基于通过能力随机性的入口匝道主动控制方法研究

发布时间：2018-11-05 19:33

【摘要】：近年来,随着我国高速公路汽车保有量的快速增长,交通拥挤现象不断发生,这对入口匝道控制提出了更高的挑战。由于目前入口匝道基本采取被动控制,难以事先消除未来时段发生的拥挤,也无法解决受主线通过能力随机变化所影响的控制阂值确定问题。因此,本文将交通状态预测机制和通过能力随机特性融入入口匝道控制启发式模式中,探寻新型的主动控制方法以有效解决主线拥挤：在分析比较若干时间序列预测方法的基础上,选取ARIMA(p,d,q)方法对主线交通流量和占有率参数进行短时预测；运用聚类分析中的K-means算法对预测参数进行交通状态判别,实现交通状态的短时预测。在高速公路主线交通状态划分的基础上,选取不同交通状态下的通过能力值进行统计分析,以获取其概率分布函数并给出取值方法,为主线通过量控制阈值的选取提供依据。研究结果表明：主线通过能力在不同交通状态下取值不同,在相同交通状态下呈现正态分布特性。采用启发式控制方式,局部层面选用ALINEA控制模型,协调层面融入交通状态预测和通过能力随机性,通过局部与协调层面的竞争比选生成匝道调节率,以提前分配引起未来主线拥挤的匝道流入量,达到主动控制的目的。仿真结果表明：该方法的控制效果优于ALINEA,且具有维持主线运行状态稳定和提前消除主线拥挤的功能。
[Abstract]:In recent years, with the rapid growth of freeway vehicle ownership in China, traffic congestion has occurred constantly, which poses a higher challenge to on-ramp control. Due to the passive control of the on-ramp at present, it is difficult to eliminate the congestion in the future period in advance, nor to solve the problem of determining the control threshold affected by the random change of the main line capacity. Therefore, in this paper, traffic state prediction mechanism and capacity stochastic characteristics are incorporated into the on-ramp control heuristic mode. Based on the analysis and comparison of several prediction methods of time series, ARIMA (ARIMA) method is selected to predict the main line traffic flow and occupation rate parameters in a short time. The K-means algorithm in clustering analysis is used to distinguish the traffic state of the prediction parameters and to realize the short-term traffic state prediction. On the basis of dividing the traffic state of the main line of expressway, the value of passing capacity under different traffic conditions is selected for statistical analysis, in order to obtain its probability distribution function and give the method of taking value, which provides the basis for the selection of threshold value of main line throughput control. The results show that the main line takes on different values in different traffic states and presents normal distribution characteristics in the same traffic state. The heuristic control method is adopted, the ALINEA control model is selected at the local level, the traffic condition prediction and the ability randomness are integrated into the coordination level, and the ramp adjustment rate is generated by the competition between the local level and the coordination level. The purpose of active control is to allocate ahead of time the ramp inflow which will cause future main line congestion. The simulation results show that the control effect of this method is better than that of ALINEA, and it has the function of keeping the main line running stably and eliminating the congestion of the main line ahead of time.
【学位授予单位】：东南大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：U491.54

【相似文献】