LINEAR TIME VARYING TERMINAL LAWS IN MPQP
|
|
- Nelson Johnson
- 5 years ago
- Views:
Transcription
1 LINEAR TIME VARYING TERMINAL LAWS IN MPQP JA Rossiter Dept of Aut Control & Systems Eng University of Sheffield, Mappin Street Sheffield, S1 3JD, UK B Kouvaritakis M Cannon Dept of Eng Science Parks Road Oxford OX1 3PJ basilkouvaritakis@engoxacuk Abstract: This paper shows how changing the structure of the terminal control law can give significant reductions in the complexity of the MPQP solution to predictive control, at a small cost to performance 1 Introduction MPQP (multi parametric quadratic programming) solutions [1] are becoming a popular solution to the quadratic programming (QP) problems arising within model predictive control (MPC) [16] This interest is due to two main reasons: (i) the MPQP solution gives some transparency to the control action during constraint handling and (ii) there is potential for transfering much of the online computation to offline computation and hence improving speed One could also argue (not discussed here) that it is a natural framework for handling uncertainty Unfortunately, MPQP solutions can be very complex For instance the MPQP solution to the MPC based QP problem arising for a simple 2 state system could require the definition of 5 or more regions [1] One must then consider whether storage and implementation of the MPQP solution is actually more efficient than implementing an online QP solver Such considerations suggest that parametric programming solutions will often not be easy to implement on large dimensional problems [1] This paper takes the viewpoint that irrespective of general limitations in applicability, it is a worthwhile goal to ask how much one could simplify an MPQP solution foragivenmpcproblem[3,6,5,13] Themoresimplification in complexity that can be achieved, the wider the potential number of applications Broadly speaking MPC calls for the computation of an optimal control trajectory which in at most n steps, takes the system state into a target region For a quadratic cost and linear constraints, the optimum is piece wise affine (PWA) on the state Hence the optimal solution is made up of a coolection of linear state feedback laws each being associated with a specific region The smaller the targer region and/or the larger the number of steps allowed, the more regions that need to be defined 1 1 Itcanbeshownthatforsomeproblems[7],forn large Several authors have looked at improving efficiency For instance [6] looked at defining regions as hypercubes as this allows an efficient search, [5] looked at using a combination of one step sets and an overbounding lyapunov function and [13] uses interpolation with only the control laws associated to the facets Other authors [3] accept that there can be no reduction in the number of regions and investigate more efficient ways of identifying the active region We take a different approach which has some parallels with [9] The main insight derives from the observation that the maximal admissible set (MAS) [7] may be small, not just due to over tuning but also because effective constraint handling may require a linear time varying (LTV) control law Of course, LTV control is precisely what predictive control allows for thereby making significant increases in the reachable space possible The question that seems little discussed in the literature is whether the same LTV control law could be used for the entire space MPQP suggests that the optimal LTV control law changes with the state, however this need not imply theredoesnoexistasub-optimalltvcontrollaw with far wider applicability This paper gives some background to MPC and MPQP in section 2 and then section 3 shows how one could find a single LTV control law which has a large feasible region Section 4 discusses and illustrates how this larger region can be conbined with MPQP to generate a significantly simpler MPQP solution, at the price of only small suboptimality 2 Background 21 Notation This paper assumes a model and constraints x k+1 = Ax k + Bu k ; y k = Cx k u u k u; x x k x (1) where x, u, y are the state, input and output respectively and where the above inequalities apply on an element-by-element basis Next we define the MAS enough the solution is equivalent to n = and hence the number of regions has a computable upper limit
2 [7] for several different control laws and show how a predicted trajectory, and corresponding invariant set, can be made developed througn a convex linear combination of these control laws The MAS for the system of (1) under the control law u = K i x is the largest set such that the use of control law u = K i x gives convergence and feasibility For appropriately defined M,d the MAS can be described as: S = {x : Mx d } (2) 22 Predictive control This paper is based on the algorithm of [14] implemented via the closed-loop paradigm [11], that is min C k n c i=1 c T k+iλc k+i st NC k + Mx k d (3) where C k = [c T k,, ct k+n c 1 ], u k = Kx k + c k and the definitions of Λ, N, M depend upon constraints and tuning weights in the predicted cost Under the assumption that u = Kx, is the unconstrained optimum it follows that for Mx d, the optimum solution is C = The maximum control admissible set (MCAS), or feasible region, is given by {x C, NC + Mx d } 23 MPQP for MPC Optimisation (3) is a quadratic program and is known [1] to have a piecewise affine solution That is, one can construct regions S i = {x : M i x d i } (4) such that the solution of (3) is given as: x S i C = L i x + p i (5) (Recall that u k = K(x k + c k )) Hence, instead of solving (3) using an online quadratic programming optimiser, one can find within which set S i the state lies and implement the corresponding control law Computing sets S i and the corresponding control parameters L i, p i can be a significant burden, but as this is an offline task it is of small consequence More importantly, the number of regions required to span the controllable space may be prohibitively large for some systems This implies that the data storage requirement and also the implied set membership tests are no longer efficient and this negates some of the benefit of an MPQP approach 24 Dual predictive mode interpretation of MPC The MPC law of [14] can be thought of as dual predictive mode (DPM) in that the predictions are constructed in two phases: (i) a LTV transient phase of n c steps in which u = Kx + c and (ii) a linear asymptotic phase thereafter in which u = Kx The control strategy could be reinterpreted as minimising the performance wrt C, over an infinite