基于浮动车数据的城市道路交通状态预测研究

发布时间：2018-03-23 14:10

本文选题：交通状态预测　切入点：BP神经网络　出处：《大连海事大学》2017年硕士论文

【摘要】：近些年,城市道路交通拥挤问题日益严重,影响着人们日常生活。尤其在上班高峰期、下班高峰期等用车高峰时段,道路拥堵现象会极其严重。然而此时并不是所有道路都处于拥堵状态,通过对城市道路交通状态的预测,可以掌握路况信息,能够最大效率地利用城市道路资源,缓解拥堵道路上的交通压力。因此,集合多种先进技术的智能交通系统逐渐成为解决城市道路拥堵问题的关键途径。城市道路交通状态预测是智能交通系统中交通控制和交通诱导的关键。在查阅总结大量文献与资料后,论文提出城市道路交通状态预测模型,该模型包括城市道路的速度预测以及城市道路交通状态判别。通过比较用于交通流预测的相关模型后,论文建立基于改进粒子群算法的BP神经网络的城市道路速度预测模型。在经典粒子群算法中引入遗传算法中"交叉"与"变异"概念,改进后的算法可以克服经典算法的较容易陷进局部最优值等问题。再利用改进的粒子群算法对经典BP神经网络的连接权值和阈值的赋值进行优化,使神经网络具有更高的精度。针对城市道路交通状态判别,论文建立的是基于模糊C均值聚类算法的交通状态判别模型,既可以体现出样本与各个聚类中心的归属关系,又可以进行类别划分。论文将大连市部分出租车的实际浮动车数据进行预处理后,应用到城市道路交通状态预测模型中进行验证。实验结果表明,论文的城市道路交通状态预测模型能够较好的对城市道路进行交通状态的预测,对交通诱导与控制具有一定的参考意义,使有限的城市路网资源发挥更大的效果。
[Abstract]:In recent years, the problem of urban road traffic congestion is becoming more and more serious, which affects people's daily life. However, not all roads are in a state of congestion at this time. By predicting the traffic state of urban roads, we can grasp the information of road conditions and make the most efficient use of urban road resources. Ease traffic pressure on congested roads. Intelligent Transportation system (its), which gathers many advanced technologies, is becoming the key way to solve the problem of urban road congestion. The prediction of urban road traffic condition is the key of traffic control and traffic guidance in its. After summing up a lot of literature and data, This paper puts forward a model of urban road traffic state prediction, which includes the speed prediction of urban road and the discrimination of urban road traffic state. In this paper, the prediction model of urban road speed based on BP neural network based on improved particle swarm optimization algorithm is established. The concepts of "crossover" and "mutation" in genetic algorithm are introduced into classical particle swarm optimization algorithm. The improved algorithm can overcome the problem that the classical algorithm is easily trapped in the local optimal value, and then optimizes the assignment of the connection weight and threshold value of the classical BP neural network by using the improved particle swarm optimization algorithm. In view of the traffic state discrimination of urban road, a traffic state discriminant model based on fuzzy C-means clustering algorithm is established, which can reflect the relationship between the sample and each cluster center. After preprocessing the actual floating vehicle data of some taxis in Dalian City, it is applied to the urban road traffic state prediction model to verify. The experimental results show that, The prediction model of urban road traffic state in this paper can better predict the traffic state of urban road, which has some reference significance for traffic guidance and control, and makes the limited urban road network resources play a greater role.
【学位授予单位】：大连海事大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：U491.14;TP183

【相似文献】