Fuzzy Logic-Based Urban Traffic Congestion Evaluation Models and Applications

Through establishing scientific indices and methods for evaluating traffic congestions, the performance efficiency of the urban traffic system can be determined and the changing trend of traffic, congestions can be readily captured. The existing methods for the traffic congestion evaluation either calculate traffic delays of all vehicles in the system using the traffic flow data, or evaluate the changes of long-term travel times and fuel consumption. Such methods lack the appropriate evaluation indices that can trace the real time congestion situations. This paper is intended to develop traffic congestion evaluation models and provide the respective evaluation grades using the traffic experiment-based fuzzy logic approach, which is accomplished by analyzing the traffic flow characteristics on the classified roads of Beijing, in conjunction with the perceptions of travelers about the level of congestions. The proposed evaluation models can capture the dynamic characteristics of traffic congestions on the entire network. In the case study, traffic congestion indices are evaluated for the typical time periods and the key geographical locations using the floating car data (FCD) collected for different years in Beijing. As a result, the paper summarizes the temporal and spatial distributions and the changing trends of traffic congestions in Beijing. It shows that the annual increase rate of the congestion is 6.84% for all day, 7.99% for the peak within the Fourth-Ring areas. Such an evaluation provides an analytical foundation to the relevant studies associated with mitigating traffic congestions in Beijing.