Cold waves, characterized by abrupt temperature drops and severe icing, trigger drastic, non-stationary power fluctuations in wind turbines—violating the stationarity assumption of conventional short-term wind power forecasting models and causing large prediction errors. Compounding this, the extreme rarity of historical cold-wave samples leads to insufficient training data for data-driven models. To address these challenges, this study first identifies cold-wave events using rigorous thresholds from the China Meteorological Administration to ensure meteorological validity. Then, a hybrid framework tailored to cold-wave scenarios is proposed: an Improved Time-series Generative Adversarial Network (ITimeGAN) is developed to augment scarce cold-wave data, where integrating self-attention and Temporal Convolutional Networks (TCN) allows ITimeGAN to capture the fine-grained temporal dynamics of extreme weather sequences, while Wasserstein optimization stabilizes adversarial training to avoid gradient collapse, generating synthetic samples highly consistent with real cold-wave physical characteristics. Meanwhile, a De-Stationary Attention (DSA) mechanism is introduced to adapt to the strongly non-stationary power patterns during cold waves, and a Bidirectional Gated Recurrent Unit (BiGRU) predictor leverages bidirectional gating to exploit temporal dependencies in both forward and backward directions. Finally, SHapley Additive exPlanations (SHAP) analysis decodes distinct feature interaction patterns between normal and cold-wave scenarios. Experimental results on real-world wind farm data demonstrate the proposed method’s superior accuracy, with one-step prediction R2 reaching 0.9663, and enhanced interpretability that guides reliable wind farm management during extreme cold-wave events.

Offshore wind power prediction is significantly challenged by data quality issues arising from various factors such as environmental conditions and measurement errors, which severely compromise prediction accuracy and stability. This paper reveals a previously overlooked phenomenon in offshore wind power data, referred to as sample trend inconsistency, where dynamic offsets in meteorological features distort their expected physical relationship with power output. Such inconsistencies hinder the extraction of key features, disrupt feature-power coupling, and ultimately lead to degraded prediction performance. To address this issue, a Differential Compensation Dimensionality Reduction (DCDR) method is proposed to actively detect and mitigate trend inconsistencies during the dimensionality reduction process. Following normalized preprocessing of raw multifeature meteorological data, the proposed DCDR method is employed to enhance sample trend consistency and perform dimensionality reduction by optimizing a selection coefficient to retain the most informative feature subset, which is then fed into deep learning models for training and accurate power forecasting. Experimental results demonstrate that DCDR achieves significant improvements over conventional dimensionality reduction methods, reducing RMSE and MAE by 21.1 % and 12.1 %, respectively. Furthermore, global feature importance analysis based on Shapley Additive Explanations (SHAP) confirms that the features retained by DCDR contribute more strongly to prediction accuracy and show improved consistency with the underlying physical relationships governing power output, thereby providing a more robust and interpretable framework that can enhance the operational reliability of offshore wind power forecasting models.

Using the linear approach to design a controller is still prevalent. The state feedback control (SFC) is applied in this paper to improve the dynamic response of permanent magnet synchronous machine (PMSM) speed regulation systems. First, a third-order augmented system is constructed for the reason that a higher-order system has better disturbance rejection. It can be found through analysis and comparison that the order of the proposed speed controller is increased. The parameters of SFC are selected by utilizing the linear quadratic regulator (LQR), and the influence of matrix Q on dynamic performance is detailed through the Bode diagram. Additionally, considering parametric uncertainties and unmodeled dynamics, a disturbance observer (DOB) using the Luenberger observer is designed to further boost anti-disturbance performance. Finally, plenty of experimental results verify the effectiveness of the proposed methods.

