Abstract: Early-warning approach of transport network was discussed. A whole early-warning framework was proposed. Based on analyzing of service attributes, an integrated quantitative service-based measure was defined. By considering weighted impacts of predicted service losses over total services, two early-warning mechanisms, static component and dynamic maintenance, was proposed. The approach and mechanisms have been applied into a real transport network environment, and application results show they are of great value.

Key words: transport network, early-warning, predicted services losses based