Optimization with piecewise-affine cost functions

Size: px
Start display at page:

Download "Optimization with piecewise-affine cost functions"

Transcription

1 Optimization with piecewise-affine cost functions G Ferrari Trecate, Paolo Letizia, Matteo Spedicato Automatic Control Laboratory, Swiss Federal Institute of Technology (ETH), Zurich, Switzerland 21 June Technical Report AUT01-13

2 Abstract This report shows how it is possible to solve the problem of the minimization of an objective function composed by the sum of piece-wise affine functions as a Mixed Integer Linear Programming (MILP) problem The optimization problem min x m i=1 k i p i (x i ) Ax B, x = where p i (x i ) are piece-wise affine functions and A, B are the constraint matrices, will be reformulated as a MILP problem: x 1 x m min V KV F V G where K = [ k 1 k v ], V = and F, G are the constraint matrices In the first chapter of this report a simple case (cost function composed of only one piece-wise affine function) will be considered The generalization to the case of a cost function composed of m piece-wise affine functions will be contained in the second chapter v 1 v v

3 Chapter 1 Simple case Consider the following optimization problem: min x k p(x) Ax B (11) where k is a scalar coefficient, p(x) is a piece-wise affine function defined on the real variable x, A and B are suitable column vectors which define the feasible region for x The piece-wise affine function p(x) is composed by n affne functions, each one defined on a different domain: p(x) = a 1 x + b 1 if x [l 1, u 1 ] a 2 x + b 2 if x ]u 1, u 2 ] a n x + b n if x ]u n 1, u n ] (12) where the lower bound l 1 and the upper bound u n of the variable x are defined by the inequalities Ax B Note that p(x) could have jump discontinuity for x = u i, i = 1,, n 1, as shown in the Figure 11 assuming n = 3 We are interested to reformulate the optimization problem (11) as a Mixed Integer Linear Programming (MILP) problem: min V KV F V G (13) where K is the weights vector; V is the optimization variables (both continuous and logic) vector; F and G are the matrices which define the constraints of the vector V Note that: 1

4 Figure 11: Example of a piece-wise affine function with equation (2) assuming n = 3 K will contain the coefficient k; V will contain the variable x; F and G will contain the matrices A and B, where k, x, A and B are the same defined in (11) To derive a MILP problem we need to introduce some auxiliary logic and real variables First of all introduce n 1 binary variables, each one defined in the following way: δ 1 = 1 x u 1 δ 2 = 1 x u 2 (14) δ n 1 = 1 x u n 1 2

5 So, by giving a binary value to the variables δ i, i = 1,, n 1, it is possible to choose a particular region for the variable x For example, for the piecewiseaffine function reported in Figure 11 we will introduce two binary variables defined as below: δ1 = 1 x 4 δ2 = 1 x 12 and assuming for instance (δ1 = 1) (δ2 = 0), the region 4 x < 12 is defined Note that the variables δ i, i = 1,, n 1, are related by the logical equations δ i = 1 = δ j = 1 j < i, i = 2,, n 1 (15) which imply that not all the combinations of the variables δ i, i = 1,, n 1, are allowed Note also that the number of the relations (15) is equal to (n 2) + (n 3) Define now n 1 real variables z as below: z 1 = z 2 = z n 1 = if δ1 = 1 a 2 x + b 2 (16) else a 1 x + b 1 if δ2 = 1 (a 3 a 2 )x + (b 3 b 2 ) (17) else 0 if δn 1 = 1 (a n a n 1 )x + (b n b n 1 ) (18) else 0 and define the real variable z given by the sum of the variables z i, i = 1,, n 1: z = z 1 + z z n 1 (19) It s easy to verify that, depending on the feasible combination of the variables δ i, i = 1,, n 1, z will be equal to one of the n pieces of the function (12) 3

6 11 Inequalities The definition of the binary variables δ (14) and their relations (15), of the real variables z (16-18) and z (19) can be translated into mixed integer linear inequalities, as shown in [1], [2] and reported in Table 11, 12 inequalities n o ineq (14) D1 δ 1 (u 1 x) + D 1 (d 1 ε)δ 1 (u 1 x) ε D2 δ 2 (u 2 x) + D 2 (d 2 ε)δ 2 (u 2 x) ε Dn 1 δ n 1 (u n 1 x) + D n 1 (d n 1 ε)δ n 1 (u n 1 x) ε 2 (n 1) (15) δ i δ j 0 j < i, i = 2,, n 1 (n 2) + (n 3) Table 11: Translation in inequalities of the definition of the variables δ and of their logical relations In the Table 11 ε is a small tolerance (typically the machine precision); D i and d i, i = 1,, n 1, represent respectively the maximum and the minimum of (u i x), where x [l 1, u n ] Theoretically, an over[under] estimate of the maximum[minimum] suffices for our purpose, so that one can assume i = 1,, n 1 D i =max i (u i x) = max(u i ) min(x) = u n l 1 i = 1,, n 1 d i = min(u i x) = min(u i ) max(x) = u 1 u n In the Table 12 M 1, m 1 are respectively the maximum and the minimum of a 1 x + b 1, where x [l 1, u n ] ; M 2, m 2 are respectively the maximum and the minimum of a 2 x + b 2, where x [l 1, u n ] ; M j, m j, j = 3,, n, are respectively the maximum and the minimum of (a j a j 1 )x + (b j b j 1 ), where x [l 1, u n ] ; 4

