Enhancements On The Hyperplanes Arrangements in Mixed-Integer Programming Techniques

J Optim Theory Appl (2012) 154:549572
DOI 10.1007/s10957-012-0022-9
Enhancements on the Hyperplanes Arrangements

in Mixed-Integer Programming Techniques
Ionela Prodan Florin Stoican Sorin Olaru
Silviu-Iulian Niculescu
Received: 4 May 2011 / Accepted: 6 March 2012 / Published online: 24 March 2012
Springer Science+Business Media, LLC 2012
Abstract This paper is concerned with improvements in constraints handling for

mixed-integer optimization problems. The novel element is the reduction of the number of binary variables used for expressing the complement of a convex (polytopic)
region. As a generalization, the problem of representing the complement of a possibly not connected union of such convex sets is detailed. In order to illustrate the
benefits of the proposed improvements, a typical control application, the control of
multiagent systems using receding horizon optimization techniques, is considered.
Keywords Mixed integer programming (MIP) Not convex constraints
Hyperplane arrangements Cell merging
Communicated by Panos M. Pardalos.

I. Prodan () S. Olaru
Automatic Control Department, SUPELEC Systems Sciences (E3S), Gif sur Yvette, France
e-mail: ionela.prodan@supelec.fr
S. Olaru
e-mail: sorin.olaru@supelec.fr
F. Stoican
Department of Engineering Cybernetics, Norwegian University of Science and Technology,
Trondheim, Norway
e-mail: florin.stoican@itk.ntnu.no
I. Prodan S.-I. Niculescu
Laboratory of Signal and Systems, CNRS-SUPELEC, Gif sur Yvette, France
I. Prodan
e-mail: silviu.niculescu@lss.supelec.fr
S.-I. Niculescu
e-mail: ionela.prodan@lss.supelec.fr
550
1 Introduction
Collision avoidance plays an important role in the context of managing multiple
agents (see [1, 2], where different frameworks for the cooperative control of multiple agents are described). In the same time, it is known to be a difficult problem,
since certain constraints are not convex, [3]. For example, the evolution of a dynamical system in an environment presenting obstacles can be modeled in terms of a not
convex feasible region. More precisely, it is possible to set up an optimization problem such that the agent state trajectory avoids a convex region, in fact, representing
an obstacle (static constraints) or another agent (dynamic constraintsleading to a
parameterization of the set of constraints with respect to the current state).
A popular framework for the treatment of such an optimization problem is represented by Mixed-Integer Programming (MIP), described in [4]. This method can be
very useful in several fields of applications, due to its ability to include not convex
constraints and discrete decisions in the optimization problem (see, for instance [5]).
There is a growing literature about optimization problems, which can be formulated
through the use of MIP techniques. For example, [68] focused their work on optimization of agent trajectories. Multivehicle target assignment and intercept problems
are studied by [3, 9]. MIP was also useful to coordinating the efficient interaction of
multiple agents in scenarios with many sequential tasks and tight timing constraints
(see [10, 11]). In [12, 13], the authors used a combination of MIP and Model Predictive Control (MPC) (see, for instance, [14, 15], for basic notions in MPC) to stabilize general hybrid systems around equilibrium points. [16] introduced MIP in a
predictive control framework to plan short trajectories around nearby obstacles. The
mixed-integer formulation has also proven to be useful for the convergence of multiple agents toward a tight formation; see [17]. Finally, an application for different
control design problems was reported in [18], where a feasible reference signal which
permits set membership testing for fault detection was computed over a not convex
region, leading to a MIP formulation.
However, despite its modeling capabilities and the availability of versatile solvers,
MIP has serious numerical drawbacks. As stated in [19], mixed-integer techniques
are classified as NP-hard, i.e., the computational complexity increases exponentially
with the number of binary variables used in the problem formulation. Consequently,
these methods may not be fast enough for real-time control of systems with large
problem formulations. There has been a number of attempts in the literature to reduce the computational requirements of MIP formulations in order to make them
attractive for real-time applications. In [20], an iterative method for including the obstacles in the best path generation is provided. Other references, like [21], consider
a predefined path constrained by a sequence of convex sets. In all of these papers,
the binary variables reduction is not tackled at the MIP level, but instead the original
decision problems are reformulated in a simplified MIP form. The negative influence
of the large number of binary variables in the problem formulation highlights the
importance of reducing them. In [22, 23], we introduced a novel linear constraint
expression for reducing the number of binary variables necessary in describing the
exterior of convex sets.
551
In the present paper, we resume these results in a unitary description of not connected convex sets (or their complement) using auxiliary binary variables. The optimization problems where the binary variables are used to express a feasible not
convex region over which a (usually quadratic) cost function has to be minimized,
are treated in a first stage. The problem is reformulated using a reduced number of
binary variables through a compact encoding of the inequalities describing the feasible region. Thus, the problem complexity will require only a polynomial number of
subproblems (Linear Programming (LP) or Quadratic Programming (QP) problems)
that have to be solved with obvious benefits for the computational effort. In the second part of the paper, the technique is extended for the treatment of not connected
not convex regions. In a similar manner, a reduced number of binary variables suffices in describing a not convex and not connected region. Some of the noteworthy
aspects of the approach which resumes also the main contributions of this paper are
the following:
a convex representation in the extended space of state plus binary variables using
the associated hyperplane arrangement;
reduced complexity of the problem upon Boolean algebra techniques;
a notable property of optimal association between regions and their binary representation leading to the minimization of the number of constraints.
The methods presented here can be used in several fields of application. We choose
to exemplify here with the control of an agent operating in a dynamic environment
with obstacles. More precisely, the agent, which may have an associated polyhedral
safety region, is required to maneuver successfully in a hostile environment. The
obstacles are designed as convex polyhedral regions. In this context, the proposed
reduction technique is embedded within an MPC path planning for multiple agents.
The rest of the paper is organized as follows. In Sect. 2, the preliminaries are presented; the main idea being detailed in Sect. 3. Furthermore, in Sect. 4, the method
is extended to not connected, not convex regions. The improvements in the computational time for the approach are detailed in Sect. 6. Discussions based on an example
are presented in Sect. 7, while the conclusions are drawn in Sect. 8.
2 Preliminaries
The following notation will be used throughout the paper. The closure of a set S,
cl(S), is the intersection of all closed sets containing S. The collection of all possible
combinations of N binary variables will be noted {0, 1}N := {(b1 , . . . , bN ) : bi
{0, 1}, i = 1, . . . , N}. For a scalar x R, we denote by x the upper integer part,
and by Bnp := {x Rn : xp 1} the unit ball of norm p, where xp is the p-norm
of vector x Rn . Notations lp(n, d) and qp(n, d) represent the complexity of solving
a linear program, quadratic program, respectively, with n constraints and d variables.
For safety and obstacle avoidance problems (to take just a few examples), the
feasible region in the space of solutions is a not convex set. Usually, this region is
considered as the complement of a convex region, which describes an obstacle and/or
a safety region. Due to their versatility and relative low computational complexity,
the polyhedra are the instrument of choice in characterizing these regions.
552
In the following, we define a polytope, P Rn , through its implicit half-space

