Haibo Pen

The identity security of massive metering terminals in low-voltage power distribution systems is a foundational requirement for smart grids. These devices, deployed in physically exposed user-side environments, face threats of identity impersonation, data tampering, and future quantum attacks. Conventional public-key certificates and pre-shared key schemes lack quantum resistance and scalability, respectively, while existing post-quantum solutions do not address the combination of hierarchical key management and batch authentication that power metering at scale demands. This paper proposes a lattice-based hierarchical identity authentication method that maps the three-tier power system architecture onto a three-level key derivation chain of hierarchical identity-based signatures, compresses multiple terminal lattice signatures into a compact aggregate verifiable in a single equation, and embeds key agreement into the authentication flow so that terminals and the substation derive independent session keys without extra round trips. A prototype was implemented and tested across multiple security levels. The signature communication volume remained quasi-constant as the terminal count increased, reducing the signature transmission delay on narrowband PLC links. All terminals successfully established independent session keys with exact agreement between both sides. Security reduces to the average-case hardness of the small integer solution problem on lattices, satisfying mutual authentication, conditional anonymity, and unlinkability. The proposed scheme trades higher signing latency for security based on quantum-hard lattice assumptions, group batch authentication, and protocol integration, offering a viable pathway for identity security upgrades in power metering systems during the quantum migration window.

This paper proposes a phase-compensated enhanced second-order generalized integrator phase-locked loop (ESOGI-PLL) to suppress the frequency coupling effect (FCE) and its associated power quality degradation in grid-following inverters (GFLIs). An output impedance model incorporating FCE dynamics is formulated via signal perturbation analysis to quantify grid-current harmonic amplification and weak-grid instability boundaries. To suppress the FCE, the proposed ESOGI-PLL structurally embeds a phase-lead compensator, directly neutralizing the inherent phase lag of conventional filters. Simulation results demonstrate that the proposed ESOGI-PLL effectively mitigates PLL-induced frequency coupling, thereby improving the output-current quality and stability of GFLIs under weak-grid conditions. Compared with the synchronous reference frame phase-locked loop (SRF-PLL), the total harmonic distortion (THD) of the grid-current decreases from 24.33% to 2.47% under harmonic disturbances, while the transient settling time is significantly reduced, confirming the effectiveness of the proposed approach in enhancing both dynamic performance and output-current quality.

With the deployment of intelligent electricity collection devices, high-frequency and multi-dimensional electricity measurement data enable detailed analysis of the electricity consumption behaviors in low-voltage distribution areas. However, data missing due to factors like device failures or network congestion can significantly compromise analysis accuracy, consequently affecting the reliability of data-driven applications such as user profiling, load forecasting, and demand response. To address this, an electricity measurement missing data recovery method based on low-rank tensor completion with triangular canonical polyadic decomposition (LRTC-TCP) is proposed. This method breaks through the limitations of traditional one-dimensional data recovery and constructs a third-order observation tensor for multi-dimensional electricity measurement data. A domain-knowledge-enhanced LRTC model is then developed by introducing the constraint based on the electrical correlation among measured parameters, thereby simultaneously leveraging temporal similarity within single-dimensional data and inter-dimensional correlations for accurate completion. Subsequently, TCP decomposition is employed to alleviate the computational complexity of singular value decomposition during tensor rank estimation, reducing the overall computational cost. The equivalence of this transformation is rigorously proven. Finally, the proposed method is evaluated on the public electricity load monitoring iAWE dataset. Experimental results show that the method improves recovery accuracy and stability, even under high missing rates and continuous missing segments.

To address the inefficiency and inaccuracy of traditional offline sampling calibration and online calibration technologies for large-scale verification of electric vehicle off-board chargers (EVObCs), this paper proposes a remote verification method for EVObC measurement errors based on the charging interactions between electric vehicles (EVs) and EVObCs. Firstly, by leveraging the principles of infectious diseases transmission mode, a progressive metrological verification strategy for massive EVObCs is introduced. Additionally, a method for calculating benchmark accumulated charging energy, based on the temporal convolutional neural network-attention mechanism (TCN-Attention), is proposed to enhance the accuracy of EVObC measurement error identification. Furthermore, based on the energy decay model, a method for evaluating the intensity degradation of EVObC metrological performance is presented to objectively reflect the degradation of EVObC metrological accuracy and the dynamic variability of metering states in the proposed model. Finally, an experiment based on real data from 192 off-board chargers distributed over an area of 40 square kilometers was conducted. The results demonstrate that the proposed method can accurately identify standard non-compliant metering states of EVObCs, demonstrating its speed, accuracy, and strong generalization capabilities for large-scale verification and error identification across different types of EVObCs.

The application of federated learning in Virtual Power Plants (VPPs) addresses the data silo issue between VPPs and enhances their ability to cope with nonlinear and stochastic scheduling characteristics, which enables VPPs better accommodate distributed energy resources and flexible loads while participating in frequency regulation services. However, although existing federated learning methods strive to solve privacy protection issues, the plaintext transmission of gradients still exposes sensitive data to the threat of curious power control centers and external inference attacks. Therefore, a privacy-protected horizontal federated reinforcement learning approach for multi-VPP optimal scheduling is proposed in this paper. Firstly, a cost-based global optimization scheduling model for multiple VPPs is constructed, modeling the internal scheduling process of VPPs as a Markov decision process. Then, an improved secure horizontal federated multi-VPP collaborative training method is presented, and local models are trained using the Deep Transformer Q-Network algorithm, with local differential privacy and CKKS homomorphic encryption implemented to ensure privacy protection. Finally, a case study is conducted using frequency regulation ancillary service market data and the IEEE-39 bus system structure. Simulation results show that the proposed approach outperforms similar algorithms, achieving high levels of privacy protection and economic operation for VPPs.

The traditional centralized optimization method encounters challenges in representing the interaction among multi-agents and cannot consider the interests of each agent. In traditional low-carbon scheduling, the fixed carbon quota trading price can easily cause arbitrage behavior of the trading subject, and the carbon reduction effect is poor. This paper proposes a two-layer dynamic community integrated energy system (CIES) low-carbon collaborative optimization operation method. Firstly, a multi-agent stage feedback carbon trading model is proposed, which calculates carbon trading costs in stages and introduces feedback factors to reduce carbon emissions indirectly. Secondly, a two-layer CIES low-carbon optimal scheduling model is constructed. The upper energy seller (ES) sets energy prices. The lower layer is the combined cooling, heating, and power (CCHP) system and load aggregator (LA), which is responsible for energy output and consumption. The energy supply and consumption are determined according to the ES energy price strategy, which reversely affects the energy quotation. Then, the non-dominated sorting genetic algorithm embedded with quadratic programming is utilized to solve the established scheduling model, which reduces the difficulty and improves the solving efficiency. Finally, the simulation results under the actual CIES example show that compared with the traditional centralized scheduling method, the total carbon emission of the proposed method is reduced by 16.34%, which can improve the income of each subject and make the energy supply lower carbon economy.