An improved minimum-cost spanning tree for optimal planning of distribution networks

An improved minimum-cost spanning tree algorithm (MCST) for optimal planning of distribution network is put forward. The proposed method is based on Kruskal algorithm with improvements such as the shortest paths between each two load points are regarded as the candidate paths, and that once a path is determined, the connection nodes on it are turned into load points. Consequently, the conventional MCST's shortcoming of not able to handle the connection nodes on paths is over come. Further more, the proposed method is extended into meshed network planning by determining the shortest routes between each source point before hand and then executing the improved MCST under the condition of no more loops being formed. The way to modify the size of the cross section of conducting wires for looped routes, trunk routes and branches are dealt with. Further optimization to obtain less losses and voltage drops are also discussed. The feasibility of the proposed method is shown by a practical example.