description:

(1)
P := x Rn : hi x ki , i = 1, . . . , N ,
with (hi , ki ) R1n R, and its complement as
CX (P ) := cl(X \ P ),
(2)
with the reduced notation C(P ) whenever X is presumed known or is considered to

be the entire space Rn .
By definition, every affine subspace, which is a support hyperplane for P ,
Hi := {x : hi x = ki },
(3)
will partition the space into two regions:1

R+ (Hi ) := {x : hi x ki },
R (Hi ) := {x : hi x ki },
(4)
with i = 1, . . . , N .
The not convex region C(P ), denoted by (2), may then be described as a union of
regions that cover all space except P :

(5)
C(P ) := R (Hi ), i = 1, . . . , N.
i
Therefore, we note that the complement of a bounded polyhedra (1) is covered by the
union of N overlapping regions denoted as R
i (a simplified notation for region (4)
associated to the ith inequality of (1)).
In order to obtain a tractable problem formulation, one has to use mixed integer
techniques with the aim of defining a polyhedra in the extended space of state +
auxiliary binary variables of the form:
hi x ki + Mi ,
i=N

i N 1,
i = 1, . . . , N,
(6)
(7)
i=1
with M a constant chosen appropriately (that is, significantly larger2 than the rest
of the variables and playing the role of a relaxation constant), and (1 , . . . , N )
{0, 1}N the auxiliary binary variables (which can activate or not the relaxation).
Remark 2.1 The set of solutions for (6)(7) can be projected in the original space Rn ,
leading to a coverage of the not convex region, which corresponds to the implicit
1 The relative interiors of these regions do not intersect, but their closures have as a common boundary the
affine subspace Hi .
2 There exists a finite M sufficiently large if and only if the polyhedra of type (1) are bounded. In the
remaining portion of the paper, all the polyhedra of type (1) are assumed to be bounded for this reason.
553
definition in (2). A region R

i can be obtained from (6) with an adequate choice of
binary variables

i := 1, . . . , 1,
0 , 1, . . . , 1 .
(8)
i
Note also that the converse is false since no choice of binary variables can lead to the
description of a region R+
i as in (4). Indeed, if a binary variable is 1, the corresponding inequality degenerates such that it covers any point x Rn (this represents
the limit case for M ). The condition (7) is thus required to ensure that at least
one binary value be 0, and consequently, at least one inequality be verified.
As it can be seen in the representation (6)(7), a binary variable is associated to
each inequality in the description of the polytope (1). Obviously, for a big number of
inequalities, the number of binary variables becomes exceedingly large. Since their
number exponentially affects the resolution of any mixed integer algorithm (usually
based on branch and cut techniques, and thus very sensitive to the number of binary
terms), the goal to reduce their number is worthwhile. A first step would be to eliminate from the half-space representation of the polytope all the redundant constraints;
see [24]. We suppose that this pretreatment be performed, and we are dealing with a
nonredundant description of the polyhedral set in (1).
3 Basic Ideas
By preserving a linear structure of the constraints, we propose in the present section
a generic solution toward the binary variables reduction.
To each of the regions in (5), we have associated in (6) a unique binary variable.
Consequently, the total number of binary variables is N , the number of supporting
hyperplanes (see (1)). However, a basic calculus shows that the minimum number of
binary variables necessary to distinguish between these regions is
N0 = log2 N.
(9)
The value of N0 is reached by observing that each of the N regions in (6) can be
linked to a unique number. Taking the numbers successively (starting from zero), it
follows that we need N0 bits to codify them into a binary representation.
The question that arises is the following:
How to describe the regions in a linear formulation similar to (6) through a
reduced number of binary variables?
We impose firstly that the binary expression appearing in the inequalities has to
remain linear for computational advantages related to the optimization solvers. This
structural constraint is equivalent with saying that any variable i should be described
by a linear mapping in the form:
i () = a0i
N0

k=1
aki k ,
(10)
554
where
(1 , . . . , N0 ) := {0, 1}N0 .
In the reduced space of , we will arbitrarily associate a tuple

i := i1 . . . iN0
(11)
(12)
to each region R
i . Note that this association is not unique, and various possibilities
can be considered: in the following, unless otherwise specified, the tuples will be
appointed in lexicographical order.
The problem of finding a mapping in , which describes region R
i , reduces then
i ) for which (i ) = 0 and (j ) 1,
to finding the coefficients (a0i , a1i , . . . , aN
i
i
0
j = i under mapping (10). This translates into the following conditions for any
i , j {0, 1}N0 :

a i + N0 a i i = 0,
k=1 k k
0
(13)
a i + N0 a i j 1, j = i,
k=1 k k
0
with ik the kth component of the tuple, i , associated to R
i .
Remark 3.1 Note that, in (13), the equality constraints for j = i were relaxed to
inequalities since the value of Mi (j ) needs only to be sufficiently large (any
i (j ) 1 being a feasible choice). Furthermore, the condition 1 can be relaxed
to an arbitrary small positive constant by means of counterbalancing through an increase in constant M.
Nothing is said a priori about the nonemptiness of the set described by (13). We
need at least a point in the coefficient space (a0 , a1 , . . . , aN0 ), which verifies conditions (13) in order to prove the nonemptiness. To this end, we present the following
proposition.
Proposition 3.1 A mapping i () : {0, 1}N0 {0} [1, [ which verifies (13), is
given by