horizon, subject to the n c step ahead prediction being within the MAS of (2) The volume and shape of the stabilisable set is determined by two main tuning parameters: (i) the shape of the MAS (or terminal region) and (ii) the number of dof, that is, n c To increase the size of the stabilisable set one must either increase the volume of the MAS or increase n c However, as the shape of the MAS is dependent on the implied terminal control law, one can only change this shape by changing the control law, in essence that is by accepting some suboptimality in the predicted asymptotic behaviour This may be an acceptable compromise although there are no obvious systematic mechanisms for choosing a state feedback which balances the volume of the associated MAS with performance The option of increasing n c may also be undesirable as this can lead to an increase in the complexity of the QP of (3) and the number of regions in the MPQP solution 2 3 Fixed LTV terminal law This paper assumes that a DPM type of strategy is a well conditioned basis for setting up the MPQP problem, but asks the question whether a nonlinear asymptotic feedback gain could be used to give a systematic compromise between performance and the volume of the feasible region This section shows how such a feedback gain can be computed and integrated into an MPQP algorithm 31 LTV feedback gain It is assumed that the predicted control law is given as follows: u k Kx k u k+1 Kx k+1 u = k+n 1 Kx +C k ; C k = k u k+n Kx k+1 L 1 L 2 L n (6) It is noted that this feedback differs in structure from an optimal constrained feedback of (5) only due to lack of the constant term p The task in hand is to compute c k and hence L i,i=1, 2,, n such that the predicted cost (3) is minimised and the associated feasible set is maximised in volume 2 It is recognised that MPQP can be used to find infinite horizon solutions for some cases and hence there would be an upper limit on complexity However, for many practical cases this upper limit is prohibitively high x k
3 32 Evaluating the cost The cost function can be written as n J = x T i Qx i + u T i Ru i + x T n P x n (7) k=1 From (6) it is clear that the predictions x i are linear in the coefficients of L i Hence, it is easy to show that is quadratic in L i ; the proof is straightforward and will be omitted For SISO case) we can write J = L T Λ L L + c; L =[L T 1,L T 2, ] (8) where c does not depend upon L Clearly the unconstrained minimisation of J wrt L gives L = 33 Constraint inequalities Substitution of prediction (6) into constraints gives the inequalities {N L 1 L 2 L n + M}x d (9) For a given value of x, this can be rearranged to: F (x)l d(x) (1) Hence the inequalities have an explicit dependence upon the non-linear feedback terms L i Let v i,i = 1, 2, be points on the boundary of the MAS given in (2) It is known that for the choice L =, (1) must hold true for x = v i The target region is larger if one instead can satisfy F (v i )L d(v i ) e i (11) where e i > An obvious objective therefore is to max L, V = i e2 i as this, in some sense, maximises the volume of the feasible region implied by (9) 34 Choosing the best LTV feedback The region of attraction and the predicted performance both depend upon the parameters of the proposed LTV feedback in a simple way Hence one can choose L by setting up a weighted average of predicted performance and set enlargement For instance using unit weights: J L = L T Λ L (x)l + i e 2 i (12) The implied constraint that e i could be included in addition to inequalities (11) and hence the overall optimisation is a quadratic program In summary, minimisation of J L subject to (11) and e i gives some compromise between the volume of the terminal region and the implied predicted performance To the authors knowledge, this compromise is more difficult to achieve (and hence less systematic) if one minimises wrt to linear feedback gains only [15] It should be noted that LMI methods tackling this, eg [8], are restricted to ellipsoidal sets The notable advantage of choosing the structure implicit in (6) is simplicity The target region is simply defined by (1) and one need only store a single LTV feedback that is applicable inside the region Moreover, as there is a single feedback law, even though LTV, the predicted cost (8) is known exactly - no overbounding is needed It is noted that the implied cost can be rewritten as: J L = x T Λ x (L)x (13) 4 Forming an MPQP solution with LTV feedback 41 The cost function Because the nonlinear feedback is suboptimal, it is necessary to redefine the cost from (3) so that it is applicable to a control law with three phases: Phase 1: Use u = Kx + c for n c steps Phase 2: Use nonlinear feedback (6) for steps n c +1ton c + n Phase 3: Use optimal feedback u = Kx for steps n c + n +1on We note that, using (13) for phases 2 and 3, the implied cost can be written as n c J = x T i Qx i + u T i Ru i + x T n c Λ x (L)x nc (14) k=1 Straightforward manipulation then gives J =[C C ] T W [C C ]+f; C = T x (15) and f is an unimportant scalar Notably, the optimum unconstrained C is no longer zero; a consequence of the fact that the LTV feedback is suboptimal 42 Constraints and the QP Taking the basic structure of the constraint equations to be NC + Mx d, if one increases n c by n, one could write [ ] C [N 1,N 2 ] + Mx d (16) C2 where C 2 was given by the C in (6) and C are the dof Hence substituting in for C 2 using the known values of L i and a prediction for x nc gives N t C + M t x d (17)
4 Hence the optimisation can be represented as: min J =[C C ] T W [C C ] st N t C+M t x d C (18) where C = T x and the control law to be implemented is given as u = Kx + c IfM t x d, then the LTV feedback (6) is feasible Optimisation takes the form of a conventional quadratic program and as such MPQP can be used to give a solution of the form given in (4,5) We will not repeat the steps here as they are well documented in the literature 43 Unusual cases Because the LTV feedback was formed to give a compromise between optimality and feasibility, for a fixed control strategy, the solution it gives will not, in general, be the same as the solution of (18) even on the boundaries of the region {x : M t x d} This is of little consequence except in the case where: MPQP algorithms with only minimal impact on performance First we will illustrate comparisons in the volume of terminal region, secondly the total number of regions required to cover a given region in space and third we will give some performance comparisons Due to space constraints, a single 2 state