an evolutionary approach to space layout planning using genetic algorithm by: hoda homayouni

Post on 22-Dec-2015

226 Views

Category:

Documents

0 Downloads

Preview:

Click to see full reader

TRANSCRIPT

An Evolutionary Approach To Space Layout Planning Using

Genetic Algorithm

By: Hoda Homayouni

Introduction to Space Layout Planning

• What is Space Layout Planning?• Motivation• Challenges:

– Solving ill defined problems– Addressing qualitative constraints– Having creativity– Compatibility with architects

Introduction to Genetic Algorithm

Computer Algorithm that resides on principles of genetic and evolution.

4

Why Genetic Algorithm?

• Hill climbing

locallocal

globalglobal

Why Genetic Algorithm?

• Multi-climbers

Why Genetic Algorithm?

• Genetic algorithm

I am not at the top.I am not at the top.My high is better!My high is better!

I am at the I am at the toptop

Height is ...Height is ...

I will continueI will continue

Why Genetic Algorithm?

• Genetic algorithm

few microseconds after

Encoding Chromosomes

• The chromosome should in some way contain information about solution which it represents

Crossover

• Crossover selects genes from parent chromosomes and creates a new offspring

Mutation

• This is to prevent falling all solutions in population into a local optimum of solved problem

Fitness Function

• Fitness function is evaluation function,that determines what solutions are better than others.

• Fitness is computed for each individual.

• Fitness function is application depended.

Algorithmic Phases

Initialize the populationInitialize the population

Select individuals for the mating poolSelect individuals for the mating pool

Perform crossoverPerform crossover

Insert offspring into the populationInsert offspring into the population

The EndThe End

Perform mutationPerform mutation

yesyes

nono

Stop?Stop?

Genetic Engineering Approach

M1

M3

M2

M4

• An object can be described by the location of units and can be ‘grown’ by locating a required number of such units, one at a time in sequence.

Genetic Engineering

Evolving Complex Design Genes Using a Hierarchical Growth

Approach

HOUSE

ZONE1 ZONE2 ZONEz• • • • • • • • • •

ROOM1 ROOM2 ROOMm• • •  

S_UNIT1 S_UNIT2 S_UNITs• • •  

Generating Units

W1

S2

N1 N2

E1

S1

W1

S2

N1

W2

N2

E1

E2

S1

W1

S1

N1

E1

P1(p) = (W1,N1,E1,S1)P2(g) = (P1,P1,E1|W1) P2(p) = (W1,N1,N2,E1,S1,S2)

P3(g) = (P2,P1,N2|S1) P3(p) = (W1,N1,W2,N2,E1,E2,S1,S2)

Crossover at Room Level

R1 = (N1|S1, E1|W1, E1 |W1, S1|N1, S1|N1)

1 2 3R2 = (E1 |W1, N1 |S1, W 1|E1, N2|S1, W 3|E1)

1 2 34 4

N1

W1

W2S1

W3

N2 N3

E1

E2

E3

S2S3S1

W1

W2

N1

S3

E4

E1

N2

S2

N3

W4E2

E3W3

S4

N4

R1 = (N1|S1, E1|W1, E1|W1, S1|N1, S1 |N1) R2 = (E1 |W1, N1|S1, W1|E1, N2 |S1, W3 |E1)

R3 = (N1|S1, E1|W1, W1|E1, N2 |S1, W3 |E1) R4 = (E1 |W1 , N1|S1, E1|W1, S1|N1, S1 |N1)

2

Crossover at Room Level

R1 = (N1|S1, E1|W1, E1 |W1, S1|N1, S1|N1)

1 2 3R2 = (E1 |W1 , N1 |S1, W 1|E1 , N2 |S1, W 3|E1)

1 2 34 4

N1

W1

W2S1

W3

N2 N3

E1

E2

E3

S2S3S1

W1

W2

N1

S3

E4

E1

N2

S2

N3

W4E2

E3W3

S4

N4

R1 = (N1|S1, E1|W1, E1|W1, S1|N1, S1 |N1) R2 = (E1 |W1, N1|S1, W1|E1, N2 |S1, W3 |E1)

R5 = (N1|S1, E1|W1, E1|W1, S1|N1, W3 |E1) R6 = (E1 |W1, N1|S1, W1|E1, N2 |S1, S1 |N1)

4

Crossover at Site LevelLiving Room EntranceDining Room

D1 E2D2L1 L2

E1

Z1 = ((L1, D2, E4|W2), E1, N5|S3)

1 2 3 4

E1

D2L1

Z2 = ((L2, D1, E3|W2), E2, S2|N3)

1 2 3 4

L2 D1

E2

Z5 = ((L1, D2, E3|W2), E2, S2|N3) Z6 = ((L2, D1, E4|W2), E1, N5|S3)

SITE 2 - CROSSOVER

D2

E2

L1 E1L2

D1

Initial Living Zone Population

Evolved Population

Initial House population

Evolved Population

Discussion

• More Fitness Functions

• Architects Role?

References

• Rosenman, M.A. (1997). The Generation of form using evolutionary approach”. Evolutionary algorithms in Engineering Applications. Springer, 1997.

• Rosenman, M.A. and Gero, J.S. (1999) Evolutionary designs by generating useful complex gene structures. Evolutionary Design by Computers, Morgan Kaufmann, San Francisco, pp.345-364.

• http://galeb.etf.bg.ac.yu/~vm/GenAlgo.ppt

top related