Mixed-integer Programming

for Control
Arthur Richards
and
Jonathan How
 Motivation
• What is Mixed-integer Programming?
– MIP is an optimization method that combines
continuous and discrete variables

• Why is it useful?
– MIP can model complex planning and control
problems involving both continuous and discrete
decisions

• Why now? Is MIP new?

– MIP is not a new concept, BUT online use has only
arrived with fast computers and software

Mixed-integer Programming for Control 2/63

 Problem Classes
Nonlinear Non-convex Logical
Convex

T F

Mixed-integer Programming for Control 3/63

 Problem Classes
Nonlinear Non-convex Logical
Convex

T F

Nonlinear opt.
No MIP
Mixed-integer Programming for Control 4/63
 Problem Classes
Nonlinear Non-convex Logical
Convex

T F

Nonlinear opt. MIP

No MIP
Mixed-integer Programming for Control 5/63
 Problem Classes
Nonlinear Non-convex Logical
Convex

T F

Nonlinear opt. Approximate MIP

No MIP by MIP
Mixed-integer Programming for Control 6/63
 Problem Classes
Nonlinear Non-convex Logical
Convex

T F
Two main goals of MIP approach:
capture non-convexity and logic
T

Nonlinear opt. Approximate MIP

No MIP by MIP
Mixed-integer Programming for Control 7/63
 Problem Classes
Nonlinear Non-convex Logical
Convex

T F
Two main goals of MIP approach:
capture non-convexity and logic
T
• Encode non-convexity
F
using
integers
• Then apply powerful integer
Nonlinear opt. Approximate MIP
optimization tools to non-convex
No MIP by MIP
problems
Mixed-integer Programming for Control 8/63
 Session Outline (1/2)
1. Introduction to MIP
– MIP definition
– Modeling
• Assignment
• Non-convex constraints
• Piecewise affine systems
– Solving MIP
– Using optimization for feedback control
– Techniques for online solution of MIP
– Examples

Mixed-integer Programming for Control 9/63

 Session Outline (2/2)
2. Projected Variable Metric Algorithm

3. MILP Assignment for Multi-Vehicle

Systems

4. Real-time Path-Planning for Tactical

UAV

5. Receding Horizon Implementation of

MILP for Vehicle Guidance
Mixed-integer Programming for Control 10/63
 Web Resources
• Slides and sample codes available online
• Web resources
acl.mit.edu/MILP
seis.bris.ac.uk/~aeagr
• Email
[email protected]
[email protected]

Mixed-integer Programming for Control 11/63

 Formal Definition
• Mixed-integer Linear Program (MILP)

• Inherently non-convex
• NP-complete
BUT with good software, can find global-
optimum in many useful instances
Mixed-integer Programming for Control 12/63
Modelling using MIP
 MIP Modelling 1: PWA
[Example from Bersimas and Tsitsiklis]

• Constrain a = f(b) where f(.) is PWA

a

(B5,A5)
(B2,A2)

(B3,A3) (B4,A4)
(B1,A1)

b Binary zi = 1 if b ∈ [Ai, A(i+1)]

Mixed-integer Programming for Control 14/63
 MIP Modelling 1: PWA
[Example from Bersimas and Tsitsiklis]

• Constrain a = f(b) where f(.) is PWA

e.g. choose z3 = 1
all other zi = 0

Binary zi = 1 if b ∈ [Ai, A(i+1)]

Mixed-integer Programming for Control 15/63
 MIP Modelling 1: PWA
[Example from Bersimas and Tsitsiklis]

• Constrain a = f(b) where f(.) is PWA

e.g. choose z3 = 1
all other zi = 0

Binary zi = 1 if b ∈ [Ai, A(i+1)]

Mixed-integer Programming for Control 16/63
 MIP Modelling 1: PWA
[Example from Bersimas and Tsitsiklis]

• Constrain a = f(b) where f(.) is PWA

e.g. choose z3 = 1
all other zi = 0

Interpolation over [A3, A4]

