| Traffic assignment(TA)is an important part of the classic traffic planning method——the four-stage method,whose purpose is to allocate the known traffic demand in the traffic network to each path reasonably.The Quasi-Dynamic Traffic Assignment(QDTA),which is improved on the basis of Static Traffic Assignment(STA),considers the capacity constraints,and can determine the location of the queue according to the supply and demand interaction between the links,and has higher computing efficiency.In recent years,an important branch of QDTA——System Optimal Quasi-Dynamic Traffic Assignment(SO-QDTA)has continuously attracted the attention of relevant scholars and formed a certain scale of research results.Most existing studies use assumptions that are contrary to reality to simplify the SO-QDTA model to reduce the complexity of the model and improve the feasibility of the algorithm.But these assumptions that do not conform to the actual situation seriously affect the practical applicability of the model and are not conducive to application promotion.Therefore,this paper attempts to change some unreasonable assumptions and optimize the SO-QDTA model so that it can be more effectively used in the real traffic network.This article mainly optimizes the model from the following aspects:in terms of the optimization and improvement of the node model,for most of the existing node models in SOQDTA research,only a single link capacity reduction factor is considered to express the constraints between the supply and demand of the upstream and downstream links of the node.Based on the conflict theory,this paper incorporates the constraints of the internal capacity of the node on the flow into the node model,proposes the optimized node model and algorithm,and proves the accuracy of the algorithm.By observing the case,it is found that the internal capacity of the node will partially impose stricter restrictions on the flow.In the calculation of the core of the SO-QDTA model——path travel time function and Path Marginal Cost(PMC),based on the assumption that the existing SO-QDTA model regards the link storage capacity as infinite,this paper uses the optimized node model for the Quasi-Dynamic Network Loading(QDNL)process considering the queue overflow effect,and proposes the calculation method of path travel time under queue overflow effect.Numerical experiments show that this method can more accurately reflect the impact of queue generation and overflow on link flow.Based on this,the calculation formulas of PMC’s externality share and internality share under queue overflow effect are derived,and the processing of the calculation process is explained in detail.In order to verify the effect of the optimized SO-QDTA model,the improved node model and the calculation method of path travel time and PMC under the overflow effect are incorporated into the SO-QDTA model,and the equivalence of the model is proved.The model is applied to the Sioux Falls test network to analyze the accuracy and convergence of the model and algorithm,as well as the impact of the node’s internal capacity and queue overflow effect on the SO-QDTA results.The analysis results show that the optimized model can achieve effective convergence and obtain more accurate results.At the same time,the node’s internal capacity constraints and whether to consider the queue overflow effect will directly affect the correctness of the model assignment results. |
