The extensive integration of renewable energy resources (RERs) into modern power grids (MPGs), which severely lowers system inertia, standing severe frequency variations. These grids are vulnerable to cyber- attacks and natural uncertainties, creating a crucial need for resilient control schemes. This article introduces a novel resilience-coordinated scheme for interconnected MPGs. The scheme integrates Accelerating Virtual Rotor Control (AVRC) with Load Frequency Control (LFC), both governed by a modified Active Disturbance Rejection Controller (ADRC-IR). To optimize this controller's performance, a modified Horned Lizard Optimization Algorithm (MHLOA) is developed, enhancing its global search capabilities and convergence. MHLOA effectiveness is confirmed using a standard benchmark function. The ADRC-IR controller significantly reduces frequency and tie-line power deviations in the LFC system by 87.07%, 83.82%, and 76.54% respectively, compared to the TID, ADRC, and ADRC-PR controllers under high RESs penetration. Furthermore, the proposed AVRC/I & LFC strategy relied on ADRC-IR, which improves MPGs performance by 62.37%, 35.86%, and 30.22% when compared to (i) MPGs without AVRC/I, (ii) MPGs with LFC & AVRC/I, and (iii) MPGs with LFC & AVRC/I relied on a PI controller under high RESs penetration and cyber-attacks. The proposed scheme effectively preserves frequency stability within acceptable boundaries, demonstrating its resilience against MPGs challenges.
Cutting-edge engine alternatives have become essential to address the limitations of conventional transportation systems. Electric vehicles and hybrid electric vehicles provide efficient pathways for decarbonizing the transport sector, enabling a transition toward smart infrastructure and electrified mobility. This shift has created demand for data-driven tools to assess and optimize NEV energy efficiency. Accurate classification of NEV energy efficiency levels is essential for optimizing vehicle performance and battery utilization. Although traditional model-based and machine learning approaches exist, their limitations necessitate improved performance and reliability.To address this, the study proposes a novel framework called ANEVREMSFGDL, designed to enhance energy efficiency classification in new energy vehicles. The framework incorporates preprocessing techniques such as cleaning and transformation to ensure consistent data quality. Hybrid feature selection methods, including minimum redundancy maximum relevance and ReliefF, identify the most informative features. A dual-branch convolutional graph attention neural network performs classification into high, medium, and low efficiency categories. Finally, the AdamW optimizer is applied to improve model performance. Experimental results demonstrate that the proposed model outperforms existing techniques across multiple evaluation metrics. These findings demonstrate the effectiveness of the proposed model for NEV energy efficiency classification and highlight its potential for further improvement using datasets and adaptive learning approaches.
Smart urban building management systems require decision support under high-dimensional, heterogeneous indicators that depend on weather conditions, energy consumption patterns, and occupant behavior, while reliable ground-truth labels for optimal interventions are typically unavailable. Conventional soft set and multi-attribute decision frameworks assume parameter independence and rely on correlation- or variance-based dimensionality reduction, which can distort decisions when prerequisite relationships exist among building energy indicators. This paper presents a dependency-aware soft-set framework that explicitly models prerequisite-type relationships among energy parameters and enforces them through an order-preserving, non-inflating dependency propagation operator. The framework extends classical soft sets by integrating a directed dependency graph and dependency functions, ensuring logical consistency and convergence to a unique dependency-consistent closure for each initial membership matrix. It further enables decision-preserving redundancy reduction, in which parameters are removed only when their omission leaves dependency-consistent memberships, rankings, and final decisions unchanged. The framework is evaluated on 300 commercial buildings using 50 indicators derived from electricity consumption and outdoor temperature data, grouped into thermal, behavioral, and flexibility-related dimensions, and validated through repeated out-of-sample train-test splits in which dependency graphs learned on training buildings are applied unchanged to held-out test buildings. Dependency relations are acquired directly from data through a combined criterion of statistical association and prerequisite-violation constraints, without expert-defined assumptions. Results show stable, interpretable, and dependency-consistent building-energy prioritization with substantial dimensionality reduction and no loss of decision integrity, providing an interpretable portfolio-screening tool for building-energy decision making under uncertainty and limited data.