N0

k ,
if ik = 0,
i () =
tk , where tk =
(14)
1 k , if ik = 1,
k=1
where k denotes the kth variable and ik its value for the tuple, i , associated to
region R
i .
i ) of the linear mapping (10) can be then obtained as
The coefficients (a0i , . . . , aN
0
a0i
N0

k=1

ik ,
aki
1,
1,
if ik = 0,
if ik = 1,
k = 1, . . . , N0 .
(15)
Proof The claim is constructive; by introducing mapping (15) in (13), it can be verified by simple inspection that the conditions are fulfilled.
555
Remark 3.2 Note that the problem of finding parameters i is independent of the
actual shape of the polytope P . The coefficients obtained in (14) can be used for any
topologically equivalent polytope (that is, with the same number of half-spaces).
3.1 Interdicted Tuples
By the choice of the cardinal N0 as in (9), the number of tuples allowed by the reduced
set of binary variables (11) may be greater than the actual number of regions.
The tuples left unallocated will be labeled as interdicted, and additional inequalities will have to be added to the extended set of constraints (6). These restrictions are
justified by the fact that, under construction (15), an unallocated tuple will not enforce
the verification of any of the constraints of (6) (see Remark 2.1). It then becomes evident that the single constraint of (7) has to be substituted by a set of constraints that
implicitly make all the unallocated tuples infeasible.
The next corollary of Proposition 3.1 provides the means to construct an inequality
which renders a tuple infeasible:
Corollary 3.1 Let there be a tuple i {0, 1}N0 . The N-dimensional point that it
describes, and exclusively this one by the combinations in (12), is made infeasible
with respect to the constraint:
N0

tki ,
(16)
k=1
with tki defined as in Proposition 3.1 and ]0, 1[ a scalar.

Proof The left side of the inequality (16) will vanish only at tuple i , and for the rest
of the tuples in the discrete set {0, 1}N0 will give values greater than or equal to 1.

Thus, the only point made infeasible by inequality (16) is i .
The number of unallocated tuples may be significant, an upper bound is given by
0 Nint 2log2 N 2log2 N 1 1 = 2log2 N 1 1,
(17)
with the bound reached for the most unfavorable case of N = 2log2 N 1 + 1.
If we associate to each of the unallocated tuples an inequality as in Corollary 3.1,
we negatively influence the speed of the associated optimization algorithm. This can
be alleviated by noting (as previously mentioned) that the association between regions and tuples is arbitrary. One could then choose favorable associations which
will permit more than one tuple to be removed through a single inequality. To this
end, we present the next proposition.
Proposition 3.2 Let there be a collection of tuples {i }i1,...,2d {0, 1}N0 , which
completely spans a d-facet3 of hypercube {0, 1}N0 . Let I be the set of the N0 d
3 d denotes the degree of the facet, ranging from 0 for extreme points to N 1 for faces of the hypercube.
0
556
indices, which retains a constant value over all the tuples {i }i1,...,2d composing the
facet. Then there exists the constraint

tk ,
(18)
kI
which renders the tuples of the given facet (and only these ones) infeasible.
Variables tk and are taken as in Corollary 3.1 with tk associated to k , the
common value of variable k over the set of tuples {i }i1,...,2d .
Proof Geometrically, the tuples are extreme points on the hypercube {0, 1}N0 and
the inequalities we are dealing with are half-spaces, which separate the points of the
hypercube. If a set of tuples completely spans a d-facet, it is always possible to isolate
a half-space that separates the points of the d-facet from the rest of the hypercube.
By a suitable association between feasible cells and tuples, we may label as unallocated the extreme points which compose entire facets on the hypercube {0, 1}N0 ,
which permits to apply Proposition 3.2 in order to obtain constraints (26).
Remark 3.3 By writing Nint as a sum of consecutive powers of 2, i.e.,
log2 Nint 1
Nint =
bi 2i ,
i=0
an upper bound Nhyp for the number of inequalities (26) can be computed:
log2 Nint 1
Nhyp =
bi log2 Nint ,
(19)
i=0
where bi {0, 1}.

Remark 3.4 Note that (19) offers an upper bound for the number of inequalities, but
practically the minimal value can be improved depending on the method used for
constructing the separating hyperplanes and of the partitioning of the tuples between
the allocated and unallocated subsets.
3.2 Illustrative Example
As an illustration of the notions described in Sect. 3, we take the following square:

0
1
1
0 1 x1
1
(20)
1
1.
0 x2
1 0
1
As stated in this section, the number of binary variables (similar to the formulation
(6)) is N = 4, equal with the number of half-spaces described in (20). The reduced
557
Fig. 1 Outer regions and their associated tuples
number of variables will be N0 = log2 4 = 2, according to (9). Following the problem formulation (14), the variables i can be expressed, as in (10), by
i () = a0i + a1i 1 + a2i 2 .
We associate to each region a tuple of two values (1 , 2 ) in lexicographical order.
The case of the 2nd half-space, associated to tuple (21 , 22 ) = (0, 1), is detailed
in Fig. 1(a). Using (13), we obtain, as depicted in Fig. 1(b), the feasible set of the
coefficients described by
a02 + a22 = 0,
a02 1,
a12 1.
This represents a polytopic region in the coefficients space (a0 , a1 , a2 ) R3 and,

according to (14), the nonemptiness is assured by the existence of at least a feasible
combination of coefficients leading to the mapping
2 () = 1 + 1 2 .
This means that the region R2 is projected from

x1
01
1 + M(1 + 1 2 ),
x2
by taking (21 , 22 ) = (0, 1) (see Remark 2.1).
Furthermore, the same computations will be performed for the rest of the regions,
resulting in an extended system of linear inequalities over mixed decision variables:
0 1
1 + M(1 + 2 )
0
1 + M(1 1 + 2 )
1
x1
1 0 x2 1 + M(1 + 1 2 ) .
1
0
1 + M(2 1 2 )
As an exemplification of the considerations in Sect. 3.1, let there be a polytope
with 5 hyperplanes. This means that the number of binary variables has to be N0 =
558
Fig. 2 Exemplification of separating hyperplanes techniques
log2 5 = 3 and then, Nint = 23 5 = 3 tuples will remain unallocated; we choose