example, with input and state limits, is used [ ] [ 25 x k + 5 ] ; y k =[1 2]x k x k+1 = (21) 1 u k 1; 5 x 1,2 5 (22) The optimal control law K =[ ] is computed with weights Q = C T C, R =1 Forthis example we selected: n c = n c1 = n c2 = 6 (23) The MCAS is computed with n c = 6 dof, the nonlinear feedback with n c1 = 6 dof and the NLMCAS ({x C, N t C + M t x d }) withn c2 =6dof M t x d & N t C + M t x d (19) That is, the unconstrained optimal is feasible outside of the set S t The feasible space (NLMAS) of the unconstrained optimal to (18) is given as S f = {x :[N t T + M t ]x d} (2) NLMCAS MCAS NLMAS MAS Lemma 41 A feasible predicted control trajectory can be computed by inspection if either x S t or x S f However, in general this feasible region will not be convex x Proof: Feasibility is obvious as for x S t one can use feedback (6) and for x S f one can use C = T x The union of two convex sets is not convex in general unless one is a subset of the other x 1 Figure 1 Feasible regions and target sets 44 MPQP with a non-convex terminal region The MPQP algorithms in the literature assume a convex terminal region and as such some modification is required The usual procedure inverts each facet of the terminal region (and subsequently other regions) in turn to search for new regions The first check is whether the proposed new point is a member of a known region In the case of non-convex terminal regions one would invert, in turn, both the facets of S f and S t The check for membership of known regions will eliminate any double computation that could arise due to non-convexity 5 Examples This section will illustrate the potential of the proposed non-linear feedback in reducing complexity of 51 Terminal regions S and S t The terminal regions MAS/NLMAS and corresponding feasible regions MCAS/NLMCAS are plotted in figure 1 Figure 2 overlays the NLMAS, the MCAS and the MPQP regions for solving (18); this is to illustrate the number of regions required by the proposed algorithm to obtain the same (or larger) feasible region as the original MPQP algorithm The inner region is S and the outer region is S t 52 Complexity of MPQP For the QP algorithm of (18), we allow the same number of dof as given to the algorithm (5) as this seems to be as fair as can be allowed Then, we find the feasible region for (18) and find out how many regions of the new algorithm are needed to cover the same space Illustrations are given in figure 2 and also summarised in table 1
5 Example 1 Regions for (5) 117 Regions for (18) 29 Table 1 No of MPQP regions x x 1 Figure 2 Feasible regions 53 Closed-loop performance It is necessary to compare the closed-loop performance of the algorithms to demonstrate that despite the significant reduction in complexity, there has been only a small loss in optimality For this, points were taken on the boundary of S (one per facet) and closed-loop simulations were performed for each point; marked with circles in figure 1 The runtime cost was computed for each simulation and summed and then divided by the runtime cost for the algorithm of (3) with a large n c (the global optimum) The comparisons are given in table 2 Example 1 Sum of run time costs (5) 15 Sum of run time costs (18) 12 Table 2 Normalised runtime costs 54 Summary and conclusions It is clear that the NLMAS may be far larger than the MAS Hence it is also unsurprising that the NLM- CAS is larger than the MCAS, for the same n c More significantly, the MPQP solution to (18) needs fewer regions to cover the MCAS, that is, the complexity is much reduced The closed-loop simulations demonstrate that rather than causing a deterioration in performance, on average the performance is also better! For this example, deploying a fixed LTV feedback simplifies MPQP and gives better performance References [1] Bemporad, A, M Morari, V Dua, and EN Pistikopoulos The explicit linear quadratic regulator for constrained systems Automatica, 38(1):3 2, January 22 [2] Blachini, F, 1999, Set invariance in control, Automatica, 35, [3] Borrelli, F, M Baotic, A Bemporad, and M Morari Efficient on-line computation of constrained optimal control In Proc 4th IEEE Conf on Decision and Control, December 21 [4] Clarke, DW, Mohtadi, C and Tuffs, PS, 1987, Generalised predictive control, Parts 1 and 2, Automatica, 23, [5] Grieder, P, F Borrelli, FD Torrisi, and M Morari Computation of the constrained infinite time linear quadratic regulator ACC 23 [6] Johansen, TA and A Grancharova, Approximate explicit constrained linear model predictive control via orthogonal search tree, IEEE Trans AC, 48, 5, , 23 [7] Gilbert, EG and K T Tan, 1991, Linear systems with state and control constraints: the theory and application of maximal output admissable sets, IEEE Trans AC, 36, 9, [8] Kothare, MV, Balakrishnan, V and Morari, M, 1996, Robust constrained model predictive control using linear matrix inequalities, Automatica, 32, [9] Kouvaritakis, B, JA Rossiter and J Schuurmans, 2, Efficient robust predictive control, IEEE Trans AC, 45, 8, pp [1] Mayne, DQ, Rawlings, JB, Rao, CV and Scokaert, POM, 2, Constrained model predictive control: stability and optimality, Automatica, 36, [11] Rossiter,JA, MJ Rice, and B Kouvaritakis A numerically robust state-space approach to stable predictive control strategies Automatica, 38(1):65 73, 1991 [12] Rossiter, JA, Kouvaritakis, B and Cannon, M, 21, Computationally efficient algorithms for constraint handling with guaranteed stability and near optimality, IJC, 74, 17, [13] Rossiter, JA and Grieder, P, Using Interpolation to Simplify Explicit Model Predictive Control, to appear ACC4 [14] Scokaert, POM and Rawlings, JB, 1998, Constrained linear quadratic regulation, IEEE Trans AC, 43, 8, pp [15] Tan, KT and EG Gilbert, 1992, Multimode controllers for linear discrete time systems with general state and control constraints, Optimisation techniques and applications (World Scientific Pub Co, Singapore), pp [16] Tsang, TTC and Clarke, DW, 1988, Generalised predictive control with input constraints, IEE Proceedings Pt D, 6,
Stochastic Tube MPC with State Estimation
Proceedings of the 19th International Symposium on Mathematical Theory of Networks and Systems MTNS 2010 5 9 July, 2010 Budapest, Hungary Stochastic Tube MPC with State Estimation Mark Cannon, Qifeng Cheng,
More informationEFFICIENT MODEL PREDICTIVE CONTROL WITH PREDICTION DYNAMICS
EFFICIENT MODEL PREDICTIVE CONTROL WITH PREDICTION DYNAMICS Stian Drageset, Lars Imsland and Bjarne A. Foss Dept. of Eng. Cybernetics, Norwegian Univ. of Science and Technology, 7491 Trondheim, Norway.
