杨凯茜　刘玮蔚

摘 要：随着社会的进步道路交通状况越来越拥挤，交通拥堵几乎成为了所有发达城市所面临的问题。因此改善交通状况变得尤为重要。文章旨在从数据的角度建立模型，来进行交通流预测。在阅读了大量文献的基础上，首先对数据挖掘技术的国内外研究情况进行基本了解，对所有数据进行预处理。然后通过线性回归分析，分别研究了单一因素和多因素对交通流预测的影响，得到各因素的影响因子，建立了模型。并进行了实例分析，结果拟合良好，验证了模型的准确性。最后，对本次模拟进行了总结，为今后的改进方向提供了思路。

关键词：交通流预测;线性回归分析;单一因素分析;多因素分析中图分类号：U4691.1+12  文献标识码：B  文章编号：1671-7988（2020）11-87-03

Abstract： Along with the progress of the society， the traffic situation is getting more and more crowded， the traffic jam has become the problem that all developed cities have to face. Crossroads is important in urban transportation network nodes， the accurate prediction of intersection traffic flow can help to improve the intersection cluttered traffic， urban traffic congestion problem solving， optimize the urban road network operation， also in the field of intelligent transportation research， applica -tion， implementation and promote the urban healthy， harmonious and stable has important significance. This paper aims to build a model from the perspective of data to predict traffic flow. On the basis of ·reading a large number of literature， first of all， we have a basic understanding of the research situation of data mining technology and preprocess all the data. Existing prediction methods analysis include： based on neural network prediction method， based on the statistical theory， based on the method of multiple linear regression model， based on the prediction method of wavelet model method and new technology. This paper used linear regression analysis. Through linear regression analysis， the influence of single factor and multi factor on traffic flow prediction is studied， and the influencing factors of each factor are obtained， and the model is established. This paper also gave an example to verify the accuracy of the model. Finally， the simulation is summarized， which provides ideas for future improvement.

Keywords： Traffic flow prediction; Single factor analysis; Multi factor analysisCLC NO.： U4691.1+12  Document Code： B  Article ID： 1671-7988（2020）11-87-03

引言

基于數据挖掘的交通流预测系统研究就是将大量的数据运用到交通流预测模型中，为决策者提供帮助来引导交通系统的畅通。从近几年国内交通治理的情况来看，单纯增加交通道路及路面硬化的方法已经不可能从根源上解决交通压力[1]。交通流数据采集系统在很长一段时间内通过人工和自动数据收集、积累了大量数据，使用数据挖掘技术人们不仅可以存储历史数据，还可以为决策系统和导航系统提供数据等等。现在已经有模型应用于交通流预测中。为了提高预测的精度和可靠性，应结合其他可靠的方法和模型来研究适合我国的交通流预测模型。智能交通系统研究的目的是使人、车、路与环境和谐共处[2]。……