Spatiotemporal evolution game: a general framework for analysing traffic flow

现有运输研究分析的许多现象可以解释为两个或两个以上移动主体暂时共享空间资源的现象。本研究旨在运用进化博弈论构建时空进化博弈模型来模拟上述现象。构建了原子流模型、信息收集模型和不完全信息下的学习与行为模型。除了对这些模型进行理论/数学研究外，本研究还采用了虚拟现实技术等实验方法来研究所提出的模型在多大程度上描述了实际的人类行为。结合这些发现，构建了时空演化博弈的框架。

Many phenomena that have been analysed by existing transport studies can be interpreted as a phenomenon in which two or more moving subjects share spatial resources temporally. This study aimed to construct 'spatiotemporal evolution game' that models the above-mentioned phenomenon by applying evolution game theory. Three models, called atomic-flow model, information collection model, and learning and behavioural model under imperfect information, were constructed. In addition to theoretical / mathematical investigations of these models, this study also adopted experimental approaches such as virtual reality technique to investigate how far the proposed models describe actual human behaviour. The framework of the spatiotemporal evolution game was then constructed by combining these findings.