摘要
提出一种基于荟萃分析的旅客出行时间价值计算方法,包括文献收集与筛选、数据处理、异质性检验、荟萃回归,以及结果分析五个步骤.异质性分析得到,数据总体异质性为37.7%;荟萃回归结果显示,调查地点、交通方式,以及调查对象分别产生31%,25%和28%的异质性来源;随后,分别建立了固定/随机效应模型和可行广义最小二乘模型,经过检验分析最终选择可行广义最小二乘模型进行参数估计,目的在于为出行者的出行决策提供合理依据和方案指导.
A meta-analysis based method for calculating the travel time value of passengers was proposed, which includes five steps: document collection and screening, data processing, heterogeneity testing, meta- regression and result analysis. Heterogeneity analysis showed that the overall heterogeneity of the data was 37.7%. The meta-regression results showed that 31%, 25% and 28% of the sources of heterogeneity were found in the survey sites, transportation modes and survey objects respectively. The fixed/random effect model and the feasible generalized least squares model were established respectively. After testing and analysis, the feasible generalized least squares model was finally selected for parameter estimation, which provides reasonable basis and scheme guidance for travelers to make travel decisions.
作者
白雪莲
羊钊
BAI Xuelian;YANG Zhao(National Key Laboratory of Air Traffic Flow Management,Nanjing University of Aeronautics and Astronautics,Nanjing 210000,China)
出处
《武汉理工大学学报（交通科学与工程版）》
2019年第4期712-717,共6页
Journal of Wuhan University of Technology（Transportation Science & Engineering）
基金
国家自然科学基金项目(51608268)
江苏省自然科学基金项目(BK20150747)资助