these to be (0, 0, 1), (1, 0, 1), and (1, 1, 1).
By applying Corollary 3.1, we observe in Fig. 2(a) the 3 inequalities that separate
the unallocated tuples from the rest (for simplicity, in the rest of the subsection, we
will use = 0.5):
(1 + 1 + 2 3 ) 0.5,
(2 1 + 2 3 ) 0.5,
(3 1 2 3 ) 0.5.
We observe in Fig. 2(b) that the tuples are positioned onto 2 edges, and consequently,
using Proposition 3.2, 2 inequalities suffice for separation:
(1 + 2 3 ) 0.5,
(2 1 3 ) 0.5.
Lastly, recalling Remark 3.4, we note that, in this particular case, a single inequality
(as seen in Fig. 2(c)), is enough for separating the unallocated tuples from the rest:
(0.321 + 1.762 + 2.133 ) 0.5.
4 Description of the Complement of a Union of Convex Sets

In the previous section, the basic reduction method was applied for treatment of the
complement of a convex set. A generic case will be detailed in the following by
considering the complement of a union of convex (bounded polyhedral) sets P :=

l Pl :
CX (P) := cl(X \ P),
(21)
559
with4
Kl

Pl :=
R + (Hkl )
kl =1
and
N :=
Kl .
This type of regions arises naturally in the context of obstacle/collision avoidance

when there is more than a single object to be taken into account. In order to deal with
the complement of a not convex region in the context of mixed-integer techniques,
several additional theoretical tools related to the arrangements of the hyperplanes will
be introduced in the following. The noninitiated reader may consult on this concept
of the well-known books [25] and [26], some recent related articles [27, 28], and the
references therein.
Definition 4.1 A collection of hyperplanes H = {Hi }i=1:N will partition the space in
a union of cells5 defined as follows:
A(H) =

l=1,..., (N )
N

R l (i) (Hi ) ,
i=1
(22)
Al
where sign tuple l {, +}N denotes feasible combinations of regions (4) obtained
for the hyperplanes in H, and (N ) denotes the number of feasible cells.
Several computational aspects are of interest. The number of feasible cells, (N ),
(in relation with the space dimensiond and the number of hyperplanesN ) is
bounded by Bucks formula [29]:
(N )
d ! "

N
i=0
(23)
with equality satisfied if the hyperplanes are in general position6 and X = Rn .

An efficient algorithm for describing (22), based on reverse search that runs in
O(N (N ) lp(N, d)) time and O(N, d) space, was presented in [30] and implemented
in [31].
4 The + superscript was chosen for the homogeneity of notation, equivalently one could have chosen any
combination of signs in the half-space representation (4) in order to describe the polyhedral regions Pl .
5 The relative interior of any two cells is disjoint, but their boundaries may have one of the hyperplanes H
i
as a common element.
6 We call a hyperplane arrangement to be in general position whenever any small change in the position
of the composing hyperplanes does not change the number of cells.
560
Note that there exists a subset {Bl }l=1,..., b (N ) of feasible polyhedral cells from
(22) (with b (N ) (N )) which describes region (21):
CX (P) =
(24)
Bl ,
l=1,..., b (N )
such that, for any l, there exists a unique i for which, Bl = Ai and Ai P = .
In (6), a single binary variable was associated to a single inequality, but the mechanism can be applied similarly to more inequalities (e.g., the ones describing one of the
cells of (24)). Thus, one can describe (22) in an extended space of state + auxiliary
binary variables as follows:
..
.
l (1)h1 x l (1)k1 + Ml
..
.
l (N )hN x l (N )kN + Ml
..
.
Bl
(25)
and condition
b
l=
(N )
l b (N ) 1,
(26)
l=1
which implies that at least a set of constraints will be verified.

