Artificial Biochemical Networks: Computational Models and Architectures

我们和其他人以前的工作已经表明，生物有机体的结构和组织如何激励计算机硬件和软件的设计，目的是捕获有用的特性，如复杂的信息处理和对环境扰动的抵抗力。这个建议集中在生物系统中发现的最复杂的结构之一：生化网络。这些结构对于生物有机体的发育、功能和进化至关重要，也是高等生物体内部复杂性的主要因素。以前试图建立硬件和软件系统的动机，这些结构导致了一组计算机架构，我们统称为人工生物化学网络模型。其中最著名的是人工遗传网络，它已被证明是表达复杂计算行为的有效手段，特别是在机器人控制中。然而，这一领域的研究得到了相对较少的关注，并鲜为人知的是，这些架构的计算性能。拟议工作的目的是开发更好的人工生物化学网络模型，我们将通过整合现有工作和引入生物科学的新理解来实现这一目标。我们还将开发一个理论框架，以更好地理解这些计算架构的能力，并展示如何将这些模型应用于控制机器人在真实的世界环境中的难题。预计这项工作也将产生对生物系统的功能和进化的见解，这些生物系统是建立在这些结构的基础上的。

Previous work by ourselves and others has shown how the structure and organisation of biological organisms can motivate the design of computer hardware and software, with the aim of capturing useful properties such as complex information processing and resistance to environmental perturbation. This proposal focuses upon one of the most complex sets of structures found in biological systems: biochemical networks. These structures are fundamental to the development, function and evolution of biological organisms, and are the main factor underlying the complexity seen within higher organisms. Previous attempts to build hardware and software systems motivated by these structures has led to a group of computer architectures which we collectively refer to as artificial biochemical network models. The best known of these is the artificial genetic network, which has shown itself to be an effective means of expressing complex computational behaviours, particularly within robotic control. Nevertheless, this field of research has received relatively little attention, and little is known about the computational properties of these architectures. The aim of the proposed work is to develop better artificial biochemical network models, which we will do by both bringing together existing work and introducing new understanding from the biological sciences. We will also develop a theoretical framework to better understand what these computational architectures are capable of, and show how how these models can be applied to the difficult problem of controlling a robot in real world environments. It is expected that this work will also produce insights into the function and evolution of the biological systems on which the architectures are modelled.