7 (16) (17) (18) inequalities (M 2 m 1 )δ 1 z 1 a 2 x + (M 2 m 1 b 2 ) (M 1 m 2 )δ 1 + z 1 a 2 x + (M 1 m 2 + b 2 ) (m 2 M 1 )δ 1 z 1 a 1 x b 1 (m 1 M 2 )δ 1 + z 1 a 1 x + b 1 M 3 δ 2 + z 2 0 m 3 δ 2 z 2 0 m 3 δ 2 + z 2 [(a 3 a 2 )x + (b 3 b 2 )] m 3 M 3 δ 2 z 2 [(a 3 a 2 )x + (b 3 b 2 )] + M 3 M n δ n 1 + z n 1 0 m n δ n 1 z n 1 0 m n δ n 1 + z n 1 [(a n a n 1 )x + (b n b n 1 )] m n M n δ n 1 z n 1 [(a n a n 1 )x + (b n b n 1 )] + M n n o ineq 4 (n 1) (19) z z1 + z z n 1 z z 1 + z z n 1 2 Table 12: Translation in inequalities of the definition of the variables z and z ε is once more the machine precision As explained before, also in this case an over[under] estimate of the maximum[minimum] suffices for our purpose, so that one can assume a global maximum M = max(m i ) and a global minimum m = min(m i ), i = 1,, n and replace all the M i, m i, i = 1,, n listed before with them 12 Derivation of the MILP problem It is now possible to reformulate the initial problem min x k p(x) Ax B as the MILP problem 5

8 min V KV F V G where: V = z z 1 z n 1 x δ 1 δ n 1 ; K = [ k 0 0 ], where the number of 0 is equal to 2(n 1) + 1); F and G are matrices of suitable sizes such that they specify the inequalities Ax B and those contained in Tables 11, 12 6

9 Chapter 2 Generalization Based on the simple case explained before, it s easy to generalize the results to the case of a cost function composed of m piece-wise affine functions The extended optimization problem will be: min x m i=1 k i p i (x i ) Ax B, x = x 1 x m (21) where k i, i = 1,, m, are scalar coefficients, p i (x i ), i = 1,, m, are piecewise affine functions each one defined on the real variable x i, A and B are suitable matrices which define the feasible region for the real variables x i, i = 1, m The i th piece-wise affine function p i (x i ) is composed by n i affine functions: p i (x i ) = a i,1 x i + b i,1 if x i [l i,1, u i,1 ] a i,2 x i + b i,2 if x i ]u i,1, u i,2 ] a i,ni x i + b i,n if x i ]u i,n 1, u i,n ], i = 1,, m (22) Our aim is once more to obtain a MILP formulation for the problem (21) For this purpose we need to introduce auxiliary logic and real variables In particular, analogously to what has just been explained for the simple case, n i 1 logic and n i 1 real variables are needed for each p i (x i ) 7

10 Define the logic variables δ as follow: δ i,1 = 1 x i u i,1 δ i,2 = 1 x i u i,2 δ i,ni 1 = 1 x i u i,ni 1, i = 1,, m (23) The logic variables δ i,j, i = 1,, m, j = 1,, n i 1 are related by the following logical equations: δ i,j = 1 = δ i,s = 1 s < j, j = 2,, n i 1 i = 1,, m (24) The same conclusions of the simple case are valid i, i = 1,, m Define now the real variables z as follows: z i,1 = z i,2 = z i,ni 1 = if δi,1 = 1 a i,2 x i + b i,2 (25) else a i,1 x i + b i,1 if δi,2 = 1 (a i,3 a i,2 )x i + (b i,3 b i,2 ) (26) else 0 if δi,ni 1 = 1 (a i,ni a i,ni 1)x i + (b i,ni b i,ni 1) (27) else 0 and define the real variable z i given by the sum of the variables z i,j, j = 1,, n i 1: z i = z i,1 + z i,2 + + z i,ni 1 (28) 21 Inequalities Analogously to the simple case the definition of the binary variables δ (23) and their relations (24), of the real variables z (25-27) and z (28) can be translated into mixed integer linear inequalities The results of this procedure are shown in the Tables 21, 22 8

11 inequalities n o ineq (23) Di,1 δ i,1 (u i,1 x i ) + D i,1 (d i,1 ε)δ i,1 (u i,1 x i ) ε Di,2 δ i,2 (u i,2 x i ) + D i,2 (d i,2 ε)δ i,2 (u i,2 x i ) ε Di,ni 1δ i,ni 1 (u i,ni 1 x i ) + D i,ni 1 (d i,ni 1 ε)δ i,ni 1 (u i,ni 1 x i ) ε 2 (n i 1) (24) δ i,j δ i,s 0 s < j, j = 2,, n i 1 (n i 2) + (n i 3) Table 21: Translation in inequalities of the definition of the variables δ and of their logical relations (25) (26) (27) (28) inequalities (M i,2 m i,1 )δ i,1 z i,1 a i,2 x i + (M i,2 m i,1 b i,2 ) (M i,1 m i,2 )δ i,1 + z i,1 a i,2 x i + (M i,1 m i,2 + b i,2 ) (m i,2 M i,1 )δ i,1 z i,1 a i,1 x i b i,1 (m i,1 M i,2 )δ i,1 + z i,1 a i,1 x i + b i,1 M i,3 δ i,2 + z i,2 0 m i,3 δ i,2 z i,2 0 m i,3 δ i,2 + z i,2 [(a i,3 a i,2 )x i + (b i,3 b i,2 )] m i,3 M i,3 δ i,2 z i,2 [(a i,3 a i,2 )x i + (b i,3 b i,2 )] + M i,3 M i,ni δ i,ni 1 + z i,ni 1 0 m i,ni δ i,ni 1 z i,ni 1 0 m i,ni δ i,ni 1 + z i,ni 1 [(a i,ni a i,ni 1)x i + (b i,ni b i,ni 1)] m i,ni M i,ni δ i,ni 1 z i,ni 1 [(a i,ni a i,ni 1)x i + (b i,ni b i,ni 1) + M i,ni z i z i,1 + z i,2 + + z i,ni 1 z i z i,1 + z i,2 + + z i,ni 1 Table 22: Translation in inequalities of the definition of the variables z and z 9

12 Fixed i, the number of inequalities which translate the equations (25-27) is 4 (n i 1), and the number of inequalities which translate the equations (28) is 2 Moreover, fixed i, the definitions and the considerations about the coefficients ε, M, m, D, d are the same reported for the simple case 22 Derivation of the MILP problem It is now possible to reformulate the initial problem min x m i=1 k i p i (x i ) Ax B, x = x 1 x m as the MILP problem where: [ ] V V =, V min V KV F V G 10

