46 A Parameter Free Genetic Algorithm for a Multiobjective Network Design Problem

In this research, an algorithm is applied for a multiobjective network design problem. It is known that Dijkstra method can solve single-objective network design problem in polynomial time. However, this method is hard to solve a multiobjective problem. We suppose that obtaining Pareto-optimal solutions is important, for example, in the case of application to a navigation system. Driver can select the most desirable route from the Pareto-optimal solutions. Some criteria of route search are assumed, for example, time, distance, and total expenses.

We define three objective network design problem. To solve this problem which has multiple criteria, we apply a multiobjective genetic algorithm, aiming at simplification of parameter tuning and efficiency improvement of maintaining diversity. The priority-based encoding method will be used to represent chromosome. We will use this algorithm to calculate a Pareto-optimal or near Pareto-optimal solutions. Numerical experiments with simple network design problem will show the feasibility of our approach.