交通方式选择的Logit模型

Logit Model of Traffic Pattern Choice

着重用离散型的出行需求模型，即 Ｍｕｌｔｉｎｏｍｉａｌ Ｌｏｇｉｔ模型，对城市交通结构迸行分析。在概率的意义下推导、论证和解释模型的合理性，说明对选择模型的变量和参数的限制条件。提出一种稳定性好、收敛速度快的参数估计算法——ＢＦＧＳ算法。Ｌｏｇｉｔ模型实用性强，便于对交通结构预测。

Thearticle mainly analyses city's traffic pattern by means of discretic travel demand model, i. e. Multinomial Logit Model.rFom the inference of probability, reasonable explanations of the model are given, limited conditi- ons of varibles and parameter for the model choice are illustrated. An esti- mated logarithm of parameter with good stability and quick convergence BFGS logarithm is suggested. The Logit Model is very feasible and convenient for the prediction of traffic pattern.