13 V = z 1 z m z 1,1 z 1,n1 1 z 2,1 z 2,n2 1 z m,1 z m,nm 1 ; V = x 1 x m δ 1,1 δ 1,n1 1 δ 2,1 δ 2,n2 1 δ m,1 δ m,nm 1 ; K = [ k 1 k m 0 0 ], where the number of 0 is equal to m i=1 2(n i 1) + 1; F and G are matrices of suitable sizes such that they specify the inequalities Ax B and those contained in the tables 21 and References [1] Bemporad, A, & Morari, M Control of systems integrating logic, dynamics, and constraints Automatica, 35(3): , 1999 [2] Mignone, D The Really Big Collection Tecnical Report - IfA,

Lecture 2. MATH3220 Operations Research and Logistics Jan. 8, Pan Li The Chinese University of Hong Kong. Integer Programming Formulations

Lecture 2. MATH3220 Operations Research and Logistics Jan. 8, Pan Li The Chinese University of Hong Kong. Integer Programming Formulations Lecture 2 MATH3220 Operations Research and Logistics Jan. 8, 2015 Pan Li The Chinese University of Hong Kong 2.1 Agenda 1 2 3 2.2 : a linear program plus the additional constraints that some or all of

More information

Computational Integer Programming. Lecture 2: Modeling and Formulation. Dr. Ted Ralphs

Computational Integer Programming. Lecture 2: Modeling and Formulation. Dr. Ted Ralphs Computational Integer Programming Lecture 2: Modeling and Formulation Dr. Ted Ralphs Computational MILP Lecture 2 1 Reading for This Lecture N&W Sections I.1.1-I.1.6 Wolsey Chapter 1 CCZ Chapter 2 Computational

More information

X X 1. x (t,x0,u y ) x2 (t,x0,u y ) x-space x u(t)

X X 1. x (t,x0,u y ) x2 (t,x0,u y ) x-space x u(t) Piecewise Linear Optimal Controllers for Hybrid Systems Alberto Bemporad, Francesco Borrelli, Manfred Morari Automatic Control Laboratory ETH Zentrum, ETL I 6 CH-9 Zurich, Switzerland bemporad,borrelli,morari@aut.ee.ethz.ch

More information

An improved method for solving micro-ferry scheduling problems

An improved method for solving micro-ferry scheduling problems Delft University of Technology Delft Center for Systems and Control Technical report 12-028 An improved method for solving micro-ferry scheduling problems M. Burger, B. De Schutter, and H. Hellendoorn

More information

Gestion de la production. Book: Linear Programming, Vasek Chvatal, McGill University, W.H. Freeman and Company, New York, USA

Gestion de la production. Book: Linear Programming, Vasek Chvatal, McGill University, W.H. Freeman and Company, New York, USA Gestion de la production Book: Linear Programming, Vasek Chvatal, McGill University, W.H. Freeman and Company, New York, USA 1 Contents 1 Integer Linear Programming 3 1.1 Definitions and notations......................................

More information

MODEL PREDICTIVE CONTROL - IDEAS FOR THE NEXT GENERATION

MODEL PREDICTIVE CONTROL - IDEAS FOR THE NEXT GENERATION MODEL PREDICTIVE CONTROL - IDEAS FOR THE NEXT GENERATION A. Bemporad, G. Ferrari-Trecate, D. Mignone, M. Morari, F. D. Torrisi Institut für Automatik ETH Swiss Federal Institute of Technology CH-892 Zürich,

More information

Modeling and control of a multi-agent system using mixed integer linear programming

Modeling and control of a multi-agent system using mixed integer linear programming Modeling and control of a multi-agent system using mixed integer linear programming Matthew G. Earl 1 Raffaello D Andrea Abstract The RoboFlag competition was proposed by the second author as a means to

More information

A 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 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 information

Multicopter Design Optimization and Validation

Multicopter Design Optimization and Validation Modeling, Identification and Control, Vol 36, No 2, 2015, pp 67 79, ISSN 1890 1328 Multicopter Design Optimization and Validation Øyvind Magnussen Morten Ottestad Geir Hovland University of Agder, Faculty

More information

The ϵ-capacity of a gain matrix and tolerable disturbances: Discrete-time perturbed linear systems

The 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 information

Model Predictive Control for continuous piecewise affine systems using optimistic optimization

Model Predictive Control for continuous piecewise affine systems using optimistic optimization Delft University of Technology Model Predictive Control for continuous piecewise affine systems using optimistic optimization Xu, Jia; van den Bm, Ton; Busoniu, L; De Schutter, Bart DOI.9/ACC.26.752658

More information

Observability and Controllability of Piecewise Affine and Hybrid Systems

Observability and Controllability of Piecewise Affine and Hybrid Systems 1864 IEEE TRANSACTIONS ON AUTOMATIC CONTROL, VOL. 45, NO. 10, OCTOBER 2000 Observability and Controllability of Piecewise Affine and Hybrid Systems Alberto Bemporad, Giancarlo Ferrari-Trecate, and Manfred

More information

Output-Feedback Model Predictive Control of Sewer Networks Through Moving Horizon Estimation

Output-Feedback Model Predictive Control of Sewer Networks Through Moving Horizon Estimation Output-Feedback Model Predictive Control of Sewer Networks Through Moving Horizon Estimation Bernat Joseph-Duran, Carlos Ocampo-Martinez, Senior Member, IEEE, and Gabriela Cembrano Abstract Based on a

More information

Announcements. Review: Lyap. thms so far. Review: Multiple Lyap. Fcns. Discrete-time PWL/PWA Quadratic Lyapunov Theory

Announcements. Review: Lyap. thms so far. Review: Multiple Lyap. Fcns. Discrete-time PWL/PWA Quadratic Lyapunov Theory EECE 571M/491M, Spring 2007 Lecture 11 Discrete-time PWL/PWA Quadratic Lyapunov Theory http://www.ece.ubc.ca/~elec571m.html moishi@ece.ubc.ca Meeko Oishi, Ph.D. Electrical and Computer Engineering University

More information

x y = x + y. For example, the tropical sum 4 9 = 4, and the tropical product 4 9 = 13.

x y = x + y. For example, the tropical sum 4 9 = 4, and the tropical product 4 9 = 13. Comment: Version 0.1 1 Introduction Tropical geometry is a relatively new area in mathematics. Loosely described, it is a piece-wise linear version of algebraic geometry, over a particular structure known

More information

Integer Programming: Cutting Planes

Integer Programming: Cutting Planes OptIntro 1 / 39 Integer Programming: Cutting Planes Eduardo Camponogara Department of Automation and Systems Engineering Federal University of Santa Catarina October 2016 OptIntro 2 / 39 Summary Introduction

More information

Model Predictive Control of Hybrid Systems by A. Bemporad Controllo di Processo e dei Sistemi di Produzione A.a.