Binary zi = 1 if b ∈ [Ai, A(i+1)]
Mixed-integer Programming for Control 17/63
 MIP Modelling 2: Disjunction
[Bersimas and Tsitsiklis, Morari, Brailsford and Williams]
• Require either a1x ≤ b1 or a2x ≤ b2

• MILP form

• M > 0, M >> a1x - b1, M >> a2x - b2

• Known as “big-M” method
Mixed-integer Programming for Control 18/63
 MIP Modelling 2: Disjunction
[Bersimas and Tsitsiklis, Morari, Brailsford and Williams]
• Require either a1x ≤ b1 or a2x ≤ b2
e.g. choose z1 = 1
• MILP form

requires z2 = 0

• M > 0, M >> a1x - b1, M >> a2x - b2

• Known as “big-M” method
Mixed-integer Programming for Control 19/63
 MIP Modelling 2: Disjunction
[Bersimas and Tsitsiklis, Morari, Brailsford and Williams]
• Require either a1x ≤ b1 or a2x ≤ b2
e.g. choose z1 = 1
• MILP form

requires z2 = 0

• M > 0, M >> a1x - b1, M >> a2x - b2

• Known as “big-M” method
Mixed-integer Programming for Control 20/63
 MIP Modelling 2: Disjunction
[Bersimas and Tsitsiklis, Morari, Brailsford and Williams]
• Require either a1x ≤ b1 or a2x ≤ b2
e.g. choose z1 = 1
• MILP form

requires z2 = 0

• M > 0, M >> a1x - b1, M >> a2x - b2

• Known as “big-M” method
Mixed-integer Programming for Control 21/63
 MIP Modelling 2: Disjunction
[Bersimas and Tsitsiklis, Morari, Brailsford and Williams]
• Require either a1x ≤ b1 or a2x ≤ b2
OR choose z2 = 1
• MILP form

requires z1 = 0

• M > 0, M >> a1x - b1, M >> a2x - b2

• Known as “big-M” method
Mixed-integer Programming for Control 22/63
 MIP Modelling 3: Avoidance
[Schouwenaars et al, Richards et al]

• Similar to disjunction
• Point (x, y) must be outside obstacle
x − x max ≥ − Mc 1
and x min − x ≥ − Mc 2 c1=0 ⇒
y (xmax , ymax)
and y − y max ≥ − Mc 3 c =0 ⇒
2
and y min − y ≥ − Mc 4 (xmin , ymin)
4 c3=0 ⇒ x
and ∑c
k =1
k ≤3
c4=0 ⇒

Mixed-integer Programming for Control 23/63

 MIP Modelling 4: Assignment
• Assign N tasks to N agents.
• Cost of assigning agent i to task j is cij

Binary zij = 1 if agent i

assigned to task j

• Special case – LP works

– Constrain 0 ≤ zij ≤ 1 : all vertices are integer

Mixed-integer Programming for Control 24/63

 MIP Modelling 4: Assignment
• Assign N tasks to N agents.
• Cost of assigning agent i to task j is cij

Binary zij = 1 if agent i

assigned to task j

• Add resource constraint – MILP needed

Mixed-integer Programming for Control 25/63

 MIP Modelling 5: Modes
• System has two modes
Mode 1

Mode 2

Mixed-integer Programming for Control 26/63

 MIP Modelling 5: Modes
• System has two modes
Mode 1

Mode 2

• MIP representation

Mixed-integer Programming for Control 27/63

 MIP Modelling 6 : Speed Limits
• 2-norm approximation
||v|| ≤ vmax : convex, easily handled

||v|| ≥ vmax : non-convex,

needs binaries

Mixed-integer Programming for Control 28/63

 MIP Modelling: Remarks
• Examples span many problem classes
– Combinations and extensions possible
• Joint assignment/path planning with avoidance
• PWA systems with disjunction constraints
• Logical constraints – “if A and B then C”

• There are often multiple ways of expressing a

problem using MIP
– Rule of thumb: big-M is nearly always an option, but
look for something better
– “Most existing MILP formulations that employ big-M
constraints do suffer from the poor relaxation
(relaxed MILP), which is a notorious feature of big-
M.”
• Improving Mixed Integer Linear Programming Formulations
A. Khurana, A. Sundaramoorthy and I. Karimi, AIChE, 2005

Mixed-integer Programming for Control 29/63

 Modelling References
• C. Floudas Nonlinear and Mixed-Integer Programming -
Fundamentals and Applications Oxford University
Press, 1995.

