Evaluation of regional traffic level of service based on adaptive neuro-fuzzy inference

At present, the research on evaluation of urban traffic level of service (LOS) mainly aims at the basic road section and intersection. However, it still lacks unified evaluation index and practical approach. Therefore, in this paper, we propose a method that the regional traffic LOS is evaluated by adaptive neuro-fuzzy inference system (ANFIS). ANFIS is used to emulate human's expertise on specifying four levels of traffic service within Shenzhen Metropolitan Area. This paper takes average delay per mileage of regional road network and scaling exponent which characterizes the uneven degree of traffic flow distribution as evaluation indices. We verify the accuracy of our system by comparing outputs of the system with opinions of experts. The result shows that the method achieves the traffic LOS accurately. It also can reflect the change of regional traffic LOS in the evaluation time interval.