In this paper, a corrosion risk assessment model of grounding device based on Improved Analytic Hierarchy Process (IAHP) is proposed, aiming to improve the safety of power tower operation and the operational reliability of power grid. The article first analyzes the factors affecting grounding device corrosion, including soil corrosion factor, stray current corrosion factor and anti-corrosion coating factor. Then, the improved hierarchical analysis method (IAHP) is introduced, which simplifies the construction of judgment matrix and consistency checking process by reducing the influence of subjective judgment of experts, and improves the calculation accuracy and applicability. Using the IAHP method, this paper constructs a hierarchical structure model of corrosion risk indicators of grounding devices and calculates the weights of each risk factor. Then, the corrosion risk of grounding device was evaluated by fuzzy comprehensive evaluation method, by determining the evaluation index set and fuzzy comment set, using the affiliation function to describe the fuzziness of the factors affecting the corrosion rate, and carrying out single-factor and multi-factor fuzzy comprehensive evaluation of each factor. Finally, the effectiveness of the proposed evaluation method is verified through case study. The results show that the method can effectively assess the corrosion risk of grounding devices and provide a new corrosion risk assessment tool for grounding devices in the power industry, which helps to realize the transformation from reactive repair to proactive prevention, and enhances the safety management and operational reliability of the power grid.

Integrated energy optimization scheduling (IEOS) is a complex problem aiming to minimize the total cost while the requirements of load balance is met. Due to the non-convex, non-differentiable and high-dimensional characteristics, there are many difficulties in solving the problem. Based on a regional integrated energy system (RIES), a decoupled distributed crisscross optimization with opposition-based learning and neighborhood re-dispatch strategy (DDCSO-OBL-NR) is proposed to solve IEOS problem by distributed method with different energy types as the scale. Initially, the CSO with excellent global search ability is firstly used to solve the complicated IEOS problem. Then, based on the distributed structure, distributed parallel computing can be achieved by DDCSO, which contributes to 1) protect the privacy of different energy data, 2) reduce the solving dimensions and 3) relieve the heavy communication burden. The total optimal cost is achieved by minimizing the cost of each portion without centralized controller. Furthermore, the opposition-based learning (OBL) strategy and the neighborhood re-dispatch (NR) strategy are combined into DDCSO aiming to optimize initial population location and enhance local search ability. Eventually, the DDCSO-OBL-NR is realized, and the effectiveness of which in solving the distributed IEOS problems is verified by the experimental results of three cases.

The classical Jiles-Atherton hysteresis model is widely used in the parameters identification for transformer core, which has the problem of large demagnetization curve error. To optimize this phenomenon, this paper proposes a new idea to revise the J-A hysteresis model based on non-hysteresis magnetization correction. Base on the Langevin equation and the energy conservation equation in ferromagnetic materials, the errors and its influencing factors of applying the classical J-A theoretical model to the identification of hysteresis loops are summarized. In the revision of the model, the hysteresis loss caused by the draw point at the moment of magnetization and demagnetization in loop curves are concluded. The equation of energy conservation that is more in line with the actual physical mechanism of magnetization are derived, which exists more efficient for the J-A parameters identification. After the model parameters are obtained by the particle swarm optimization algorithm, simulation studies are carried out for different samples, its comparative analysis with the actual measurement data is summarized. The results show that, compared with the classical J-A model, the simulation results of the revised model proposed in this paper are in high agreement with the measured magnetic property data of the material, and the hysteresis loop tends to be better fitted in the saturated section, which is more accurately adapted to the physical process of magnetization of the iron core material.

Factors such as high penetration of renewable energy, load, geographic location, and interactions between price zones make accurate electricity price forecasting (EPF) very challenging, especially day-ahead electricity price forecasting (DAEPF). To address the issue, A spatio-temporal graph neural network prediction model based on multi-view fusion is proposed in this paper, which learns and analyzes distance relationships, price correlations, and similarities in price distributions across multiple regions, four kinds of graph matrix are constructed to represent the complex spatio-temporal interaction in electricity market. To realize information aggregation between multiple perspectives, a novel multi-view fusion module (MVF) is proposed, which actively mines and utilizes the correlation between nodes within the graph and nodes across the graph through spatial attention and graph attention mechanism, and a temporal embedding module is proposed. The temporal information between nodes is represented by multi-head temporal attention mechanism and the time dependence of multiple receptive fields is obtained by multi-scale gated convolution. Massive experiments are conducted on multiple price zones in the European power market with a high proportion of new energy sources. The results show that MVF can effectively integrate multiple scenario information and improve the prediction accuracy of the network, and the proposed combined network has significant advantages over other models involved in this study.