• A. Bemporad and M. Morari “Control of systems

integrating logic, dynamics, and constraints,”
Automatica, 35:407-427, 1999

• H. Williams and S. Brailsford, ``Computational Logic

and Integer Programming," in Advances in Linear and
Integer Programming, Editor J. E. Beasley, Clarendon
Press, 1996, pp.249-281.

• D. Bertsimas and J. N. Tsitsiklis, Introduction to Linear

Optimization, Athena Scientific, 1997.

Mixed-integer Programming for Control 30/63

MIP Solution
 MIP Solution: Branch & Bound
• Finds global optimum by tree search
1. Relax binary constraints zi ∈ {0,1}→0 ≤ zi ≤ 1
2. Solve relaxed problem (bounding)
3. Choose an i and (branching)
a) Fix zi = 0; go to 2;
b) Fix zi = 1; go to 2;

• Recursive tree search of binary options

• Can stop “early” by fathoming, if relaxation
– is infeasible, or
– gives binary result, or
– has worse cost than best binary so far
Mixed-integer Programming for Control 32/63
 MIP Solution: Branch & Bound
• Finds global optimum by tree search
1. Relax binary constraints zi ∈ {0,1}→0 ≤ zi ≤ 1
2. Solve relaxed problem (bounding)
3. Choose an i and (branching)
a) Fix zi = 0; go to 2;
b) Fix zi = 1; go to 2;

• Recursive tree search of binary options

• Can stop “early” by fathoming, if relaxation
– is infeasible, or
– gives binary result, or
– has worse cost than best binary so far.
Mixed-integer Programming for Control 33/63
 MIP Solution: AMPL
• AMPL easily translates models
– A Mathematical Programming Language
– Helps sort out indexing
– Interfaced to many solver codes
PWA var x{i in 1..N};
var z{i in 1..(N-1)} binary;
example
subject to acon: a = sum{i in 1..N} A[i]*x[i];
subject to bcon: b = sum{i in 1..N} B[i]*x[i];
subject to xsum: sum{i in 1..N} x[i] = 1;
subject to xpos{i in 1..N}: x[i] >= 0;
subject to x1: x[1] <= z[1];
subject to xi{i in 1..(N-1)}: x[i] <= z[i-1] + z[i];
subject to xN: x[N] <= z[N-1];
subject to zsum: sum{i in 1..(N-1)} = 1;

Mixed-integer Programming for Control 34/63

 MIP Solution: CPLEX
• Commercial solver code from ILOG

• Implements branch-and-bound in
conjunction with tried and tested
branching heuristics

• Interfaces
– AMPL
– Matlab MEX
– C API

Mixed-integer Programming for Control 35/63

 Other Software
• Decision tree for Optimization with discrete variables Online

• Modeling examples: MILP Model for Short Term Scheduling of Multistage

Batch Plants by Grossmann et al. Online
• Good comparison of non-commercial MILP software:

• MINOPT: A Modeling Language and Algorithmic Framework for Linear,

Mixed-Integer, Nonlinear, Dynamic, and Mixed-Integer Nonlinear
Optimization by Floudas et al. Online
• The Hybrid Systems Group
– Multi-Parametric Toolbox.

• Interface Software and example (Matlab ÅÆ AMPL ÅÆ CPLEX)

• AMPL: R. Fourer, D. M. Gay, and B. W. Kernighar, AMPL, A modeling

language for mathematical programming, The Scientific Press, 1993.