Construction (25)(26) will permit, through projection along the binary variables
l (see (8)), to obtain any of the cells of the union (24).
Analogously to Sect. 3, we propose, in the following, the reduction of the number of binary variables by associating to each of the cells a unique tuple. The binary
part will be computed following the constructive result in Proposition 3.1 and used
accordingly in (25). Additional inequalities that render infeasible the unallocated tuples, are introduced as in Proposition 3.2.
A few remarks relating to the number of hyperplanes and their corresponding arrangement are in order.
Remark 4.1 The number of inequalities in (25) can be reduced by observing that not
all the hyperplanes of H are active in a particular cell, and thus they can be discarded
from the final representation.
Remark 4.2 Note that if we discard the linear structure and allow a nonlinear formulation involving products of binary variables, the hyperplane arrangements (22) can
561
be represented as
..
.
hi x ki + M
'
l ,
l=1,..., b (N )
l (i)=
hi x ki + M
'
(27)
l ,
l=1,..., b (N )
l (i)= +
..
.
for all sign tuples l associated to cells Bl from covering (24). We have used the fact
that the cells of (24) use the same half-spaces (up to a sign), and thus they can be
concatenated. The method presented in [32] transforms an inequality with nonlinear
binary components into a set of inequalities with linear binary components. However,
this can be made only at the expense of introducing additional binary variables, which
in the end gives a larger problem than the one presented in (25)(26).
4.1 Exemplification of Hyperplane Arrangements
Consider the following illustrative example depicted in Fig. 3, where the complement of the union of two triangles (P = P1 P2 ) represents the feasible region. We
take H = {Hi }i=1:4 the collection of N = 4 hyperplanes (given as in (3)) which define P1 , P2 .
We observe that the bound given in (23) is reached, that is, we have 11 cells (obtained as in the arrangement (22)). From them, a total of 9, which we denote here
as B1 , . . . , B9 , describe the not convex region (21). To each of them, we associate a
unique tuple from {0, 1}N0 as seen in Fig. 3 with N0 = log2 9 = 4.
Fig. 3 Exemplification of hyperplane arrangement
562
As per Proposition 3.1 and (25), we are now able to write the set of inequalities (28).
h3 x k3
h4 x k4
(
+ M(
1 + 2 + 3 + 4 ) B1 ,
h3 x k3 + M(1 + 1 + 2 + 3 4 ) B2 ,
h4 x k4
h1 x k1
h2 x k2 + M(1 + 1 + 2 3 + 4 ) B3 ,
h3 x k3
(
h1 x k1
+ M(2 + 1 + 2 3 4 ) B4 ,
h3 x k3
h1 x k1
h2 x k2
+ M(1 + 1 2 + 3 + 4 ) B5 ,
h3 x k3
h4 x k4
(
h2 x k2
+ M(2 + 1 2 + 3 4 ) B6 ,
h4 x k4
h1 x k1
h2 x k2 + M(2 + 1 2 3 + 4 ) B7 ,
h4 x k4
h1 x k1
h3 x k3 + M(3 + 1 2 3 4 ) B8 ,
h4 x k4
h1 x k1
h2 x k2
+ M(1 + 1 2 + 3 + 4 ) B9 .
h3 x k3
h4 x k4
h2 x k2
(28)
Note that, in the above set, we have simplified the description by cutting the redundant hyperplanes in a cell representation (e.g., for cell A1 , 2 hyperplanes suffice
for a complete description).
563
Since only 9 tuples, from a total number of 16, are associated to cells, we need to
add constraints to the problem such that the remaining 7 unallocated tuples will never
be feasible. Using Corollary 3.1, we obtain:
(2 1 2 + 3 + 4 ) 0.5,
(3 1 2 3 + 4 ) 0.5,
(3 1 2 + 3 4 ) 0.5,
(4 1 2 3 4 ) 0.5,
(29)
(2 1 + 2 3 + 4 ) 0.5,
(3 1 + 2 3 4 ) 0.5,
(2 1 + 2 + 3 4 ) 0.5.
We observe that for the 7 unallocated tuples, 4 of them, (1, 1, 0, 0), (1, 1, 0, 1),
(1, 1, 1, 0), and (1, 1, 1, 1), form a 2-facet of the hypercube {0, 1}4 . Tuples (1, 0, 1, 0)
and (1, 0, 1, 1) form an edge and (1, 0, 0, 1) is on a vertex. We can now apply Proposition 3.2 and obtain the following constraints:
(2 1 2 ) 0.5,
(2 1 + 2 3 ) 0.5,
(30)
(2 1 + 2 + 3 4 ) 0.5.
Note that we were able to diminish the number of inequalities from 7 in (29) to only
3 in (30): the first 4 constraints of (29) are replaced by the 1st constraint of (30). The
same holds for the next 2 that correspond to the 2nd, and for the last that is identical
with the 3rd.
5 Refinements for the Complement of a Union of Convex Sets

As seen in [22], palliatives for reducing the computational load exist but ultimately,
the computation time is in the worst case scenario exponentially dependent on the
number of binary variables, which in turn depends on the number of cells of the
hyperplane arrangements (see (23)). We conclude then that the problem becomes
prohibitive for a relatively small number of polyhedra in P, and that any reduction in
the number of cells is worthwhile and should be pursued.
This can be accomplished in two complementary ways. Firstly, we note that bound
(23) is reached for a given number of hyperplanes if and only if they are in general
position. As such, particular classes of polyhedra may somewhat reduce the actual
number of cells in arrangement (22), and consequently, the number of auxiliary binary variables. In increasing order of their versatility, we may mention hypercubes,
orthotopes, parallelotopes, and zonotopes as classes of interest (for a computation of
the number of cells (see, [33])).
The other direction, which we chose to pursue in the rest of the paper, is reducing
the number of cells that describe (21).
564
In [22], we have proposed a hybrid scheme which permits to express (21) as a

union of the cells of (24) which intersect P , and of the regions (in the sense of (4))
which describe C(P ), where P denotes the convex hull of P.
Alternatively, the merging of adjacent cells into possibly overlapping regions,
which describe (21), is discussed in [23]. This results in a reduced representation,
both in number of cells and of interdicting constraints. In the next subsection, we
detail the merging techniques used and show how the complexity of the problem is
reduced.
5.1 Cell Merging
Recall that any of the cells of (24) is described by a unique sign tuple (Bl l ).
As such, we obtain that the cells are disjunct and cover the entire feasible space. For
our purposes, we are satisfied with any collection of regions not necessarily disjoint
which covers the feasible space. In this context, we may ask if it is not possible to
merge the existing cells of (24) into a reduced number of regions which will still
cover region (21). Note that by reducing the number of regions, the number of binary
auxiliary variables will also decrease substantially.
We can formally represent the problem by requiring the existence of a collection
of regions,

Ck ,
(31)
CX (P) =
k=1,..., c (N )
which verifies next conditions:

the new polyhedra are formed as unions of the old ones (i.e., for any
k there exists
a set Ik which selects indices from 1, . . . , b (N ) such that Ck = iIk Bi ),
the union is minimal, that is, the number c (N ) of regions is minimal.
Existing merging algorithms are usually computationally expensive, but here we
can simplify the problem by noting two properties of the cells in (24):
the sign tuples l describe an adjacency graph since any two cells whose sign tuples
differ at only one position are neighbors,
the union of any two adjacent cells is a polyhedra.
In order to construct (31), we may use merging algorithms (see, for example, [34],
which adapts a branch and bound algorithm to merge cells of a hyperplane arrangement), or we can pose the problem in the Boolean algebra framework. The
merging problem of regions from (24) is functionally identical to the minimization
of a Boolean function given in the sum of products form. A cell describing the
(in)feasible region (21) corresponds to a 1 (0) value in the truth table at the position determined by its associated sign tuple, whereas infeasible sign tuples correspond to do not care values. It is then straightforward to apply minimization algorithms (Karnaugh maps, the QuineMcCluskey algorithm, or the Espresso heuristic
logic minimizer) in order to obtain Boolean minterms who describe the merged cells
of (31). We note that a similar approach was proposed in [35] in order to deal with
polyhedral piecewise affine systems.
565
Remark 5.1 Note that a region Ck is described by at most N f hyperplanes, where

