Risk-Aware CPS Operation Scheme for Self-organizing Balancing-Power Pools

One of the challenges in cyber-physical systems (CPS) is to guarantee system safety, i.e. system functionality has to ensure that system state paths are allowed also in presence of environmental uncertainty and decentralized actions by autonomous components (aka agents) in the edge. Particular interest is in scalable adaptive control schemes where a system observer just broadcasts one or few guiding quantities that agents then use to adjust their actions in a self-organizing manner. In the electricity system, challenge is to scale up reliable operation of growing numbers of flexible decentralized energy resources (DER) within their individual constraints under volatile renewable generation for various smart grid use cases. In this paper, we present a case study for a balancing power use case. Currently, an aggregator controls each flexible DER in his pool individually, to fulfill the volatile power demand signal. Here we develop a system operation scheme with broadcasting, for simplicity, a single demand quantity using just aggregation of each energy agent’s behavior prognosis, which is based on utility function taking operational constraints of its flexible DER into account. System simulations for historic aFRR signals show that the proposed scalable adaptive control allows for safe CPS operation within the limits of nominal DER characteristics. In particular, heterogeneous pools with an agent utility function for risk-aware operation significantly extend region of safe operation close to the physical system limits.