Artificiel Bee Colony (ABC) Algorithme Isfahan University of Technology Fall 2010 1 Elham...

39
Artificiel Bee Colony (ABC) Algorithme Isfahan University of Technology Fall 2010 1 Elham Seifossadat Faegheh Javadi

Transcript of Artificiel Bee Colony (ABC) Algorithme Isfahan University of Technology Fall 2010 1 Elham...

Page 1: Artificiel Bee Colony (ABC) Algorithme Isfahan University of Technology Fall 2010 1 Elham Seifossadat Faegheh Javadi.

Artificiel Bee Colony (ABC) Algorithme

Isfahan University of Technology Fall 2010

1

Elham SeifossadatFaegheh Javadi

Page 2: Artificiel Bee Colony (ABC) Algorithme Isfahan University of Technology Fall 2010 1 Elham Seifossadat Faegheh Javadi.

A BEE COLONY OPTIMIZATION ALGORITHM TO JOB SHOP SCHEDULING

Job shop scheduling problems are considered to bea member of a large class of intractable numerical problemsknown as NP-hard.

Job shop scheduling is concerned with finding a sequentialallocation of competing resources that optimizes a particularobjective function.

Isfahan University of Technology Fall 2010

2

Page 3: Artificiel Bee Colony (ABC) Algorithme Isfahan University of Technology Fall 2010 1 Elham Seifossadat Faegheh Javadi.

Isfahan University of Technology Fall 2010

3

A BEE COLONY OPTIMIZATION ALGORITHM TO JOB SHOP SCHEDULING

Isfahan University of Technology Fall 2010

3

A finite set J of n jobs to be processed on a finite set M of m machines.

Each job Ji must be processed on every machine and consists of a chain of mi operations Oi1, Oi2,…,Oim which have to be scheduled in a pre-determined given order.

Oij is the jth operation of job Ji which has to be processed on a machine Mx for a processing time period of τij without interruption and preemption.

Page 4: Artificiel Bee Colony (ABC) Algorithme Isfahan University of Technology Fall 2010 1 Elham Seifossadat Faegheh Javadi.

Isfahan University of Technology Fall 2010

4 Isfahan University of Technology Fall 2010

4

A BEE COLONY OPTIMIZATION ALGORITHM TO JOB SHOP SCHEDULING

Each machine can process only one job and each job can be processed by only one machine at a time.

The longest duration in which all operations of all jobs are completed is referred to as the makespan Cmax.

Page 5: Artificiel Bee Colony (ABC) Algorithme Isfahan University of Technology Fall 2010 1 Elham Seifossadat Faegheh Javadi.

Isfahan University of Technology Fall 2010

5 Isfahan University of Technology Fall 2010

5

A BEE COLONY OPTIMIZATION ALGORITHM TO JOB SHOP SCHEDULINGAi be the set of ordered pairs of operations

constrained by the precedence relations for each job Ji.

For each machine Mx, the set Ex describes the set of all pairs of operations to be performed on the machine.

For each operation Oij, let its earliest possible process start time be Tij.

Page 6: Artificiel Bee Colony (ABC) Algorithme Isfahan University of Technology Fall 2010 1 Elham Seifossadat Faegheh Javadi.

Isfahan University of Technology Fall 2010

6

A BEE COLONY OPTIMIZATION ALGORITHM TO JOB SHOP SCHEDULING

Page 7: Artificiel Bee Colony (ABC) Algorithme Isfahan University of Technology Fall 2010 1 Elham Seifossadat Faegheh Javadi.

A BEE COLONY OPTIMIZATION ALGORITHM TO JOB SHOP SCHEDULING

Isfahan University of Technology Fall 2010

7

Page 8: Artificiel Bee Colony (ABC) Algorithme Isfahan University of Technology Fall 2010 1 Elham Seifossadat Faegheh Javadi.

