Thesis Abstract

<p> There is a multitude of software available for mathematical simulation, however there is no tool aimed specifically at roboticists. This project provides such a tool, designed around the terminology and research methodologies used by those interested in using biologically inspired models of neural networks to create a richer breed of robots. The simulation environment created offers tools for visualising, manipulating and recording experiments in real time, offering insight into modelling possibilities which may be suitable for robot based control systems.<br>Also presented is some research, using the simulator created, aimed at suggesting some new possibilities for cause of a phenomenon found in many real neural cells. Neural cells are often found to control the activity of other neural cells, and the mechanisms which permit that control are investigated.<br>The results show that previous conclusions about the behaviour of neurons in such scenarios may have been restricted to only a subset of cases. They also demonstrate the difficulty facing roboticists who wish to use complex models to control artificial hardware. <br></p>

Thesis Overview

<p> 1. INTRODUCTION __________________________________________________________ 1<br>1.1. Motivation ______________________________________________________________ 2<br>1.2. Report Structure __________________________________________________________ 2<br>2. BACKGROUND_____________________________________________________________ 3<br>2.1. Autonomous Robotics _____________________________________________________ 3<br>2.2. First Efforts _____________________________________________________________ 3<br>2.3. ‘Engineered’ Robotic Learning Methods ________________________________________ 4<br>2.4. The Influence of Cognitive Neuroscience_______________________________________ 5<br>2.5. Bioinspired Neural Paradigms________________________________________________ 7<br>2.6. Using Neuron Models in Biological Research ___________________________________ 12<br>2.7. The Nature of Robotic Investigations_________________________________________ 14<br>2.8. Tools Assisting Theoretical Robotics _________________________________________ 17<br>3. METHODOLOGY _________________________________________________________ 20<br>3.1. Requirements ___________________________________________________________ 20<br>3.2. Program Structure________________________________________________________ 22<br>3.3. Design Detail ___________________________________________________________ 24<br>4. IMPLEMENTATION_______________________________________________________ 32<br>4.1. Interactive Simulation_____________________________________________________ 32<br>4.2. Batch Mode ____________________________________________________________ 37<br>5. TESTING _________________________________________________________________ 39<br>5.1. Code Validation _________________________________________________________ 39<br>5.2. Research carried out using the simulator_______________________________________ 40<br>5.3. Conclusions of the Entrainment Study ________________________________________ 45<br>6. EVALUATION_____________________________________________________________ 47<br>6.1. Potential Future Work ____________________________________________________ 49<br>7. REFERENCES ____________________________________________________________ 51<br>Appendix A REFLECTION____________________________________________________ 53<br>Appendix B PROJECT SCHEDULE_____________________________________________ 55<br>B.1. Diary Extract ___________________________________________________________ 56<br>IV<br>Appendix C TEST OUTPUT ___________________________________________________ 58<br>Appendix D EXAMPLE SIMULATION SCREENSHOT____________________________ 61<br>Appendix E SCRIPTING LANGUAGE STRUCTURE______________________________ <br></p>