Research on Actual Capacity of Urban Roads Based on Magnetic Sensor Data

Study of actual road capacity through traffic flow data collected by traffic flow detectors is of great significance for traffic planning, design and engineering practice. However, there are still research gaps on methodologies and practical issues on this topic, including how to determine the time interval of the traffic flow data for statistics and how to avoid the influences of missing or abnormal traffic flow data on the results, and the research on the statistical distributions of the capacity of different classes of urban roads is also not thorough enough. Based on the traffic flow data collected by the magnetic sensors in Wuhan, the influence of different time intervals of statistics on the characteristics of traffic flow parameters is analyzed, a method of estimating the actual traffic capacity of urban roads based on the percentiles of the traffic flow statistics is proposed, and the method is analyzed and verified in combination with the data of typical roads in Wuhan. The results show that the 15-minute time interval and the 95% percentile of traffic flow data are more appropriate for estimating the actual traffic capacity. On the basis of analyzing the data of 856 magnetic sensors in Wuhan, it is found that there are obvious differences of the actual capacities between roads with the same road class. Therefore, the roads of each class are further divided into four different levels of service (LOS) in general according to the traffic saturation levels, and the statistical distribution patterns of the actual capacities of roads of each class under different general LOSs are obtained. The results of this study are useful for more in-depth and systematic understanding of the characteristics of the actual traffic capacity of the city roads of different levels.