f denotes the number of indices in the sign tuples which flip the sign. It makes sense
then to, not only reduce the number of regions, but also to maximize the number of
cells that go into the description of a region from (31).
In Algorithm 1, we sketch the notions presented in this section.
Algorithm 1 Scheme for representing C(P)
Input P
1 obtain the cell arrangement as in (22) for H
2 obtain the feasible cells (24) and merge them in representation (31)
3 get the number c (N ) of feasible regions and the number N0 of auxiliary binary
variables
4 partition the tuples of {0, 1}N0 such that Proposition 3.2 can be efficiently applied
5 create the extended polyhedron (25) and add the constraints (26)
Note that the steps 1 and 2 are the most computationally expensive. For the first
step, an efficient algorithm for describing (22), based on reverse search that runs in
O(N (N ) lp(N, d)) time and O(N, d) space, was detailed in [30] and implemented
in [31]. For the second step, Boolean algebra techniques have been used: for a small
number of hyperplanes, exact methods like Karnaugh maps prove to be effective.
For higher numbers, the heuristic Espresso logic minimizer can be used. The latter
provides solutions very close to the optimum while being more efficient and reducing memory usage and computation time by several orders of magnitude relative to
classical methods (see [36]).
5.2 Exemplification of Hyperplane Arrangements with Cell Merging
We revisit here the example provided in Sect. 4.1 and apply the results presented in
Sect. 5.1 in order to show the improvements.
For this simple case, we apply, as seen in Fig. 4, a Karnaugh diagram and obtain
that the feasible region (21) is expressed by a union as in (25).
As seen from the Karnaugh diagram from Fig. 4, we obtain 4 overlapping regions:
C1 = B1 B2 B3 B4 , C2 = B4 B5 B6 , C3 = B6 B7 B8 B1 and C4 =
B8 B9 B1 B2 , which we depict in Fig. 5. Consequently, we note that N0 = 2
Fig. 4 Karnaugh diagram for
obtaining the reduced cell
representation
566
Fig. 5 Exemplification of
hyperplane arrangement with
merged regions
auxiliary binary variables suffice in coding the regions. As for Proposition 3.1 and
(25), we are now able to write the following set of inequalities (we attach to each of
the regions a tuple in lexicographical order):

h3 x k3 + M(
1 + 2 )
C1 ,
(
h1 x k1
+ M(1 + 1 2 )
C2 ,
h4 x k4
(32)

h4 x k4 + M(1 1 + 2 )
C3 ,
(
h1 x k1
+ M(2 1 2 )
C4 .
h2 x k2
Note that, in addition to reducing the number of regions in (32) comparative with
(28), we also have reduced the number of hyperplanes appearing in the regions halfspace representation (see Remark 3.3).
6 Numerical Considerations
In this section, we will test the computation time improvements for our approach
versus the standard technique encountered in the literature. As previously mentioned,
a MI problem is NP-hard in the number of binary variables (due to the fact that the
algorithms implement branch-and-bound techniques and as such, in the worst case
scenario, they need to iterate through all the branches defined by the binary search
tree variables). Therefore, even a small reduction will render sensible improvements.
The complexity of the MI algorithm with constraints in the classical form (6)(7)
will be of the order of O(2N p(N + 1, d)), where p(n, d) denotes either lp(n, d) or
qp(n, d), as it is required. Using the alternative formulation proposed in Sect. 3, we
567
Fig. 6 Comparative test for

computation time for classical
and enhanced methodtime
axis in logarithmic scale
Table 1 Numerical values for

the solving of an MI
optimization problem under
classical and enhanced methods
No. of
hyperplanes
10
15
Classical
9.91 64.06 91.74 511.47 306.04
Enhanced
1.14 0.81
0.59
20
4.84
25
4.18
obtain the complexity as

O 2log2 N p N + log2 N 1 , d = O N p(N, d) .
50
100
3.66 2.94
(33)
We consider the worst case scenario for the mixed-integer problem, that is, the
optimization algorithm will need to pass through each of the possible combinations
provided by the binary variables. In the case of formulation (6)(7), this leads to a
NP complexity in the number of lp/qp problems to be solved. On the other hand,
the formulation shown in (33), through the reduction of binary variables, provides a
polynomial complexity.
To illustrate these computational gains, we will compare the times of execution
for both schemes as follows: the computational time will be measured and averaged
for 10 samples of 2D polytopes with the same number of support hyperplanes; furthermore, the procedure will be iterated by changing the number of hyperplanes from
4 to 25. The results are depicted in Fig. 6 on a semilogaritmic scale and, as it can be
seen, there are significant improvements. In fact, the differences may be even more
pronounced since, under default settings, the MI algorithm over the classical method
stopped computing the optimum value after a maximum number of iterations was
reached (the MI algorithm used was the one described in [12]).
Similar results are shown in Table 1, where we observe the evident improvement
of our method relative to the classical technique.
In Sect. 4, a method for describing in the MI formalism of the complement of
a possibly not connected union of polytopes was presented. The main drawback is
that in both classical and reduced formulation, the problem depends on the number
568
of cells. Supposing that the hyperplanes from the hyperplane arrangement (24) be in
random position, we obtain, for formulation (25)(26), a complexity of order

b
O 2 (N ) p N N d + 1 , d ,
(34)
which can be further reduced, using the techniques from Sect. 3, to

b
O 2log2 (N ) p N N d + 1, d = O b (N ) p N d+1 , d .
(35)
Again, we observe that the mixed-integer problem becomes polynomial in the number of branches: in the worst case, we have b subproblems to solve, and b
which in turn is given by the polynomial expression (23). However, the problem is
still challenging due to the number of cells (see (23)). By reducing the number of cells
as in Sect. 5, it is possible to significantly reduce the computation time. For exemplification, take the example depicted in Figs. 3 and 5. We observe that in this particular
case, we were able to reduce the representation from 9 cells to only 4. Presumably,
for a higher number of hyperplanes, the gain will be even more pronounced.
7 Collision Avoidance Example