A BEE COLONY OPTIMIZATION ALGORITHM TO JOB SHOP SCHEDULING

Isfahan University of Technology Fall 2010

8

Page 9: Artificiel Bee Colony (ABC) Algorithme Isfahan University of Technology Fall 2010 1 Elham Seifossadat Faegheh Javadi.

A BEE COLONY OPTIMIZATION ALGORITHM TO JOB SHOP SCHEDULING

The challenge is to adapt the self-organization behavior of the colony for solving job shop scheduling problems.

There are two major characteristics of the bee colony in searching for food sources: waggle dance and forage (or nectar exploration).

Isfahan University of Technology Fall 2010

9

Page 10: Artificiel Bee Colony (ABC) Algorithme Isfahan University of Technology Fall 2010 1 Elham Seifossadat Faegheh Javadi.

A BEE COLONY OPTIMIZATION ALGORITHM TO JOB SHOP SCHEDULING

waggle dance• A forager fi on return to the hive from nectar

exploration will attempt with probability p to perform waggle dance on the dance floor with duration D = di A,

• it will also attempt with probability ri to observe and follow a randomly selected dance.

Isfahan University of Technology Fall 2010

10

Page 11: Artificiel Bee Colony (ABC) Algorithme Isfahan University of Technology Fall 2010 1 Elham Seifossadat Faegheh Javadi.

A BEE COLONY OPTIMIZATION ALGORITHM TO JOB SHOP SCHEDULING

Isfahan University of Technology Fall 2010

11

Profitability rating for a forager:

Page 12: Artificiel Bee Colony (ABC) Algorithme Isfahan University of Technology Fall 2010 1 Elham Seifossadat Faegheh Javadi.

A BEE COLONY OPTIMIZATION ALGORITHM TO JOB SHOP SCHEDULING

Isfahan University of Technology Fall 2010

12

The bee colony’s average profitability rating:

Page 13: Artificiel Bee Colony (ABC) Algorithme Isfahan University of Technology Fall 2010 1 Elham Seifossadat Faegheh Javadi.

A BEE COLONY OPTIMIZATION ALGORITHM TO JOB SHOP SCHEDULING

Isfahan University of Technology Fall 2010

13

The dance duration:

Page 14: Artificiel Bee Colony (ABC) Algorithme Isfahan University of Technology Fall 2010 1 Elham Seifossadat Faegheh Javadi.

A BEE COLONY OPTIMIZATION ALGORITHM TO JOB SHOP SCHEDULING

Isfahan University of Technology Fall 2010

14

Adjusting Probability of Following a Waggle Dance:

Page 15: Artificiel Bee Colony (ABC) Algorithme Isfahan University of Technology Fall 2010 1 Elham Seifossadat Faegheh Javadi.

A BEE COLONY OPTIMIZATION ALGORITHM TO JOB SHOP SCHEDULING

Forage (Nectar Exploration)

Isfahan University of Technology Fall 2010

15

• A population of L foragers is defined in the colony.

• When a forager is at a specific node, it can only move to next node that is defined in a list of presently allowed nodes, imposed by precedence constraints of operations.

Page 16: Artificiel Bee Colony (ABC) Algorithme Isfahan University of Technology Fall 2010 1 Elham Seifossadat Faegheh Javadi.

A BEE COLONY OPTIMIZATION ALGORITHM TO JOB SHOP SCHEDULING

Isfahan University of Technology Fall 2010

16

A forager chooses the next node from the list according to the state transition rule:

Page 17: Artificiel Bee Colony (ABC) Algorithme Isfahan University of Technology Fall 2010 1 Elham Seifossadat Faegheh Javadi.

A BEE COLONY OPTIMIZATION ALGORITHM TO JOB SHOP SCHEDULING

Isfahan University of Technology Fall 2010

17

The rating ρij of the edge (directed) between node I and node j is given by:

Page 18: Artificiel Bee Colony (ABC) Algorithme Isfahan University of Technology Fall 2010 1 Elham Seifossadat Faegheh Javadi.

