The Study On Macroscopic Fundamental Diagrams Based On Automatic Number Plate Recognition Data

With the booming economy of China,there are increasing amount of construction ongoing in the metropolitan region and the speed of urban-rural integration continues to accelerate.As the important links between cities and the major arteries within cities,transportation is playing a critical role in urban development.However,medium-size and big cities in China are suffering from the traffic congestion during the peak periods.So,the proposed scientific and effective traffic management and control methods are critical to alleviate the traffic congestion problems.Traffic flow theory,as the core theory of traffic engineering,is a theoretic basis for solving practical traffic problems.However,the classical traffic flow models mainly focused on the vehicles operation on the road sections and intersections.There are limitations for applying these models to network-wide analysis.The macroscopic fundamental diagram,developed over the past decade,is a model describing traffic operation on a regional network.This model has attracted extensive attention from domestic and foreign scholars.This thesis has conducted the research on the macroscopic fundamental diagram,and the main achievements are shown below:(1)This thesis proposed a new method of creating macroscopic fundamental diagram,which considers the mid-link sources and sinks rate and bases on automatic number plate recognition data(MLSSR method).Compared with the macroscopic fundamental diagram created by the CUPRITE model,the effectiveness of the method is proved.The Edie definition was used to create a macroscopic fundamental diagram that did not consider the sources and sinks rate of the road section.As a benchmark of macroscopic fundamental diagram,the macroscopic fundamental diagram created by the MLSSR method and the CUPRITE model was compared with it.The results showed that the amount of the sources and sinks rate of the road section is large,and the maximum flow and critical density of a macroscopic fundamental diagram will be underestimated without considering it.(2)The accuracy of a macroscopic fundamental diagram was compared between the arithmetic mean method and the harmonic mean method.The results show that the harmonic mean method is more accurate.By adjusting the probe penetration rate,we found that the higher the probe penetration rate,the more accurate the macroscopic fundamental diagram be estimated.When the probe penetration rate is 10%,the precision reaches 95.9%.The box diagram of the statistical analysis shows that the estimation of a macroscopic fundamental diagram in peak period is higher than that in non-peak period.The construction of intelligent traffic systems in China is booming and automatic license plate recognition systems are widely used.It is of great theoretical and practical value to establish a method based on automatic recognition of license plate data to create a macroscopic fundamental diagram.The MLSSR method uses the automatic license plate recognition data collected on the border of the regional road network to calculate the sources and sinks of a regional road network.It is applicable to the situation that some sections lack the automatic license plate recognition equipment,and has good practicability.