A number of commonly found situations in the control related to Multi-Agent Systems imply a cost function that has to be minimized, while in the same time, the agent
avoids collision with obstacles and other agents. To solve this problem, there exists
various methods. Arguably, they can be gathered in methods which penalize through
the cost function as the violation of the constraints (e.g., Potential Field Method [37]
and Navigation Functions [38]), and methods which impose hard constraints that may
not be broken. The latter group usually employs receding horizon techniques as they
naturally take into account constraints [7, 15]. In this illustrative example, we will
describe an agent that has to navigate its way around a group of fixed obstacles. We
consider the dynamics of the agent described by a LTI system as follows:
k+1 = Ak + Buk .
(36)
The agent model is used in a predictive control (see [14]) context which permits
the use of not convex state constraints for obstacle avoidance behavior.
u is obtained from the control sequence u :=

An optimal control action
uk|k , uk+1|k , . . . , uk+N 1|k as a result of the optimization problem:

T
u = arg min k+N
|k P k+N |k +
u

s.t.:
N
1

T
k+l|k
Qk+l|k +
l=1
k+l|k = Ak+l1|k + Buk+l1|k ,

k+l|k C(P),
l = 1, . . . , N.
N
1

uTk+l|k Ruk+l|k ,
l=0
(37)
Here, Q 0, R > 0 are the weighting matrices, P 0 defines the terminal cost and
P is an union of polytopes describing the obstacles.
569
Fig. 7 Interdicted region in the

state space
As a practical application, we consider a linear system (vehicle, pedestrian or agent

in general form), whose dynamics are described by
0 0
0 0
1
0
0 0
0 0
0
1
,
B =
(38)
A=
1 0 ,
0 0
0
m
m
1
0 0
0
m
0 m
where = [x y vx vy ]T , u = [ux uy ]T are the state and the input of the system. With
the components of the state being (x, y) the position, and (vx , vy ) the velocities of
the agent, m is the mass of the agent and its damping factor.
We consider the position component of the agent state to be constrained by 4
obstacles as shown in Fig. 7.
Considering the 14 hyperplanes which describe the polyhedra associated with the
obstacles, we have (14) = 106 regions obtained as in (22). Additionally, we observe
that 10 of the cells will describe the interdicted regions, and the rest, b (14) = 96
will describe the feasible region, as shown in (24). Furthermore, we apply the notions
from Sect. 5.1 to obtain a reduced representation for the feasible region as in (31).
We observe that the number of cells is substantially reduced, from b (14) = 96 to
c (14) = 11, which warrants in turn a reduction of the auxiliary binary variables from
7 to 4 that, for a worst case scenario, equals to an eightfold speed up. In Fig. 8(a),
we depict the cells of (24) and the obstacles, while in Fig. 8(b) we show the covering
(31) of merged cells.
We apply the predictive control strategy for horizon N = 3 and cost matrices Q =
105 I4 , R = I2 and P = 105 I4 , and obtain the trajectory depicted in Fig. 9.
A last aspect of interest is the scalability of the problem for a larger number of
agents/obstacles. Obviously, the results depend greatly on the number of sets, the
number of hyperplanes defining them and their relative positions. We depict in Table 2
a few cases (showing the number of obstacles, hyperplanes, and regions (6)(8) and
(19), respectively) and observe that the improvements are (at least for the examples
considered) significant.
570
Fig. 8 Cell partitionings of the feasible region