As an important part of the grounding system, the detection of grounding leads is of great significance to the safety of operation and maintenance. In view of the problem that the faults of grounding leads cannot be efficiently recognized when they occur during actual operation, this paper proposes a method for recognizing defects in welded seams of grounding leads based on the ultrasonic guided wave detection method and the improved Subtraction-Average-Based Optimizer (ISABO) algorithm. Firstly, the ultrasonic guided wave detection of weld damage in soil-free/soil-containing grounding leads lap structure is analyzed, and numerical simulation and signal acquisition are performed; secondly, Subtraction-Average-Based Optimizer (SABO) is improved by circle mapping with inverse learning and Cauchy variation strategy, and the optimization of variational mode decomposition (VMD) parameter finding based on the improved SABO algorithm is proposed and compared with the support vector machine (SVM) combination of a grounding leads defect detection method; Finally, the application of the improved SABO algorithm is verified by the detection signal data set, the improved SABO algorithm in the parameter optimization is better than the traditional method, through the detection of multiple defects in the absence of soil-free/soil-containing detection comparison, the defect identification accuracy in the two conditions has reached 93.3% and 86.67% respectively, indicating that compared with the SABO algorithm, the method is more effective in detecting defect types and has good practical value.

This paper proposes a composite sliding mode control (SMC) to optimize the tracking performance and the anti-disturbance performance of permanent magnet synchronous machine (PMSM) speed regulation systems. The differential term in the control law can magnify the measurement noise, resulting in more discontinuity. To filter out the high frequency noise and make the control law smoother, the first-order differentiator (FOD) is employed to estimate the speed error and its derivative. Since the feedforward compensation can improve the robustness of the system, a disturbance observer (DOB) based on the sliding mode observer (SMO) is designed to reinforce the dynamic performance under disturbance variation. Under the effect of the feedforward compensation, chattering can be further weakened by decreasing the switching gain appropriately. Finally, the effectiveness of the proposed methods is confirmed by various experimental results.

Aiming at the accurate prediction of remanent magnetization for the transformer core, a model for the predicting of remanent magnetization based on approximate entropy criterion is presented in this paper, which can be used for the prediction of remanent in a wide range of magnetization. In its modeling, variational mode decomposition (VMD) is used to obtain the approximate entropy, and the intrinsic mode functions (IMF) group of current signals are analyzed. Results of the approximate entropy are used to set its prediction of remanent magnetization in the transformer core. The simulation results demonstrate the validity of the proposed model. Results from the simulation show that, the approximate entropy has a good linear relationship with the remanent magnetic size of the core, and the accuracy of the remanent magnetization is 85% when its error is ±0.05 per unit. Also, the model can realize the rapid evaluation of the remanent magnetization within 0.2s, which will be beneficial for developing of the control phase closing device taking the remanent magnetization into account.

为解决现有压电能量俘获电路效率低、开关辅助电路结构复杂且能耗高等问题,基于串联同步开关电感(Series Synchronized Switch Harvesting on Inductor,S-SSHI)电路输出功率峰值高和同步电荷提取(Synchronous Electric Charge Extraction,SECE)电路负载范围宽的优点,提出一种具有自供电功能的能量俘获电路.为实现S-SSHI电路和SECE电路混合,基于二者开关动作均在压电元件开路电压的极值点,提出了正极值点电压翻转、负极值点电荷提取的策略;设计了两个非对称无源正/负峰值检测电路检测正/负极值,使正/负半周期的工作模式分别为S-SSHI和SECE,从而达到输出功率和负载范围的平衡.通过仿真建模研究所提电路的可行性,验证了所提电路俘能性能优越,输出功率峰值可达SEH电路的4倍,同时具有较宽的负载范围.

Purpose This paper aims to present the design and fabrication of a rotary magnetostrictive energy generator, using to harvest the rotation energy of human knee joint. Design/methodology/approach A rotary magnetostrictive energy generator is presented in this paper. The harvester consists of six movable flat Terfenol-D rods, surround by the picked-up coils respective, and alternate permanent magnet (PM) array fixed in the upper cover of the stator. The harvester rotates like as a stepper motor, which has rotary electromagnetic power generating effect and impacted magnetostrictive power generating effect in its rotation. Modeling and simulation are used to validate the concept. A prototype of harvester is fabricated and subjected to the experimental characterization. Findings The size of proposed structure is control as 77 cm 3 , and its mass is about 0.21 kg. Huge induced voltage generated in the short-time impact situation, and that induced voltage in the harvester can up to 18.6 V at 0.32 s stepper rotation. Also, the presented harvester has good harvesting effects at low frequency human walking situation, which is suitable to be used for future researches of wearable knee joint applications. Originality/value A new concept of magnetostrictive harvester is presneted, which will be benefit for the application of human knee joint wearable. Also, this concept will give us more idea for collection of human movement energy.