A BEE COLONY OPTIMIZATION ALGORITHM TO JOB SHOP SCHEDULING

Isfahan University of Technology Fall 2010

18

Page 19: Artificiel Bee Colony (ABC) Algorithme Isfahan University of Technology Fall 2010 1 Elham Seifossadat Faegheh Javadi.

A BEE COLONY OPTIMIZATION ALGORITHM TO JOB SHOP SCHEDULING

The performance of the honey bee colony scheduling approach is studied by evaluating them on the following 82 job shop problem instances.

The sizes of these problems range from 6 to 50 jobs and 5 to 20 machines.

Isfahan University of Technology Fall 2010

19

Page 20: Artificiel Bee Colony (ABC) Algorithme Isfahan University of Technology Fall 2010 1 Elham Seifossadat Faegheh Javadi.

A BEE COLONY OPTIMIZATION ALGORITHM TO JOB SHOP SCHEDULING

Isfahan University of Technology Fall 2010

20

Page 21: Artificiel Bee Colony (ABC) Algorithme Isfahan University of Technology Fall 2010 1 Elham Seifossadat Faegheh Javadi.

A BEE COLONY OPTIMIZATION ALGORITHM TO JOB SHOP SCHEDULING

Isfahan University of Technology Fall 2010

21

Page 22: Artificiel Bee Colony (ABC) Algorithme Isfahan University of Technology Fall 2010 1 Elham Seifossadat Faegheh Javadi.

Bee Colony Optimization (BCO)

Isfahan University of Technology Fall 2010

22

Page 23: Artificiel Bee Colony (ABC) Algorithme Isfahan University of Technology Fall 2010 1 Elham Seifossadat Faegheh Javadi.

Bee Colony Optimization (BCO)

There are two alternating phases (forward pass and backward pass) constituting single step in the BCO algorithm.

The hive is an non-natural object, with no precise location and does not influence the algorithm execution.

Isfahan University of Technology Fall 2010

23

Page 24: Artificiel Bee Colony (ABC) Algorithme Isfahan University of Technology Fall 2010 1 Elham Seifossadat Faegheh Javadi.

Bee Colony Optimization (BCO)In each forward pass, every artificial bee

visits NC solution components, creates partial solution, and after that returns to the hive.

Isfahan University of Technology Fall 2010

24

Page 25: Artificiel Bee Colony (ABC) Algorithme Isfahan University of Technology Fall 2010 1 Elham Seifossadat Faegheh Javadi.

Bee Colony Optimization (BCO)

Isfahan University of Technology Fall 2010

25

Page 26: Artificiel Bee Colony (ABC) Algorithme Isfahan University of Technology Fall 2010 1 Elham Seifossadat Faegheh Javadi.

Bee Colony Optimization (BCO)In the backward pass, all artificial bees share

information about the quality of their partial solutions. Having all solutions evaluated, each bee decides with a certain probability whether it will stay loyal to its solution or not.

Isfahan University of Technology Fall 2010

26

Page 27: Artificiel Bee Colony (ABC) Algorithme Isfahan University of Technology Fall 2010 1 Elham Seifossadat Faegheh Javadi.

Bee Colony Optimization (BCO)

Isfahan University of Technology Fall 2010

27

Page 28: Artificiel Bee Colony (ABC) Algorithme Isfahan University of Technology Fall 2010 1 Elham Seifossadat Faegheh Javadi.

Bee Colony Optimization (BCO)

Isfahan University of Technology Fall 2010 28

Page 29: Artificiel Bee Colony (ABC) Algorithme Isfahan University of Technology Fall 2010 1 Elham Seifossadat Faegheh Javadi.

Bee Colony Optimization (BCO)When all solutions are completed the best

one is determined, it is used to update global best solution and an iteration of the BCO is accomplished.

At this point all B solutions are deleted, and the new iteration could start. The BCO runs iteration by iteration until a stopping condition is met.

Isfahan University of Technology Fall 2010

29

Page 30: Artificiel Bee Colony (ABC) Algorithme Isfahan University of Technology Fall 2010 1 Elham Seifossadat Faegheh Javadi.

