交通状态分类方法研究与设计

发布时间：2018-01-10 00:04

本文关键词：交通状态分类方法研究与设计　出处：《浙江工业大学》2014年硕士论文　论文类型：学位论文

【摘要】：交通出行与人们的日常生活息息相关,而交通拥堵已成为很多大、中型城市共同面临的问题。如何缓解城市交通拥堵、优化道路通行已成为我们亟待解决的重要问题。在硬件方面,可以通过加快和完善道路基础设施建设提高车辆的通行能力；在软件建设方面,可以通过运用计算机技术和数据处理技术,从海量的交通数据中获取有用的交通信息,发现交通出行规律,提高交通运行与管理的水平。数据挖掘技术可从大量的数据中发掘对我们有帮助意义的信息或知识。这些信息和知识可以应用于各种领域,起到指导和辅助决策的作用,如经济形势分析、生产目标设定、市场定位等。本文运用数据挖掘技术,对交通数据进行处理,得出了一系列相关的分析结果,对缓解城市交通拥堵和优化道路通行具有一定的应用价值。本文提出了一种短时交通状态分类方法。此方法是一种适用于5分钟时间间隔、基于交通参数统计特性的交通状态分类方法。该方法主要有三个步骤：(1)首先是根据交通参数的统计特性,利用聚类算法分别对交通参数进行分类；(2)然后利用交通参数的模糊组合划分交通状态,建立交通状态分类模型；(3)最后利用所得到的分类模型判别实时数据的交通状态。通过一些实例,验证了算法的有效性。
[Abstract]:Traffic and people's daily lives, and traffic congestion has become many large and medium-sized city, facing the problem. How to alleviate the city traffic congestion, road traffic optimization has become an important problem to be solved. In the hardware aspect, can speed up and improve the road infrastructure construction and improve the vehicle capacity; in the software construction that can be through the use of computer technology and data processing technology, to obtain useful traffic information from the traffic data, found that the transportation regulation, improve traffic operation and management level.
Data mining technology can help us to explore the significance of information or knowledge from large amounts of data. These information and knowledge can be applied in various fields, play a guiding and supporting role in decision-making, such as the analysis of the economic situation, production goal setting, market positioning and so on. This paper use data mining technology to deal with the traffic data. Got a series of related analysis results, and has certain application value to alleviate the city traffic congestion and traffic optimization.
This paper proposes a classification method of short-term traffic state. This method is applicable to a 5 minute time interval, traffic state classification method based on the statistical properties of the traffic parameters. This method has three main steps: (1) according to the statistical characteristics of traffic parameters, respectively, to classify the traffic parameters by using clustering algorithm (2); and then use the traffic parameter fuzzy combination division of traffic state classification model for traffic state; (3) the classification model with the real-time traffic state identification data. Through some examples, verify the validity of the algorithm.

【学位授予单位】：浙江工业大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：U491

【参考文献】