Genetic Algorithms

(C) Andres Harris

Genetic algorithms in design work similarly to genetic codes in science.  Each phenome is ma(C) wikipedia commonsde up from a genetic code defined by combiations of four proteins.

 

 

 

 

 

 

 

These combinations define a genome; the way an organism will look like.In design we can assign basic codes to mutations of a simple shape, that subsequentially after a number of combinations would form a different morphological system in a macro-scale.

 

(C) Andres Harris

 

The combination of two genomes (“parents”) who are products of mutations themselves, would generate a new “offspring” which combined with other offsprings will generate a new generation.

Fitness criteria (in this case a rational selection done by the designer) would determine the fittest “parents” that would generate a new generation.  The fitness criteria in design could be rules of selection that would allowan expedite  genetic form finding process, and also higher options for mutations.

In this example the fintess criteria is determined by flat edges that would allow the shapes combine and attach to each other, leaving the ones that show the higher of exposed vertices out of the selection process.

(C) Andres Harris

 

 

Hierarchies appear based on organisations at different scales of the system; based on orientation of the elements, direction, sizes,etc.

These organisations generate patterns which can be distinguished in different scales.

(C) Andres Harris

 

 

 

 

 

 

(C) Andres Harris

(C) Andres Harris

(C) Andres Harris

 

 

 

 

 

 

 

© Andres Harris – All rights reserved