Stability of road networks in relation to both usuall and unusuall traffic crowds have been noticed by traortation researchers these days. Increasing the value of time in the result of development of economic activities in various human societies, have made unexpected traffic delays in road networks to create a large amount of costs for users of these networks. So these networks must be design in a way in wich drivers be sure about their smouth and uniform movement. We must think of designing and improving additional paths between different origins and destinations, so the network can work continuously with providing an acceptable service level. In this thesis identifying important links has been investigated. Important links are those that cutting them makes most of origin-destination trips to be inappropriative in the network. For achieving this porpuse an index namely k-link connected index has been used, This index originally has been introduced for measuring the efficiency of communication networks. So with altering this index, we have made it usable for measuring the efficiency of traortation networks in relation to occurring stochastic incidents and important links have been identified by using this alterd index. Also an effective approach based on huristic methods, for identifying unimportant links is proposed. The efficiency of proposed approach have been evaluated with implementing it on a real mesdium size network (Sioux Falls). It is applicability in relation to large networks has been shown with implementing it on Isfahan traortation network.
