毕业设计（论文）《基于粒子群的快速路小时交通量预测》.doc

五邑大学本科毕业设计 PAGE PAGE IV 摘 要 本文主要目的是对城市快速路的小时交通量进行智能预测，由于城市快速路交通量的具有不确定特、非线性等特性，因而对城市快速路的交通流预测需要通过构建预测模型方法进行研究。人们在不断地探索交通流量预测模型的历程中，在最近发现并提出了一种基于粒子群优化的BP神经网络的新型的交通量预测模型。这个预测模型的基础为BP神经网络，辅以用粒子群优化算法对其网络的权值和阀值进行优化，以此来提高神经网络对交通流预测的精度和准确度。经过研究和数据分析，粒子群-BP神经网络预测模型在预测交通流量应用中具有较好的效果也比较精确。本文下面分章节主要系统地讨论现代智能交通系统的发展概况，现代城市快速路与交通交通流预测模型，粒子群优化算法的起源与其原理等，粒子群优化算法与神经网络结合的原理和应用，以及PSO-BP神经网络的交通预测模型的实验仿真等等内容。 关键词：PSO算法；BP神经网络；交通量预测；城市快速路；智能交通系统 Abstract Main purpose of this article is hour traffic volume of urban expressway intelligent forecast, because of urban expressway traffic flow with uncertainty and nonlinear characteristics, so for the urban expressway traffic flow forecast need to build forecast model methods are studied. People in constant exploration in the course of a traffic flow prediction model, the recently discovered and put forward a kind of BP neural network based on particle swarm optimization of the new model of the traffic volume forecast. This prediction model is the basis of the BP neural network, by using particle swarm optimization algorithm for the network weights and thresholds are optimized, in order to improve the accuracy of neural network to traffic flow prediction and accuracy. Through research and data analysis, the particle swarm - BP neural network prediction model in the prediction of traffic flow in the application has a good effect is also more accurate. In this paper, the following chapters mainly discuss systematically the development of modern intelligent transportation system, the modern urban expressway traffic flow prediction model and traffic, such as the origin of the particle swarm optimization algorithm and its principle, the principle of particle swarm optimization algorithm combined with neural network and its application, and the PSO - BP neural network traffic prediction model of experimental simulation and so on. Key words：PSO algorithm；BP neural network；Traffic volume forecast；Urban Expressway；Intelligent transportation