Research On Traffic State Discrimination And Active Guidance Method For Regional Expressway Network

Based on the continuous expansion of the demand for high-speed highway transportation as the proportion of motorized travel increases year by year,the growth rate of high-speed highway transportation demand in many areas far exceeds the construction speed of highways,resulting in an imbalance between traffic demand and supply,which has led to an increasingly prominent traffic congestion problem,seriously affecting the operating efficiency and traffic safety of the highway system.Studies have shown that the vast majority of congestion is caused by uneven traffic flow distribution within the road network,and balancing traffic flow distribution at the network level is an effective and economical way to alleviate congestion.Therefore,studying the traffic state discrimination and active induction methods at the network level has important practical significance and application value.Taking the definition and classification of traffic congestion as the starting point,this paper analyzes the three diffusion forms of congestion on the road network based on the spatiotemporal distribution characteristics of congestion on the expressway.The diffusion characteristics of expressway traffic congestion are clarified,providing a theoretical basis for the development of subsequent active induction methods.Based on the ETC data and vehicle detector data on highways,this paper starts from the concept and characteristics of traffic states and selects representative traffic state indicators from both the section and network levels,establishing a system for identifying traffic states on highways.By using a fuzzy clustering algorithm based on the entropy weighting method,the traffic states are accurately classified,and a traffic state index model for highways is constructed based on the clustering results.The neural network model is introduced to make short-term predictions of the traffic state index,providing technical support for the subsequent implementation of active response to congestion.Following the concept of active induction,this paper has determined the starting conditions for inducing diversion based on the traffic state index model and short-term prediction technology.Based on the diffusion form of congestion,the impact range of congestion on highways has been analyzed,and a method for selecting induction diversion areas and nodes in the highway network has been proposed,which is helpful for highway managers to quickly develop induction plans.Based on the traffic simulation model,taking the local road network of Hangzhou-JinhuaQuzhou Expressway as an example,the feasibility and effectiveness of the active guidance method proposed in the previous paper were verified by setting different guidance schemes through simulation.The simulation results show that compared with the traditional passive response guidance method,the active guidance method has a significant effect on improving the overall operation efficiency of the road network.