Sürekli Düzlemsel Uzayda Ağaç-Yıldız Bağlantılı Elektrik Dağıtım Ağı Tasarımı

Abstract

In this thesis, a two-level distribution network design problem is studied in order to serve the demand points of rural areas where there is difficulty in accessing electrical energy. At the lower level, demand points are connected to generators or converters in a star pattern within a certain distance to receive service. At the higher level, the generators feed the converters over tree networks. The problem includes constraints that enable converters to serve without exceeding their capacities and prevent voltage drop due to the distance of demand points from the converter. All points on the plane are candidate locations for the converters and generators. It is aimed to design a distribution network by minimizing generator, converter and connection costs. The discrete problem is also difficult to solve for large-scale examples. Therefore, three different heuristic methods are proposed to solve the discrete expression of the problem. In the first heuristic method, candidate facility locations are found and local networks are created that provide a direct connection between demand points and the facilities. Then, the higher-level problem is solved and the connections between the facilities are added to the network. In the other two heuristics, we aim to obtain the solutions of the lower-level problem in more reasonable times. Finally, location adjustments on the continuous space are proposed to improve the obtained results. The effects of all methods are demonstrated by numerical experiments.