支路汇入型道路车流量模型构建与预测研究

Research on the Construction and Prediction of Traffic Flow Models for Roads Incorporating Branch Connections

摘要: 针对支路汇入型道路因支路与主路交通流交互频繁、时空分布波动显著，数据采集效率偏低等问题，本文开展支路汇入型道路车流量模型构建与预测研究。论文结合实测交通数据，系统分析支路车流汇入主路的演化规律与特点，构建了两类支路汇入型道路的函数关系模型。在此基础上，运用分段线性回归和SLSQP算法对相关模型参数加以估计，并对相关时刻的车流量加以预测。进一步，论文对车流量采样时刻进行优化，利用贪心算法确定最少的观测采样时刻。优化后的采样策略在确保模型可靠性的同时，可降低支路交通监测成本。因此，论文的研究成果为交通管理提供了一定的理论价值和参考。

Abstract: This paper addresses the issue of low data collection efficiency at roads incorporating branch connections, which is caused by frequent interactions between traffic flows on slip roads and main roads, as well as significant spatio-temporal distribution fluctuations. It investigates the construction and prediction of traffic flow models for road incorporating branch connections. By integrating actual traffic data, the study systematically analyses the evolutionary patterns and characteristics of traffic merging from slip roads onto main roads. This establishes functional relationship models for the two main categories of roads incorporating branch connections. Based on these models, segmented linear regression and the SLSQP algorithm are employed to estimate relevant parameters and predict traffic volumes at specific times. Furthermore, the paper optimizes traffic sampling intervals by utilizing a greedy algorithm to determine the minimum number of observation points required. This optimized sampling strategy reduces monitoring costs for slip roads while maintaining model reliability. Consequently, the research findings offer both theoretical value and practical guidance for traffic management.