More informationEE C128 / ME C134 Feedback Control Systems
EE C128 / ME C134 Feedback Control Systems Lecture Additional Material Introduction to Model Predictive Control Maximilian Balandat Department of Electrical Engineering & Computer Science University of
More informationRobust Explicit MPC Based on Approximate Multi-parametric Convex Programming
43rd IEEE Conference on Decision and Control December 4-7, 24 Atlantis, Paradise Island, Bahamas WeC6.3 Robust Explicit MPC Based on Approximate Multi-parametric Convex Programming D. Muñoz de la Peña
More informationRobustly stable feedback min-max model predictive control 1
Robustly stable feedback min-max model predictive control 1 Eric C. Kerrigan 2 and Jan M. Maciejowski Department of Engineering, University of Cambridge Trumpington Street, Cambridge CB2 1PZ, United Kingdom
More information! # %& ( ) ) + (,./ & ,.0.4 %./ % 95 : 1.../9 1! # ; &+ ; + 3,. 0 < < : 1.../9 = < % = + 7 % 1 , < = =+> 7 %9 5 0
! # %& ( ) ) + (,./ 1 2 1 1 3&+ + 3 45,..4 %./4 6778. 4% 95 : 1.../9 1! # ; 1 2 1 1 3&+ ; + 3,. < < : 1.../9 = < % = + 7 % 1, < = =+> 7 %9 5? OPTIMAL CONTROL APPLICATIONS AND METHODS Optim. Control Appl.
More informationDisturbance Rejection and Set-point Tracking of Sinusoidal Signals using Generalized Predictive Control
Proceedings of the 47th IEEE Conference on Decision and Control Cancun, Mexico, Dec 9-, 28 Disturbance Rejection and Set-point Tracking of Sinusoidal Signals using Generalized Predictive Control Liuping
More informationA Globally Stabilizing Receding Horizon Controller for Neutrally Stable Linear Systems with Input Constraints 1
A Globally Stabilizing Receding Horizon Controller for Neutrally Stable Linear Systems with Input Constraints 1 Ali Jadbabaie, Claudio De Persis, and Tae-Woong Yoon 2 Department of Electrical Engineering
More informationIMPROVED MPC DESIGN BASED ON SATURATING CONTROL LAWS
IMPROVED MPC DESIGN BASED ON SATURATING CONTROL LAWS D. Limon, J.M. Gomes da Silva Jr., T. Alamo and E.F. Camacho Dpto. de Ingenieria de Sistemas y Automática. Universidad de Sevilla Camino de los Descubrimientos
More informationFurther results on Robust MPC using Linear Matrix Inequalities
Further results on Robust MPC using Linear Matrix Inequalities M. Lazar, W.P.M.H. Heemels, D. Muñoz de la Peña, T. Alamo Eindhoven Univ. of Technology, P.O. Box 513, 5600 MB, Eindhoven, The Netherlands,
More informationEnlarged terminal sets guaranteeing stability of receding horizon control
Enlarged terminal sets guaranteeing stability of receding horizon control J.A. De Doná a, M.M. Seron a D.Q. Mayne b G.C. Goodwin a a School of Electrical Engineering and Computer Science, The University
More informationFeasibility for non-square linear MPC
Feasibility for non-square linear MPC L.R.E. Shead, J.A Rossiter Abstract This paper investigates how the attributes of nonsquare systems affects feasibility in linear (MPC). Following on from the foundation
More informationImproved MPC Design based on Saturating Control Laws
Improved MPC Design based on Saturating Control Laws D.Limon 1, J.M.Gomes da Silva Jr. 2, T.Alamo 1 and E.F.Camacho 1 1. Dpto. de Ingenieria de Sistemas y Automática. Universidad de Sevilla, Camino de
More informationTheory in Model Predictive Control :" Constraint Satisfaction and Stability!
Theory in Model Predictive Control :" Constraint Satisfaction and Stability Colin Jones, Melanie Zeilinger Automatic Control Laboratory, EPFL Example: Cessna Citation Aircraft Linearized continuous-time
More informationFINITE HORIZON ROBUST MODEL PREDICTIVE CONTROL USING LINEAR MATRIX INEQUALITIES. Danlei Chu, Tongwen Chen, Horacio J. Marquez
FINITE HORIZON ROBUST MODEL PREDICTIVE CONTROL USING LINEAR MATRIX INEQUALITIES Danlei Chu Tongwen Chen Horacio J Marquez Department of Electrical and Computer Engineering University of Alberta Edmonton
More informationExplicit Robust Model Predictive Control
Explicit Robust Model Predictive Control Efstratios N. Pistikopoulos Nuno P. Faísca Konstantinos I. Kouramas Christos Panos Centre for Process Systems Engineering, Department of Chemical Engineering, Imperial
More informationIntroduction to Model Predictive Control. Dipartimento di Elettronica e Informazione
Introduction to Model Predictive Control Riccardo Scattolini Riccardo Scattolini Dipartimento di Elettronica e Informazione Finite horizon optimal control 2 Consider the system At time k we want to compute
More informationNonlinear Model Predictive Control for Periodic Systems using LMIs
Marcus Reble Christoph Böhm Fran Allgöwer Nonlinear Model Predictive Control for Periodic Systems using LMIs Stuttgart, June 29 Institute for Systems Theory and Automatic Control (IST), University of Stuttgart,
More informationLearning Model Predictive Control for Iterative Tasks: A Computationally Efficient Approach for Linear System
Learning Model Predictive Control for Iterative Tasks: A Computationally Efficient Approach for Linear System Ugo Rosolia Francesco Borrelli University of California at Berkeley, Berkeley, CA 94701, USA
More informationRegional Solution of Constrained LQ Optimal Control
Regional Solution of Constrained LQ Optimal Control José DeDoná September 2004 Outline 1 Recap on the Solution for N = 2 2 Regional Explicit Solution Comparison with the Maximal Output Admissible Set 3
More informationA new low-and-high gain feedback design using MPC for global stabilization of linear systems subject to input saturation
