基于灰色神经网络的城市交通流量预测方法研究

发布时间：2018-03-21 22:30

本文选题：智能交通　切入点：交通流预测　出处：《沈阳大学》2014年硕士论文　论文类型：学位论文

【摘要】：随着我国城市化进程的加快，城市机动车保有量持续增长，交通问题日益突出。智能交通系统已成为解决城市交通问题的主要途径。交通流预测是智能交通系统中的一项关键技术，主要用来实现城市交通路网中的各个节点和线路交通流量状态的预测与分析。交通流预测的准确性是实现城市交通流控制的前提条件。交通流是一个多变量、时变的、结构复杂的非线性系统。传统的单预测模型只能概括系统的部分特性，预测精度受限。鉴于此，本文提出了基于相关路口分析和灰色神经网络的城市道路交通流量的组合预测方法，具体针对交通流预测模型、方法和实现进行了较系统的研究，主要工作包括以下几个方面：（1）提出了较系统的交通数据预处理方法，给出了数据错误、丢失、冗余的判断方法和处理过程。利用该方法可以有效去除噪声数据的干扰，减少数据冗余，从而提高后续交通流预测的效率和准确性。（2）结合灰色系统和神经网络理论的各自优势，建立了灰色神经网络模型。利用该模型对城市道路交通流量进行预测，仿真结果表明该模型可有效提高预测的精度和实时性。（3）提出了一种路网路口流量相关性分析的方法。基于历史数据，通过主成分分析法对路网目标路口进行流量相关性分析。建立了一种组合预测模型，利用路网相关路口数据预测数据缺失路口交通流量。以城市道路路口为单位，，根据实际路况，同时考虑路口间流量大小及路口间隔长度，以沈阳市部分行政区域路口为例建立局域路网。利用相关路口和目标路口历史数据，对预测模型、方法和实现进行仿真实验，验证了其准确性和实时性。
[Abstract]:With the acceleration of urbanization in China, the number of motor vehicles in urban areas continues to grow. Traffic problems are becoming more and more prominent. Intelligent Transportation system (its) has become the main way to solve urban traffic problems. Traffic flow forecasting is a key technology in Intelligent Transportation system (its). It is mainly used to predict and analyze the traffic flow state of each node and route in the urban traffic network, and the accuracy of the traffic flow prediction is the precondition to realize the urban traffic flow control. Traffic flow is a multivariable, time-varying and complex nonlinear system. The traditional single prediction model can only generalize some of the characteristics of the system, and the prediction accuracy is limited. In this paper, the combined forecasting method of urban road traffic flow based on correlation intersection analysis and grey neural network is put forward. The model, method and realization of traffic flow forecasting are studied systematically. The main work includes the following aspects:. In this paper, a systematic method of traffic data preprocessing is put forward, and the judgment method and processing process of data error, loss and redundancy are given. By using this method, the interference of noise data can be effectively removed and the data redundancy can be reduced. In order to improve the efficiency and accuracy of subsequent traffic flow prediction. Combining the respective advantages of grey system and neural network theory, the grey neural network model is established. The model is used to forecast the urban road traffic flow. The simulation results show that the model can effectively improve the accuracy and real-time performance of the prediction. Based on historical data and principal component analysis (PCA), a combined forecasting model is established. Using road network related intersection data to predict the traffic flow of missing intersection. Taking urban road junctions as units, according to the actual road conditions, considering the volume of intersections and the length of intersections, taking some administrative district junctions in Shenyang as an example, the local road network is established. The historical data of relevant intersections and target junctions are used. Simulation experiments on prediction model, method and implementation are carried out to verify its accuracy and real-time performance.
【学位授予单位】：沈阳大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：U491.14;TP183

【参考文献】