目录摘要..................................................................IIAbstract................................................................III前言...................................................................1第一章绪论..........................................................11.1研究背景及意义.....................................................11.2短时交通流预测国内外研究现状.......................................2第二章时间序列预测模型相关理论........................................22.1时间序列模型预测概述...............................................22.2时间序列模型预测原理...............................................32.3时间序列预测算法...................................................4第三章短时交通流预测..................................................93.1短时交通流预测概述.................................................93.2短时交通流预测的评价指标...........................................93.3交通流数据的选择..................................................103.4数据性质..........................................................11第四章运用时间序列模型进行短时交通流预测.............................114.1运用三种时间序列算法进行短时交通流预测............................114.1.1趋势拟合法...................................................114.1.2平滑法.......................................................164.1.3ARIMA方法...................................................194.2预测结果对比分析..................................................214.2.1对一次指数平滑法进行拟合.....................................22第五章结束语.........................................................235.1论文的优点........................................................235.2论文的不足........................................................23参考文献..............................................................24致谢................................................................25附录：Matlab软件程序.....................................................26摘要随着现代社会的高速发展，城市车辆越来越多。随之而来的，道路阻塞和交通事故层出不穷。这些交通问题严重影响着人们的出行,急需解决。解决这些问题就需要对某一路段未来某一时间段的交通状况进行科学的预测，从而为交通规划、运输业、交通管理、交通控制提供重要技术保障，实现交通智能化。本文从预防的角度出发，在考虑成本的条件下，利用已有的历史数据（车流量、车道占有率）运用时间序列模型对路段的未来短时交通流数据进行预测，为争取更多的时间解决将要发生的交通问题做好充分准备。本文以贵阳市的实际测量交通流数据作为训练样本，利用MATLAB编制程序，使用线性拟合、曲线拟合、移动平均法、指数平滑法和ARIMA模型进行预测、分析、比较。并对指数平滑法得到的结果进行了改进，提高了预测精度。关键词：交通智能化,时间序列,短时交通流,预测分析AbstractWiththerapiddevelopmentofmodernsociety,moreandmorecarsappearincities,whichmakethetrafficcongestionandthetrafficaccidentsemergeendlessly.Allthesetrafficproblemsarenotgoodforpeople'stravel,andneedtobesolvedimmediately.Tosolvetheseproblems,weneedtopredictthetrafficconditionsofaroadinacertainperiodoftimeinfuture.Wecangiveimportanttechnicalsupporttothetrafficandtransportationplanning,thetransportation,thetrafficmanagementandtrafficcontrol,torealizeintelligenttraffic.Fromtheperspectiveofprevention,ourpaperusedtheexistinghistoricaldata(thenumberofcars,laneoccupancy)andthetimeseriesmodeltoforecastashort-termfuturetrafficflowdataofar...
