Spatio-temporal enabled urban decision-making process modeling and visualization under the cyber-physical environment

Abstract

Urban management is increasingly dependent on the construction of cyber-physical systems, which have been widely deployed in multiple fields, including agriculture, transportation, and health care. Cyberphysical systems provide a large volume of big sensory data. Thus, the application of big data for urban emergency decision-making has become a challenge. Traditional decision-making models are unable to overcome several technical problems, including spatio-temporal information support and access to real-time service for disaster management and emergency response. In this paper, we propose a spatio-temporal enabled urban decision-making process modeling method in the cyber-physical environment and discuss the development off our basic metadata components, the definition of eleven-tuple information description structure. A prototype system is designed for modeling registering, management, visualization, and other functions. Decision-making processes are implemented as composite process chains through three phases, i.e., decision-making process modeling, instantiation of the decision-making process chain and execution of the decision-making process chain. For application to gas leak events, the decision-making process is executed to confirm the feasibility of the proposed decision-making process modeling method and the flexibility of the eleven-tuple information structure.

创新点

1) 可共享的时空决策过程

当前决策模型往往是孤立的、分散的和难以复用的。本文提出的决策过程链通过基于元模型的11元信息结构来描述了多种决策过程。决策过程链通过标准元数据信息和描述被设计成可共享的、可延伸的和易于管理的决策过程模型。此外, 传统的决策模型缺乏时空描述。城市中的突发事件可能会因为不同的时空属性而对应不同的决策过程。因此时空描述对信息物理环境下的决策过程有着重要的作用。

2) 决策过程的拓展性

本文提出的决策过程11元信息结构包括决策通用属性、决策特殊属性和时空属性。本文以燃气泄漏事件为例证明提出的决策过程的通用信息描述结构的可行性, 同时该信息描述结构可以拓展到其他突发事件如洪涝灾害事件等。同时说明了本文提出的方法具有良好的拓展性。

3) 动态过程链决策支持

当前城市应急决策模型通常是静态模型, 这些静态模型用于已经发生的突发事件, 并不能实现事件发生前的监测和实时快速的应急响应。而本文的决策过程是基于观测的决策过程, 可以快速发现突发事件并进行高效的事件决策过程。此外, 本文提出的决策过程链实现了突发事件的分阶段的动态决策过程。决策过程链将决策过程划分成不同的阶段, 在不同的阶段分配不同的任务, 并且动态推演突发事件的发展过程。此外, 根据基于SensorML的决策过程链, 用户可以高效的执行决策过程。