Mixed-integer Programming for Control 36/63

 MIP Solution: Remarks
• AMPL is good for prototyping, but more
direct interfaces are faster

• Other discrete optimization tools are

available
– Heuristic methods for general MIP
• Genetic algorithms
• Simulated annealing
– Special cases
• Dynamic programming for knapsack problem

Mixed-integer Programming for Control 37/63

MIP for Control
 Control using Optimization
• Have seen that MIP can find optimal solutions
for complex planning problems
• “Control” also considers uncertainty
– Introduce feedback to compensate
– Update plans to include new information
• Concept is the same as Model Predictive
Control (MPC)
1. Use numerical optimization to design an open-loop
control sequence for the future
2. Execute some initial portion of that sequence
3. Go to 1.

Mixed-integer Programming for Control 39/63

 MIP and MPC
• MPC optimization has three key features
– Dynamics model
– Constraints

– Cost

• MIP can enter all three

– PWA dynamics, logic states (e.g. ‘task done’)
– Non-convex constraints (e.g. avoidance)
– PWA cost

Mixed-integer Programming for Control 40/63

 Properties of MPC
• Start with a recursion
– Solution is {u(k0|k0) u(k0+1|k0) … u(k0+N|k0)}
at k0
– Then {u(k0+1|k0) … u(k0+N|k0) ???} feasible
at k0+1

Mixed-integer Programming for Control 41/63

 Properties of MPC
• Start with a recursion
– Solution is {u(k0|k0) u(k0+1|k0) … u(k0+N|k0)}
at k0 “The tail” Some control here
– Then {u(k0+1|k0) … u(k0+N|k0) ???} feasible
at k0+1

Mixed-integer Programming for Control 42/63

 Properties of MPC
• Start with a recursion
– Solution is {u(k0|k0) u(k0+1|k0) … u(k0+N|k0)}
at k0 “The tail” Some control here
– Then {u(k0+1|k0) … u(k0+N|k0) ???} feasible
at k0+1

• Use recursion to prove cost decrease

– J(k+1) ≤ J(k) – a(x(k))

• Use cost J as Lyapunov function

– Stage cost must be positive definite in x
Mixed-integer Programming for Control 43/63
 Properties of MPC
• Start with a recursion
– Solution is {u(k0|k0) u(k0+1|k0) … u(k0+N|k0)}
at k0 “The tail” Some control here
– Then {u(k0+1|k0) … u(k0+N|k0) ???} feasible
at k0+1

• Very
Use general
recursion to prove
results, cost
easily decrease
handling MIP
dynamics, at setpoint, J(k+1)
– If notconstraints ≤ J(k)
and cost – a(x(k))
[Bemporad and Morari, 1999]

• Use cost J as Lyapunov function

– Stage cost must be positive definite in x
Mixed-integer Programming for Control 44/63
 MIP/MPC Example: Modes
• Recall earlier example
Mode 1

Mode 2

Mixed-integer Programming for Control 45/63

 MIP/MPC Example: Modes
1. Optimize

2. Apply u(k) = u*(k|k)

3. Repeat
Mixed-integer Programming for Control 46/63
 MIP/MPC Example: Modes
1. Optimize
Cost

Dynamics
Initial condition
Terminal
Control limits

Mode logic

2. Apply u(k) = u*(k|k)

3. Repeat
Mixed-integer Programming for Control 47/63
Using MIP Online
 Computation Time
• Need to solve MIP online
– Good computer and solver handle most cases
– Still NP-hard problem – some instances of large
problems can be slow

• Tricks to accelerate solution time are often

problem-specific
– More to follow in example talk

• Three general approaches

– Use prior knowledge
– Approximate cost-to-go
– Multi-parametric integer programming
Mixed-integer Programming for Control 49/63
 Prior Knowledge I : Iteration
Space Station Rendezvous with Plume
Impingement (PI) Constraints

• Firing expected only

at start and end:
“bang-off-bang”
• Remove constraints
on PI around center
of maneuver
• Check results for PI
• Iterate, adding
constraints back in,
until no PI

