居民出行方式选择非集计模型的建立

Development of a Multinomial Logit Model for Trave Mode Choice of Residents

基于非集计离散选择模型的基本理论与建模方法,结合2003年北京居民出行调查数据,对影响居民出行方式选择的因素进行了分析,选择包括公交车、出租车、私人小汽车等在内的5种日常生活中较为常用的交通方式作为居民出行的方式选择肢,确定了影响居民出行方式选择的特性变量及相应的取值方法,建立了交通方式选择MNL模型。应用Matlab优化工具箱中的无约束最优化函数对所建的MNL模型的参数进行了标定,并通过命中率的计算验证了模型的有效性。结果表明,非集计建模方法能够较全面的考虑居民出行选择的各方面影响因素,尤其是将出行者的个人特性影响因素引入模型,提高了模型的预测精度和实用性。

Based on the theory of disaggregate discrete choice and its modeling method, referring to the survey data of residents＇ daily travel of Beijing City in 2003, five daily travel ways including bus, taxi, private car, etc. were selected as a alternative set, and characteristic variables affecting mode choice and the method of getting its values were determined, then a Muhinomial Logit Model （MNL） was developed finally. The model＇s parameters were calibrated by applying the function of fminunc in Optimization Toolbox of Matlab, and the model＇s validity was testified by calculating the hit ratio. The result shows that the disaggregate model considers more influence factors, especially the individual character factors, it can help improve the model＇s forecasting accuracy.