城市交通与土地利用一体化规划模型

Integrated Model of Urban Transportation and Land Use

从研究基础和模型建立的理论基础角度对城市交通与土地利用之间的关系进行综合评述,并提出已有四种代表性模型中普遍存在的缺陷.在此基础上,介绍加拿大学者近年来的研究成果,即基于24小时个人行为和活动的非集计动态ILUTE模型.应用2008年北京市居民出行2809份调查样本数据对出行方式选择行为进行模拟.参数估计采用遗传算法和蒙特卡洛模拟法.结果显示:小汽车、公共交通、自行车、步行、出租车以及搭乘等六种出行方式的样本数据观测值与模型预测值的绝对误差百分比分别为0.52%,1.95%,0.77%,0.79%,0.16%,2.29%;模型拟合度达到0.759,精确度较高,能够很好地描述出行者的出行方式选择行为.最后提出对此模型的本土化研究展望.

A review of integrated urban transportation and land use models from the aspects of research base and basic theory is summarized. Then the new model achievements of Canadian researchers in recent years were described, which is 24 hours agent activity based, disaggregated, dynamic ILUTE model. The 2809 sample data from Beijing Travel Survey in 2008 are used for model validation. Parameters are estimated by Monte Carlo method and genetic algorithm. The results have indicated that the absolute errors between measured values and statistical values of those modes such as auto, transit, bike, walk, taxi and passenger are 0.52%, 1.95% ,0.77% ,0.79% ,0.16% and 2.29% ,respectively. The model accuracy, with a fitting de-gree of 0. 759, was high. It means the model can describe the choice behavior of trips. At last, a research prospect for localization of these models is put forward..