Mixed-integer Programming for Control 50/63

 Prior Knowledge II : Grouping
• Minimum fuel Space Station fly-by with PI
constraints

Mixed-integer Programming for Control 51/63

 Prior Knowledge II : Grouping
• Minimum fuel Space
Station fly-by with PI
constraints Time
– 9600 binary variables (secs)
– Impractical to find
No plume
optimal solution 8
constraints

• Strategy:
All plume 1800
– Group time-steps
constraints (limit)

Groups of three 587

Mixed-integer Programming for Control 52/63

 Approximate Cost-to-Go
• Replace tail of long plan with approximate
plan
– Re-plan online as goal approached

Mixed-integer Programming for Control 53/63

 Multi-Parametric MIP
• Calculate solution offline as a piecewise
affine function of the initial condition
– [Bemporad et al 2004]
• Replace online MIP with PWA look-up
Online Offline Online

Solve Solve PWA

MIP mpMIP lookup

Execute Solution Execute

first step map first step

Mixed-integer Programming for Control 54/63

Task Assignment Example
 Example: Task Assignment

• Different types of UAV’s

• Different types of targets
HVT

SAM

• Example timing constraints

– BDA after Strike after Recon.
– SAM site visited before HVT
– Synchronized strike of HVT
Æ All can be expressed in a canonical form
• Select assignments & trajectories to optimize desired mission
objectives and satisfy dynamic/timing constraints
– Costs for demonstration based on time, but extends to scores / risks
– Hard because problems are tightly coupled

Mixed-integer Programming for Control 56/63

 Experimental Scenario
• 4 vehicles
• 16 tasks
• Full horizon assignment
approach
not scalable for real-time
assignment
• Environment with dynamic
changes
– Original assignment
– Vehicle death
– Pop-up target
– Vehicle death
– Target location update
Mixed-integer Programming for Control 57/63
 Experimental Scenario
• 4 vehicles
• 16 tasks
• Full horizon assignment
approach
not scalable for real-time
assignment
• Environment with dynamic
changes
– Original assignment
– Vehicle death
– Pop-up target
– Vehicle death
– Target location update
Mixed-integer Programming for Control 58/63
UAV Example
 UAV Example
• Very simple trial example
– One UAV avoiding one obstacle
– Uses MPC and MIP

• Problem
– Matlab sim script
– AMPL data file
– AMPL state file
– AMPL script
– AMPL model file

• Download: http://seis.bris.ac.uk/~aeagr

Mixed-integer Programming for Control 60/63

 UAV Example
• Very simple trial example
– One UAV avoiding one obstacle
– Uses MPC and MIP

• Implementation
– Matlab sim script
– AMPL data file
– AMPL state file
– AMPL script
– AMPL model file

• Download: http://seis.bris.ac.uk/~aeagr

Mixed-integer Programming for Control 61/63

 Session Outline
• Projected Variable Metric Algorithm for Mixed Integer Optimization
Problem
– Ali Ahmadzadeh (Univ. of Pennsylvania),
– Bijan Sayyarrodsari (Pavilion Tech. Inc),
– Abdollah Homaifar (North Carolina A\&T State Univ.)

• MILP Assignment Problems for Multi-Vehicle Systems

– Matthew Earl (BAE)
– Raffaello D'Andrea (Cornell Univ.)

• Real-Time MILP Path-Planning for Tactical UAV Applications

– Cedric Ma (Northrop Grumman Corp.)
– Robert Miller (Northrop Grumman Corp.)

• Receding Horizon Implementation of MILP for Vehicle Guidance

– Yoshiaki Kuwata (MIT)
– Jonathan P. How (MIT)

Mixed-integer Programming for Control 62/63

 Questions?

• Web resources
acl.mit.edu/MILP
seis.bris.ac.uk/~aeagr
• Email
[email protected]
[email protected]

Mixed-integer Programming for Control 63/63