A new low-and-high gain feedbac design using MPC for global stabilization of linear systems subject to input saturation Xu Wang 1 Håvard Fjær Grip 1; Ali Saberi 1 Tor Arne Johansen Abstract In this paper,
More informationPrinciples of Optimal Control Spring 2008
MIT OpenCourseWare http://ocw.mit.edu 6.33 Principles of Optimal Control Spring 8 For information about citing these materials or our Terms of Use, visit: http://ocw.mit.edu/terms. 6.33 Lecture 6 Model
More informationPiecewise-affine functions: applications in circuit theory and control
Piecewise-affine functions: applications in circuit theory and control Tomaso Poggi Basque Center of Applied Mathematics Bilbao 12/04/2013 1/46 Outline 1 Embedded systems 2 PWA functions Definition Classes
More informationIMPLEMENTATIONS OF TRACKING MULTIPARAMETRIC PREDICTIVE CONTROLLER. Pregelj Boštjan, Gerkšič Samo. Jozef Stefan Institute, Ljubljana, Slovenia
IMPLEMENTATIONS OF TRACKING MULTIPARAMETRIC PREDICTIVE CONTROLLER Pregelj Boštjan, Gerkšič Samo Jozef Stefan Institute, Ljubljana, Slovenia Abstract: With the recently developed multi-parametric predictive
More informationPrashant Mhaskar, Nael H. El-Farra & Panagiotis D. Christofides. Department of Chemical Engineering University of California, Los Angeles
HYBRID PREDICTIVE OUTPUT FEEDBACK STABILIZATION OF CONSTRAINED LINEAR SYSTEMS Prashant Mhaskar, Nael H. El-Farra & Panagiotis D. Christofides Department of Chemical Engineering University of California,
More informationPredictive control of hybrid systems: Input-to-state stability results for sub-optimal solutions
Predictive control of hybrid systems: Input-to-state stability results for sub-optimal solutions M. Lazar, W.P.M.H. Heemels a a Eindhoven Univ. of Technology, P.O. Box 513, 5600 MB Eindhoven, The Netherlands
More informationStatic Output Feedback Stabilisation with H Performance for a Class of Plants
Static Output Feedback Stabilisation with H Performance for a Class of Plants E. Prempain and I. Postlethwaite Control and Instrumentation Research, Department of Engineering, University of Leicester,
More informationMATH4406 (Control Theory) Unit 6: The Linear Quadratic Regulator (LQR) and Model Predictive Control (MPC) Prepared by Yoni Nazarathy, Artem
MATH4406 (Control Theory) Unit 6: The Linear Quadratic Regulator (LQR) and Model Predictive Control (MPC) Prepared by Yoni Nazarathy, Artem Pulemotov, September 12, 2012 Unit Outline Goal 1: Outline linear
More informationA CHANCE CONSTRAINED ROBUST MPC AND APPLICATION TO HOT ROLLING MILLS
A CHANCE CONSTRAINED ROBUST MPC AND APPLICATION TO HOT ROLLING MILLS Il Seop Choi Anthony Rossiter Peter Fleming University of Sheffield, Department of Automatic Control & Systems Engineering, Sheffield
More informationOptimal and suboptimal event-triggering in linear model predictive control
Preamble. This is a reprint of the article: M. Jost, M. Schulze Darup and M. Mönnigmann. Optimal and suboptimal eventtriggering in linear model predictive control. In Proc. of the 25 European Control Conference,
More informationA SIMPLE TUBE CONTROLLER FOR EFFICIENT ROBUST MODEL PREDICTIVE CONTROL OF CONSTRAINED LINEAR DISCRETE TIME SYSTEMS SUBJECT TO BOUNDED DISTURBANCES
A SIMPLE TUBE CONTROLLER FOR EFFICIENT ROBUST MODEL PREDICTIVE CONTROL OF CONSTRAINED LINEAR DISCRETE TIME SYSTEMS SUBJECT TO BOUNDED DISTURBANCES S. V. Raković,1 D. Q. Mayne Imperial College London, London
More informationA Stable Block Model Predictive Control with Variable Implementation Horizon
American Control Conference June 8-,. Portland, OR, USA WeB9. A Stable Block Model Predictive Control with Variable Implementation Horizon Jing Sun, Shuhao Chen, Ilya Kolmanovsky Abstract In this paper,
More informationMPC for tracking periodic reference signals
MPC for tracking periodic reference signals D. Limon T. Alamo D.Muñoz de la Peña M.N. Zeilinger C.N. Jones M. Pereira Departamento de Ingeniería de Sistemas y Automática, Escuela Superior de Ingenieros,
More informationPostface to Model Predictive Control: Theory and Design
Postface to Model Predictive Control: Theory and Design J. B. Rawlings and D. Q. Mayne August 19, 2012 The goal of this postface is to point out and comment upon recent MPC papers and issues pertaining
More informationChapter 2 Optimal Control Problem
Chapter 2 Optimal Control Problem Optimal control of any process can be achieved either in open or closed loop. In the following two chapters we concentrate mainly on the first class. The first chapter
More informationComplexity Reduction in Explicit MPC through Model Reduction
Proceedings of the 17th World Congress The International Federation of Automatic Control Seoul, Korea, July 6-11, 28 Complexity Reduction in Explicit MPC through Model Reduction Svein Hovland Jan Tommy
More informationNonlinear Reference Tracking with Model Predictive Control: An Intuitive Approach
onlinear Reference Tracking with Model Predictive Control: An Intuitive Approach Johannes Köhler, Matthias Müller, Frank Allgöwer Abstract In this paper, we study the system theoretic properties of a reference
More informationOn the Inherent Robustness of Suboptimal Model Predictive Control
On the Inherent Robustness of Suboptimal Model Predictive Control James B. Rawlings, Gabriele Pannocchia, Stephen J. Wright, and Cuyler N. Bates Department of Chemical & Biological Engineering Computer
