Non-Equilibrium Statistical Mechanics of Co-Evolving Complex Systems

共同演化复杂系统的非平衡统计力学

• 批准号: 1507371
• 负责人: Kevin Bassler
• 金额: $ 32.4万
• 依托单位: University of Houston
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2016
• 资助国家: 美国
• 起止时间: 2016-01-15 至 2019-12-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1507371&HistoricalAwards=false
• 关键词: Non; Equilibrium; Statistical; Mechanics; Co

NONTECHNICAL SUMMARYThis award supports research and education aimed to find fundamental principles that govern the behavior of complex systems, an important and challenging problem. A complex system often consists of components that each behave in a well-defined way, but when combined display new distinctly different behavior that is more than the sum of the individual behaviors. The new behavior is said to be emergent. The components of complex systems often have heterogeneous properties and interactions. Examples abound in nature, ranging from the brain and the regulatory systems of cells in biology, to social networks and ecosystems, and from interdependent infrastructure networks in engineering to atmospheric and oceanic dynamics. A principal difficulty in understanding many complex systems is that their dynamics is driven by external mechanisms that prevent them from reaching the steady state of equilibrium. Thus, the foundational principles that govern equilibrium systems, discovered over a century ago, typically do not apply to them. The analogous principles of driven, non-equilibrium systems are poorly understood, necessitating the need for finding the foundational principles of the behavior of non-equilibrium systems and for understanding the range of their possible behavior. This award supports research to address this problem by focusing on three different types of non-equilibrium complex systems and their generalizations. Each system is motivated as a model that captures the essence of real neural, biological or social behavior, and so, understanding their specific properties is important. The goal is either to develop or enhance mathematical methods and tools that can be used generally to study co-evolving complex systems, and thus, to gain transformational insight into complicated systems such as the human brain.The work on each of the systems will build on recent preliminary results obtained by the investigators in their ongoing research effort. The research will be both analytical and computational, and will range from fundamental problems in mathematical physics, to the analysis and application of the fundamental ideas to real - world experimental data. The funded activities will significantly contribute to ongoing efforts at the University of Houston in both Computational and Network Science. These efforts foster interdisciplinary collaboration within the University and with local industry. The PI and his group will participate in outreach activities for the community each year. The PI will also continue to teach multidisciplinary graduate courses he has designed to broadly educate students about recent advances in both Computation and Network Science. The grant will also support the PI's service as a faculty facilitator in the continuing education training each year for 20 teachers without a physics background that teach physics to approximately 3000 students in high-need school districts in the Houston area. It will also support the Co-PI's collaboration on a new text that teaches physics to students in the biological sciences. His focus is on explaining the central role that understanding systems far from equilibrium plays in all biological organisms. The graduate students supported by the grant will be trained in broadly applicable analytic and computational skills that will prepare them for a wide range of career opportunities. They will also contribute to the planned educational and outreach activities. Involving people from under-represented groups in this work will also be a focus. TECHNICAL SUMMARYThis award supports research and education aimed to find fundamental principles that govern the behavior of complex systems. A principal difficulty in understanding many complex systems is that their dynamics is driven by exogenous mechanisms, preventing them from reaching thermal equilibrium. Unlike equilibrium statistical mechanics, the principles of nonequilibrium statistical mechanics remain poorly understood. Yet, such systems are ubiquitous, ranging from the neural, genetic, and metabolic regulatory systems in biology, to population dynamics and competition in social systems and ecosystems, and from interdependent infrastructure networks to atmospheric/oceanic dynamics at the global scale. Establishing the foundational principles of nonequilibrium statistical mechanics as they apply to complex systems would transform not only our understanding of nature, but also our ability to control it. In order to establish the principles of nonequilibrium statistical mechanics, it is helpful to understand the range of possible behavior in complex systems, driven far from equilibrium. A particularly challenging set of such systems to understand is those in which two or more distinct components co-evolve in an interdependent manner. Systems structured as complex networks that have interdependent dynamics in both node- and link- degrees of freedom are prime examples of co-evolving complex systems. The research will involve three sets of simple model network systems that capture the essence of non-equilibrium behavior found in nature. These models are also chosen because they can be efficiently simulated numerically, and have some aspects that can be understood analytically. The PIs will combine careful simulations with analytic calculations, to gain insight into both fundamental issues in non-equilibrium statistical mechanics, and to develop deeper understanding of the intriguing phenomena displayed in real-world systems. The first set of projects will study the co-evolutionary dynamics of social network models consisting of interacting agents endowed with a temperament as either introverts or extroverts. Game-theoretic node dynamics will be combined interdependently with link dynamics due to the temperament of the nodes. These models merge two paradigms of non-equilibrium statistical mechanics, one for the node dynamics and the other for the link dynamics, to explore new co-evolutionary phenomena. The second set of projects will study the adaptive dynamics of Boolean networks in which both nodes and links co-evolve. They are prototypical examples of heterogeneous complex systems and as such present distinct challenges to their understanding. They were originally proposed as simple models of genetic regulatory systems, and recently have become widely used as simple models of neural systems. Understanding the co-evolution of the network structure and the rules of node behavior of these models may unlock keys to understanding neural function. Finally, the third set of projects will extend the PI's recent work on the co-evolutionary dynamics leading to Emergent Speciation, a novel method of biological speciation that may be a root cause of the origin of the species.