Hybrid microgrids (HMGs) have emerged as a critical framework for integrating renewable distributed generation, where the intermittency of sources such as solar PV necessitates robust energy storage and advanced control strategies. This paper proposes a high-performance hybrid controller that combines fuzzy logic with a fractional-order proportional–integral–derivative (FOPID) and conventional proportional–integral (PI) schemes for regulating a bidirectional DC–DC converter (BDC) interfaced with a battery energy storage (BES) system. The hybrid controller significantly enhances converter adaptability, optimises charge–discharge coordination, and ensures stable DC-link voltage under variable renewable energy conditions. To further strengthen microgrid stability, the study conducts a comprehensive evaluation of state-of-the-art grid-forming (GFM) inverter control strategies. Four GFM methods are investigated: droop control, virtual synchronous machine (VSM), standard virtual oscillator control (VOC), and two nonlinear oscillator-based extensions—dead-zone oscillator VOC (VOC-DZO) and Van der Pol oscillator VOC (VOC-VPO). Unlike droop and VSM techniques, which are reliant on phasor-based active and reactive power estimation, VOC methods utilise instantaneous current feedback, yielding superior transient and dynamic performance. The proposed control framework and inverter strategies are rigorously validated through extensive simulations in MATLAB/Simulink and real-time experiments on the OPAL-RT platform. Results demonstrate clear performance improvements in dynamic response, voltage regulation, and renewable energy utilisation, establishing the effectiveness of the hybrid control architecture for next-generation HMG applications.
Behind-the-meter energy networks (BTMENs) have emerged as cyber–physical infrastructures that integrate sustainable hardware platforms, edge devices, gateways, digital-twin energy networks, and intelligent control architectures to support real-time monitoring and optimized operation of distributed energy resources. However, the rapid proliferation of IoT-enabled sensing devices and high-density service information traffic introduces substantial stress on communication congestion, overloads embedded hardware, degrades digital twin modeling performance, and increases retransmission overhead. Congestion-induced packet loss can further lead to erroneous switching operations in power electronic hardware and unnecessary energy consumption during retransmission. To address these challenges, this study approaches a cloud-edge computing-enabled behind-the-meter congestion avoidance and energy utilization (BTMCAEU) mechanism that jointly considers communication efficiency, energy flow, and sustainable hardware utilization. Real-world data collected from a 20.8kW behind-the-meter solar Photovoltaic (PV) farm are leveraged to develop event-driven situational awareness and carbon-aware energy utilization. A two-layer indicator prioritized replay memory is introduced to reduce edge computational burden and accelerate learning convergence, serving as a guideline to set the transition priority. Benchmarking against RNN-, CNN-, and gated recurrent unit (GRU)-based models, the proposed BTMEN architecture provides a scalable, sustainable, and low-carbon communication–energy co-management solution suitable for smart buildings, intelligent microgrids, and Industry 4.0 environments.
Scientific workflows automate complex data analysis processes and are increasingly employed to handle large-scale datasets. Scientific communities leverage heterogeneous computing clusters, consisting of interconnected resources with diverse computational capabilities, to execute their large-scale data analysis workflows due to privacy and economic factors. Consequently, they often impose high computational demands and lead to significant energy consumption. Designing an effective energy-aware scheduling approach is therefore crucial for heterogeneous computing systems, as it enables high processing efficiency while minimizing overall energy usage. However, achieving both generality in workflow structures and solvability of the corresponding optimization models remains a major challenge in workflow scheduling. In this paper, we introduce a novel mathematical approach that models the non-overlapping execution of tasks successfully by representing disjoint time intervals as linear constraints. By applying this technique, we formulate a mixed-integer linear programming model that can be solved within a reasonable time, providing exact optimal solutions for arbitrary DAG workflow structures whenever computationally tractable. We analyze the trade-off between energy consumption and makespan. In addition, post-optimality analyses, including sensitivity analysis and relaxation of optimal energy consumption for makespan improvement, are investigated. Experimental results demonstrate that the proposed MILP approach consistently outperforms competing methods, achieving up to approximately 6%–7% reduction in energy consumption compared with the best competing approach and up to about 10%–11% compared with the worst-performing method.
Efficient maximum power point tracking (MPPT) in grid-connected wind energy conversion systems (WECS) remains challenging under stochastic wind conditions due to the slow convergence and oscillatory behavior of conventional algorithms. Although artificial intelligence based MPPT techniques improve tracking capability, their black-box nature may lead to physically inconsistent predictions under turbulent wind dynamics. This paper proposes a physics-informed long short-term memory (PI-LSTM) based MPPT framework that integrates aerodynamic constraints derived from the wind turbine power equation and the tip speed ratio relationship into the learning process. By embedding turbine physics within the learning structure, the proposed approach enables physically consistent optimal rotor speed references while preserving the temporal learning capability of LSTM networks. To further enhance system stability, 2DOF-FOPI controllers are implemented in both the machine-side and grid-side converters of a vector control permanent magnet synchronous generator WECS. The proposed framework is validated through MATLAB/Simulink simulations and OPAL-RT real-time simulator under random, real, and gust wind profiles. Comparative analysis with P&O, optimal torque control, and LSTM based MPPT methods demonstrates that the proposed PI-LSTM approach achieves MPPT efficiencies up to 99.95% and increases harvested energy by up to 4.4% compared with the P&O under real wind conditions. Furthermore, the integration of the 2DOF-FOPI controller significantly improves system dynamics by reducing DC link voltage overshoot from 7.47% to 1.20%, decreasing torque ripple from 29.82% to 16.82%, and lowering grid current total harmonic distortion from 4.57% to 1.58%. The integration of PI-LSTM with fractional-order control provides improved energy extraction, enhanced disturbance rejection capability.