More information4F3 - Predictive Control
4F3 Predictive Control - Lecture 3 p 1/21 4F3 - Predictive Control Lecture 3 - Predictive Control with Constraints Jan Maciejowski jmm@engcamacuk 4F3 Predictive Control - Lecture 3 p 2/21 Constraints on
More informationarxiv: v1 [cs.sy] 28 May 2013
From Parametric Model-based Optimization to robust PID Gain Scheduling Minh H.. Nguyen a,, K.K. an a a National University of Singapore, Department of Electrical and Computer Engineering, 3 Engineering
More informationESC794: Special Topics: Model Predictive Control
ESC794: Special Topics: Model Predictive Control Discrete-Time Systems Hanz Richter, Professor Mechanical Engineering Department Cleveland State University Discrete-Time vs. Sampled-Data Systems A continuous-time
More informationAn SVD based strategy for receding horizon control of input constrained linear systems
An SVD based strategy for receding horizon control of input constrained linear systems Osvaldo J. Rojas, Graham C. Goodwin, María M. Serón and Arie Feuer School of Electrical Engineering & Computer Science
More informationThe ϵ-capacity of a gain matrix and tolerable disturbances: Discrete-time perturbed linear systems
IOSR Journal of Mathematics (IOSR-JM) e-issn: 2278-5728, p-issn: 2319-765X. Volume 11, Issue 3 Ver. IV (May - Jun. 2015), PP 52-62 www.iosrjournals.org The ϵ-capacity of a gain matrix and tolerable disturbances:
More informationECE7850 Lecture 8. Nonlinear Model Predictive Control: Theoretical Aspects
ECE7850 Lecture 8 Nonlinear Model Predictive Control: Theoretical Aspects Model Predictive control (MPC) is a powerful control design method for constrained dynamical systems. The basic principles and
More informationA FAST, EASILY TUNED, SISO, MODEL PREDICTIVE CONTROLLER. Gabriele Pannocchia,1 Nabil Laachi James B. Rawlings
A FAST, EASILY TUNED, SISO, MODEL PREDICTIVE CONTROLLER Gabriele Pannocchia, Nabil Laachi James B. Rawlings Department of Chemical Engineering Univ. of Pisa Via Diotisalvi 2, 5626 Pisa (Italy) Department
More informationFinite horizon robust model predictive control with terminal cost constraints
Finite horizon robust model predictive control with terminal cost constraints Danlei Chu, Tongwen Chen and Horacio J Marquez Department of Electrical & Computer Engineering, University of Alberta, Canada,
More informationOnline Model Predictive Torque Control for Permanent Magnet Synchronous Motors
Online Model Predictive Torque Control for Permanent Magnet Synchronous Motors Gionata Cimini, Daniele Bernardini, Alberto Bemporad and Stephen Levijoki ODYS Srl General Motors Company 2015 IEEE International
More informationCourse on Model Predictive Control Part II Linear MPC design
Course on Model Predictive Control Part II Linear MPC design Gabriele Pannocchia Department of Chemical Engineering, University of Pisa, Italy Email: g.pannocchia@diccism.unipi.it Facoltà di Ingegneria,
More informationModel Predictive Control Based on Linear Programming The Explicit Solution
1974 IEEE TRANSACTIONS ON AUTOMATIC CONTROL, VOL. 47, NO. 12, DECEMBER 2002 Model Predictive Control Based on Linear Programming The Explicit Solution Alberto Bemporad, Francesco Borrelli, and Manfred
More informationAdaptive Nonlinear Model Predictive Control with Suboptimality and Stability Guarantees
Adaptive Nonlinear Model Predictive Control with Suboptimality and Stability Guarantees Pontus Giselsson Department of Automatic Control LTH Lund University Box 118, SE-221 00 Lund, Sweden pontusg@control.lth.se
More informationPREDICTIVE CONTROL OF NONLINEAR SYSTEMS. Received February 2008; accepted May 2008
ICIC Express Letters ICIC International c 2008 ISSN 1881-803X Volume 2, Number 3, September 2008 pp. 239 244 PREDICTIVE CONTROL OF NONLINEAR SYSTEMS Martin Janík, Eva Miklovičová and Marián Mrosko Faculty
More informationRobust Observer for Uncertain T S model of a Synchronous Machine
Recent Advances in Circuits Communications Signal Processing Robust Observer for Uncertain T S model of a Synchronous Machine OUAALINE Najat ELALAMI Noureddine Laboratory of Automation Computer Engineering
More informationConstrained Control of Uncertain, Time-varying Linear Discrete-Time Systems Subject to Bounded Disturbances
Constrained Control of Uncertain, ime-varying Linear Discrete-ime Systems Subject to Bounded Disturbances Hoaï-Nam Nguyen, Sorin Olaru, Per-Olof Gutman, Morten Hovd o cite this version: Hoaï-Nam Nguyen,
More informationCOMPUTATIONAL DELAY IN NONLINEAR MODEL PREDICTIVE CONTROL. Rolf Findeisen Frank Allgöwer
COMPUTATIONAL DELAY IN NONLINEAR MODEL PREDICTIVE CONTROL Rolf Findeisen Frank Allgöwer Institute for Systems Theory in Engineering, University of Stuttgart, 70550 Stuttgart, Germany, findeise,allgower
More informationExplicit constrained nonlinear MPC
Tor A. Johansen Norwegian University of Science and Technology, Trondheim, Norway Main topics ffl Approximate explicit nonlinear MPC based on orthogonal search trees ffl Approximate explicit nonlinear
More informationConstrained Linear Quadratic Optimal Control
5 Constrained Linear Quadratic Optimal Control 51 Overview Up to this point we have considered rather general nonlinear receding horizon optimal control problems Whilst we have been able to establish some
More informationOutline. 1 Linear Quadratic Problem. 2 Constraints. 3 Dynamic Programming Solution. 4 The Infinite Horizon LQ Problem.
