An evolutionary approach to decision-making in distribution planning

The problem of distribution planning is a large-scale, very complex, difficult-to-define problem. This paper attempted to formulate the problem in a realistic fashion. Upon identifying the key issues of the problem, an approach was proposed in which the most advanced technologies and algorithms, such as GIS (geographical information systems), evolutionary algorithms and hedging algorithms, were used. The most innovative aspects of the proposed evolutionary algorithms, as they relate to the specific problem of distribution planning, were presented. We believe that it is possible to obtain solutions that satisfy the network operational constraints, that are realistic, feasible and practical, that account for the costs in accordance with the geographical reality, and that meet the uncertainties associated with equipment costs, the growth of present loads and the appearing of new ones.