Abstract:

In order to study the dynamic decision making processes of travelers, based on decision field theory (DFT), automobile travelers’microcosmic and macroscopic dynamic decision making behavior under multiple attribute are studied. The basic model of DFT is analyzed, and travelers’dynamic decision making scene and frame are constructed. Historical travel time, forecast travel time and comprehensive time of the both are considered as route attributes, and calculation method of attention weight matrix is produced according to the difference of attributes and actual travel time. An example is studied, and we find that, freedom travel time is the primacy factor influencing route choice result, travelers rely on route attributes intensely, and preference reversal exists during the decision making, which denotes decision making is continuously changing with the decision time and travelers route choice is the result of time accumulating.

Key words: traffic engineering, dynamic decision making, decision field theory；route choice, multiple attribute, preference reversal

摘要：

为了研究出行者进行路径选择时的动态决策过程，本文采用决策场理论，研究 了多属性条件下，机动车出行者的微观和宏观动态决策行为.分析了决策场理论的基本模 型，构建了出行者动态决策行为场景和框架，以历史出行时间、预测出行时间和两者综合 值作为出行者决策的路径属性，并根据属性与实际出行时间的差值为依据，提出了出行 者注意力权重矩阵计算方法.通过算例分析发现，自由流出行时间对出行者决策影响较 大，出行者对路径属性的依赖性非常强，决策过程中存在路径偏好反转的现象，表明出行 者动态决策过程是随时间不断变化的，路径选择是时间累计的结果.

关键词: 交通工程, 动态决策, 决策场理论, 路径选择, 多属性, 偏好反转