Scaling in Complex Systems

这项研究将解决这样一个问题，即一个相互作用的多组件系统的哪些特性对经常观察到的非常健壮的伸缩负有责任。稳健性例如通过以下事实来指示：尽管当一个人从层次结构中的一个级别到下一个级别(例如从细胞器到器官再到有机体)时，尽管功能发生了变化，但缩放仍然有效。将被解决的其他明显的开放问题包括：1)为什么像生物有机体和人类城市这样不同的系统都表现出缩放规律？2)为什么这些缩放规律经常呈现相反的趋势。例如，大生物体中的细胞比小生物体中的细胞消耗更少的能量，而大城市的人比小城镇的人消耗更多的能量，更有创造力？牛津小组主要关注社会系统的社会物理模型，而帝国小组则使用统计力学来开发进化生态学的模型。在与韦斯特教授的合作中，我们的目标是加深对标度律在异质复杂系统中有多普遍的理解，以及这些律背后的原理。韦斯特教授将由伦敦帝国理工学院和牛津大学的复杂性小组接待。在该计划的头两年，韦斯特教授每年将访问英国两个半月。在韦斯特教授访问英国期间，他将参加公开会议和研讨会，以确保整个英国复杂性社区有可能受益于韦斯特教授的专业知识。

The research will address the question concerning what properties of an interacting many components system are responsible for the often observed very robust scaling. The robustness is e.g. indicated by the fact that the scaling remains valid despite the change in functionality as one crosses from one level in the hierarchy to the next, say from organelles to organs to organism etc. Other obvious open questions that will be addressed include: 1) Why systems as different as biological organisms and human cities both exhibit scaling laws?2) Why these scaling laws often exhibit opposite trends. E.g. a cell in a large organism consumes less energy than a cell in a small organism, whereas a person in a big city consumes more energy and is more creative than a person in a small town? The Oxford group is mainly concerned with socio-physics modelling of social systems, whereas the Imperial group uses statistical mechanics to developed models of evolutionary ecology. In collaboration with Prof West we aim to develop our understanding of how widespread scaling laws are in heterogeneous complex systems and what principles are behind them. Professor West will be hosted by the complexity groups at Imperial College London and Oxford University. Prof West will visit the UK for a total of two and a half months each year during the first two-years of the proposal. During Prof West's visit to UK he will participate in open meetings and give seminars to ensure that the entire UK complexity community will have the possibility to benefit from Prof West's expertise.