Reinforcement Learning Based Variable Speed Limit Control For Expressway In Icy And Snowy Weather

Icy and snowy weather has a negative impact on highway traffic safety and efficiency and it also brings challenges to highway traffic control.Speed limit control is an important measure to ensure the highway traffic safety in icy and snowy weather,however,the existing speed limit control method cannot adjust the speed limit strategy independently according to the changes of road environment and traffic conditions,which is easy to produce unreasonable speed limit control and leads to traffic congestion.Furthermore,it affects the highway traffic efficiency and safety.Reinforcement learning adjusts and updates speed limiting strategies by interacting with the road environment,and can adopt corresponding optimal speed limiting strategies according to changes in real-time traffic conditions,so as to effectively improve traffic conditions.Therefore,this paper proposes a variable speed limit control method based on reinforcement learning algorithm for highway traffic safety and traffic efficiency under icy and snowy weather,which not only guarantees the traffic safety of highway,but also further improves traffic efficiency.The main research contents are as follows:(1)Traffic flow modeling based on improved cellular transmission model under icy and snowy weather.This paper analyzes the traffic flow characteristics and the mechanism of variable speed limit control in icy and snowy weather conditions,and establishes a traffic flow model suitable for describing the highway traffic characteristics by improving the traditional traffic cellular transmission model,which lays a theoretical foundation for the subsequent study of variable speed limit control methods.(2)Variable speed limit control based on reinforcement learning algorithm in icy and snowy weather.On the premise of ensuring driving safety and for the purpose of improving traffic efficiency,a variable speed limit control method of expressway under icy and snowy weather is designed,so that it can take traffic safety under icy and snowy weather conditions into account,and effectively improve the traffic efficiency of bottleneck areas,so as to reduce the negative impact of icy and snowy weather on highway traffic.(3)Variable speed limit control considering the uncertainty of traffic flow in icy and snowy weather.Analyze the uncertainty of traffic flow caused by model observation error and sudden traffic event intervention,further investigates the mechanism of traffic flow uncertainty on the evolution of traffic flow,proposes a traffic prediction algorithm under traffic flow uncertainty,and combines it with reinforcement learning algorithm to design a variable speed limit control method considering traffic flow uncertainty under snow and ice to ensure the efficiency of highway traffic under sudden traffic event intervention.The proposed algorithm is combined with a reinforcement learning algorithm to design a variable speed limit control method that takes into account the uncertainty of traffic flow under snow and ice conditions,so as to ensure the efficiency of highway traffic under sudden traffic events.(4)Development and application of variable speed limit control system on expressways in icy and snowy weather.On the basis of theoretical research,the overall design of the freeway variable speed limit control system under icy and snowy weather is carried out,and the hardware equipment selection,function module design and software platform development are carried out.Finally,the functional completeness of the system is verified through deployment testing and practical application,and the systematic management and control of the expressway under icy and snowy weather is realized.