Model Predictive Control Short Course Regulation James B. Rawlings Michael J. Risbeck Nishith R. Patel Department of Chemical and Biological Engineering Copyright c 217 by James B. Rawlings Outline 1 Linear
More informationOn robustness of suboptimal min-max model predictive control *
Manuscript received June 5, 007; revised Sep., 007 On robustness of suboptimal min-max model predictive control * DE-FENG HE, HAI-BO JI, TAO ZHENG Department of Automation University of Science and Technology
More informationOn the stability of receding horizon control with a general terminal cost
On the stability of receding horizon control with a general terminal cost Ali Jadbabaie and John Hauser Abstract We study the stability and region of attraction properties of a family of receding horizon
More informationCONTROL SYSTEMS, ROBOTICS, AND AUTOMATION Vol. XI Model-Based Predictive Control - Edoardo Mosca
COTROL SYSTEMS, ROBOTICS, AD AUTOMATIO Vol. XI Model-Based Predictive Control - Edoardo Mosca MODEL-BASED PREDICTIVE COTROL Edoardo Mosca University of Florence, Italy Keywords: optimal control, linear
More informationDecentralized and distributed control
Decentralized and distributed control Centralized control for constrained discrete-time systems M. Farina 1 G. Ferrari Trecate 2 1 Dipartimento di Elettronica, Informazione e Bioingegneria (DEIB) Politecnico
More informationPredictive control for linear systems with delayed input subject to constraints
Proceedings of the 17th World Congress The International Federation of Automatic Control Predictive control for linear systems with delayed input subject to constraints S. Olaru, S.-I. Niculescu SUPELEC,
More informationOn the Stabilization of Neutrally Stable Linear Discrete Time Systems
TWCCC Texas Wisconsin California Control Consortium Technical report number 2017 01 On the Stabilization of Neutrally Stable Linear Discrete Time Systems Travis J. Arnold and James B. Rawlings Department
More informationA Model Predictive Control Scheme with Additional Performance Index for Transient Behavior
A Model Predictive Control Scheme with Additional Performance Index for Transient Behavior Andrea Alessandretti, António Pedro Aguiar and Colin N. Jones Abstract This paper presents a Model Predictive
More informationDistributed and Real-time Predictive Control
Distributed and Real-time Predictive Control Melanie Zeilinger Christian Conte (ETH) Alexander Domahidi (ETH) Ye Pu (EPFL) Colin Jones (EPFL) Challenges in modern control systems Power system: - Frequency
More informationThe Partial Enumeration Method for Model Predictive Control: Algorithm and Examples
2 TWMCC Texas-Wisconsin Modeling and Control Consortium 1 Technical report number 2006-01 The Partial Enumeration Method for Model Predictive Control: Algorithm and Examples Gabriele Pannocchia Department
More informationIEOR 265 Lecture 14 (Robust) Linear Tube MPC
IEOR 265 Lecture 14 (Robust) Linear Tube MPC 1 LTI System with Uncertainty Suppose we have an LTI system in discrete time with disturbance: x n+1 = Ax n + Bu n + d n, where d n W for a bounded polytope
More informationExplicit Model Predictive Control for Linear Parameter-Varying Systems
Proceedings of the 47th IEEE Conference on Decision and Control Cancun, Mexico, Dec. 9-11, 2008 Explicit Model Predictive Control for Linear Parameter-Varying Systems Thomas Besselmann, Johan Löfberg and
More informationThe Explicit Solution of Model Predictive Control via Multiparametric Quadratic Programming
Proceedings of the American Control Conference Chicago, Illinois June 2000 The Explicit Solution of Model Predictive Control via Multiparametric Quadratic Programming Alberto Bemporad t, Manfred Morari
More informationRobust Model Predictive Control of Heat Exchangers
A publication of CHEMICAL EGIEERIG RASACIOS VOL. 9, 01 Guest Editors: Petar Sabev Varbanov, Hon Loong Lam, Jiří Jaromír Klemeš Copyright 01, AIDIC Servizi S.r.l., ISB 978-88-95608-0-4; ISS 1974-9791 he
More informationStability analysis of constrained MPC with CLF applied to discrete-time nonlinear system
. RESEARCH PAPER. SCIENCE CHINA Information Sciences November 214, Vol. 57 11221:1 11221:9 doi: 1.17/s11432-14-5111-y Stability analysis of constrained MPC with CLF applied to discrete-time nonlinear system
More informationGLOBAL STABILIZATION OF THE INVERTED PENDULUM USING MODEL PREDICTIVE CONTROL. L. Magni, R. Scattolini Λ;1 K. J. Åström ΛΛ
Copyright 22 IFAC 15th Triennial World Congress, Barcelona, Spain GLOBAL STABILIZATION OF THE INVERTED PENDULUM USING MODEL PREDICTIVE CONTROL L. Magni, R. Scattolini Λ;1 K. J. Åström ΛΛ Λ Dipartimento
More informationGiulio Betti, Marcello Farina and Riccardo Scattolini
1 Dipartimento di Elettronica e Informazione, Politecnico di Milano Rapporto Tecnico 2012.29 An MPC algorithm for offset-free tracking of constant reference signals Giulio Betti, Marcello Farina and Riccardo
More informationLinear Model Predictive Control via Multiparametric Programming
3 1 Linear Model Predictive Control via Multiparametric Programming Vassilis Sakizlis, Konstantinos I. Kouramas, and Efstratios N. Pistikopoulos 1.1 Introduction Linear systems with input, output, or state
More informationModel Predictive Controller of Boost Converter with RLE Load
Model Predictive Controller of Boost Converter with RLE Load N. Murali K.V.Shriram S.Muthukumar Nizwa College of Vellore Institute of Nizwa College of Technology Technology University Technology Ministry
More informationLecture Note 5: Semidefinite Programming for Stability Analysis
ECE7850: Hybrid Systems:Theory and Applications Lecture Note 5: Semidefinite Programming for Stability Analysis Wei Zhang Assistant Professor Department of Electrical and Computer Engineering Ohio State
More informationStability of linear time-varying systems through quadratically parameter-dependent Lyapunov functions
Stability of linear time-varying systems through quadratically parameter-dependent Lyapunov functions Vinícius F. Montagner Department of Telematics Pedro L. D. Peres School of Electrical and Computer
More informationOn the Inherent Robustness of Suboptimal Model Predictive Control
On the Inherent Robustness of Suboptimal Model Predictive Control James B. Rawlings, Gabriele Pannocchia, Stephen J. Wright, and Cuyler N. Bates Department of Chemical and Biological Engineering and Computer
