Abstract Computational models have been described as exceptionally adept at examining the complex relationships of human and crowd behaviour, with a significant portion dedicated to investigating spatial behaviour in defined environments. Within this context, this paper presents an agent-based model (ABM) for simulating activity in public spaces at the level of the individual user. Although other ABMs of individuals' spatial activity exist, they are often found to simulate specific building-related activities, and fewer still are found to examine activity in public spaces, in a systematic manner. This research provides a generalized formalization of human spatial behaviour incorporating stationary activities and social interaction within a 3D environment, and is presented using a widely accepted framework for describing ABM, the Overview, Design Concepts, and Details (ODD) protocol. A sample study using a synthetic environment is used to demonstrate applicability, and the model is tested extensively to establish robustness. Furthermore, model output is compared to observed activity patterns in other studies of similar spaces, and simulated spatial patterns of activity are found to match those observed in real-world scenarios, providing insight into the dynamics of the processes, and highlighting the potential of this approach for studying the complexities of human spatial behaviour.

中文翻译：

---

基于代理的公共空间使用模型

摘要 计算模型被描述为非常擅长检查人类和人群行为的复杂关系，其中很大一部分专门用于研究特定环境中的空间行为。在此背景下，本文提出了一种基于代理的模型 (ABM)，用于在个人用户层面模拟公共空间中的活动。尽管存在其他关于个人空间活动的 ABM，但他们经常被发现模拟特定的建筑相关活动，而很少有人以系统的方式检查公共空间中的活动。这项研究提供了人类空间行为的广义形式化，结合了 3D 环境中的静止活动和社会互动，并使用广泛接受的框架来描述 ABM，概述，设计概念和细节 (ODD) 协议。使用合成环境的样本研究用于证明适用性，并广泛测试模型以建立稳健性。此外，将模型输出与在类似空间的其他研究中观察到的活动模式进行比较，发现模拟的活动空间模式与在现实世界场景中观察到的活动模式相匹配，从而提供对过程动态的洞察，并突出了这种活动的潜力研究人类空间行为复杂性的方法。