For the constrained multi‑objective power flow optimization (MOOPF), an adaptive intelligent migration multi-task multi-objective evolutionary algorithm (MTCMO-AIMS) was proposed for optimal power flow (OPF) problems with wind-solar photovoltaic fields and FACTS devices. A mathematical model is established to describe the OPF formulation while explicitly accounting for the variability and unpredictability of wind and solar generation. The essential constraints include the nodal power balance relationships, the permissible operating ranges of conventional generator active power outputs, the allowable rating bounds of shunt reactive power compensation equipment together with the operational limits imposed on Flexible AC Transmission System (FACTS) apparatus. The adaptive migration probability mechanism (AMPM) incorporates additional modules for adaptive migration probability calculation and dynamic updating, enabling knowledge migration to be dynamically adapted to the real-time state of the population. Replacing blind random migration with directional elite-guided transfer, the intelligent elite-guided migration strategy (IEGMS) makes better use of complementary information across distinct optimization tasks. MTCMO-AIMS is applied to complex constrained MOOPF problems that integrate wind farms, solar farms and FACTS equipment, with detailed analysis on the IEEE 30-bus system and IEEE 57-bus system. The obtained results are compared with those of the AFSEA, TPCMaO, TSTI and CCMO algorithms. Experimental results demonstrate that it can find excellent solutions and generate a satisfactory Pareto frontier in both system scales, verifying its effectiveness and scalability in larger-scale power systems.
Smart grid operation strongly depends on Accurate Distribution System State Estimation (DSSE), which is highly affected by noisy measurements, data gap and malicious data manipulation. The current paper introduces an Energy Efficient DSSE framework that uses Sparse Deep Unrolling Networks (SDUN) and data provenance enforcement through blockchain. SDUN architecture converts the iterative optimization-based state estimation into a learnable deep model which can converge fast with a low amount of computational energy. Blockchain provenance also guarantees that the data is only admitted to the estimation pipeline with verified measurement data of IoT sensors and PMUs, avoiding data poisoning attacks. The joint design helps a great deal to decrease the redundant computation and maintain the accuracy of the estimation. Adversarial data injection under simulation shows a State Estimation Error Reduction of 36.2, Computational energy savings of 28.5% compared with the conventional WLS-DSSE baseline., Data Trust Assurance rate of 99.3, Estimation Convergence Acceleration of 41.7 and Improved Robustness Index of 33.9. The presented solution offers the safe, power-efficient, and scalable DSSE solution of the next-generation distribution networks.
Energy management is a major challenge in Internet of Things (IoT) networks, particularly in large-scale and dynamic environments where node behavior, traffic patterns, and energy consumption are uncertain. Many existing approaches rely on static control policies or fully control-driven deep reinforcement learning (DRL) schemes, which may suffer from limited adaptability, sensitivity to parameter tuning, or unnecessary computational overhead under rapidly changing conditions. Consequently, achieving stable and energy-efficient operation remains an open problem.This paper proposes a predictive and adaptive energy management framework for IoT networks that explicitly incorporates behavioral uncertainty into the decision-making process. A hybrid forecasting model combining Prophet and long short-term memory (LSTM) networks is employed to capture both long-term trends and short-term temporal variations in node-level energy consumption. The deviation between predicted and observed energy behavior is interpreted as a dynamic behavioral signal rather than a fixed anomaly. This residual information is then integrated into a deep reinforcement learning–based decision layer to adapt energy management policies online without predefined thresholds or rule-based interventions.The effectiveness of the proposed framework is evaluated through simulations conducted in MATLAB and NS-3 under varying network densities and traffic dynamics. Performance is assessed in terms of average energy consumption, network lifetime, and energy usage stability. Results show consistent improvements over non-adaptive methods, rule-based strategies, and conventional DRL baselines, achieving approximately 3–4% reductions in average energy consumption and corresponding extensions in network lifetime, confirming the practical effectiveness of the proposed approach.