Model Predictive Control of Hybrid Systems by A. Bemporad Controllo di Processo e dei Sistemi di Produzione A.a. Model Predictive Control of Hybrid Systems 2009 by A. Bemporad Controllo di Processo e dei Sistemi di Produzione A.a. 2008/09 1/52 Hybrid Control Problem hybrid process continuous inputs continuous states

More information

Observation of Hybrid Guarded Command Programs

Observation of Hybrid Guarded Command Programs Observation of Hybrid Guarded Command Programs Domitilla Del Vecchio 1 Eric Klavins 2 Division of Engineering and Applied Science California Institute of Technology {ddomitilla,klavins}@cds.caltech.edu

More information

Computational Integer Programming Universidad de los Andes. Lecture 1. Dr. Ted Ralphs

Computational Integer Programming Universidad de los Andes. Lecture 1. Dr. Ted Ralphs Computational Integer Programming Universidad de los Andes Lecture 1 Dr. Ted Ralphs MIP Lecture 1 1 Quick Introduction Bio Course web site Course structure http://coral.ie.lehigh.edu/ ted/teaching/mip

More information

Mingbin Feng, John E. Mitchell, Jong-Shi Pang, Xin Shen, Andreas Wächter

Mingbin Feng, John E. Mitchell, Jong-Shi Pang, Xin Shen, Andreas Wächter Complementarity Formulations of l 0 -norm Optimization Problems 1 Mingbin Feng, John E. Mitchell, Jong-Shi Pang, Xin Shen, Andreas Wächter Abstract: In a number of application areas, it is desirable to

More information

Complementarity Formulations of l 0 -norm Optimization Problems

Complementarity Formulations of l 0 -norm Optimization Problems Complementarity Formulations of l 0 -norm Optimization Problems Mingbin Feng, John E. Mitchell, Jong-Shi Pang, Xin Shen, Andreas Wächter May 17, 2016 Abstract In a number of application areas, it is desirable

More information

Supplementary Material for MTH 299 Online Edition

Supplementary Material for MTH 299 Online Edition Supplementary Material for MTH 299 Online Edition Abstract This document contains supplementary material, such as definitions, explanations, examples, etc., to complement that of the text, How to Think

More information

Complementarity Formulations of l 0 -norm Optimization Problems

Complementarity Formulations of l 0 -norm Optimization Problems Complementarity Formulations of l 0 -norm Optimization Problems Mingbin Feng, John E. Mitchell,Jong-Shi Pang, Xin Shen, Andreas Wächter Original submission: September 23, 2013. Revised January 8, 2015

More information

PART 4 INTEGER PROGRAMMING

PART 4 INTEGER PROGRAMMING PART 4 INTEGER PROGRAMMING 102 Read Chapters 11 and 12 in textbook 103 A capital budgeting problem We want to invest $19 000 Four investment opportunities which cannot be split (take it or leave it) 1.

More information

Predictive 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 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 information

Probabilistic Controllability Analysis of Sampled-Data/Discrete-Time Piecewise Affine Systems

Probabilistic Controllability Analysis of Sampled-Data/Discrete-Time Piecewise Affine Systems Probabilistic Controllability Analysis of Sampled-Data/Discrete-Time Piecewise Affine Systems Shun-ichi Azuma Jun-ichi Imura Toyo Institute of Technology; 2-12-1, O-oayama, Meguro-u, Toyo 152-8552, Japan

More information

ESSENTIALS OF ALGEBRA II

ESSENTIALS OF ALGEBRA II ESSENTIALS OF ALGEBRA II Grades 11-12 Draft: January 2003 Killingly Public Schools Essentials of Algebra II Grades 11-12 Mathematical Models and Matrices CONTENT STANDARD 11-12 EAII 1: The student will

More information

MVE165/MMG630, Applied Optimization Lecture 6 Integer linear programming: models and applications; complexity. Ann-Brith Strömberg

MVE165/MMG630, Applied Optimization Lecture 6 Integer linear programming: models and applications; complexity. Ann-Brith Strömberg MVE165/MMG630, Integer linear programming: models and applications; complexity Ann-Brith Strömberg 2011 04 01 Modelling with integer variables (Ch. 13.1) Variables Linear programming (LP) uses continuous

More information

What is an integer program? Modelling with Integer Variables. Mixed Integer Program. Let us start with a linear program: max cx s.t.

What is an integer program? Modelling with Integer Variables. Mixed Integer Program. Let us start with a linear program: max cx s.t. Modelling with Integer Variables jesla@mandtudk Department of Management Engineering Technical University of Denmark What is an integer program? Let us start with a linear program: st Ax b x 0 where A

More information

Preprocessing. Complements of Operations Research. Giovanni Righini. Università degli Studi di Milano

Preprocessing. Complements of Operations Research. Giovanni Righini. Università degli Studi di Milano Preprocessing Complements of Operations Research Giovanni Righini Università degli Studi di Milano Preprocessing Computational complexity theory classifies problems. However, when we run algorithms, this

More information

arxiv: v2 [math.oc] 27 Aug 2018

arxiv: v2 [math.oc] 27 Aug 2018 From Uncertainty Data to Robust Policies for Temporal Logic Planning arxiv:181.3663v2 [math.oc 27 Aug 218 ABSTRACT Pier Giuseppe Sessa Automatic Control Laboratory, ETH Zurich Zurich, Switzerland sessap@student.ethz.ch

More information

Identification of Piecewise Affine Systems via Mixed-Integer Programming

Identification of Piecewise Affine Systems via Mixed-Integer Programming Identification of Piecewise Affine Systems via Mixed-Integer Programming Jacob Roll, Alberto Bemporad, Lennart Ljung Division of Automatic Control Department of Electrical Engineering Linköpings universitet,

