir Hovland, Mechatronics Group, Grimstad, Norway June 30, 2008 C. Repeat question B, but plot the phase curve

More information

Control Systems II. ETH, MAVT, IDSC, Lecture 4 17/03/2017. G. Ducard

Control Systems II. ETH, MAVT, IDSC, Lecture 4 17/03/2017. G. Ducard Control Systems II ETH, MAVT, IDSC, Lecture 4 17/03/2017 Lecture plan: Control Systems II, IDSC, 2017 SISO Control Design 24.02 Lecture 1 Recalls, Introductory case study 03.03 Lecture 2 Cascaded Control

More information

Prüfung Regelungstechnik I (Control Systems I) Übersetzungshilfe / Translation aid (English) To be returned at the end of the exam!

Prüfung Regelungstechnik I (Control Systems I) Übersetzungshilfe / Translation aid (English) To be returned at the end of the exam! Prüfung Regelungstechnik I (Control Systems I) Prof. Dr. Lino Guzzella 29. 8. 2 Übersetzungshilfe / Translation aid (English) To be returned at the end of the exam! Do not mark up this translation aid

More information

Lecture 6 Classical Control Overview IV. Dr. Radhakant Padhi Asst. Professor Dept. of Aerospace Engineering Indian Institute of Science - Bangalore

Lecture 6 Classical Control Overview IV. Dr. Radhakant Padhi Asst. Professor Dept. of Aerospace Engineering Indian Institute of Science - Bangalore Lecture 6 Classical Control Overview IV Dr. Radhakant Padhi Asst. Professor Dept. of Aerospace Engineering Indian Institute of Science - Bangalore Lead Lag Compensator Design Dr. Radhakant Padhi Asst.

More information

ECE317 : Feedback and Control

ECE317 : Feedback and Control ECE317 : Feedback and Control Lecture : Routh-Hurwitz stability criterion Examples Dr. Richard Tymerski Dept. of Electrical and Computer Engineering Portland State University 1 Course roadmap Modeling

More information

Fundamental Design Limitations of the General Control Configuration

Fundamental Design Limitations of the General Control Configuration IEEE TRANSACTIONS ON AUTOMATIC CONTROL, VOL 48, NO 8, AUGUST 2003 1355 Fundamental Design Limitations of the General Control Configuration Jim S Freudenberg, Fellow, IEEE, C V Hollot, Senior Member, IEEE,

More information

Attitude Regulation About a Fixed Rotation Axis

Attitude Regulation About a Fixed Rotation Axis AIAA Journal of Guidance, Control, & Dynamics Revised Submission, December, 22 Attitude Regulation About a Fixed Rotation Axis Jonathan Lawton Raytheon Systems Inc. Tucson, Arizona 85734 Randal W. Beard

More information

ECE317 : Feedback and Control

ECE317 : Feedback and Control ECE317 : Feedback and Control Lecture : Steady-state error Dr. Richard Tymerski Dept. of Electrical and Computer Engineering Portland State University 1 Course roadmap Modeling Analysis Design Laplace

More information

Design and Tuning of Fractional-order PID Controllers for Time-delayed Processes

Design and Tuning of Fractional-order PID Controllers for Time-delayed Processes Design and Tuning of Fractional-order PID Controllers for Time-delayed Processes Emmanuel Edet Technology and Innovation Centre University of Strathclyde 99 George Street Glasgow, United Kingdom emmanuel.edet@strath.ac.uk

More information

IMC based automatic tuning method for PID controllers in a Smith predictor configuration

IMC based automatic tuning method for PID controllers in a Smith predictor configuration Computers and Chemical Engineering 28 (2004) 281 290 IMC based automatic tuning method for PID controllers in a Smith predictor configuration Ibrahim Kaya Department of Electrical and Electronics Engineering,

More information

(Continued on next page)

(Continued on next page) (Continued on next page) 18.2 Roots of Stability Nyquist Criterion 87 e(s) 1 S(s) = =, r(s) 1 + P (s)c(s) where P (s) represents the plant transfer function, and C(s) the compensator. The closedloop characteristic

