基于主线密度预测的快速路入口匝道控制方法

Expressway Entrance Ramp Control Method Based on Mainline Traffic Density Prediction

针对快速路入口匝道区域拥堵问题,本文提出一种基于交通密度预测的入口匝道控制方法。通过分析快速路匝道区域的交通特性,定义交通流缓行状态,利用经典宏观交通流模型对快速路主线的历史密度进行统计,基于速度、密度参数进行聚类分析量化缓行状态交通参数特征。基于城市交通流的日变化特性,利用时间序列模型对交通密度进行短期预测,基于预测结果识别缓行状态。针对交通缓行状态,提出一种快速路主线交通密度与入口匝道排队长度的协同控制模型,通过控制入口匝道进而控制主线下游密度不超过临界密度,利用PSO-PID(Particle Swarm Optimization-Proportional Integral Derivative Control)算法对控制模型进行求解。以南昌市洪都大道快速路为例,对晚高峰缓行时段进行仿真分析,并与经典的匝道控制方案进行对比分析。结果表明:本文模型相较ALINEA模型,将匝道调节率提高11.0%,匝道平均排队长度和平均延误分别降低了14.8%和13.5%。

Aiming at the congestion problem in the expressway entrance ramp area,this paper proposes an entrance ramp control method based on traffic density prediction.By analyzing the traffic characteristics of the expressway ramp area,this study defines the slow traffic flow state,use the classic macro traffic flow model to make statistics on the historical density of the expressway mainline,and perform cluster analysis based on speed and density parameters to quantify the traffic parameter characteristics of the slow travel state.Based on the diurnal variation characteristics of urban traffic flow,a time series model is used to make short-term predictions of traffic density,and slow traffic conditions are identified based on the prediction results.In view of the slow traffic state,a collaborative control model is proposed for the traffic density of the expressway mainline and the queue length of the entrance ramp.By controlling the entrance ramp,the density downstream of the mainline is controlled not to exceed the critical density,and the PSO-PID algorithm is used to solve the control model.Taking the Hongdu Avenue Expressway in Nanchang City as an example,a simulation analysis was conducted on the evening peak slowdown period,and a comparative analysis was conducted on the classic ramp control scheme.The results show that compared with the ALINEA model,the proposed model increases the ramp adjustment rate by 11.0%,and reduces the average queue length and average delay of the ramp respectively by 14.8%and 13.5%.