基于决策场理论的公交出行多阶段动态决策

Multi-stage Dynamic Decision of Bus Travel Based on Decision Field Theory

针对城市公交出行决策问题,采用改进决策场模型从心理微观偏好进化的角度对两阶段出行决策行为进行了研究。提取了不同公交方式的同质属性构建属性矩阵;结合出行者个体特征和交通属性状态的波动构建注意力权重矩阵,同时优化了注意力转移概率的计算方法。通过问卷调查与计算机仿真相结合的方式对偏好进化过程进行分析,结果显示:出行者存在偏好震荡现象,且在短时思考时间内更加明显;在出行过程中进行决策时,换乘行为的触发需要更多的偏好进化时间;同一时刻下,个人对不同的交通方式的属性注意力是独立分布的;反馈矩阵中的记忆效应和方案竞争性对决策的形成时间有显著影响,强记忆效应和弱竞争性下的偏好进化速度更快。决策场模型可对公交出行动态决策行为进行描述。

Considering the decision-making problem of urban public transit,an improved model of decision field theory was applied to study the two-stage travel decision-making behavior from the perspective of psychological micro-preference evolution.The homogeneous attributes of different modes of public transport were extracted to construct attribute matrix;the attention weight matrix was constructed by combining the individual characteristics of traveler and the attribute fluctuation of traffic guidance information;the proposed attention transfer matrix optimized the calculation method of transfer probability.The preference evolution process was analyzed through questionnaire survey and computer simulation.Results show that travelers have the phenomenon of preference oscillation,especially in a short think time.In the process of decision-making,the triggering of transfer behavior requires more preference evolutionary time;at the same moment,the attributes attention to different modes of transportation is distributed independently;the memory effect and the competition among schemes in the feedback matrix have significant influence on the formation time of decision-making:preferences evolve faster under strong memory effect and weak competitiveness.It is proved that the model of decision field theory can describe the dynamic decision-making behavior in public transit.