Research And Application Of Timing Optimization Scheme For Short-term Traffic Flow At Urban Intersection

Accurate and efficient traffic flow prediction is the core part of intelligent transportation system and the premise of guidance and control,while improving signal timing at intersections is the key to improve the efficiency of intelligent transportation system.Based on short-term traffic flow forecasting,the signal timing optimization of intersections is studied,which has certain theoretical significance and practical application value.On the basis of studying the existing short-term traffic flow forecasting models,this thesis proposes a hybrid traffic flow forecasting model and a multi-objective intersection signal timing optimization model,which integrates the traditional machine learning and deep learning models,and achieves the intersection signal timing optimization based on forecasting traffic flow.The main contents and contributions of this thesis are as follows:1.The main characteristics of traffic flow and its spatial-temporal correlation,as well as the characteristics of traffic flow data sets are analyzed,and the characteristics of traffic flow are analyzed from different dimensions.On this basis,a CNN-XGBoost hybrid forecasting model is proposed for short-term traffic flow forecasting.By comparing CNN-XGBoost model with other four prediction models,the advantages of the model are verified.2.Based on short-term traffic flow forecasting,a multi-objective optimization model of intersection signal timing is established,and an improved non-dominated sorting genetic algorithm is proposed.3.Based on the geometric conditions and phase sequence of Yucai Street and Yuhua Road intersections in Hebei Province,and the data generated from actual traffic flow,the simulation of the intersection is carried out using VISSIM,and the performance indexes are compared with those of Webster method.The validity of the multi-objective timing optimization model and the algorithm proposed in this thesis is verified.4.Based on the multi-objective intersection signal timing optimization model and solution scheme,a subsystem of intersection signal timing optimization calculation is designed and implemented.As a subsystem of urban traffic signal intelligent control system,the subsystem mainly includes two modules:parameter analysis and simulation optimization.Among them,the parameter analysis mainly inquires and predicts the traffic flow,and the simulation optimization mainly calculates the timing scheme and performance indicators.At present,the main functions of the subsystem have been basically achieved.It is very convenient to manage and control the signal controller,and it can also generate signal timing schemes for designers for reference.