Performance Evaluation Of Green Wave Control For Urban Arterials Based On Vehicle Arrival-departure Characteristics

Traffic signal performance evaluation is the foundation of the problem diagnosis and efficiency investigation in the field of arterial signal coordination(e.g.green wave control).To improve the accuracy and effectiveness of the traditional signal performance evaluation methods including the HCM analytical models and travel time based methods,innovative methods based on high-resolution data have gained lots of attentions in literature and field applications recently.This study focused on utilizing vehicle license plate identification data,which is currently widely available in most cities in China,to investigate traffic flow characteristics at signalized intersections and evaluate traffic progression on arterials.To investigate traffic flow characteristics at signalized intersections,the classification problem of the saturated flow and unsaturated flow at signalized intersections was studied.Due to the uncertainty issues using single static threshold,an improved method was proposed,which is able to tolerate noise when identifying the first vehicle of the unsaturated flow and select the optimal thresholds for classification.In terms of identifying the first vehicle of the unsaturated flow,the proposed rule could tolerate one or two unexpected larger headway between following vehicles.As for selecting the optimal thresholds,a non-parametric test based method was proposed.Rank sum test was applied to evaluate the statistical significant difference in headways between saturated and unsaturated flow.The optimal thresholds were selected among various reasonable candidates based on the minimum P-value.The proposed method was implemented using field collected data from Kunshan City.Furthermore,the influence factors of saturation flow rate,including region types,analysis period and lane distribution,were also investigated.To evaluate traffic progression on arterials,this study aimed to overcome the drawbacks of arrival on green based method.The effectiveness of arrival on green based method could be significantly affected by the detector location,travel speed selection and queuing vehicles at intersection.In terms of application issues,the high-resolution signal event-based data,which is the fundamental data source to derive arrival on green rate,is still not available in most cities in China.Thus,a stop rate based method using vehicle license plate identification data was proposed in this study.First,the stop rate was considered as the ratio between saturated and unsaturated flow based on the LWR theory.Second,the stop rate calculation method was designed to account any movements on arterial by matching vehicle license plate.The proposed method was implemented on two arterials in Kunshan City.The evaluation results showed the proposed method well addressed the issues of arrival on green method,and was able to find opportunities in applications such as diagnosing the suitability and utilization of conducting green wave control and analyzing temporal and spatial correlation of arterial traffic flow.The study was a preliminary trial on traffic signal performance evaluation method based on China specific high-resolution traffic flow data,which also provided certain technical support for applications in the field of signal performance evaluation and optimization.