Kumar, Sanjeev;
(2004)
Investigating Computational Models of Development for the Construction of Shape and Form.
Doctoral thesis (Ph.D), UCL (University College London).
Preview |
Text
Investigating computational models of development for the construction of shape and form.pdf Download (28MB) | Preview |
Abstract
A central problem in Artificial Life is that of construction. Researchers have been investigating methods of constructing solutions to problems that overcome fundamental issues of adaptation, scalability and repetition. For example, what is the correct configuration of thousands of basic elements needed to give rise to a complex adaptive solution? Such questions and the sophisticated developmental processes and mechanisms of construction evolved by natural selection provide the general source of inspiration and motivation for the field of Computational Development (CD), and consequently, the subject matter of this thesis. This thesis investigates computational models of development for the construction of shape and form. To this end, seven sets of experiments using different computational models of development are reported. A comparison of the performance and behaviour of three classes of developmental encodings for the construction of shape and form is made, and their properties, advantages and disadvantages are identified. Traditional genetic encodings are shown to suffer from severe problems of scalability, while computational models of development are shown to offer plausible solutions. A computer software testbed, the Evolutionary Developmental System, is used to investigate the self-organising construction of 3D shape and form, such as a planes, spheres, cubes and cylinders. The EDS uses a novel genetic representation inspired by developmental genetics to evolve genetic regulatory networks (GRNs). The EDS additional mechanism of receptor-mediated signal transduction is shown to construct designs that exhibit phenomena resembling biological cellular differentiation. This thesis also shows that computational models of development are useful for biologists. It presents the first ever study of the evolution of robust mechanisms of morphogen gradient maintenance for positional information in biology using genetic algorithms. Evolved mechanisms of morphogen gradient maintenance over 3- to 10-cells are shown to provide positional information leading to the emergence of spatial patterns of cellular differentiation.
| Type: | Thesis (Doctoral) |
|---|---|
| Qualification: | Ph.D |
| Title: | Investigating Computational Models of Development for the Construction of Shape and Form |
| Open access status: | An open access version is available from UCL Discovery |
| Language: | English |
| Additional information: | Thesis digitised by ProQuest |
| Keywords: | Applied sciences; Artificial Life |
| URI: | https://discovery.ucl.ac.uk/id/eprint/10100954 |
Archive Staff Only
 |
View Item |
