Spectral clustering for designing robust and reliable multi‐MG smart distribution systems

Abstract

Reliability has become a key design aspect in modern energy system's planning. Owing to the higher fault rate in power distribution systems (PDSs), comparing with generation and transmission systems, considering reliability in PDSs' planning is very crucial. This study presents a novel robust approach to cluster the existing PDSs with intermittent distributed generators (DGs) into a set of reliable microgrids (MGs). For this purpose, first, a new reliability index is defined to evaluate the reliability of MGs in terms of real and reactive power adequacy as well as frequency and duration of interruptions. Then, the k -means algorithm, based on weighted graph partitioning method, is proposed for changing the system into a multi-MG system. Furthermore, a modified version of particle swarm optimisation approach is proposed and the Silhouette technique is used to determine the optimal location and sizes of DGs as well as the number of MGs. The design and sensitivity analysis performed by the proposed multi-objective optimisation algorithm on the well known IEEE 69-bus distribution system show the effectiveness and robustness of the proposed algorithms for constructing reliable MGs in modern PDSs.