More information

1 The linear algebra of linear programs (March 15 and 22, 2015)

1 The linear algebra of linear programs (March 15 and 22, 2015) 1 The linear algebra of linear programs (March 15 and 22, 2015) Many optimization problems can be formulated as linear programs. The main features of a linear program are the following: Variables are real

More information

Piecewise Defined Functions

Piecewise Defined Functions Piecewise Defined Functions Most of the functions that we ve looked at this semester can be expressed as a single equation. For example, fx) =3x 2 5x +2,orgx) = p x 1, or hx) =e 3x 1. Sometimes an equation

More information

A hybrid MPC approach to the design of a Smart adaptive cruise controller

A hybrid MPC approach to the design of a Smart adaptive cruise controller Delft University of Technology Delft Center for Systems and Control Technical report 6-9 A hybrid MPC approach to the design of a Smart adaptive cruise controller D. Corona, M. Lazar, B. De Schutter, and

More information

2014 IEEE. Digital Object Identifier:

2014 IEEE. Digital Object Identifier: 214 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other users, including reprinting/ republishing this material for advertising or promotional purposes,

More information

ECE7850 Lecture 9. Model Predictive Control: Computational Aspects

ECE7850 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 information

2007/87. Valid inequalities for the single-item capacitated lot sizing problem with step-wise costs. Ayse Akbalik and Yves Pochet

2007/87. Valid inequalities for the single-item capacitated lot sizing problem with step-wise costs. Ayse Akbalik and Yves Pochet 2007/87 Valid inequalities for the single-item capacitated lot sizing problem with step-wise costs Ayse Akbalik and Yves Pochet CORE DISCUSSION PAPER 2007/87 Valid inequalities for the single-item capacitated

More information

Selected Examples of CONIC DUALITY AT WORK Robust Linear Optimization Synthesis of Linear Controllers Matrix Cube Theorem A.

Selected Examples of CONIC DUALITY AT WORK Robust Linear Optimization Synthesis of Linear Controllers Matrix Cube Theorem A. . Selected Examples of CONIC DUALITY AT WORK Robust Linear Optimization Synthesis of Linear Controllers Matrix Cube Theorem A. Nemirovski Arkadi.Nemirovski@isye.gatech.edu Linear Optimization Problem,

More information

Explicit Robust Model Predictive Control

Explicit 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 information

Section Notes 9. IP: Cutting Planes. Applied Math 121. Week of April 12, 2010

Section Notes 9. IP: Cutting Planes. Applied Math 121. Week of April 12, 2010 Section Notes 9 IP: Cutting Planes Applied Math 121 Week of April 12, 2010 Goals for the week understand what a strong formulations is. be familiar with the cutting planes algorithm and the types of cuts

More information

Improved Big-M Reformulation for Generalized Disjunctive Programs

Improved Big-M Reformulation for Generalized Disjunctive Programs Improved Big-M Reformulation for Generalized Disjunctive Programs Francisco Trespalacios and Ignacio E. Grossmann Department of Chemical Engineering Carnegie Mellon University, Pittsburgh, PA 15213 Author

More information

3.7 Strong valid inequalities for structured ILP problems

3.7 Strong valid inequalities for structured ILP problems 3.7 Strong valid inequalities for structured ILP problems By studying the problem structure, we can derive strong valid inequalities yielding better approximations of conv(x ) and hence tighter bounds.

More information

19. Logic constraints, integer variables

19. Logic constraints, integer variables CS/ECE/ISyE 524 Introduction to Optimization Spring 2016 17 19. Logic constraints, integer variables If-then constraints Generalized assignment problems Logic constraints Modeling a restricted set of values

More information

Error Analysis for Solving a Set of Linear Equations

Error Analysis for Solving a Set of Linear Equations Error Analysis for Solving a Set of Linear Equations We noted earlier that the solutions to an equation Ax = b depends significantly on the matrix A. In particular, a unique solution to Ax = b exists if

More information

arxiv: v1 [math.oc] 11 Jan 2018

arxiv: v1 [math.oc] 11 Jan 2018 From Uncertainty Data to Robust Policies for Temporal Logic Planning arxiv:181.3663v1 [math.oc 11 Jan 218 ABSTRACT Pier Giuseppe Sessa* Automatic Control Laboratory, ETH Zürich Zürich, Switzerland sessap@student.ethz.ch

More information

Integer Programming ISE 418. Lecture 16. Dr. Ted Ralphs

Integer Programming ISE 418. Lecture 16. Dr. Ted Ralphs Integer Programming ISE 418 Lecture 16 Dr. Ted Ralphs ISE 418 Lecture 16 1 Reading for This Lecture Wolsey, Chapters 10 and 11 Nemhauser and Wolsey Sections II.3.1, II.3.6, II.3.7, II.5.4 CCZ Chapter 8

More information

A Numerically Robust Mixed-Integer Quadratic Programming Solver for Embedded Hybrid Model Predictive Control

A Numerically Robust Mixed-Integer Quadratic Programming Solver for Embedded Hybrid Model Predictive Control Preprints, 6th IFAC Conference on Nonlinear Model Predictive Control A Numerically Robust Mixed-Integer Quadratic Programming Solver for Embedded Hybrid Model Predictive Control Alberto Bemporad Vihangkumar

More information

IMES - Center of Mechanics ETH Zentrum, Tannenstrasse 3 CH-8092 Zürich, Switzerland

IMES - Center of Mechanics ETH Zentrum, Tannenstrasse 3 CH-8092 Zürich, Switzerland Chapter 16 THE GEOMETRY OF NEWTON S CRADLE Christoph Glocker and Ueli Aeberhard IMES - Center of Mechanics ETH Zentrum, Tannenstrasse 3 CH-8092 Zürich, Switzerland glocker@imes.mavt.ethz.ch Abstract A

More information

Ma/CS 6a Class 19: Group Isomorphisms

Ma/CS 6a Class 19: Group Isomorphisms Ma/CS 6a Class 19: Group Isomorphisms By Adam Sheffer A Group A group consists of a set G and a binary operation, satisfying the following. Closure. For every x, y G x y G. Associativity. For every x,