Fig. 9 Simulations of agent
trajectories
Table 2 Complexity analysis

with increased numbers of
obstacles
#P
#Hi
#Ai
#Bi
#Ci
19
210
191
13
24
295
266
16
30
496
449
23
8 Concluding Remarks
In this paper, we revisit a technique, which transforms a not convex and possibly
not connected region into a polyhedra in an augmented space (state and auxiliary
binary variables) through the use of hyperplane arrangements. With respect to previous results, we minimized the number of cells describing the feasible region through
merging methods and discussed an improvement of the optimization problem such
that the number of additional constraints is minimized. These numerical improve-
571
ments were presented, tested, and discussed on applications concerning the obstacle
avoidance control problem.
Acknowledgements The research of Ionela Prodan is financially supported by the EADS Corporate
Foundation (091-AO09-1006). Florin Stoicans work was carried out in Supelec, during the tenure of
a CARNOT C3S fellowship. The authors would like to thank Professor Panos M. Pardalos, as well as
the anonymous reviewers for their useful comments and remarks that helped in improving the overall
presentation of this paper.
References
1. Grundel, D., Murphey, R., Pardalos, P.: Cooperative Systems, Control and Optimization, vol. 588.
Springer, Berlin (2007)
2. Grundel, D., Pardalos, P.: Theory and Algorithms for Cooperative Systems, vol. 4. World Scientific,
Singapore (2004)
3. Earl, M., DAndrea, R.: Modeling and control of a multi-agent system using mixed integer linear programming. In: Proceedings of the 40th IEEE Conference on Decision and Control, Orlando, Florida,
USA, vol. 1, pp. 107111 (2001)
4. Osiadacz, A.: Integer and combinatorial optimization. In: Nemhauser, G.L., Wolsey, L.A. (eds.)
Wiley-Interscience Series in Discrete Mathematics and Optimization. Wiley, New York (1988). International Journal of Adaptive Control and Signal Processing, 4(4), 333344 (1990)
5. Jnger, M., Liebling, T., Naddef, D., Nemhauser, G., Pulleyblank, W., Reinelt, G., Rinaldi, G.,
Wolsey, L.: 50 Years of Integer Programming 19582008: From the Early Years to the State-of-theArt. Springer, Berlin (2009)
6. Schouwenaars, T., De Moor, B., Feron, E., How, J.: Mixed integer programming for multi-vehicle
path planning. In: Proceedings of the 2nd IEEE European Control Conference, Porto, Portugal, pp.
26032608. (2001)
7. Richards, A., How, J.: Model predictive control of vehicle maneuvers with guaranteed completion
time and robust feasibility. In: Proceedings of the 24th American Control Conference, Portland, Oregon, USA, vol. 5, pp. 40344040 (2005)
8. Ousingsawat, J., Campbell, M.: Establishing trajectories for multi-vehicle reconnaissance. In: Proceedings of AIAA Guidance, Navigation and Control Conference, Providence, Rhode Island, USA,
pp. 21882199 (2004)
9. Beard, R., McLain, T., Goodrich, M., Anderson, E.: Coordinated target assignment and intercept for
unmanned air vehicles. Proc. IEEE Trans. Robot. Autom. 18(6), 911922 (2002)
10. Richards, A., Bellingham, J., Tillerson, M., How, J.: Coordination and control of multiple UAVS. In:
Proceedings of AIAA Guidance, Navigation and Control Conference. Monterey, CA (2002)
11. Schumacher, C., Chandler, P., Pachter, M.: UAV task assignment with timing constraints. In: Proceedings of AIAA Guidance, Navigation and Control Conference, Austin, Texas (2003)
12. Bemporad, A., Morari, M.: Control of systems integrating logic, dynamics, and constraints. Automatica 35, 407428 (1999)
13. Bemporad, A., Borrelli, F., Morari, M.: Optimal controllers for hybrid systems: stability and piecewise
linear explicit form. In: Proceedings of the 39th IEEE Conference on Decision and Control, Sydney,
NSW, Australia, pp. 18101815 (2000)
14. Mayne, D., Rawlings, J., Rao, C., Scokaert, P.: Constrained model predictive control: stability and
optimality. Automatica 36, 789814 (2000)
15. Camacho, E.F., Bordons, C.: Model Predictive Control. Springer, Berlin (2004)
16. Bellingham, J., Richards, A., How, J.: Receding horizon control of autonomous aerial vehicles. In:
Proceedings of the 21th American Control Conference, Anchorage, Alaska, USA, vol. 5, pp. 3741
3746 (2002)
17. Prodan, I., Olaru, S., Niculescu, S.: Predictive control for tight group formation of multi-agent systems. In: Proceedings of the 18th IFAC World Congress, Milano, Italy, pp. 138143 (2011)
18. Stoican, F., Olaru, S., Seron, M., De Don, J.: Reference governor for tracking with fault detection
capabilities. In: Proceedings of the 2010 Conference on Control and Fault Tolerant Systems, Nice,
France, pp. 546551 (2010)
572
19. Garey, M., Johnson, D.: Computers and intractability. A Guide to the Theory of NP-Completeness.
A Series of Books in the Mathematical Sciences. Freeman, San Francisco (1979)
20. Earl, M., DAndrea, R.: Iterative MILP methods for vehicle-control problems. IEEE Trans. Robot.
21(6), 11581167 (2005)
21. Vitus, M., Pradeep, V., Hoffmann, G., Waslander, S.L., Tomlin, C.J.: Tunnel-milp: path planning with
sequential convex polytopes. In: Proceedings of AIAA Guidance, Navigation and Control Conference,
Honolulu, Hawaii, USA (2008)
22. Stoican, F., Prodan, I., Olaru, S.: On the hyperplanes arrangements in mixed-integer techniques. In:
Proceedings of the 30th American Control Conference, San Francisco, CA, USA, pp. 18981903
(2011)
23. Stoican, F., Prodan, I., Olaru, S.: Enhancements on the hyperplane arrangements in mixed integer
techniques. In: 50th IEEE Conference on Decision and Control and European Control Conference,
Orlando, FL, USA, pp. 39863991 (2011)
24. Olaru, S., Dumur, D.: Avoiding constraints redundancy in predictive control optimization routines.
IEEE Trans. Autom. Control 50(9), 14591465 (2005)
25. Orlik, P., Terao, H.: Arrangements of Hyperplanes, vol. 300. Springer, Berlin (1992)
26. Ziegler, G.: Lectures on Polytopes. Springer, Berlin (1995)
27. Orlik, P.: Hyperplane arrangements. In: Floudas, C.A., Orlik, P.M. (eds.) Encyclopedia of Optimization, pp. 15451547. Springer, New York (2009)
28. Pardalos, P.: Hyperplane arrangements in optimization. In: Floudas, C.A., Pardalos, P.M. (eds.) Encyclopedia of Optimization, pp. 15471548. Springer, New York (2009)
29. Buck, R.: Partition of Space. Am. Math. Mon. 50(9), 541544 (1943)
30. Avis, D., Fukuda, K.: Reverse search for enumeration. Discrete Appl. Math. 65(1), 2146 (1996)
31. Ferrez, J.A., Fukuda, K.: Implementations of lp-based reverse search algorithms for the zonotope
construction and the fixed-rank convex quadratic maximization in binary variables using the zram and
the cddlib libraries. Tech. rep., Mcgill (2002)
32. Kobayashi, K., Imura, J.: Modeling of discrete dynamics for computational time reduction of Model
Predictive Control. In: Proceedings of the 17th International Symposium on Mathematical Theory of
Networks and Systems, Kyoto, Japan, pp. 628633 (2006)
33. Zaslavsky, T.: Counting the faces of cut-up spaces. Am. Math. Soc. 81, 5 (1975)
34. Geyer, T., Torrisi, F., Morari, M.: Optimal complexity reduction of piecewise affine models based on
hyperplane arrangements. In: Proceedings of the 23th American Control Conference, Boston, Massachusetts, USA, vol. 2, pp. 11901195 (2004)
35. Geyer, T., Torrisi, F., Morari, M.: Optimal complexity reduction of polyhedral piecewise affine systems. Automatica 44(7), 17281740 (2008)
36. Rudell, R., Sangiovanni-Vincentelli, A.: Espresso-mv: algorithms for multiple-valued logic minimization. In: Proceedings of the IEEE Custom Integrated Circuits Conference, pp. 230234 (1985)
37. Olfati-Saber, R., Murray, R.: Distributed cooperative control of multiple vehicle formations using
structural potential functions. In: Proceedings of the 15th IFAC World Congress, Barcelona, Spain
pp. 346352 (2002)
38. Rimon, E., Koditschek, D.: Exact robot navigation using artificial potential functions. IEEE Trans.
Robot. Autom. 8(5), 501518 (1992)

Enhancements On The Hyperplanes Arrangements in Mixed-Integer Programming Techniques

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Enhancements On The Hyperplanes Arrangements in Mixed-Integer Programming Techniques

Uploaded by

Copyright:

Available Formats

J Optim Theory Appl (2012) 154:549572