多源信息环境下的路径决策模型  

Route Decision Model under the Environment of Multi-source Information

作　　者：刘澜[1] 骆晨[1] 尹俊淞[1] 马亚峰[1] 

机构地区：[1]西南交通大学交通运输与物流学院,四川成都610031

出　　处：《西南交通大学学报》2015年第5期891-897,共7页Journal of Southwest Jiaotong University

基　　金：四川省科技支撑计划资助项目(2014GZ0019-1);四川省重点实验室研究基金资助项目(szjj2011-031)

摘　　要：针对路径选择过程中主观因素和客观因素共同作用且无法量化的问题,根据直觉模糊集理论,建立了双隶属度函数,分别构造了主观和客观直觉模糊函数;利用直觉模糊函数中的隶属度、犹豫隶属度建立了决策函数;借鉴证据理论的证据冲突思想,改进犹豫隶属度函数投影法,用该方法进行了犹豫隶属度分配;并运用对数增长型权重模型,量化了时序多源信息的重要程度,建立了分析道路决策选择结果的驾驶决策行为模型.算例分析结果表明:该模型在考虑个人历史经验信息、交通管制信息、道路物理环境信息与道路拥堵信息形成的多元信息环境下,选择驶入和不驶入的决策近似隶属度分别为0.767和0.233.Aimed at the route choice process integrated with the subjective perception and objective analysis that cannot be quantified,the intuitionistic fuzzy set theory is applied to quantify the selection results of route decision, in which the subjective and objective intuitionistic fuzzy numbers are constructed by the establishment of the dual membership function,and decision-making functions are set up with the membership degree and hesitating membership degree of intuitionistic fuzzy numbers.Drawing lessons from evidence conflicting ideas of evidence theory,a function projection method is proposed to solve the distribution problem of hesitating degree function. Then,a logarithmic growth model is applied to determine dynamic weights and quantify temporal characteristics of multi-source information,and a route decision model is established to analyze the driving decision-making behavior.Finally,an example derived from traffic practices demonstrates that the approximate membership degrees of ＂go＂and ＂change＂decision are calculated as 0. 767 and 0. 233,respectively,with the decision-making functions under a multivariate information environment including experiences,traffic management and control,roads,and congestion.

关 键 词：智能交通 路径决策 直觉模糊集 证据论 异质信息 

分 类 号：U491[交通运输工程—交通运输规划与管理]