Quantum-dot Cellular Automata (QCA) is a nanotechnology that—due to its lower energy consumption, reduced area, and superior speed and latency characteristics compared to CMOS technology—could serve as a viable future alternative to it. QCA technology can be employed to design logic circuits. The 2:1 multiplexer (MUX) and the D-latch have diverse applications, such as serving as memory elements and facilitating the simultaneous transmission of data. For this reason, in this paper, a 2:1 multiplexer is designed in QCA technology, which has 10 QCA cells, a delay of 0.25 clock cycles, an area of 0.01µm2, stability of 87.5%, and also its energy loss at the level of 0.5EK is equal to 10.85 meVs. Multiplexers are widely used in digital circuits. They can also be used to design circuits such as latches, flip-flops, counters, shift registers, and larger multiplexers. Therefore, using the proposed multiplexer, two new D-latches in QCA technology were designed and proposed. The first latch has 24 QCA cells, a delay of 0.5 clock cycles, an area of 0.01 µm2, and a robustness of 73.38%. The second latch has 22 QCA cells, a delay of 0.25 clock cycles, an area of 0.01 µm2 and a robustness of 75%. Also, the energy loss of the first and the second latch at the level of 0.5EK is equal to 42.04 meVs and 33.43 meVs, respectively. Also a 4:1 MUX have proposed using the proposed 2:1 MUX which have 41 QCA cells, 0.75 cycle latency and are of 0.03 µm2.
Hybrid microgrids with renewable energy sources provide reliable, efficient and sustainable energy supply systems. But current control techniques face challenges in smooth transition, stability and power quality issues during dynamic operation. This research paper a Model Predictive Control (MPC)-based approach for seamless transition and power quality improvement in a hybrid AC/DC islanded microgrid. The proposed system, simulated using MATLAB/Simulink, comprises of photovoltaic (PV) panels, wind energy conversion system (WECS), battery storage system (BSS), AC/DC loads and power electronic converters. The microgrid is designed to operate in grid-connected, islanded, and transition modes for seamless operation. The main control challenges involve regulating the DC bus voltage, AC voltage and frequency, improving power quality, and managing the battery state-of-charge (SOC) while avoiding overcharging and over-discharging. The developed MPC is designed to anticipate system behavior and determine switching states to reduce voltage and current deviations, while managing power flow. Harmonic filtering with a LCL filter and low total harmonic distortion (THD) are achieved. The controller is capable of compensating for unbalanced loads and intermittent renewables, resulting in stable and sinusoidal output. A special mode control strategy ensures seamless mode transition without dip or frequency drop. The proposed approach effectively reduces THD from 28.4% to 0.43% with MPC and LCL filter. The proposed system ensures stability under different load and renewable energy conditions, such as 20% load increase and grid outage, with zero energy not supplied (ENS), due to the coordination of battery storage.
Smart grids, despite their advanced monitoring and communication capabilities, they are increasingly vulnerable to cyberattacks. Among the critical components of smart grids, phasor measurement units (PMUs) are essential for monitoring power systems. While existing research has explored hybrid neural networks for attack detection, their reliance on extensive training data limits adaptability. This study proposes a case-based reasoning (CBR) method for detecting anomalies and cyberattacks in PMU data. The developed CBR system identifies patterns of normal and attack states by comparing real-time input with historical cases stored in a structured, incrementally updated library. This work is a simulation-based study, using data from a General Electric-validated smart grid model, emulating realistic PMU environments. The framework includes a multi-stage feature extraction pipeline that transforms voltage–frequency signals into pseudo-color images, followed by many steps to compute several statistical indices. The proposed method achieved 100% accuracy on clean data and 94.44% accuracy with a 94.47% F1-score for mixed datasets. Thus, it shows improved accuracy and adaptability compared to existing methods and offers a practical and effective solution for enhancing cyber–physical security in smart grid PMU monitoring.
Large-scale, heterogeneous Internet of Things (IoT) networks, such as those in supply chain management, healthcare, and agriculture, rely heavily on optimization and intelligent technologies. However, this reliance introduces significant challenges in control and security. Blockchain (BC) technology, with its distributed ledger, immutable records, and independence from centralized authorities, offers a promising solution by enhancing security, reliability, and distributed trust in IoT applications. Despite its potential, IoT devices are constrained by limited energy resources, computational power, storage, and bandwidth, which complicate the implementation of consensus algorithms and BC data storage. Furthermore, traditional BC consensus algorithms face limitations such as low throughput, high energy consumption, and inherent privacy concerns. This paper presents a scalable and efficient consensus mechanism designed to meet the unique requirements of IoT environments. The proposed mechanism leverages Zero-Knowledge Proof (ZKP) to provide robust security and privacy while maintaining compatibility with IoT constraints. In addition, it utilizes a multi distributed ledger combining Directed Acyclic Graphs (DAGs) and BC to enhance scalability. The comparative results indicate that the proposed method outperforms existing state-of-the-art mechanisms, providing enhanced security and privacy while achieving reductions in bandwidth, occupied memory, and energy consumption by 28%, 34%, and 22%, respectively. Additionally, the method demonstrates a 15-fold improvement in scalability.
