Abstract: 

The performance of a road pricing scheme varies greatly by its actual design and implementation. The design of the scheme is also normally constrained by several practicality requirements. One of the practicality requirements which is tackled in this paper is the topology of the charging scheme. The cordon shape of the pricing scheme is preferred due to its user-friendliness (i.e. the scheme can be understood easily). This has been the design concept for several real world cases (e.g. the schemes in London, Singapore, and Norway). The paper develops a methodology for defining an optimal location of a multi-concentric charging cordons scheme using Genetic Algorithm (GA). The branch-tree structure is developed to represent a valid charging cordon scheme which can be coded using two strings of node numbers and number of descend nodes. This branch-tree structure for a single cordon is then extended to the case with multi-concentric charging cordons. GA is then used to evolve the design of a multi-concentric charging cordons scheme encapsulated in the two-string chromosome. The algorithm developed, called GA-AS, is then tested with the network of the Edinburgh city in UK. The results suggest substantial improvements of the benefit from the optimised charging cordon schemes as compared to the judgemental ones which illustrate the potential of this algorithm.