A model for Optimal Continuous Network Design Problem with Elastic Demand by Simulated Annealing Algorithm

One way to reduce traffic congestion is to increase network capacity. The solution is applicable by either adding new links to the network or improving the existing links. The purpose of this study is to provide a model able to determine the optimum capacity of network links considering an elastic demand. The problem is known as Continuous Network Design Problem (CNDP) in transportation. The model has been developed as a bi-level program. The objective function at the upper level is to maximize the consumer surplus for which Simulated Annealing (SA) algorithm has been used to determine the optimal solution. The lower level problem is formulated as assigning traffic considering elastic demand for which Frank–Wolf method has been used. Two networks have been used to compare the results. Also, sensitivity analysis on the parameters of SA was done to find optimum values for the parameters. Finally, after accomplishing another sensitivity analysis on the budget, consumer surplus changes have been reported against budget variations.