基于动态因素的居民租车出行选择行为研究

摘要

To study the residents` car rental intention and model selection, to improve the accuracy of the prediction model for the residents` car rental travel choice, so correlation analysis was used to choose influence factors which were closely related to the mode choice of residents car travel to determine the car rental travel choice of dynamic factors (i.e., seasons and passing car rental choice factors).Finally, the dynamic model for residents` car rental choice was established that based on the MNL model.On the basis of a large number of questionnaires, the reference model and dynamic model parameters were calibrated and predicted.The results showed that the dynamic factors of residents travel choice has a significant impact on residents` car rental travel choice, the established dynamic model fitting Ryobi degree and forecasting accuracy rate is greatly improved compared with the MNL model, which can more accurately predict the residents` car choice travel choice behavior.%为研究居民租车出行意愿以及车型选择,提高租车出行选择行为的预测模型精度,使用相关性分析筛选与居民出行租车选择行为密切相关的影响因素,确定租车出行选择的动态因素(即季节和过往租车选择因素),在MNL模型的基础上建立居民租车出行选择动态模型.在大量问卷调查基础上,对参照模型和动态模型进行参数标定和对比预测.结果表明,动态因素对居民租车出行选择具有显著影响,所建立的动态模型的拟合优比度和预测准确率较MNL模型有较大提高,能够较准确地预测居民租车出行选择行为.