Distribution planning with fuzzy loads and independent generation

The classical long range distribution network planning problem consists on deciding network investments to meet future demands at a minimum cost, while attending technical constraints. The decisions whether to construct or reinforce substations and branches lead to a mixed integer programming problem with a great number of decision variables. Besides, the network injections have a fuzzy nature on account of the nonavailability of statistical data in what concerns future loads namely in new areas and due to the presence of independent producers. Moreover several objectives must be taken into account. This leads to a fuzzy multiobjective, mixed integer problem. This paper presents a case study, adapted from a real network, that illustrates the application of an integrated methodology to deal with the planning problem. This case study aims into illustrating the proposed methodology and to point out its flexibility to adapt to the planner's needs.