Smart home energy management system under false data injection attack

Abstract

Modern smart home energy management system (SHEMS) is naturally prone to cyber attack, hence it demands cyber attack resilient scheduling schemes. Current scenario of SHEMS may result in increased charging and discharging cycles deteriorating the battery life. Therefore, demand scheduling formulations also need to cater the effect of battery degradation cost along with user comfort. The present work attempts to formulate a comprehensive scheduling problem in terms of energy cost minimization considering the battery degradation cost. Further, a cyber attack resilient scheduling model is proposed in this study. This article investigates the effect of demand scheduling on the life span of battery as well as the energy cost. Further, false data injection attack (FDIA) has been modeled using machine learning techniques, and its effects on the scheduling has also been incorporated in the objective function. Scenario tree based stochastic bill generation has been also formulated to develop an FDIA resilient scheduling. Optimisation results of the study have established that the resulting formulation is robust against FDI attacks. © 2020 John Wiley & Sons Ltd