Smart reconfiguration using fuzzy graphs

Distribution Systems traditionally have a tree-like structure with several branches. They supply loads that vary through the day and as a result, some branches are loaded more than others are. By reconfiguring the system, loads from the overloaded branches may be moved to under-loaded branches. Consequently, loads in the branches can be balanced so that the real power losses are reduced and the voltage profile is improved. While Smart grid technologies in the future will facilitate real-time reconfiguration of distribution systems, it requires the use of efficient and fast methods. In that direction, this paper proposes a smart reconfiguration method without using load flow. It models the distribution system as a fuzzy graph using a data structure. It quantifies membership functions of edges of fuzzy graph using line impedance values and uses approximate MVA flow in the lines to characterize the fuzzy graph. Using the membership functions and line flow values, the method computes a fuzzy system participation function. The proposed method minimizes this function to balance the loads in the branches and consequently minimizes real power losses. The paper reports results from a sample system study to demonstrate the proposed method.