文档介绍：长沙理工大学硕士学位论文基于BP神经网络的LOGIT交通方式划分模型研究姓名:尹逸云申请学位级别:硕士专业:交通运输规划与管理指导教师:黄中祥 20080320 摘要交通方式分担预测是交通规划理论“四阶段法"中的重要步骤之一,目前研究的方法主要有集计分析方法与非集计分析方法。非集计分析方法是以单个出行者作为分析对象,充分利用每个调查样本的数据,求出描述个体选择行为的概率值,是目前交通规划领域的研究热点之一。 LOGIT模型是交通方式划分研究中应用非常广泛的一种非集计分析模型。该模型是基于最大效用原则的离散选择模型,模型的求解关键是求出效用值。 LOGIT模型将效用值取为个人特性和选择枝特性的线性函数,但在实际当中,效用值往往是个人特性和选择枝特性的非线性函数。本文提出了基于BP神经网络的LOGIT模型,利用BP神经网络良好的非线性逼近能力,以及Matlab神经网络工具箱,对LOGIT模型的效用值的计算方法上述缺陷进行改进。该方法依据实际中关于个人交通出行的调查结果,把各影响因素的无量纲化后的属性值和个人交通出行的实际选择结果分别作为输入和输出代入BP人工神经网络的进行学习与训练,得到各影响因素的权重系数,从而得出各交通方式的效用值,最终确定各交通方式的分担概率。最后,本文将基于BP神经网络的LOGIT模型应用于具体算例,结果表明该方法具有良好的精度,且更加符合实际。关键词:交通方式划分LOG T模型 BP神经网络效用值 Abstract Traffic Prediction is one ofimportant phases offour-stage method of trafficplanning current study methods are mainly aggregate analysis method and disaggregate analysis latterconsiders asingle traveler as an analysis object and makes full use of the data of each sample and obtains the probability value which describes individual’S choice is one of the current focuses inthe fieldof traffic planning. Logit model is one of disaggregate models applied widely on the research of traffic is a discrete choice model based on RUM(Random utility maximization).The key tosolve thismodel is toobtain the utility model considers the value as alinear function ofpersonal characteristic and choosing limb infactthevalue iscustomarily anonlinear function. The thesis puts forward Logit model based on BP work which improves the Algorithm of the utility value using work nonlinear approximation ability and Matlab-ANN substitution of the data included the dimensionless attribute values of influence factors and the actual selecting result ofpersonal trip as input and output into BP ANN to study and train, the method can obtain the weight coefficient of each influence factors,thus calculating the utility values ofeach traffic ,the probability of each trafficmodes can be