Theoretical Research For Capacity Of Signalized Intersection On Urban Road Network In Mixed Traffic Condition

Capacity of signalized intersections is a critical measure of the traffic supply provided by the intersections,and is an essential basis for performance analyses.On urban road network in China,the traffic condition is characterized by high volumes of pedestrians and bicycles.In such mixed traffic condition,pedestrians and non-motors (mainly bicycles) have significant effects on vehicular capacities of signalized intersections.Therefore,it is imperative to further investigate into the theories on and methods for capacity calculations under mixed-traffic conditions,in order to provide a theoretical basis for alleviating traffic congestion,as well as and technical guidance for traffic operations and facility designs.This dissertation centers on the theories on and methods for calculating capacity of signalized intersections with mixed-traffic,taking into account the characteristics and conflicting mechanisms of the mixed traffic flows on urban roads in China.Firstly,the existing relevant researches were reviewed,categorized,and analyzed in perspectives of methodologies and model developments,which lays groundwork for further research in this dissertation.Secondly,accounting for the conflicting mechanisms in mixed traffic conditions,two analytical models were proposed based on gap-acceptance and probability theory to quantify the effects of pedestrians and non-motors on capacity of signalized intersections.And then,the methods to aggregate the effects of pedestrians and non-motors were detailed.Thirdly,to contribute to the existing "deterministic" capacity methodologies,a methodology was presented for analyzing the stochastic nature of signalized intersection capacity.The methods were proposed to estimate capacity distributions,and to calculate capacity reliability considering vehicle random arrivals.Finally,in view of traffic operations of signalized intersections,an approach was developed based on fuzzy neural network to predict users' perceptions,and evaluate the signalized intersection LOS.The relationship between capacity and signalized intersection LOS was also investigated.What it follows contains the detailed innovations of this dissertation:1.An analytical model for the capacity of signalized intersections influenced by pedestrians at signalized intersections was proposed.The model takes into account pedestrian gap-acceptance behaviors,pedestrian-vehicle conflict mechanisms,and pedestrian arrivals in platoons.The conflict zone capacity was modeled based on the conflict mechanism.The model was calibrated and used as the basis to calculate the vehicular capacities.2.This study divided the influence of bicyclists at conflict zones on vehicular flow into four types,with the time durations estimated based on probability,shock wave,and gap acceptance theory.Vehicular saturation flow rate was predicted for various conditions on the basis of the speed-flow curve for calculating the capacity of signalized intersections influenced by bicycle traffic.The model to some extent overcomes the limitations of the Highway Capacity Manual(HCM,2000) method for left-turns due to data collection,and takes into account the effect of "trapped" bicycles on the through vehicular traffic.3.In order to supplement the existing "deterministic" capacity calculation methodology,this study addresses the capacity reliability of signalized intersections in mixed traffic conditions,which comprise traffic flows of vehicles,bicycles,and pedestrians.Simulation experiments were designed using VISSIM to provide data for estimating the capacity distributions,and the effects of analysis intervals on the distributions were discussed.An assessment method for the capacity reliability was proposed considering the random arrivals of vehicles.Analyses were also presented regarding the sensitivity of capacity reliability to the conflicting pedestrian and bicycle volumes.4.An approach to evaluating signalized intersection LOS was developed based on fuzzy neural network model.A quantitative surveying method for the driver perceptions was also designed using visualized simulations,and implemented.Fuzzy Set Theory was used to model how the users perceive the attributes affecting their perceptions of LOS.Then the neural network with the fuzzy reasoning experiences stored in it was employed to combine these attributes to evaluate LOS.The relationship between driver perceptions and vehicular capacity are analysis through analysis of correlation and variance.