基于多项Logit模型的候车时间价值估计

Estimation of Waiting Time Value Based on Multi-Logit Model

候车时间是影响公共交通吸引力的重要因素之一,候车时间价值作为候车时间与出行成本间的转化媒介往往对公共交通组织优化结果产生较大影响。通过调研访问及SP调查法获取出行者在不同条件下为减少候车时间的支付意愿数据,应用非集计理论建立基于多项Logit模型的支付行为预测模型;基于出行者支付意愿,利用预测结果测算出行者在不同条件下的候车时间价值;研究候车时间价值在不同条件下的差异及随候车时长变化的规律。结果表明:候车时间价值受候车时长、收入水平、出行目的及天气条件影响显著,且与候车时长和收入水平均呈正相关;工作出行的候车时间价值要高于娱乐出行,恶劣天气条件下的候车时间价值要高于舒适天气;Logistic曲线能够较好地解析候车时间价值与候车时长的关系,拟合决定系数接近于1。

Waiting time is one of the important factors affecting the attractiveness of public transportation.Waiting time value as the medium of conversion between waiting time and travel cost often has a greater impact on the organizational optimization results of public transportation.The data of willingness to pay(WTP)for travelers to reduce waiting time under different conditions was obtained through research interviews and SP survey methods.Based on the theory of disaggregation,a prediction model of payment behavior based on multinomial Logit model was established.Based on the traveler’s willingness to pay,the value of waiting time under different conditions was calculated by using the prediction results,and the difference of waiting time value under different conditions and its variation rule changing with waiting time length were studied.The results show that the waiting time value is significantly affected by the waiting time length,income level,travel purpose and weather conditions,and it is positively related to waiting time length and income level.The waiting time value of working trip is higher than that of entertainment trip,meanwhile,the waiting time value in bad weather is higher than that in fine weather.In addition,the relationship between the waiting time value and the waiting time length can be well analyzed by Logistic curve,and the fitting coefficient is close to 1.