Research On Performance Measures For Traffic Signal System Based On License Identification Data

The development of motorlization brings benefits for travelers,however the imbalance between traffic supply and traffic demand caused by economy is becoming increasingly tough.Traffic signal control is an effective method of solving the problem of traffic delay,queuing and so on.The traffic managers focus on comparing different traffic signal strategies and managing the traffic effectively.With the advent of big data era,it is much easier to obtain traffic detection data than before.The deep mining of traffic detection data is an efficient apporach to understand the traffic operation.This dissertation compares the before and after adjustment of traffic signal control strategies with certain performance measurements based on the traffic detection data mining technology.Finally,the most suitable signal control strategy is selected in a real case.This study reviews the performance measures of common detectors of different types,and make a comparison between different types.The situation is found that the utilization of detector data is not sufficient.Therefore,this study focuses on data analysis based on detector which can identify individual vehicle.Detector data,which can identify individual vehicle,is adopted to formulate the queue model considering non-oversaturated condition and minimum headway,and also a queue model considering oversaturated condition and vehicle travel time.Besides,this study proposes a Platoon Ratio indicator based on the identification of individual vehicle,which can make contributions to the growing body of the evaluation system of traffic signal control in intersections.As for performance measures for arterial coordination control,this study reviews the method for calculating vehicle travel time taking advantage of the matching function of lisence identification data.Then methods for calculating five travel time reliability indices including buffer time index are introduced.TAD model formulated in this dissertation based on theoretical arrival time is another performance measure in evaluating progression of arterial coordination control,which takes queueing and signal information into account.The license identification data-based evaluation system is enriched and the verification of the model is conducted via a real case.After that,this dissertation explores the optimization strategies for the traffic signal control system based on the results of evaluation.The arterial coordination offset optimization strategy is proposed based on the average travel time,TAD,arrival on green.The hill climbing algorithm is applied to solve the arterial coordination offset optimization problem in a real case.Finally,this study tries to explore the development of signal control on-line evaluation system in order to make up for the lack of signal control progression evaluation platform in China.Researches on performance measures using license indentification data in this dissertation promote the building of evaluation system,as well as establish a hillclimbing optimization algorithm for arterial coordination control,which is of great significance for traffic managers in dealing with practical traffic problems.