More information

A Solution of a Tropical Linear Vector Equation

A Solution of a Tropical Linear Vector Equation A Solution of a Tropical Linear Vector Equation NIKOLAI KRIVULIN Faculty of Mathematics and Mechanics St. Petersburg State University 28 Universitetsky Ave., St. Petersburg, 198504 RUSSIA nkk@math.spbu.ru

More information

MATHEMATICS (MIDDLE GRADES AND EARLY SECONDARY)

MATHEMATICS (MIDDLE GRADES AND EARLY SECONDARY) MATHEMATICS (MIDDLE GRADES AND EARLY SECONDARY) l. Content Domain Mathematical Processes and Number Sense Range of Competencies Approximate Percentage of Test Score 0001 0003 24% ll. Patterns, Algebra,

More information

The Ellipsoid (Kachiyan) Method

The Ellipsoid (Kachiyan) Method Yinyu Ye, MS&E, Stanford MS&E310 Lecture Note: Ellipsoid Method 1 The Ellipsoid (Kachiyan) Method Yinyu Ye Department of Management Science and Engineering Stanford University Stanford, CA 94305, U.S.A.

More information

BMO Round 2 Problem 3 Generalisation and Bounds

BMO Round 2 Problem 3 Generalisation and Bounds BMO 2007 2008 Round 2 Problem 3 Generalisation and Bounds Joseph Myers February 2008 1 Introduction Problem 3 (by Paul Jefferys) is: 3. Adrian has drawn a circle in the xy-plane whose radius is a positive

More information

April 25 May 6, 2016, Verona, Italy. GAME THEORY and APPLICATIONS Mikhail Ivanov Krastanov

April 25 May 6, 2016, Verona, Italy. GAME THEORY and APPLICATIONS Mikhail Ivanov Krastanov April 25 May 6, 2016, Verona, Italy GAME THEORY and APPLICATIONS Mikhail Ivanov Krastanov Games in normal form There are given n-players. The set of all strategies (possible actions) of the i-th player

More information

Randomly switching max-plus linear systems and equivalent classes of discrete event systems

Randomly switching max-plus linear systems and equivalent classes of discrete event systems Delft University of Technology Delft Center for Systems and Control Technical report 8-8 Randomly switching max-plus linear systems and equivalent classes of discrete event systems T. van den Boom and

More information

Minimal Element Interpolation in Functions of High-Dimension

Minimal Element Interpolation in Functions of High-Dimension Minimal Element Interpolation in Functions of High-Dimension J. D. Jakeman, A. Narayan, D. Xiu Department of Mathematics Purdue University West Lafayette, Indiana Random Model as a function We can consider

More information

3.8 Strong valid inequalities

3.8 Strong valid inequalities 3.8 Strong valid inequalities By studying the problem structure, we can derive strong valid inequalities which lead to better approximations of the ideal formulation conv(x ) and hence to tighter bounds.

More information

where u is the decision-maker s payoff function over her actions and S is the set of her feasible actions.

where u is the decision-maker s payoff function over her actions and S is the set of her feasible actions. Seminars on Mathematics for Economics and Finance Topic 3: Optimization - interior optima 1 Session: 11-12 Aug 2015 (Thu/Fri) 10:00am 1:00pm I. Optimization: introduction Decision-makers (e.g. consumers,

More information

Technical Notes and Correspondence

Technical Notes and Correspondence 1108 IEEE RANSACIONS ON AUOMAIC CONROL, VOL. 47, NO. 7, JULY 2002 echnical Notes and Correspondence Stability Analysis of Piecewise Discrete-ime Linear Systems Gang Feng Abstract his note presents a stability

More information

3.10 Lagrangian relaxation

3.10 Lagrangian relaxation 3.10 Lagrangian relaxation Consider a generic ILP problem min {c t x : Ax b, Dx d, x Z n } with integer coefficients. Suppose Dx d are the complicating constraints. Often the linear relaxation and the

More information

Lecture 2: Haar Multiresolution analysis

Lecture 2: Haar Multiresolution analysis WAVELES AND MULIRAE DIGIAL SIGNAL PROCESSING Lecture 2: Haar Multiresolution analysis Prof.V. M. Gadre, EE, II Bombay 1 Introduction HAAR was a mathematician, who has given an idea that any continuous

More information

Higher Order ODE's, (3A)

Higher Order ODE's, (3A) Higher Order ODE's, (3A) Initial Value Problems, and Boundary Value Problems Copyright (c) 2011-2015 Young W. Lim. Permission is granted to copy, distribute and/or modify this document under the terms

More information

3. Convex functions. basic properties and examples. operations that preserve convexity. the conjugate function. quasiconvex functions

3. Convex functions. basic properties and examples. operations that preserve convexity. the conjugate function. quasiconvex functions 3. Convex functions Convex Optimization Boyd & Vandenberghe basic properties and examples operations that preserve convexity the conjugate function quasiconvex functions log-concave and log-convex functions

More information

Integer Programming ISE 418. Lecture 12. Dr. Ted Ralphs

Integer Programming ISE 418. Lecture 12. Dr. Ted Ralphs Integer Programming ISE 418 Lecture 12 Dr. Ted Ralphs ISE 418 Lecture 12 1 Reading for This Lecture Nemhauser and Wolsey Sections II.2.1 Wolsey Chapter 9 ISE 418 Lecture 12 2 Generating Stronger Valid

More information

3.3 Easy ILP problems and totally unimodular matrices

3.3 Easy ILP problems and totally unimodular matrices 3.3 Easy ILP problems and totally unimodular matrices Consider a generic ILP problem expressed in standard form where A Z m n with n m, and b Z m. min{c t x : Ax = b, x Z n +} (1) P(b) = {x R n : Ax =

More information

6.252 NONLINEAR PROGRAMMING LECTURE 10 ALTERNATIVES TO GRADIENT PROJECTION LECTURE OUTLINE. Three Alternatives/Remedies for Gradient Projection

6.252 NONLINEAR PROGRAMMING LECTURE 10 ALTERNATIVES TO GRADIENT PROJECTION LECTURE OUTLINE. Three Alternatives/Remedies for Gradient Projection 6.252 NONLINEAR PROGRAMMING LECTURE 10 ALTERNATIVES TO GRADIENT PROJECTION LECTURE OUTLINE Three Alternatives/Remedies for Gradient Projection Two-Metric Projection Methods Manifold Suboptimization Methods

More information

LINEAR INTERVAL INEQUALITIES

LINEAR INTERVAL INEQUALITIES LINEAR INTERVAL INEQUALITIES Jiří Rohn, Jana Kreslová Faculty of Mathematics and Physics, Charles University Malostranské nám. 25, 11800 Prague, Czech Republic 1.6.1993 Abstract We prove that a system

More information

Algebra II. A2.1.1 Recognize and graph various types of functions, including polynomial, rational, and algebraic functions.

