基于多项Logit模型下胶州居民出行行为选择研究

Research on Jiaozhou residents ′ travel behavior choice based on multinomial Logit model

针对青岛胶州市城区交通规划中出行方式分析的复杂性问题,提出基于多项Logit模型下影响青岛胶州市居民的出行行为选择的方法,探讨模型中的交通参数,运用SPSS数据分析处理交通出行行为。根据当地居民出行调查数据,利用多项Logit模型建模分析影响胶州市居民出行行为的变量,调查统计出模型选择枝为非机动车、公交车和私家车,三种选择枝作为对数模型参数进行系数标定。结果表明,性别、年龄、出行目的、户年收入以及出行方式属于影响胶州市民行为选择的因素;男性更倾向于私家车,户年收入与倾向私家车出行程度成正相关,胶州市居民在选择非机动车或私家车上的概率更高。模型验证结果与老城区规划定性分析结果一致,验证结果具有一定的借鉴作用。

In view of the complexity of travel mode analysis in the urban traffic planning of Jiaozhou city in Qingdao,a Jiaozhou residents′travel behavior choice method based on multinomial Logit model is proposed.The traffic parameters of the model are discussed.The SPSS data are used to analysis and process their travel behaviors.According to the data of travel survey of the local residents,the variables that affect the travel behaviors of residents in Jiaozhou city are modeled and analyzed by multinomial Logit model.According to the investigation and statistics,the choice branches of the model are non⁃motor vehicles,buses and private cars.The three choice branches are used as the logarithmic model parameters for coefficient calibration.The results show that the gender,age,travel purpose,household annual income and travel mode are the factors that affects the behavior choices of the residents in Jiaozhou city;among the respondents,males are more inclined to private cars;annual income per household is positively correlated with the inclination degree of traffic by private car;the probability of choosing non⁃motor vehicles or private cars is higher.The results of model verification are consistent with those of the qualitative analysis of the old city planning,so the verified results can be used for reference to some extent.