Bee Colony Optimization (BCO)B - The number of bees in the hive;

NC - The number of constructive moves during one forward pass.

Isfahan University of Technology Fall 2010

30

Page 31: Artificiel Bee Colony (ABC) Algorithme Isfahan University of Technology Fall 2010 1 Elham Seifossadat Faegheh Javadi.

Bee Colony Optimization (BCO)(1) Initialization: an empty solution is assigned to each bee;(2) For each bee: // (the forward pass)

(a) Set k = 1; // (count constructive moves in the forward pass)(b) Evaluate all possible constructive moves;(c) Choose one move using the roulette wheel;(d) k = k + 1; If k ≤ NC Goto step (b).

(3) All bees are back to the hive; // (backward pass starts)(4) Evaluate (partial) objective function value for each bee;(5) Each bee decides randomly whether to continue its own exploration and become a recruiter,

or to become a follower;(6) For each follower, choose a new solution from recruiters by the roulette wheel;(7) If solutions are not completed Goto step 2;(8) Evaluate all solution and find the best one;(9) If the stopping criteria is not met Goto step 2;(10) Output the best solution found.

Isfahan University of Technology Fall 2010 31

Page 32: Artificiel Bee Colony (ABC) Algorithme Isfahan University of Technology Fall 2010 1 Elham Seifossadat Faegheh Javadi.

Bee Colony Optimization (BCO)

Loyalty decision

Isfahan University of Technology Fall 2010

32

Page 33: Artificiel Bee Colony (ABC) Algorithme Isfahan University of Technology Fall 2010 1 Elham Seifossadat Faegheh Javadi.

Bee Colony Optimization (BCO)

Isfahan University of Technology Fall 2010

33

Recruiting process

Page 34: Artificiel Bee Colony (ABC) Algorithme Isfahan University of Technology Fall 2010 1 Elham Seifossadat Faegheh Javadi.

Scheduling Independent Tasks by BCO

Let T = {1, 2, . . . , n} be a given set of independent tasks, and P = {1, 2, . . .,m} a set of identical machines.

The processing time of task i (i = 1, 2, . . . , n) is denoted by li.

Isfahan University of Technology Fall 2010

34

Page 35: Artificiel Bee Colony (ABC) Algorithme Isfahan University of Technology Fall 2010 1 Elham Seifossadat Faegheh Javadi.

Scheduling Independent Tasks by BCO

Isfahan University of Technology Fall 2010

35

Probability pi that specific bee chooses task i was equal:

Page 36: Artificiel Bee Colony (ABC) Algorithme Isfahan University of Technology Fall 2010 1 Elham Seifossadat Faegheh Javadi.

Scheduling Independent Tasks by BCO

Isfahan University of Technology Fall 2010

36

Probability pj that specific bee chooses processor j was calculated as:

Page 37: Artificiel Bee Colony (ABC) Algorithme Isfahan University of Technology Fall 2010 1 Elham Seifossadat Faegheh Javadi.

Scheduling Independent Tasks by BCO

Isfahan University of Technology Fall 201037

Page 38: Artificiel Bee Colony (ABC) Algorithme Isfahan University of Technology Fall 2010 1 Elham Seifossadat Faegheh Javadi.

References A BEE COLONY OPTIMIZATION ALGORITHM TO JOB SHOP

SCHEDULING- Chong, Low, Sivakumar, and Gay-Proceedings of the 2006 Winter Simulation Conference.

Bee Colony Optimization: The Applications Survey - DUˇSAN TEODOROVI´C TATJANA DAVIDOVI´C and MILICA ˇSELMI´C- ACM Transactions on Computational Logic,2011.

Isfahan University of Technology Fall 2010

38

Page 39: Artificiel Bee Colony (ABC) Algorithme Isfahan University of Technology Fall 2010 1 Elham Seifossadat Faegheh Javadi.

Thanks for your attention!