Algebra II. A2.1.1 Recognize and graph various types of functions, including polynomial, rational, and algebraic functions. Standard 1: Relations and Functions Students graph relations and functions and find zeros. They use function notation and combine functions by composition. They interpret functions in given situations.

More information

MPC for continuous piecewise-affine systems

MPC for continuous piecewise-affine systems Delft University of Technology Fac. of Information Technology and Systems Control Systems Engineering Technical report CSE02-004 MPC for continuous piecewise-affine systems B. De Schutter and T.J.J. van

More information

Stability and Stabilizability of Switched Linear Systems: A Short Survey of Recent Results

Stability and Stabilizability of Switched Linear Systems: A Short Survey of Recent Results Proceedings of the 2005 IEEE International Symposium on Intelligent Control Limassol, Cyprus, June 27-29, 2005 MoA01-5 Stability and Stabilizability of Switched Linear Systems: A Short Survey of Recent

More information

Nonlinear convex and concave relaxations for the solutions of parametric ODEs

Nonlinear convex and concave relaxations for the solutions of parametric ODEs OPTIMAL CONTROL APPLICATIONS AND METHODS Optim. Control Appl. Meth. 0000; 00:1 22 Published online in Wiley InterScience (www.interscience.wiley.com). Nonlinear convex and concave relaxations for the solutions

More information

Range of Competencies

Range of Competencies MATHEMATICS l. Content Domain Range of Competencies Mathematical Processes and Number Sense 0001 0003 19% ll. Patterns, Algebra, and Functions 0004 0007 24% lll. Measurement and Geometry 0008 0010 19%

More information

XMA2C011, Annual Examination 2012: Worked Solutions

XMA2C011, Annual Examination 2012: Worked Solutions XMA2C011, Annual Examination 2012: Worked Solutions David R. Wilkins 1. (a) Let A, B and C be sets. Prove that A (B \ C) = (A B) \ (A C). We show that every element of A (B \ C) is an element of (A B)

More information

Model predictive control for max-min-plus-scaling systems

Model predictive control for max-min-plus-scaling systems Delft University of Technology Fac. of Information Technology and Systems Control Systems Engineering Technical report bds:00-03 Model predictive control for max-min-plus-scaling systems B. De Schutter

More information

Mathematics for Engineers. Numerical mathematics

Mathematics for Engineers. Numerical mathematics Mathematics for Engineers Numerical mathematics Integers Determine the largest representable integer with the intmax command. intmax ans = int32 2147483647 2147483647+1 ans = 2.1475e+09 Remark The set

More information

Analysis and Control of Nonlinear Actuator Dynamics Based on the Sum of Squares Programming Method

Analysis and Control of Nonlinear Actuator Dynamics Based on the Sum of Squares Programming Method Analysis and Control of Nonlinear Actuator Dynamics Based on the Sum of Squares Programming Method Balázs Németh and Péter Gáspár Abstract The paper analyses the reachability characteristics of the brake

More information

6.854J / J Advanced Algorithms Fall 2008

6.854J / J Advanced Algorithms Fall 2008 MIT OpenCourseWare http://ocw.mit.edu 6.854J / 18.415J Advanced Algorithms Fall 2008 For information about citing these materials or our Terms of Use, visit: http://ocw.mit.edu/terms. 18.415/6.854 Advanced

More information

Disturbance Attenuation Properties for Discrete-Time Uncertain Switched Linear Systems

Disturbance Attenuation Properties for Discrete-Time Uncertain Switched Linear Systems Disturbance Attenuation Properties for Discrete-Time Uncertain Switched Linear Systems Hai Lin Department of Electrical Engineering University of Notre Dame Notre Dame, IN 46556, USA Panos J. Antsaklis

More information

A Polytope for a Product of Real Linear Functions in 0/1 Variables

A Polytope for a Product of Real Linear Functions in 0/1 Variables Don Coppersmith Oktay Günlük Jon Lee Janny Leung A Polytope for a Product of Real Linear Functions in 0/1 Variables Original: 29 September 1999 as IBM Research Report RC21568 Revised version: 30 November

More information

X-MA2C01-1: Partial Worked Solutions

X-MA2C01-1: Partial Worked Solutions X-MAC01-1: Partial Worked Solutions David R. Wilkins May 013 1. (a) Let A, B and C be sets. Prove that (A \ (B C)) (B \ C) = (A B) \ C. [Venn Diagrams, by themselves without an accompanying logical argument,

More information

Subgradient methods for huge-scale optimization problems

Subgradient methods for huge-scale optimization problems Subgradient methods for huge-scale optimization problems Yurii Nesterov, CORE/INMA (UCL) May 24, 2012 (Edinburgh, Scotland) Yu. Nesterov Subgradient methods for huge-scale problems 1/24 Outline 1 Problems

More information

Determine the size of an instance of the minimum spanning tree problem.