More informationPredictive control for general nonlinear systems using approximation
Loughborough University Institutional Repository Predictive control for general nonlinear systems using approximation This item was submitted to Loughborough University's Institutional Repository by the/an
More informationJUSTIFICATION OF INPUT AND OUTPUT CONSTRAINTS INCORPORATION INTO PREDICTIVE CONTROL DESIGN
JUSTIFICATION OF INPUT AND OUTPUT CONSTRAINTS INCORPORATION INTO PREDICTIVE CONTROL DESIGN J. Škultéty, E. Miklovičová, M. Mrosko Slovak University of Technology, Faculty of Electrical Engineering and
More informationOffset Free Model Predictive Control
Proceedings of the 46th IEEE Conference on Decision and Control New Orleans, LA, USA, Dec. 12-14, 27 Offset Free Model Predictive Control Francesco Borrelli, Manfred Morari. Abstract This work addresses
More informationarxiv: v2 [cs.sy] 29 Mar 2016
Approximate Dynamic Programming: a Q-Function Approach Paul Beuchat, Angelos Georghiou and John Lygeros 1 ariv:1602.07273v2 [cs.sy] 29 Mar 2016 Abstract In this paper we study both the value function and
More informationAdaptive Dual Control
Adaptive Dual Control Björn Wittenmark Department of Automatic Control, Lund Institute of Technology Box 118, S-221 00 Lund, Sweden email: bjorn@control.lth.se Keywords: Dual control, stochastic control,
More informationA Method for Determining Stabilizeability of a Class of Switched Systems
Proceedings of the 7th WSEAS International Conference on Systems Theory and Scientific Computation Athens Greece August 4-6 007 7 A Method for Determining Stabilizeability of a Class of Switched Systems
More informationESTIMATES ON THE PREDICTION HORIZON LENGTH IN MODEL PREDICTIVE CONTROL
ESTIMATES ON THE PREDICTION HORIZON LENGTH IN MODEL PREDICTIVE CONTROL K. WORTHMANN Abstract. We are concerned with model predictive control without stabilizing terminal constraints or costs. Here, our
More informationRobust Model Predictive Control through Adjustable Variables: an application to Path Planning
43rd IEEE Conference on Decision and Control December 4-7, 24 Atlantis, Paradise Island, Bahamas WeC62 Robust Model Predictive Control through Adjustable Variables: an application to Path Planning Alessandro
More informationFast Model Predictive Control with Soft Constraints
European Control Conference (ECC) July 7-9,, Zürich, Switzerland. Fast Model Predictive Control with Soft Constraints Arthur Richards Department of Aerospace Engineering, University of Bristol Queens Building,
More informationScenario-Based Approach to Stochastic Linear Predictive Control
Scenario-Based Approach to Stochastic Linear Predictive Control Jadranko Matuško and Francesco Borrelli Abstract In this paper we consider the problem of predictive control for linear systems subject to
More informationCourse on Model Predictive Control Part III Stability and robustness
Course on Model Predictive Control Part III Stability and robustness Gabriele Pannocchia Department of Chemical Engineering, University of Pisa, Italy Email: g.pannocchia@diccism.unipi.it Facoltà di Ingegneria,
More informationECE7850 Lecture 9. Model Predictive Control: Computational Aspects
ECE785 ECE785 Lecture 9 Model Predictive Control: Computational Aspects Model Predictive Control for Constrained Linear Systems Online Solution to Linear MPC Multiparametric Programming Explicit Linear
More informationModel Predictive Control Short Course Regulation
Model Predictive Control Short Course Regulation James B. Rawlings Michael J. Risbeck Nishith R. Patel Department of Chemical and Biological Engineering Copyright c 2017 by James B. Rawlings Milwaukee,
More informationTube Model Predictive Control Using Homothety & Invariance
Tube Model Predictive Control Using Homothety & Invariance Saša V. Raković rakovic@control.ee.ethz.ch http://control.ee.ethz.ch/~srakovic Collaboration in parts with Mr. Mirko Fiacchini Automatic Control
More informationESC794: Special Topics: Model Predictive Control
ESC794: Special Topics: Model Predictive Control Nonlinear MPC Analysis : Part 1 Reference: Nonlinear Model Predictive Control (Ch.3), Grüne and Pannek Hanz Richter, Professor Mechanical Engineering Department
More informationWE CONSIDER linear systems subject to input saturation
440 IEEE TRANSACTIONS ON AUTOMATIC CONTROL, VOL 48, NO 3, MARCH 2003 Composite Quadratic Lyapunov Functions for Constrained Control Systems Tingshu Hu, Senior Member, IEEE, Zongli Lin, Senior Member, IEEE
More informationA Condensed and Sparse QP Formulation for Predictive Control
211 5th IEEE Conference on Decision and Control and European Control Conference (CDC-ECC) Orlando, FL, USA, December 12-15, 211 A Condensed and Sparse QP Formulation for Predictive Control Juan L Jerez,
More informationEfficient robust optimization for robust control with constraints Paul Goulart, Eric Kerrigan and Danny Ralph
Efficient robust optimization for robust control with constraints p. 1 Efficient robust optimization for robust control with constraints Paul Goulart, Eric Kerrigan and Danny Ralph Efficient robust optimization
More informationFast Algorithms for SDPs derived from the Kalman-Yakubovich-Popov Lemma
Fast Algorithms for SDPs derived from the Kalman-Yakubovich-Popov Lemma Venkataramanan (Ragu) Balakrishnan School of ECE, Purdue University 8 September 2003 European Union RTN Summer School on Multi-Agent
More informationModel Predictive Control of Nonlinear Systems: Stability Region and Feasible Initial Control
International Journal of Automation and Computing 04(2), April 2007, 195-202 DOI: 10.1007/s11633-007-0195-0 Model Predictive Control of Nonlinear Systems: Stability Region and Feasible Initial Control
More information1 Introduction 198; Dugard et al, 198; Dugard et al, 198) A delay matrix in such a lower triangular form is called an interactor matrix, and almost co
Multivariable Receding-Horizon Predictive Control for Adaptive Applications Tae-Woong Yoon and C M Chow y Department of Electrical Engineering, Korea University 1, -a, Anam-dong, Sungbu-u, Seoul 1-1, Korea
More informationDistributed Receding Horizon Control of Cost Coupled Systems
Distributed Receding Horizon Control of Cost Coupled Systems William B. Dunbar Abstract This paper considers the problem of distributed control of dynamically decoupled systems that are subject to decoupled
More information