Global Energy Interconnection最新文献

Timely inspection of defects on the surfaces of wind turbine blades can effectively prevent unpredictable accidents. To this end, this study proposes a semi-supervised object-detection network based on You Only Looking Once version 4 (YOLOv4). A semi-supervised structure comprising a generative adversarial network (GAN) was designed to overcome the difficulty in obtaining sufficient samples and sample labeling. In a GAN, the generator is realized by an encoder- decoder network, where the backbone of the encoder is YOLOv4 and the decoder comprises inverse convolutional layers. Partial features from the generator are passed to the defect detection network. Deploying several unlabeled images can significantly improve the generalization and recognition capabilities of defect-detection models. The small-scale object detection capacity of the network can be improved by enhancing essential features in the feature map by adding the concurrent spatial and channel squeeze and excitation (scSE) attention module to the three parts of the YOLOv4 network. A balancing improvement was made to the loss function of YOLOv4 to overcome the imbalance problem of the defective species. The results for both the single- and multi-category defect datasets show that the improved model can make good use of the features of the unlabeled images. The accuracy of wind turbine blade defect detection also has a significant advantage over classical object detection algorithms, including faster R-CNN and DETR.

To predict renewable energy sources such as solar power in microgrids more accurately, a hybrid power prediction method is presented in this paper. First, the self-attention mechanism is introduced based on a bidirectional gated recurrent neural network (BiGRU) to explore the time-series characteristics of solar power output and consider the influence of different time nodes on the prediction results. Subsequently, an improved quantum particle swarm optimization (QPSO) algorithm is proposed to optimize the hyperparameters of the combined prediction model. The final proposed LQPSO-BiGRU-self-attention hybrid model can predict solar power more effectively. In addition, considering the coordinated utilization of various energy sources such as electricity, hydrogen, and renewable energy, a multi-objective optimization model that considers both economic and environmental costs was constructed. A two-stage adaptive multi- objective quantum particle swarm optimization algorithm aided by a Lévy flight, named MO-LQPSO, was proposed for the comprehensive optimal scheduling of a multi-energy microgrid system. This algorithm effectively balances the global and local search capabilities and enhances the solution of complex nonlinear problems. The effectiveness and superiority of the proposed scheme are verified through comparative simulations.

Owing to the integration of energy digitization and artificial intelligence technology, smart energy grids can realize the stable, efficient and clean operation of power systems. However, the emergence of cyber-physical attacks, such as dynamic load-altering attacks (DLAAs) has introduced great challenges to the security of smart energy grids. Thus, this study developed a novel cyber-physical collaborative security framework for DLAAs in smart energy grids. The proposed framework integrates attack prediction in the cyber layer with the detection and localization of attacks in the physical layer. First, a data-driven method was proposed to predict the DLAA sequence in the cyber layer. By designing a double radial basis function network, the influence of disturbances on attack prediction can be eliminated. Based on the prediction results, an unknown input observer-based detection and localization method was further developed for the physical layer. In addition, an adaptive threshold was designed to replace the traditional precomputed threshold and improve the detection performance of the DLAAs. Consequently, through the collaborative work of the cyber-physics layer, injected DLAAs were effectively detected and located. Compared with existing methodologies, the simulation results on IEEE 14-bus and 118- bus power systems verified the superiority of the proposed cyber-physical collaborative detection and localization against DLAAs.

Achieving a balance between accuracy and efficiency in target detection applications is an important research topic. To detect abnormal targets on power transmission lines at the power edge, this paper proposes an effective method for reducing the data bit width of the network for floating-point quantization. By performing exponent prealignment and mantissa shifting operations, this method avoids the frequent alignment operations of standard floating-point data, thereby further reducing the exponent and mantissa bit width input into the training process. This enables training low-data-bit width models with low hardware-resource consumption while maintaining accuracy. Experimental tests were conducted on a dataset of real-world images of abnormal targets on transmission lines. The results indicate that while maintaining accuracy at a basic level, the proposed method can significantly reduce the data bit width compared with single-precision data. This suggests that the proposed method has a marked ability to enhance the real-time detection of abnormal targets in transmission circuits. Furthermore, a qualitative analysis indicated that the proposed quantization method is particularly suitable for hardware architectures that integrate storage and computation and exhibit good transferability.

The integration of distributed generation brings in new challenges for the operation of distribution networks, including out-of-limit voltage and power flow control. Soft open points (SOP) are new power electronic devices that can flexibly control active and reactive power flows. With the exception of active power output, photovoltaic (PV) devices can provide reactive power compensation through an inverter. Thus, a synergetic optimization operation method for SOP and PV in a distribution network is proposed. A synergetic optimization model was developed. The voltage deviation, network loss, and ratio of photovoltaic abandonment were selected as the objective functions. The PV model was improved by considering the three reactive power output modes of the PV inverter. Both the load fluctuation and loss of the SOP were considered. Three multi-objective optimization algorithms were used, and a compromise optimal solution was calculated. Case studies were conducted using an IEEE 33-node system. The simulation results indicated that the SOP and PVs complemented each other in terms of active power transmission and reactive power compensation. Synergetic optimization improves power control capability and flexibility, providing better power quality and PV consumption rate.

