Ant Colony

optimizAtion

1
 SwArm intelligenCe

“Swarm intelligence (SI) is artificial

intelligence based on the collective
behaviour of decentralized, self-organized
systems”

2
 ChArACteriStiCS of SwArmS

Composed of many individuals

Individuals are homogeneous
Local interaction based on simple rules
Self-organization
Constituting a natural model particularly
suited to distributed problem solving

3
4
5
Ant Colony optimizAtion

• Optimization Technique Proposed by Marco Dorigo in the

early ’90
• Heuristic optimization method inspired by biological
systems
• Multi-agent approach for solving difficult combinatorial
optimization problems
• Has become new and fruitful research area

6
nAturAl behAvior of Ant

7
how CAn they mAnAge SuCh
greAt tASkS ?

Ants are ,
 essentially blind, deaf and dumb.
 social creatures – behavior directed to survival of colony

Question: how can ants find the short path to food sources?

8
 ShorteSt pAth

Answer:

 Ants deposit pheromones on ground that form a trail. The

trail attracts other ants.

 Pheromones evaporate faster on longer paths.

 Shorter paths serve as the way to food for most of the other
ants.

9
 AppliCAtion
• Traveling Salesman Problem

• Quadratic Assignment Problem

• Network Model Problem

• Vehicle routing

• Graph coloring

10
ACo SyStem -pSeuDoCoDe
 Often applied to TSP (Travelling Salesman Problem): shortest
path between n nodes

 Algorithm in Pseudocode:
– Initialize Trail
– Do While (Stopping Criteria Not Satisfied) – Cycle Loop
• Do Until (Each Ant Completes a Tour) – Tour Loop
• Local Trail Update
• End Do
• Analyze Tours
• Global Trail Update
– End Do

11
 Algorithm
• Ant Colony Algorithms are typically use to solve minimum
cost problems.
• We may usually have N nodes and A undirected arcs
• There are two working modes for the ants: either forwards or
backwards
• The ants memory allows them to retrace the path it has
followed while searching for the destination node
• Before moving backward on their memorized path, they
eliminate any loops from it. While moving backwards, the ants
leave pheromones on the arcs they traversed.

12
 Algorithm
• At the beginning of the search process, a constant amount of
pheromone is assigned to all arcs. When located at a node i an
ant k uses the pheromone trail to compute the probability of
choosing j as the next node:

 τijα
 if j ∈N k

=∑ τα i
pijk l∈N ik il

0 if j ∉ N i
k

• where is the neighborhood of ant k when in node i.

N ik

13
 Algorithm
• When the arc (i,j) is traversed , the pheromone value changes
as follows:

τ ij ← τ ij + ∆τ k

• By using this rule, the probability increases that forthcoming

ants will use this arc.
• After each ant k has moved to the next node, the pheromones
evaporate by the following equation to all the arcs:

τ ij ← (1 − p )τ ij , ∀(i, j ) ∈ A

14
 StepS for Solving A
problem by ACo
2. Represent the problem in the form of sets of components
and transitions, or by a set of weighted graphs, on which
ants can build solutions
3. Define the meaning of the pheromone trails
4. Define the heuristic preference for the ant while
constructing a solution
5. If possible implement a efficient local search algorithm for
the problem to be solved.
6. Choose a specific ACO algorithm and apply to problem
being solved
7. Tune the parameter of the ACO algorithm.

15
A simple TSP example []

[]

B C
2

[]

3
A
[]
D

dAB =8;dBC = 4;dCD =15;dDA =6

16
Iteration 1
[C]
[B]

C
B 3
2

[A]

A
[D]

D 4

17
How to build next sub-solution?

[A] B C

18
Iteration 2
[A,B] [B,C]

1 C 2
B

[D,A]

A
[C,D]

D 3

19
Iteration 3
[A,B,C]
[D,A,B]

C 1
B 4

[B,C,A]

3
A
[B,C,D]

2
D

20
Iteration 4
[D,A,B,C]
[C,D,A,B]

C 4
B 3

[B,C,D,A]

2
A
[A,B,C,D]

1
D

21
Path and Pheromone Evaluation

[A,B,C,D

L1 =27
1
[B,C,D,A]

L2 =25
2
[C,D,A,B]

L3 =29
3

[D,A,B,C]

4
L4 =18 Best tour

22
End of First Run

Save Best Tour (Sequence and length)

All ants die

New ants are born

23
 ADvAntAgeS AnD
DiSADvAntAgeS
 For TSPs (Traveling Salesman Problem), relatively efficient
 for a small number of nodes, TSPs can be solved by
exhaustive search
 for a large number of nodes, TSPs are very computationally
difficult to solve exponential time to convergence
 Performs better against other global optimization techniques
such as neural net, genetic algorithms, simulated annealing
 Can be used in dynamic applications (adapts to changes such
as new distances, etc.)
 Convergence is guaranteed, but time to convergence uncertain

24