Distribution system includes medium voltage and low voltage networks. Low voltage network is usually supplied using its dedicated transformer station. Simulations and result interpretation for each of the low voltage grids is burdensome in terms of required time and computational resources. The aim of this thesis was to develop a method for categorisation of networks and to use this method with a number of low voltage networks of Novo mesto.
Matlab was used for clustering. We have chosen k-mediod method, which requires a number of the groups to be defined in advance. During the clustering, a representative grid is chosen for each of the groups. In order to check accuracy of our approach, we separately simulated all the networks in a particular group and the representative network of this group using OpenDSS environment. Accuracy of group representation by the representative network was then examined through the voltage comparison.
Networks were assorted into four groups: densely populated villages; public places and companies; sparsely populated villages; cities. Ability to represent the whole group was examined for all the representative network except for public places and companies, through network simulations.
Network clustering and examination of just representative network only speeds up the simulation process and eases the result analysis.
|