Determine the size of an instance of the minimum spanning tree problem. 3.1 Algorithm complexity Consider two alternative algorithms A and B for solving a given problem. Suppose A is O(n 2 ) and B is O(2 n ), where n is the size of the instance. Let n A 0 be the size of the

More information

Utah Secondary Mathematics Core Curriculum Precalculus

Utah Secondary Mathematics Core Curriculum Precalculus A Correlation of Trigonometry Lial To the Utah Secondary Mathematics Core Curriculum Precalculus Resource Title: Trigonometry Publisher: Pearson Education Inc Publishing as Prentice Hall ISBN (10 or 13

More information

Mixed Integer Non Linear Programming

Mixed Integer Non Linear Programming Mixed Integer Non Linear Programming Claudia D Ambrosio CNRS Research Scientist CNRS & LIX, École Polytechnique MPRO PMA 2016-2017 Outline What is a MINLP? Dealing with nonconvexities Global Optimization

More information

Optimization Exercise Set n. 4 :

Optimization Exercise Set n. 4 : Optimization Exercise Set n. 4 : Prepared by S. Coniglio and E. Amaldi translated by O. Jabali 2018/2019 1 4.1 Airport location In air transportation, usually there is not a direct connection between every

More information

Course Outline. FRTN10 Multivariable Control, Lecture 13. General idea for Lectures Lecture 13 Outline. Example 1 (Doyle Stein, 1979)

Course Outline. FRTN10 Multivariable Control, Lecture 13. General idea for Lectures Lecture 13 Outline. Example 1 (Doyle Stein, 1979) Course Outline FRTN Multivariable Control, Lecture Automatic Control LTH, 6 L-L Specifications, models and loop-shaping by hand L6-L8 Limitations on achievable performance L9-L Controller optimization:

More information

Chapter 2: Linear Programming Basics. (Bertsimas & Tsitsiklis, Chapter 1)

Chapter 2: Linear Programming Basics. (Bertsimas & Tsitsiklis, Chapter 1) Chapter 2: Linear Programming Basics (Bertsimas & Tsitsiklis, Chapter 1) 33 Example of a Linear Program Remarks. minimize 2x 1 x 2 + 4x 3 subject to x 1 + x 2 + x 4 2 3x 2 x 3 = 5 x 3 + x 4 3 x 1 0 x 3

More information

SPSS, University of Texas at Arlington. Topics in Machine Learning-EE 5359 Neural Networks

SPSS, University of Texas at Arlington. Topics in Machine Learning-EE 5359 Neural Networks Topics in Machine Learning-EE 5359 Neural Networks 1 The Perceptron Output: A perceptron is a function that maps D-dimensional vectors to real numbers. For notational convenience, we add a zero-th dimension

More information

An RLT Approach for Solving Binary-Constrained Mixed Linear Complementarity Problems

An RLT Approach for Solving Binary-Constrained Mixed Linear Complementarity Problems An RLT Approach for Solving Binary-Constrained Mixed Linear Complementarity Problems Miguel F. Anjos Professor and Canada Research Chair Director, Trottier Institute for Energy TAI 2015 Washington, DC,

More information

Optimization in Process Systems Engineering

Optimization in Process Systems Engineering Optimization in Process Systems Engineering M.Sc. Jan Kronqvist Process Design & Systems Engineering Laboratory Faculty of Science and Engineering Åbo Akademi University Most optimization problems in production

More information

Vector spaces. EE 387, Notes 8, Handout #12

Vector spaces. EE 387, Notes 8, Handout #12 Vector spaces EE 387, Notes 8, Handout #12 A vector space V of vectors over a field F of scalars is a set with a binary operator + on V and a scalar-vector product satisfying these axioms: 1. (V, +) is

More information

(1) for all (2) for all and all

(1) for all (2) for all and all 8. Linear mappings and matrices A mapping f from IR n to IR m is called linear if it fulfills the following two properties: (1) for all (2) for all and all Mappings of this sort appear frequently in the

More information

The Law of Averages. MARK FLANAGAN School of Electrical, Electronic and Communications Engineering University College Dublin

The Law of Averages. MARK FLANAGAN School of Electrical, Electronic and Communications Engineering University College Dublin The Law of Averages MARK FLANAGAN School of Electrical, Electronic and Communications Engineering University College Dublin Basic Principle of Inequalities: For any real number x, we have 3 x 2 0, with

More information

Optimization Exercise Set n.5 :

Optimization Exercise Set n.5 : Optimization Exercise Set n.5 : Prepared by S. Coniglio translated by O. Jabali 2016/2017 1 5.1 Airport location In air transportation, usually there is not a direct connection between every pair of airports.

More information

4. Convex optimization problems (part 1: general)

4. Convex optimization problems (part 1: general) EE/AA 578, Univ of Washington, Fall 2016 4. Convex optimization problems (part 1: general) optimization problem in standard form convex optimization problems quasiconvex optimization 4 1 Optimization problem

More information

Technische Universität München, Zentrum Mathematik Lehrstuhl für Angewandte Geometrie und Diskrete Mathematik. Combinatorial Optimization (MA 4502)

Technische Universität München, Zentrum Mathematik Lehrstuhl für Angewandte Geometrie und Diskrete Mathematik. Combinatorial Optimization (MA 4502) Technische Universität München, Zentrum Mathematik Lehrstuhl für Angewandte Geometrie und Diskrete Mathematik Combinatorial Optimization (MA 4502) Dr. Michael Ritter Problem Sheet 1 Homework Problems Exercise

More information

More General Functions Is this technique limited to the monomials {1, x, x 2, x 3,...}?

More General Functions Is this technique limited to the monomials {1, x, x 2, x 3,...}? More General Functions Is this technique limited to the monomials {1, x, x 2, x 3,...}? Interpolation with General Sets of Functions For a general set of functions {ϕ 1,..., ϕ n }, solve the linear system

More information

Hybrid Systems: Modeling

Hybrid Systems: Modeling Hybrid Systems: Modeling Alberto Bemporad - Model Predictive Control course - Academic year 2017/18 (C) 2018 A. Bemporad - Model Predictive Control course 04-1 Hybrid dynamical systems 0 1 0 00 01 1 1

More information

February 13, Option 9 Overview. Mind Map

February 13, Option 9 Overview. Mind Map Option 9 Overview Mind Map Return tests - will discuss Wed..1.1 J.1: #1def,2,3,6,7 (Sequences) 1. Develop and understand basic ideas about sequences. J.2: #1,3,4,6 (Monotonic convergence) A quick review:

More information