数字孪生、图灵测试和城市模型
Digital Twins, Turing Tests and Urban Models
迈克尔•巴蒂
英国伦敦大学学院教授，规划专业荣誉教授
林旭辉（译）
英国伦敦大学学院巴特莱特可持续建造学院博士研究生
摘要: 数字孪生已经成为热门话题，它是计算机在科学理论和实践发展过程中的最新阶段。如果回到20世纪20年代，“模型”一词意味着“实物的简化”或“抽象概念”，当时几乎不为人知，但计算机一出现，计算机模型首先兴起，紧随其后的是各种模型的快速发展。在科学领域，理论的概念现在几乎可以与模型的概念互换。重点阐述计算机或数字模型背景下数字孪生的定义，并提出困扰该领域的语义学悖论和难题。围绕“如何定义一个好模型”，引入图灵测试的“宽松版本”变体，即简单地提出“你能看出孪生和真实事物之间的区别吗？”这一问题。基于该测试对伦敦数字模型的应用，建议未来的发展方向是对同一地点开发数字模型的多种版本、示例和应用，从而启动数字模型的对比，并为它们的集成提供平台，以此为信息预测提供更丰富的背景。
Abstract: Digital twins have become all the rage and are simply the latest stage in the process of using computers in the development of scientific theory and practice. If you go back to the 1920s, the term "model" meaning a "simplification of the real thing", or an"abstraction" was barely visible but once computers emerged, the idea of computer models first, but fast on their heels, models of many kinds gained momentum. In science, the notion of theory has now become almost interchangeable with the idea of model. In this short note, I will focus on defining digital twins in the context of computer or digital models, raising paradoxes and conundrums that pertain to the semantics dominating the field. After some brief discussion of definitions, I will ask the question "how do we define a good model" and then introduce variants of a "relaxed version" of Turing's famous test that simply asks the question "can you see the difference between the twin and the real thing". This is one of many open question pertaining to such tests but then,using applications from our own work on digital models in London, I suggest that the way forward is to develop many variants,many examples, many applications of digital models of the same place, thereby initiating comparisons and providing platforms for their integration. We imply that many rather than one model, one twin, thus provides a much richer context for informed prediction.
关键词：数字孪生；多模型；城市模拟
Keyword: digital twins; multiple models; urban simulations
中图分类号：TU981
文献标识码: A
