Real-Time Modeling and Control of a Cyber-Physical Energy System

this paper introduces an approach for applying real-time scheduling techniques to balance electric loads in cyber-physical energy systems. The proposed methodology aims to determine, guarantee and optimize an upper bound on the peak load of electric power, which represents a desirable feature for both the electricity supplier and the user of the electrical system. For this purpose, networked electric devices are modeled using parameters derived from the real-time scheduling discipline used for computing systems. Therefore, the upper bound can be enforced by predictably and timely switching on/off the electric devices composing the electrical system. The paper contribution include: the illustration of the relevance of electric load balancing in cyber-physical energy systems, motivating the use of real-time scheduling techniques to achieve predictability of electric loads scheduling; the presentation of a novel and powerful modeling methodology of the physical system based on a set of periodically activated loads, to enable the use of traditional real-time system models and scheduling algorithms, with adequate adaptations, to man- age loads activation/deactivation. We finally derive interesting properties of real-time parameters and provide theoretical results concerning the computation of their values.