基于分散决策的拥挤型网络流及其变分不等式

The Flows in a Congested Network in View of the Decentralized Decision and Its Variational Inequalities

以交通网为背景讨论一种基于分散决策的拥挤型网络流,给出并证明了网络流的变分不等式,并进而讨论了选址于弧上的微变特性,得出了一个重要的结果.

A congested network in which the cost of travelling an arc is dependent on the number of users is described. The flows in view of the decentralized decision and the variational inequalities about the flows in a congested network are investigated. A two-level optimization location model is discussed. In this model, users are concentrated at some nodes and the destination is located on the node or on the arc such that the user's total cost is a minimum. In almost all location problems, there is at least one node on which the location is optimal. However, it is shown that in a two-level model this is not true. Based on a perturbation analysis of locating on arcs, the necessary condition for an optimal location of the destination on arcs is given.