<scp>IJRNC</scp> Guest Editorial: Control and Learning for Cyber Physical Energy Systems

The design and operation of power systems are undergoing transitions to embrace new energy sources, support new applications, and resist diverse disruptions. With the supply crisis of oil and natural gas, renewable energy sources (solar, wind, hydro, tidal, geothermal, biomass, etc.) are gaining popularity. In addition, with the Industry 4.0 revolution, new applications such as electric vehicles (EVs) and aircraft call for new transportation infrastructure and energy support. In the meantime, power systems are undergoing disruptions from both physical and cyber domains, including natural hazards, extreme weather, communication failures, maloperation, and attacks of various physical and cyber forms. New concepts for modern power systems have attracted many researchers, including microgrids, virtual power plants, and energy sharing, among many others. These challenges and opportunities for modern power systems require new control solutions to be developed in every phase of the energy journey, ranging from generation to transmission, distribution, storage, and consumption. Distributed control solutions for microgrids have the potential to improve flexible and robust operations. Consensus-based battery management solutions improve the lifetime of power supply in edge devices. Cyber-physical control systems provide co-design techniques across cyber- and physical dimensions for complex power systems. Game-theoretic approaches promise to facilitate energy distribution under cooperative and non-cooperative settings. Machine learning techniques allow optimal power system scheduling under uncertainties and unknowns. Resilient control solutions minimize the impact of attacks and allow the power system to quickly restore to its normal operation. The development of effective control solutions for modern power systems requires collaborative efforts from multiple communities from the theory to the application domains. This special issue provides the incentives to bring together researchers and engineers to collaboratively tackle the fundamental challenges in control for power systems. The 11 accepted papers are summarized as follows. In “Distributed Cooperative Control of a Flywheel Array Energy Storage System,” X. Lyu, Y. Hu, and Z. Lin study a flywheel array energy storage system (FAESS) that is formed of an array of flywheel energy storage systems (FESSs). Building on a charge/discharge model for individual FESS units and an undirected communication graph, they develop a distributed control algorithm to deliver the desired power while balancing the state-of-charge (SoC) across the array. They analyze performance and robustness, including resilience to communication disruptions (e.g., intermittent connectivity). The paper “Distributed Event-triggered Algorithms for the Management of Networked Battery Systems” by Y.-Y. Qian, T. Meng, Z. Lin, Y. Wan, and Y. Shamash investigates the distributed control of a networked battery system to achieve state-of-charge balancing across all battery units while delivering a desired total power. To reduce the communication and computational load, they design event-triggered controllers coupled with event-triggered distributed estimators. The authors also prove that Zeno behavior is excluded under the proposed schemes. In “A Neural Fictitious Self-Play Anti-Jamming Strategy for Secondary Frequency Control in Microgrids with Imperfect Observations,” Y. Li, S. Liu, L. Zhu, and B. Chen address secondary frequency control for microgrids under rational jamming attacks that target both feedback and feedforward channels, all in the presence of imperfect observations. They formulate a multiplayer, multistage security game and develop a neural fictitious self-play (NFSP) anti-jamming strategy that counters such attacks. The authors establish that the proposed NFSP approach achieves a Nash equilibrium. The paper “On the Sample Complexity of Observers for Unknown Linear Systems with Biased Dynamics Estimations” by X. Ding, H. Wang, J. He, C. Chen, and X. Guan investigates observer design for unknown, noisy, linear time-invariant systems. It analyzes the impact of the bias–variance trade-off in dynamics estimation to observer design using a support vector regression (SVR)-based estimator. The authors analyze design parameters to achieve both estimation accuracy and performance optimality. The paper “An Efficient Hierarchical Electric Vehicle Charging Control Strategy” by C. He and Z. Zhang investigates an EV charging control strategy that accounts for network-wide communication overhead, computational complexity, total energy cost, user preferences, and data-privacy protection. The proposed hierarchical framework employs centralized control at the EV-aggregator level and distributed control among EVs within each aggregator. The paper “Event-Triggered Control of Power Buffers in DC Microgrids over an Unreliable Network” by Y.-Y. Qian, A. Premakumar, Y. Wan, Z. Lin, Y. Shamash, and A. Davoudi investigates distributed event-triggered control of power buffers over an unreliable network subject to packet loss and transmission delays. The authors develop a dynamic event-triggered communication mechanism whose key feature is a designable, strictly positive minimum inter-event time, maintained even in the presence of packet losses and delays. The paper “Distributed Aperiodic Sampled-Data Implementation for Dual-Objective Control of Energy Storage Systems” by Y. Su, X. Liu, H. Cai, Y. Zhang, and X. Chen addresses a dual-objective control problem for energy storage systems comprising multiple energy storage units (ESUs). The proposed sample-and-hold controller integrates a leaderless consensus algorithm for state-of-energy (SoE) balancing and a leader-following distributed observer to estimate the power command. The paper “Distributed Adaptive Optimization for Microgrids” by Y. Yan and J. Sun investigates adaptive, distributed optimization for charge–discharge management in microgrids. The authors develop a one-critic neural network to track the dynamic equilibrium point of a battery pack's remaining energy under stochastic generation and uncertain loads. The paper “Privacy-Preserving Optimization Algorithm for Distributed Energy Management over Time-Varying Graphs: A State Decomposition Method” by M. Luan, G. Wen, and T. Yang addresses privacy in distributed energy management. It develops a privacy-preserving distributed optimization algorithm that minimizes cost via gradient tracking while satisfying local constraints. The proposed state-decomposition method protects the confidentiality of agents' cost-gradient information even in the presence of network eavesdroppers. The paper “New Safety Feedback Control Design to Guarantee Adequate Frequency Performance in Microgrids” by F. Taousser, S. Morovati, Y. Zhang, S. Djouadi, K. Tomsovic, and M. Olama analyzes safety in a diesel-wind energy system. To keep critical variables within prescribed bounds following a disturbance, the authors design a safety feedback controller to provide a supplementary input to the wind turbine generator. The paper “Cascaded Integral Minus Tilt Integral Derivative with Filter for Frequency Stabilization of V2G/G2V-Enabled Hybrid Microgrid” by S. Santra and M. De studies frequency stabilization in a hybrid microgrid with EVs. A diesel-engine generator is incorporated for load-frequency control under system frequency disturbances, using a cascaded Integral-Minus-Tilt-Integral-Derivative-with-Filter (I–TDN)–Proportional–Integral (PI), that is, (I-TDN)–PI controller. The sensitivity of the proposed controller is investigated in the presence of system nonlinearities, load perturbations, time delays, parameter variations, and renewable-energy-source (RES) power fluctuations. The authors declare no conflicts of interest.