More information

Pole placement control: state space and polynomial approaches Lecture 2

Pole placement control: state space and polynomial approaches Lecture 2 : state space and polynomial approaches Lecture 2 : a state O. Sename 1 1 Gipsa-lab, CNRS-INPG, FRANCE Olivier.Sename@gipsa-lab.fr www.gipsa-lab.fr/ o.sename -based November 21, 2017 Outline : a state

More information

Chapter 7 - Solved Problems

Chapter 7 - Solved Problems Chapter 7 - Solved Problems Solved Problem 7.1. A continuous time system has transfer function G o (s) given by G o (s) = B o(s) A o (s) = 2 (s 1)(s + 2) = 2 s 2 + s 2 (1) Find a controller of minimal

More information

Multi-Model Adaptive Regulation for a Family of Systems Containing Different Zero Structures

Multi-Model Adaptive Regulation for a Family of Systems Containing Different Zero Structures Preprints of the 19th World Congress The International Federation of Automatic Control Multi-Model Adaptive Regulation for a Family of Systems Containing Different Zero Structures Eric Peterson Harry G.

More information

João P. Hespanha. January 16, 2009

João P. Hespanha. January 16, 2009 LINEAR SYSTEMS THEORY João P. Hespanha January 16, 2009 Disclaimer: This is a draft and probably contains a few typos. Comments and information about typos are welcome. Please contact the author at hespanha@ece.ucsb.edu.

More information

EEE582 Homework Problems

EEE582 Homework Problems EEE582 Homework Problems HW. Write a state-space realization of the linearized model for the cruise control system around speeds v = 4 (Section.3, http://tsakalis.faculty.asu.edu/notes/models.pdf). Use

More information

FEL3210 Multivariable Feedback Control

FEL3210 Multivariable Feedback Control FEL3210 Multivariable Feedback Control Lecture 8: Youla parametrization, LMIs, Model Reduction and Summary [Ch. 11-12] Elling W. Jacobsen, Automatic Control Lab, KTH Lecture 8: Youla, LMIs, Model Reduction

More information

Professor Fearing EE C128 / ME C134 Problem Set 7 Solution Fall 2010 Jansen Sheng and Wenjie Chen, UC Berkeley

Professor Fearing EE C128 / ME C134 Problem Set 7 Solution Fall 2010 Jansen Sheng and Wenjie Chen, UC Berkeley Professor Fearing EE C8 / ME C34 Problem Set 7 Solution Fall Jansen Sheng and Wenjie Chen, UC Berkeley. 35 pts Lag compensation. For open loop plant Gs ss+5s+8 a Find compensator gain Ds k such that the

More information

Übersetzungshilfe / Translation aid (English) To be returned at the end of the exam!

Übersetzungshilfe / Translation aid (English) To be returned at the end of the exam! Prüfung Regelungstechnik I (Control Systems I) Prof. Dr. Lino Guzzella 9. 8. 2 Übersetzungshilfe / Translation aid (English) To be returned at the end of the exam! Do not mark up this translation aid -

More information

Control Systems. EC / EE / IN. For

Control Systems.   EC / EE / IN. For Control Systems For EC / EE / IN By www.thegateacademy.com Syllabus Syllabus for Control Systems Basic Control System Components; Block Diagrammatic Description, Reduction of Block Diagrams. Open Loop

More information

Event-Triggered Decentralized Dynamic Output Feedback Control for LTI Systems

Event-Triggered Decentralized Dynamic Output Feedback Control for LTI Systems Event-Triggered Decentralized Dynamic Output Feedback Control for LTI Systems Pavankumar Tallapragada Nikhil Chopra Department of Mechanical Engineering, University of Maryland, College Park, 2742 MD,

More information

An Adaptive LQG Combined With the MRAS Based LFFC for Motion Control Systems

An Adaptive LQG Combined With the MRAS Based LFFC for Motion Control Systems Journal of Automation Control Engineering Vol 3 No 2 April 2015 An Adaptive LQG Combined With the MRAS Based LFFC for Motion Control Systems Nguyen Duy Cuong Nguyen Van Lanh Gia Thi Dinh Electronics Faculty

More information

Chapter 6 - Solved Problems

Chapter 6 - Solved Problems Chapter 6 - Solved Problems Solved Problem 6.. Contributed by - James Welsh, University of Newcastle, Australia. Find suitable values for the PID parameters using the Z-N tuning strategy for the nominal

More information