A novel clustering-based method for reliability assessment of cyber-physical microgrids considering cyber interdependencies and information transmission errors

Abstract

In recent years, the evolution of microgrids (MGs) toward a complex interacted cyber-physical system is significant, which has received much attention. Cyber-physical MGs (CPMGs) are vulnerable to various cyber-physical interdependencies mainly. Therefore, the reliability evaluation of such systems is a crucial issue. Several research works have focused on the reliability assessment of CPMGs in the literature. However, there is a research gap in developing fast and accurate models to assess the reliability of CPMGs, simultaneously considering cyber and physical failures, cyber-power interdependencies, and information transmission errors. This article aims to fill such a knowledge gap by developing a new clustering-based method to evaluate the reliability of CPMGs. The scenarios for information transmission errors are generated by the Monte Carlo simulation (MCS), besides uncertain physical parameters, e.g., the output power of renewable distributed generations (DGs). Afterward, the clustering algorithms are used to reduce the number of scenarios. In this paper, the k-means clustering algorithm has been selected to cluster the scenarios. The introduced clustering-based method is applied to a CPMG, and test results are compared to available MCS-based studies. The inaccuracy of the proposed clustering-based method is less than 3%, while its execution time is about 22 times faster than MCS-based ones. The comparative test results under various scenarios like available methods, neglecting the cyber failures and information errors illustrate the advantages of this research. The comparative test results illustrate the accuracy of the reliability calculations, while the execution time is desired.