摘要
鉴于SML-GHK方法估计MNP模型参数具有初值敏感性,提出了一种初值点确定与参数估计的实用方法。首先应用贝叶斯方法估计MNP模型参数的初值,再结合基于GHK的模拟极大似然方法估计MNP模型参数,并通过实际算例对该方法进行了验证。结果表明:文中方法实用有效,特别是在小样本情况下效果明显;将基于该方法得到的参数值用于预测,能够较好地再现实际交通分担率;文中方法能够帮助规划师探讨某些交通政策变化时全体出行者对各种出行方式的选择概率的变化.
As estimating MNP model parameters using the SML-GHK method is sensitive to the initial values,a practical method for initial value determination and parameter estimation is proposed.It uses the Bayesian approach to define the initial values of MNP model parameters,and adopts the simulated maximum likelihood based on GHK framework to estimate the MNP model parameters.Then,a case study is performed to verify the effectiveness of the proposed method.The results show that （1 ）the proposed method is applicable and effective,especially for small-scale samples;（2）using parameters defined by the proposed approach,better trip mode share forecast results can be achieved ;and （3 ）it is a useful tool for transport planners to exam the probability variations of trip mode choice resulted by transport policy changes.
出处
《华南理工大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2014年第2期103-108,138,共7页
Journal of South China University of Technology(Natural Science Edition)
基金
国家自然科学基金资助项目(51108192
51378222)
作者简介
俞礼军（1972-），男，博士，副教授，主要从事交通运输规划与设计方法研究．E-mail：yulijun@scut．edu．cn
