The Daisyworld control system


Dyke, J. G. (2009) The Daisyworld control system. University of Sussex, Centre for Computational Neuroscience and Robotics, Doctoral Thesis .

Download

[img] PDF
Download (3226Kb)

Description/Abstract

The original Gaia Hypothesis proposed that life on Earth, along with the oceans, atmosphere and crust, forms a homeostatic system which reduces the effects of external perturbations, so that conditions are maintained to within the range that allows widespread life. Daisyworld is a simple mathematical model intended to demonstrate certain aspects of this planetary homeostasis. There have been a considerable number of extensions and developments to the original Daisyworld model. Some of this work has been produced in response to criticism of the Gaia Hypothesis and Daisyworld specifically and some has been produced by using Daisyworld as a testbed to explore a range of issues. This thesis examines the Daisyworld control system and in doing so explains how Daisyworld performs homeostasis. The control system is classified as a rein control system which is potentially applicable to a wide range of scenarios from physiological and environmental homeostasis to robotic control. A series of simple Daisyworld models are produced and aspects of the original Daisyworld are explained, in particular the inverse response to forcing: why temperature goes down on Daisyworld when the brightness of the star increases. The Daisyworld control system is evaluated within an evolutionary context. A key result is that environmental regulation emerges not despite of Darwinian evolution but because of it. Within an ecological context, it is found that increasing the complexity of a self-regulating ecosystem can increase its stability. An energy balance climate model is developed to assess the effects of non-equilibrium thermodynamic processes on the Daisyworld control system. Results are presented that support the hypothesis that when the system is in a state of maximum entropy production, homeostasis is maximised.

Item Type: Thesis (Doctoral)
Divisions: Faculty of Physical Sciences and Engineering > Electronics and Computer Science > Agents, Interactions & Complexity
ePrint ID: 272885
Date Deposited: 29 Sep 2011 14:06
Last Modified: 27 Mar 2014 20:18
Further Information:Google Scholar
URI: http://eprints.soton.ac.uk/id/eprint/272885

Actions (login required)

View Item View Item