Vibration measurements can be used to evaluate the operation status of power equipment and are widely applied in equipment quality inspection and fault identification. Event-sensing technology can sense the change in surface light intensity caused by object vibration and provide a visual description of vibration behavior. Based on the analysis of the principle underlying the transformation of vibration behavior into event flow data by an event sensor, this paper proposes an algorithm to reconstruct event flow data into a relationship correlating vibration displacement and time to extract the amplitude-frequency characteristics of the vibration signal. A vibration measurement test platform is constructed, and feasibility and effectiveness tests are performed for the vibration motor and other power equipment. The results show that event-sensing technology can effectively perceive the surface vibration behavior of power and provide a wide dynamic range. Furthermore, the vibration measurement and visualization algorithm for power equipment constructed using this technology offers high measurement accuracy and efficiency. The results of this study provide a new noncontact and visual method for locating vibrations and performing amplitude-frequency analysis on power equipment.

In a “low-carbon” context, the power load is affected by the coupling of multiple factors, which gradually evolves from the traditional “pure load” to the generalized load with the dual characteristics of “load + power supply.” Traditional time-series forecasting methods are no longer suitable owing to the complexity and uncertainty associated with generalized loads. From the perspective of image processing, this study proposes a graphical short-term prediction method for generalized loads based on modal decomposition. First, the datasets are normalized and feature-filtered by comparing the results of Xtreme gradient boosting, gradient boosted decision tree, and random forest algorithms. Subsequently, the generalized load data are decomposed into three sets of modalities by modal decomposition, and red, green, and blue (RGB) images are generated using them as the pixel values of the R, G, and B channels. The generated images are diversified, and an optimized DenseNet neural network was used for training and prediction. Finally, the base load, wind power, and photovoltaic power generation data are selected, and the characteristic curves of the generalized load scenarios under different permeabilities of wind power and photovoltaic power generation are obtained using the density-based spatial clustering of applications with noise algorithm. Based on the proposed graphical forecasting method, the feasibility of the generalized load graphical forecasting method is verified by comparing it with the traditional time-series forecasting method.

Urban energy systems (UESs) play a pivotal role in the consumption of clean energy and the promotion of energy cascade utilization. In the context of the construction and operation strategy of UESs with multiple complementary energy resources, a comprehensive assessment of the energy efficiency is of paramount importance. First, a multi-dimensional evaluation system with four primary indexes of energy utilization, environmental protection, system operation, and economic efficiency and 21 secondary indexes is constructed to comprehensively portray the UES. Considering that the evaluation system may contain a large number of indexes and that there is overlapping information among them, an energy efficiency evaluation method based on data processing, dimensionality reduction, integration of combined weights, and gray correlation analysis is proposed. This method can effectively reduce the number of calculations and improve the accuracy of energy efficiency assessments. Third, a demonstration project for a UES in China is presented. The energy efficiency of each scenario is assessed using six operational scenarios. The results show that Scenario 5, in which parks operate independently and investors build shared energy-storage equipment, has the best results and is best suited for green and low-carbon development. The results of the comparative assessment methods show that the proposed method provides a good energy efficiency assessment. This study provides a reference for the optimal planning, construction, and operation of UESs with multiple energy sources.

To address the low accuracy and stability when applying classical control theory in distribution networks with distributed generation, a control method involving flexible multistate switches (FMSs) is proposed in this study. This approach is based on an improved double-loop recursive fuzzy neural network (DRFNN) sliding mode, which is intended to stably achieve multiterminal power interaction and adaptive arc suppression for single-phase ground faults. First, an improved DRFNN sliding mode control (SMC) method is proposed to overcome the chattering and transient overshoot inherent in the classical SMC and reduce the reliance on a precise mathematical model of the control system. To improve the robustness of the system, an adaptive parameter-adjustment strategy for the DRFNN is designed, where its dynamic mapping capabilities are leveraged to improve the transient compensation control. Additionally, a quasi-continuous second- order sliding mode controller with a calculus-driven sliding mode surface is developed to improve the current monitoring accuracy and enhance the system stability. The stability of the proposed method and the convergence of the network parameters are verified using the Lyapunov theorem. A simulation model of the three-port FMS with its control system is constructed in MATLAB/Simulink. The simulation result confirms the feasibility and effectiveness of the proposed control strategy based on a comparative analysis.

In this study, a novel residential virtual power plant (RVPP) scheduling method that leverages a gate recurrent unit (GRU)-integrated deep reinforcement learning (DRL) algorithm is proposed. In the proposed scheme, the GRU- integrated DRL algorithm guides the RVPP to participate effectively in both the day-ahead and real-time markets, lowering the electricity purchase costs and consumption risks for end-users. The Lagrangian relaxation technique is introduced to transform the constrained Markov decision process (CMDP) into an unconstrained optimization problem, which guarantees that the constraints are strictly satisfied without determining the penalty coefficients. Furthermore, to enhance the scalability of the constrained soft actor-critic (CSAC)-based RVPP scheduling approach, a fully distributed scheduling architecture was designed to enable plug-and-play in the residential distributed energy resources (RDER). Case studies performed on the constructed RVPP scenario validated the performance of the proposed methodology in enhancing the responsiveness of the RDER to power tariffs, balancing the supply and demand of the power grid, and ensuring customer comfort.