基于浮动车技术的城市短时交通状态预测模型研究

发布时间：2018-04-16 03:18

本文选题：短时预测 + 交通状态　；参考：《大连海事大学》2015年硕士论文

【摘要】：近年来,随着社会经济的发展,机动车数量的飞速增长,为生活带来了交通上的便捷,但同时也加重了道路拥堵程度。由于受土地资源等限制,城市中难以通过扩建道路满足通行需求,因此,智能交通系统成为国内外解决拥堵的研究热点。短时交通状态预测是ITS中交通诱导和控制的关键,短时交通状态预测方法的应用,取得了一定的效果,但大部分方法是基于固定检测器进行预测,难以适应浮动车数据的特点。基于浮动车数据的预测模型,一般选择忽略缺失数据,预测精度较低,不能满足舒缓交通的需求。为了提高短时交通状态预测的准确性,本文在改进的BP神经网络预测模型的基础上,给出基于历史和实时数据对训练数据进行分类补缺的方法,根据输入层数据的缺失情况,选择不同的改进BP神经网络预测模型。本文选取了行程速度作为表征交通状态的参数,针对现有浮动车数据的特点,采用筛选、拟合、补缺以及降噪的预处理过程,针对补缺处理,给出了基于历史和实时数据的K-means分类补缺方法,并对预处理结果进行了验证。在分析行程速度的时空相关性的基础上,分别基于时间、空间和时空维度数据对短时交通状态进行预测,以大连市部分出租车的实际运行数据作为浮动车数据对预测结果进行了验证,给出基于输入层数据缺失的短时交通状态综合预测模型。实验结果表明,本文给出的模型可以较好地对大连市短时交通状态进行估计,具有一定的准确度和可靠性,实例验证数据结果基本符合大连市的实际交通状况,可以满足出行者对短时交通状态预测的需求。本文的研究结果,对提高城市交通拥堵预测能力具有一定的理论和实际应用价值。
[Abstract]:In recent years, with the development of social economy and the rapid growth of the number of motor vehicles, it has brought convenience to life, but also aggravated the degree of road congestion.Due to the limitation of land resources, it is difficult to meet the traffic demand by expanding roads in cities. Therefore, Intelligent Transportation system (its) has become the research hotspot in solving congestion at home and abroad.Short-time traffic state prediction is the key to traffic guidance and control in ITS. The application of short-time traffic state prediction method has achieved some results, but most of the methods are based on fixed detector to predict, so it is difficult to adapt to the characteristics of floating vehicle data.Based on the prediction model of floating vehicle data, the missing data is generally ignored, and the prediction accuracy is low, which can not meet the needs of traffic relief.In order to improve the accuracy of short-term traffic state prediction, based on the improved BP neural network prediction model, this paper presents a method of classifying and filling the training data based on historical and real-time data, according to the lack of input layer data.Different improved BP neural network prediction models are selected.In this paper, the travel speed is selected as the parameter to represent the traffic state. According to the characteristics of the existing floating vehicle data, the pre-processing process of screening, fitting, filling and noise reduction is adopted.The method of K-means classification and filling based on historical and real time data is presented, and the preprocessing results are verified.On the basis of analyzing the temporal and spatial correlation of travel speed, the short-time traffic state is predicted based on time, space and space-time dimension data, respectively.The actual operation data of some taxis in Dalian are used as floating vehicle data to verify the prediction results, and a comprehensive short-term traffic state prediction model based on the lack of input layer data is presented.The experimental results show that the model presented in this paper can estimate the traffic state of Dalian in a short time, and has certain accuracy and reliability.It can meet the demand of travelers for short-time traffic state prediction.The results of this paper have certain theoretical and practical application value to improve the ability of urban traffic congestion prediction.
【学位授予单位】：大连海事大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：U495;TP183

【参考文献】