该奖项支持旨在发现控制复杂系统行为的基本原理的研究和教育，这是一个重要而具有挑战性的问题。一个复杂的系统通常由组件组成，每个组件的行为都以一种明确定义的方式进行，但当它们组合在一起时，就会显示出新的、明显不同的行为，这种行为比单个行为的总和还要多。这种新的行为被称为涌现行为。复杂系统的组成部分通常具有异质性质和相互作用。自然界中有很多这样的例子，从生物学中的大脑和细胞调节系统，到社会网络和生态系统，从工程中相互依存的基础设施网络到大气和海洋动力学。理解许多复杂系统的一个主要困难是，它们的动力学是由阻止它们达到稳定平衡状态的外部机制驱动的。因此，一个多世纪前发现的控制平衡系统的基本原则，通常不适用于它们。人们对驱动的非平衡系统的类似原理了解甚少，因此需要找到非平衡系统行为的基本原理，并了解其可能行为的范围。该奖项支持通过关注三种不同类型的非平衡复杂系统及其概括来解决这一问题的研究。每个系统都是作为一个模型来驱动的，这个模型捕捉了真实的神经、生物或社会行为的本质，因此，理解它们的特定属性是很重要的。目标是开发或增强数学方法和工具，这些方法和工具可用于研究共同进化的复杂系统，从而获得对复杂系统（如人类大脑）的转化性见解。每个系统的工作都将建立在研究人员在他们正在进行的研究工作中获得的最近初步结果的基础上。研究将是分析性和计算性的，范围将从数学物理的基本问题，到对现实世界实验数据的基本思想的分析和应用。所资助的活动将大大有助于休斯敦大学在计算科学和网络科学方面的持续努力。这些努力促进了大学内部以及与当地工业的跨学科合作。PI和他的团队每年都会参加社区的外展活动。PI还将继续教授他设计的多学科研究生课程，以广泛教育学生关于计算和网络科学的最新进展。这笔赠款还将支持PI每年为20名没有物理背景的教师提供继续教育培训，这些教师为休斯顿地区高需求学区的约3000名学生教授物理。它还将支持Co-PI合作编写一套新教材，向生物科学专业的学生教授物理学。他的重点是解释理解远离平衡的系统在所有生物有机体中所起的核心作用。受资助的研究生将接受广泛适用的分析和计算技能的培训，为他们提供广泛的职业机会。他们还将为计划中的教育和外联活动作出贡献。让代表性不足的群体参与这项工作也将是一个重点。该奖项支持旨在发现控制复杂系统行为的基本原理的研究和教育。理解许多复杂系统的一个主要困难是它们的动力学是由外生机制驱动的，阻止它们达到热平衡。不像平衡统计力学，非平衡统计力学的原理仍然知之甚少。然而，这样的系统是无处不在的，从生物学中的神经、遗传和代谢调节系统，到社会系统和生态系统中的种群动态和竞争，从相互依存的基础设施网络到全球范围内的大气/海洋动态。建立应用于复杂系统的非平衡统计力学的基本原理，不仅会改变我们对自然的理解，还会改变我们控制自然的能力。为了建立非平衡统计力学原理，理解远离平衡状态的复杂系统的可能行为范围是有帮助的。理解这类系统的一个特别具有挑战性的集合是那些两个或更多不同的组件以相互依赖的方式共同进化的系统。结构为复杂网络的系统在节点自由度和链路自由度上都具有相互依赖的动态，这是共同进化的复杂系统的主要例子。这项研究将涉及三组简单的模型网络系统，这些系统捕获了自然界中发现的非平衡行为的本质。选择这些模型还因为它们可以有效地进行数值模拟，并且具有可以解析理解的某些方面。pi将结合仔细的模拟和分析计算，以深入了解非平衡统计力学的基本问题，并对现实世界系统中显示的有趣现象有更深的理解。第一组项目将研究社会网络模型的共同进化动力学，该模型由具有内向或外向气质的相互作用主体组成。由于节点的性质，博弈论节点动力学将与链路动力学相互依赖地结合在一起。这些模型融合了两种非平衡统计力学范式，一种用于节点动力学，另一种用于链接动力学，以探索新的共同进化现象。第二组项目将研究布尔网络的自适应动力学，其中节点和链接共同进化。它们是异质复杂系统的典型例子，因此对它们的理解提出了不同的挑战。它们最初是作为遗传调控系统的简单模型提出的，最近被广泛用作神经系统的简单模型。了解网络结构的协同进化和这些模型的节点行为规则可能是理解神经功能的关键。最后，第三组项目将扩展PI最近在导致紧急物种形成的共同进化动力学方面的工作，这是一种新的生物物种形成方法，可能是物种起源的根本原因。