Jilin Electric Power Research Institute (China)

This study presents a specific analytical solution procedure to the local-buckling problem in angle steels using a two-dimensional improved Fourier-series method (2D-IFSM). The effect of coupling between the sub-plates of an angle steel on its local-buckling behaviour is studied by incorporating rotational spring constraints between them. The proposed solution procedure enables one to convert the local-buckling problem of angle steels into solving sets of linear algebraic equations, thereby effectively simplifying its solution process. The critical load and related buckling-mode results obtained in this study are in good agreement with the existing analytical solutions and finite-element-method numerical data, verifying the effectiveness of the proposed method. Based on the derived solutions, a quantitative analysis is conducted to investigate the influences of aspect ratio, width–thickness ratio, and rotational constraint degree on the local-buckling behaviour of angle steels.

To address the operational optimization of distribution grids in commercial integrated-industrial parks, this study proposes a non-cooperative three-player game involving an energy operator, distributed photovoltaic (PV) prosumers, and electric-vehicle (EV) charging aggregators. A multienergy-flow scenario coupling electricity and heat is designed to define the dispatch framework and strategic logic of each entity, with all participants simultaneously determining their energy-scheduling strategies. A tractable mathematical program is employed to solve for the Nash equilibrium, where payoff functions quantify the operator’s regulation cost, the prosumer’s PV self-consumption surplus, and the aggregator’s load adjustment expenditure, while operational constraints calibrate parameters. Simulations show that at equilibrium: the operator reduces grid injection during mid-day PV peaks and increases district-heating output in office hours to ease distribution grid load; PV prosumers boost local self-consumption and supply residual energy to EV aggregators, reducing PV surplus back flow pressure; EV aggregators absorb mid-day PV surplus and shift evening charging loads, mitigating early evening distribution-network stress. The proposed model significantly improves PV utilization (system PV absorption rate reaching 97.94%), enhances multi-energy complementary, and strengthens distribution grid stability without compromising individual entities’ operational rationality.

Abstract The interaction between solar power and DC network increases the uncertainty of power prediction, and it is necessary to accurately analyze the variation patterns of solar power to achieve a more accurate prediction. Therefore, this article explores the power prediction algorithm for wind solar integration into DC power grids under trend analysis. We employ Savitzky-Golay filtering and linear interpolation to preprocess the data of wind solar power grid connection. The Newton-Raphson method is adopted to construct a DC grid power flow model, solving voltage and power distribution to support prediction; And establish a wind power prediction model based on BP neural network. The extracted features are used as inputs, and the power of wind solar power grid connection is used as output. Training the network to achieve accurate power prediction, experiments have shown that this method has high accuracy, small error, and fast operation (<5 seconds) in DC power grids, and has practical value.

Brazed type copper-aluminum transition clamps are widely applied in power transmission systems. The interface quality influences electrical conductivity and operational safety. Traditional ultrasonic detection methods have limitations in identifying defects at dissimilar metal interfaces, including insufficient resolution and difficulty in quantitative analysis. An ultrasonic phased array technology (PAUT) based detection method has been proposed in this paper. A set of artificial calibration blocks with typical artificial defects has been designed and machined. High-precision imaging and quantitative assessment of copper-aluminum brazing interface defects based on image reconstruction and signal analysis algorithms has been achieved. Experimental results confirm that this method could effectively identify defects of 2mm diameter or 0.3mm depth lack-of-fusion zone. This method shows good repeat ability and applicability.

Wind-solar power generation is inherently uncertain. These uncertainties bring considerable difficulties to the assessment of hosting capacity. To tackle these difficulties, it is essential to create typical scenarios that can precisely capture the statistical traits and interrelationships of wind-solar power. In this research, we systematically integrate various scenario generation techniques, resulting in the creation of a holistic framework grounded in kernel density estimation (KDE) and Copula functions. Our proposed approach represents the stochastic nature of wind-solar power output by constructing their respective probability density functions (PDFs). It comprehensively depicts the potential spatiotemporal complementarity between wind-solar power by utilizing Copula functions and establishing a joint probability distribution model. Through Monte Carlo simulation, we generated a large number of wind-solar output scenarios. Subsequently, we employed the K-means clustering algorithm to reduce the number of scenarios. The findings reveal that the integrated framework, which combines KDE and Copula theory, achieves higher fitting accuracy for the marginal distributions and correlation structures of wind-solar power generation. As a result, the generated scenarios are more representative and reliable, offering strong support for photovoltaic (PV) hosting capacity analysis (HCA) and the formulation of typical plans. We validate the proposed method using historical wind-solar data from several representative regions in China, such as Inner Mongolia, northern Hebei, the Beijing–Tianjin–Hebei region, and Hubei Province. This validation demonstrates the method’s applicability under various geographical and climatic conditions.

Abstract To optimize the complex nonlinear temporal relationships in wind power forecasting, we proposed an improved BAS-BiLSTM-Attention model for enhancing the accuracy and robustness of wind power predictions. The Beetle Antennae Search algorithm was utilized for obtaining hidden features across different frequency bands. The Bidirectional Long Short-Term Memory algorithm captured the bidirectional temporal associations of hidden features. Finally, an Attention mechanism was added to ensure that the weight selection of the feature components remains unaffected. Results demonstrated that our method performed excellently in wind power forecasting tasks. Compared to individual models and other common algorithms, it exhibited higher prediction accuracy and robustness across multiple evaluation metrics.

Infrared imaging plays a crucial role in substation equipment inspection by enabling non-contact identification of thermal anomalies that reveal potential defects. However, achieving high detection accuracy often requires large convolutional neural networks, which imposes significant computational and memory burdens and limits deployment on edge devices commonly used in power systems. To address this challenge, this paper proposes a lightweight and efficient infrared defect detection framework that incorporates wavelet-regularized soft channel pruning into a deep neural network. Specifically, a Wavelet-Based Channel Pruning (WCP) strategy is introduced to evaluate channel importance by analyzing the wavelet-domain coefficients of convolutional filters, allowing the model to preserve channels rich in defect-related edge information. Furthermore, to mitigate the risk of prematurely removing informative channels, a Soft Channel Reconstruction (SCR) mechanism is developed to dynamically restore selected pruned channels through cosineinterpolated parameter fusion during training. Extensive experiments demonstrate that the proposed method achieves superior recognition accuracy while reducing the number of parameters and FLOPs by more than 50%.