基于时序关联规则挖掘的交通拥堵预测技术研究

发布时间：2019-06-29 18:28

【摘要】：当前我国城市现代化进程不断推进,然而交通拥堵问题日益突显,交通拥堵已经成为困扰各个大中城市的严重问题之一。城市交通拥堵带来的危害性主要有两个方面:一是交通拥堵发生时造成的时间延误和能源浪费,给社会带来了巨大的经济损失。据中科院专家统计得出的数据,我国每天因城市交通拥堵造成的经济损失可达约10亿元。二是当车速过低时,汽车尾气污染程度大大增加,与此同时会产生大量噪声,使空气质量以及城市环境质量急剧下降,进而对市民的身心健康造成严重危害,降低了市民生活水平。所以,对复杂的交通状况进行有效的预测是当前亟须解决的重要问题。近年来,越来越多的学者开始致力于智能交通系统的研究,提出多种交通拥堵预测方法。常见的交通拥堵预测方法主要是基于各类数学模型,并且大多只对单一道路单一时刻进行预测。由于交通系统复杂多变的特性,这类方法往往考虑的参数并不全面,同时没有考虑到交通拥堵事件的时序性,无法很好地适应实际情况。在交通系统中,各个路段发生拥堵往往遵循一定的因果关系,同时考虑到交通拥堵事件的时序性,本文提出一种基于时序关联规则挖掘的交通拥堵预测方法,先利用遗传算法挖掘出时序关联规则,再将这些关联规则作为数据样本构建分类器,以达到对交通拥堵预测的目的。本方法采用进化算法的思想,有效避免了传统方法需要确定参数过多的弊端,算法更为贴近实际生活情况,能够有效预测交通拥堵,为及时缓解城市交通压力、降低道路拥堵发生率、提高道路通畅度、保障高效快捷地出行提供了参考依据。
[Abstract]:At present, the process of urban modernization in our country is advancing, but the problem of traffic congestion is becoming more and more obvious, and the traffic jam has become one of the serious problems in the large and medium-sized cities. The harmfulness of urban traffic congestion mainly includes two aspects: one is the time delay and energy waste caused by the traffic jam, and brings great economic loss to the society. According to the data from the experts of the Chinese Academy of Sciences, the economic loss caused by the traffic congestion of the city is about 1 billion yuan a day. Secondly, when the vehicle speed is too low, the pollution degree of the automobile exhaust is greatly increased, and meanwhile, a large amount of noise is generated, so that the air quality and the urban environmental quality are greatly reduced, and further serious harm to the physical and mental health of the citizen is caused, and the living standard of the citizen is reduced. Therefore, the effective prediction of complex traffic conditions is an important problem to be solved at present. In recent years, more and more scholars have begun to study the intelligent transportation system, and put forward a variety of traffic jam prediction methods. The common traffic congestion prediction method is mainly based on various mathematical models, and most of the traffic jam prediction methods are only predicted at a single time of a single road. Due to the complex and changeable nature of the traffic system, the parameters often taken into account are not comprehensive, and the timing of the traffic jam events is not taken into account, and the actual situation cannot be well adapted. In the traffic system, the traffic jam of each road section often follows a certain causal relationship, while taking into account the timing of the traffic jam event, this paper proposes a traffic jam prediction method based on time-series association rule mining, which first uses the genetic algorithm to mine the time-series association rules, The correlation rules are used as data samples to construct a classifier so as to achieve the purpose of predicting the traffic jam. The method adopts the idea of an evolutionary algorithm, effectively avoids the defect that the traditional method needs to determine the excessive parameters, the algorithm is more close to the actual living condition, the traffic jam can be effectively predicted, the traffic pressure can be relieved in time, the traffic congestion rate is reduced, the road smoothness is improved, And provides a reference basis for ensuring the high-efficiency and fast travel.
【学位授予单位】：沈阳理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：U491.14;TP311.13

【参考文献】