Multiobjective Optimization Using Evolutionary Algorithms - A Comparative Case Study
PPSN V Proceedings of the 5th International Conference on Parallel Problem Solving from Nature
A novel multiobjective optimization algorithm based on bacterial chemotaxis
Engineering Applications of Artificial Intelligence
| Hi-index | 0.00 |
Logistics network design is one of the principal parts of strategic decisions in the planning and control of production systems. It deals with determining the warehouses locations and the definition of product flow between facilities and clients. This work is focused in finding an approximation of the Pareto-optimal front for two conflicting objective functions in logistic networks design: minimize costs and maximize coverage. Since the establishing of which warehouses must be opened constitute a combinatorial optimization problem, two metaheuristic techniques, namely Improved Strength Pareto Evolutionary Algorithm - SPEA2 and a novel binary version of Bacterial Chemotaxis Multi-objective Optimization Algorithm - BCMOA, were applied. With the aim of finding the optimal flow between clients and warehouses, network flow algorithms were also used. The performances of the above techniques were evaluated by comparative analysis of the results obtained in the solution of eight randomly generated problems by means of the dominated hypervolume metric (S-metric). The hybrid methodology here developed to solve the logistics network design problem - which combines metaheuristic techniques with a network flow algorithms - showed to be competitive regarding the Pareto Optimal Front approximation, and also displayed high efficiency in execution times.