Resiliency and reliability of the power grid in the time of COVID-19: An integrated ABC-K-means model for optimal positioning of repair crew

More than one year has passed since the outbreak of a new phenomenon in the world, a phenomenon that has affected and transformed all aspects of human life, it is nothing but pandemic of COVID-19. The field of electrical energy is no exception to this rule and has faced many changes and challenges over the 2020. In this paper, by applying artificial intelligence and the integrated clustering model, by k-means technique, combined with the meta-heuristic artificial bee colony (ABC) algorithm a new methodology is presented in order to optimal positioning of the repair crew based on annual data of power grid under situation of COVID-19 to improve the reliability and resiliency of the network due to the importance of electricity for medical purposes, home quarantine, telecommuting, and electronic services. Current research benefits from real interruption data related to year 2020 in Isfahan Province (Iran), reflexing both the huge changes in patterns of power consumption and dispatching as well as novel geographical distribution of blackouts due to COVID pandemic. The temporal distribution of interruptions is very close to the uniform distribution and the geographical distribution of interruptions relative to the density of subscribers had a normal distribution. Accordingly, proposed model is implemented for clustering the spatial data of blackouts recorded during 2020. The number of clusters is equal to the number of repair teams which in this study is considered equal to three. In the next step, the average spatial coordinates of the points of each cluster are calculated, which after reviewing the geographical conditions in the geo-spatial information system (GIS), indicates the optimal point for the deployment of electrical repair crew related to that cluster. The research findings show that after using the optimal points for a month, system average interruption duration index (SAIDI) decreased by an average of 23% compared to the same period of the 2020.