Modelo híbrido não linear multiestágio aplicado ao planejamento da expansão de sistemas de distribuição de energia elétrica com alocação de geração distribuída

Abstract

The planning of expansion of distribution networks aims to set solutions for the energy growth demand in the electric power system. The expansion topologies of the electric networks must respect technical and operational criteria in order to provide quality energy, reliability and continuity to the final consumers. In recent years, the power system essentially centralized power has undergone modification due to a greater penetration of distributed systems in the network. The inclusion of these decentralized systems is a trend that contributes to the expansion solution and the future demand attendance. The planning challenge is to combine all these variables, representing them as close as possible to the real situation. Therefore, this research presents a new algorithm for the expansion planning application of distribution generation, contemplating the distributed generation allocation. The algorithm was developed from a nonlinear model for the representation of the electric network and calculation of its power flow, considering the demand growth in different stages - multistage planning. These two features (non-linear modeling and multistage planning) bring the proposed model to the real scenario, and make the results more reliable to the existing electrical pattern. The model is solved by a specialized heuristic algorithm, applied in two scenarios, the first one with a pseudodynamic sensitivity indicator and the second one with a dynamic sensitivity indicator. The algorithm suggests building a new line for each interaction, where the chosen line is the one with the lowest proportional value to the combination: lower building cost and lower losses. Once determined the topology, the second part of the model indicates the optimal point for insertion of a distributed generation, in each topology, in order to minimize the total electrical losses in the system. The installation of this distributed generation is determined from the optimum capacity established. The method for establishing variants of the distributed generation is performed by analytical models. The final model was computationally tested in three adapted test systems from the specialized literature, allowing the validation and evaluation of its performance. The results were satisfactory.