Research On State Prediction And Speed Coordinated Control Of Mixed Traffic Flow On Expressway

There will be a mixed traffic flow of Autonomous Driving Vehicles(AD Vs)and Human Driving Vehicles(HDVs)for a long time.The application of mixed traffic flow control considering traffic state prediction is not only a method to improve traffic efficiency,but also the development trend of mixed traffic flow.Therefore,this paper proposes a research based on the state prediction and speed cooperative control of mixed traffic flow on expressway.Firstly,the paper proposes a traffic state prediction method based on Q-Learning and Long Short-Term Memory(LSTM).This method not only avoids the problem of the explosion and disappearance of traffic data gradients,but also has efficient storage and analysis capabilities.The continuous training and memory storage of the training sets are used to improve system sensitivity.Subsequently,the test sets are predicted based on the accumulated experience pool and a high-precision state prediction result is obtained.Finally,the traffic congestion index is obtained,and the first part of traffic state prediction is completed.Secondly,a traffic risk prediction method considering big data technology and Stacked AutoEncoderm-Gate Recurrent Unit(SAE-GRU)is proposed under the state prediction model.The paper uses big data technology to construct a dynamic identification model to realize the identification of operating state and risk state.Subsequently,SAE-GRU is used to realize safety prediction based on the recognition results.Finally,the traffic risk index is obtained,and the second part of traffic state prediction is completed.Subsequently,the paper proposes a coordinated control model of mixed traffic speed based on the state prediction under the expressway merging area.Firstly,the paper constructs an analytical model of in-fleet and inter-fleet stability to obtain the stability interval.Secondly,a speed cooperative control model and trajectory planning scheme for mixed traffic are constructed based on the stability characteristics.Finally,numerical simulation is used to analyze the traffic conditions between used model and unused model with Time To Collision(TTC),Dynamic Space Occupancy(DSO),and Vehicle Specific Power(VSP).Finally,the paper proposes the speed control strategy of mixed traffic flow under different roadside unit deployment characteristics based on the deployment distance,interaction radius,and communication delay.This method aims to achieve the goal of safety and efficiency.Firstly,the current domestic roadside unit deployment characteristics are expressed as functions,and the adaptive efficiency functions are proposed.Secondly,the traditional mixed traffic speed control model is linearly correlated with the characteristics of the roadside unit and state prediction to obtain an improved speed control model.Finally,the corresponding numerical simulation analysis is carried out to evaluate the feasibility of the model at multiple levels from the perspective of vehicle stability and traffic efficiency.In summary,the paper effectively integrates the correlation between state prediction and control research.At the level of theoretical research,it effectively promotes the multi-scenario application technology of mixed traffic flow in consideration of traffic state,and further improves the research on the characteristics of mixed traffic flow.At the actual engineering level,the paper fully considers the change process of traffic flow under the development trend of future autonomous driving technology,and provides theoretical basis and technical support for implementing mixed traffic control.