城市居民出行空间和方式联合选择模型研究

Travel Space and Mode Combined Choice Model for Urban Travelers

考虑到出行者出行空间和出行方式之间的重要联系,本文基于非集计模型中Nested Logit模型的理论和方法,结合2005年国内某大型城市居民出行调查数据,建立了城市居民出行空间和方式的联合选择模型。模型为2层的NL模型,第1层为空间选择层,第2层为方式选择层,以此将出行空间选择和出行方式选择连接起来。模型采用统计软件STATA 9进行标定,并通过统计指标进行验证,深入分析了影响出行者出行空间选择和出行方式选择的重要因素。最后,用同样的数据源建立了方式选择的MNL模型,通过对比联合模型和MNL模型的命中率和集计结果表明,出行空间和方式联合选择模型能够在较细的层面上预测出行者出行空间和交通方式的选择情况,从而提高了模型的精度和实用性。

Considering the important relationship between travel space and mode choice, based on the theory of Nested Logit model and its modeling method for disaggregate model, using the data of a revealed preference survey of daily travel of one Chinese big city in 2005, a conbined choice model was built to analyse the travel space and mode choice of urban travelers. The model is a two level NL model which consists of space choice level and mode choice level to connect space choice and mode choice. The model was calibrated and tested by applying STATA9 statistics software. Based on the model, the important factors affecting the travel space and mode choices were revealed. At the end, a mode choice MNL model was built with the same data, and the hit ratio and aggregate result of the presented model were tested and compared with those of muhinomial Logit model at the end. The result shows that this combined model can predict travel space and mode choice behavior at a more detailed level to improve the model＇ s accuracy and tility.