中文摘要：

为研究特大城市居民出行决策规律，以居民常规出行方式（步行和乘公交车）为研究对象。以期望效用理论为理论基础，以居民出行时间价值损失最小为约束条件，构建居民出行行为的马尔科夫决策模型，研究结果显示：在特定的出行范围内，居民出行行为决策与居民的出行时问价值关系密切，出行时间价值越高，选择可靠性高、时间价值实现风险低的出行方式概率越大。

英文摘要：

The paper constructs the markov decision model of residents travel behavior in order to re- search megapolis residents travel decision rule. It takes travel mode of residents （walking and bus） as re- search object, expected utility theory as the theoretical basis and minimum loss of time value as constraint conditions for residents travel. The results are showed that： there is a close relationship between the behav- ior decision-making and time value of residents travel in the specific travel range, and the higher the time value of travel, the greater probability for residents to choose the travel mode that is the high reliability and low realization risk of time value.