Advanced Software Development for Complex Systems Modelling

Flat(Flexile Large-Scale Agents Modeling Environment)是一款方便对多个简单个体(智能体)进行模拟的软件。它为这些代理提供建模功能，允许设置变量和交互规则。一旦定义了代理模型，就可以运行模拟，由于代理之间在许多时间步长上的交互而产生紧急行为。这在每个时间步长更新代理及其变量(如有必要)时发生。可以用这种方式模拟的系统的一些例子可以在FLAME网站(http://flame.ac.uk/projects/))上找到，包括生物和经济系统。我提出的研究课题(复杂系统建模的高级软件开发)将集中在FLAME-II中使用的核心方法和算法的进一步开发，FLAM-II是一个基于FLAM的框架，用于并行基于代理的应用程序。本项目将涉及的研究问题和假设尚未确定，但它们将遵循这样的思路：“是否有可能在FLAME-II内实施或改进功能，从而允许使用FLAME-II框架进行有效和可行的模拟？”项目标题足够宽泛，让我可以在项目的第一年左右，将这个问题细化到特定学科中的特定主题。一旦问题(S)决定了，我会再分配一年，外加回旋余地来实现所需的标准，然后再用剩余的时间来完成一篇论文，并确保功能是正确的。因为我目前对这个项目感兴趣的领域是生物系统的模拟，这是我的研究最初的方向。这可能涉及对可能从模拟中受益的领域进行个别和有指导的研究，以及与专家联系，以确保这是他们所在领域的正确假设。在设计阶段和实施阶段接近尾声时，来自专家的这一输入将为创建实际使用的功能提供必要的方向。预计项目的最终结果将为特定主题提供执行模拟的高效、强大的潜力，并为该区域内的研究人员提供一个新的工具，他们可以利用该工具来可视化系统，从而洞察涉及许多不同实体交互的错综复杂的情况。资源FLAME网站-http://flame.ac.uk/Brief火焰概述-http://flame.ac.uk/docs/overview.htmlDr保罗·里士满的相关项目，火焰-图形处理器-http://www.flamegpu.com/Agent-based建模-http://www.pnas.org/content/99/suppl_3/7280.full

FLAME (Flexible Large-scale Agent Modelling Environment) is a piece of software which facilitates the simulation of many simple individuals (agents). It provides modelling capabilities for these agents, allowing variables and interaction rules to be set. Once the agent models are defined, simulations may be run, with emergent behaviour arising due to the interaction between agents over numerous time steps. This happens as each time step updates the agents, and their variables as necessary. Some examples of systems which may be simulated in such a manner can be found on the FLAME website (http://flame.ac.uk/projects/) and include both biological and economic systems.My proposed research topic (Advanced Software Development for Complex Systems Modelling) will focus on the further development of the core methods and algorithms used within FLAME-II, a framework based on FLAME, to parallelise agent based applications. The research questions and hypotheses to be covered in this project are yet to be decided, though they will follow along the lines of "Is it possible to implement or improve functionality within FLAME-II allowing for efficient and viable simulation of ... using the FLAME-II framework?". The project title is broad enough to allow me to refine this question to a specific topic within a specific discipline over the first year or so of the project. Once the question(s) are decided upon I would allot another year plus leeway to implement the functionality to the required standard before using the remaining time to finalise a thesis and ensure the functionality is correct.As my current area of interest with relation to the project is the simulation of biological systems, this is where my research would initially be directed at. This would likely involve individual and guided research into areas which may benefit from simulation as well as contact with experts to ensure that this is a correct assumption in their area. During the design and near the end of the implementation phase this input from experts would provide the necessary direction to create functionality of real use.I would expect that the end result of the project would provide a specific topic an efficient, powerful potential for performing simulations and giving researchers within the area a new tool with which they may visualise systems, providing insight into the intricacies involved with the interaction of many distinct entities.ResourcesFLAME website - http://flame.ac.uk/Brief overview of FLAME - http://flame.ac.uk/docs/overview.htmlDr Paul Richmond's related project, FLAME-GPU - http://www.flamegpu.com/Agent-based modelling - http://www.pnas.org/content/99/suppl_3/7280.full