基于模糊神经网络的公交客流时段预测  

Real-time forecast for passenger flow of bus based on fuzzy neural network

作　　者：温惠英[1] 王晓巍[1] 荣利利[1] 陈先龙 

机构地区：[1]华南理工大学交通学院,510640 [2]广州交通规划研究所,510030

出　　处：《微计算机信息》2009年第12期225-226,204,共3页Control & Automation

摘　　要：公交线路由一系列站点组成,线路客流的形成和变化是各站点共同作用的结果。本文通过相似性分析找到站点客流实时分布与线路客流实时分布之间的相关性,找出影响线路客流分布的关键站点,对这些关键站点的实段分布数据运用模糊神经网络技术预测线路客流的短时分布规律。该方法在广州公交客流数据调研的基础上加以验证,结果表明这种预测方法能够有效的对公交线路客流分布进行短时预测,预测结果满足精度要求。Bus lines is composed of series of sites. The formation and change of line passengers flows are the results of sites commonly worked. Similarity analysis is applied in this paper to find the correlation between real-time distribution of site flow and line flow and the key sites which influence the distrbution of passenger flow. And also, it forecasts short -term distribution of passenger flow by using fuzzy neural network technology on the base of real-time data of these key sites. The method is verified on the basis of guangzhou bus passenger flow data, The results showed that this method can predict the short-term bus flow distribution effectively, and meet the requirements of precision.

关 键 词：公共交通 线路客流实时分布 模糊神经网络 相似性分析 

分 类 号：U491[交通运输工程—交通运输规划与管理]