摘要
为寻求一种描述各种出行分布情况的普适性模型,运用最大信息熵原理,结合数理统计方法,建立了以原点矩均值为约束条件的新模型体系。为确定模型参数,提出了熵模型的参数标定方法,利用长春市居民出行调查数据对参数标定方法进行了验证。数据验证结果显示,对于步行、自行车、公交车和小汽车四种出行方式,最高阶原点矩的阶数分别取3、4、5时,所有模型均通过置信度为95%的2χ检验。通过χ2值和分布曲线的比较发现,不同出行方式取得最好效果的最高阶原点矩的阶数并不一致,这种差异取决于出行方式的内在特性。居民出行分布的信息熵模型可以定量描述任意出行方式的分布规律,所提出的参数标定方法简单有效。
In order to find some general models to describe all kinds of inhabitant trip distributions, according to the maximum information entropy theory and statistics, new models under the constraints of the means of origin moments were built, a parameters calibration method was put forward to ascertain the parameters of the models. They were validated by the inhabitant trip survey data in Changchun city. The trip modes of walk, bike, bus and car pass χ^2 tests with 95% confidence when their orders of the highest origin moments are 3, 4 and 5 respectively. The comparison of χ^2 values and inhabitant trip distributions curves shows that the orders are different when the best effects of all trip modes are gotten. The differences result from the inner characteristics of trip modes. Analyzing result indicates that the information entropy models can quantificationally describe the distribution principle of any trip mode, the parameters validation method is simple and effective. 6 tabs, 4 figs, 12 refs.
出处
《交通运输工程学报》
EI
CSCD
北大核心
2005年第4期106-110,共5页
Journal of Traffic and Transportation Engineering
基金
国家自然科学基金项目(70071039)
吉林省杰出青年基金项目(20010114)
高等学校骨干教师计划项目(1000)
关键词
交通规划
出行分布
最大信息熵原理
熵模型
traffic planning
trip distribution
maximum information entropy theory
entropy model
