CAJ | 학술논문

交通流微观仿真模型是目前广泛开展的智能运输系统项目开发、研究的重要内容.人们在研究、应用交通流微观仿真模型时,对模型验证过程重视不够.模型验证不足,使得模型本身不能达到其应有的精确程度,应用时常发生模拟结果与实际路况不能完全吻合的现象.结合一种基于投影寻踪回归的车辆跟驰模型,着重从微观角度,运用直观对比法和非参数统计方法对模型进行验证.验证结果表明,直观对比法和非参数统计方法用于模型验证是可行的.
교통류미관방진모형시목전엄범개전적지능운수계통항목개발、연구적중요내용.인문재연구、응용교통류미관방진모형시,대모형험증과정중시불구.모형험증불족,사득모형본신불능체도기응유적정학정도,응용시상발생모의결과여실제로황불능완전문합적현상.결합일충기우투영심종회귀적차량근치모형,착중종미관각도,운용직관대비법화비삼수통계방법대모형진행험증.험증결과표명,직관대비법화비삼수통계방법용우모형험증시가행적.
The microscopic traffic flow simulation models are important content in the study of intelligent transportation systems.The validation is not attracted attention enough in the study or application of microscopic traffic simulation models.The exactitude degree can not be achieved,because the models are short of validation.As a result,simulation results often can not match to the field ones when the models are used.In this paper,the straight comparison method and nonparametric statistical methods are used to validate the car-following model based on AHP in microcosmic level.The validation models are feasible through the test.