Application on Network Travel Time Reliability Analysis Based on Critical Travel Paths by Fourth-Moment Method with RFID Data

This paper proposes a method for evaluating the travel time reliability of the road network (NTTR) based on the crucial travel time reliability of the path (PTTR). By assessing factors such as vehicle travel path and journey time, the frequent pattern-growing (FP-growth) algorithm is used to identify the important pathways of the urban road network, and the fourth-moment approach is chosen to compute the crucial PTTR. Then, using weight factors of path flow (PF), path length (PL), and path connection degree (PC), we create the mixed weights to compute NTTR. The radio frequency identification (RFID) data set from the road network in Nanjing is used for validation. Results show that the most traveled path is often the most reliable in the road network. On workdays, the NTTR of the morning peak of commuting behavior is significantly reduced, and the day’s lowest NTTR occurs at the evening peak.