变压器空载合闸时将产生大的励磁涌流,受剩磁、合闸角、变压器参数等因素影响,涌流特性复杂,辨识困难.这一现象在包含变压器匝间短路场景下,体现尤为明显.目前尚缺少考虑匝间短路因素下励磁涌流特性的数学表征和辨识判据.本文笔者提出了一种基于二次谐波分量衰减特性的励磁涌流表征方法,采用基于变压器二次谐波分量与基波分量幅值比的涌流判定依据,可实现包含匝间短路等多场景下励磁涌流的快速辨识.

To accurately simulate the dynamic characteristics of a transformer, it is imperative to establish a precise hysteresis model for its application. The J-A hysteresis model, characterized by its fewer parameters and clear physical significance, has found widespread application in research. However, addressing the issues of low accuracy and time-consuming optimization algorithms using for the identification of parameters in existing models, this paper introduces a novel hybrid J-A parameter recognition algorithm. The hybrid approach integrates Sparrow Search Algorithm with the Crisscross Optimization Algorithm, which helps to prioritize the convergence speed for traditional Sparrow Algorithm and the precision for area search provided by the Crisscross Optimization Algorithm. Simulation are concluded to verify the presented hybrid algorithm, its results demonstrate that the proposed hybrid J-A parameter recognition algorithm exhibits a faster computation and convergence speed accompanied by a higher recognition accuracy compared to individual algorithms.

Random intermittency and high fluctuation limit the wind power prediction accuracy. Although current studies offer various solutions, the prediction accuracy and data fitting performance are not satisfactory during violently fluctuating periods. To address the issue, a novel hybrid model is proposed in this paper, which combines adaptive variational mode decomposition (VMD), temporal convolution network with convolutional block attention module (CBTCN), and gated recurrent unit (GRU). First, VMD is used for data decomposition. Due to the high fluctuation of wind power data, it is difficult to optimize parameters for VMD. Thus grey wolf crossover optimization algorithm (GSCSO) is proposed, which combines four optimization algorithms. Then, an important index, i.e., dynamic error entropy (DEE) is proposed as the fitness function to ensure decomposition integrity, mode complexity, and predictability of sub-sequences for the first time. Thereafter, by extracting deep temporal features with CBTCN, GRU is cascaded to further mine the temporal correlation of these features and predict the wind power. Multiple experiments are conducted, and the results demonstrate that the proposed hybrid model can track the peaks and troughs satisfactorily, especially during violently fluctuating periods. For instance, in three-step prediction, the RMSE reduces by over 40% compared with other advanced models.

Multi-area economic dispatch (MAED) is a non-convex non-differentiable high-dimensional optimization problem, aiming to minimize the total fuel cost and meet the requirements of load balance. Current distributed mathematical programming methods are hard to solve the valve point effect in MAED. In addition, distributed meta-heuristic algorithms so far require a centralized controller. To address these issues, a decoupled distributed crisscross optimization with population cross generation (DDCSO-PCG) algorithm is proposed, which can 1) realize decentralization, 2) protect area data privacy, 3) reduce dimensions and improve convergence ability. First, the population cross generation (PCG) strategy is integrated into the crisscross optimization (CSO) algorithm to maintain population diversity and enhance exploitation ability. Second, the DDCSO-PCG algorithm is implemented to solve the MAED problem in a fully decentralized manner. Under the distributed framework, the proposed algorithm employs CSO-PCG independently to optimize the area dispatch in parallel. The total optimal cost is achieved by minimizing the cost of each area with no centralized controller required. The experimental results on multi-area static economic dispatch (MASED) and multi-area dynamic economic dispatch (MADED) problems show that the proposed DDCSO-PCG algorithm can not only provide a distributed solution, but also achieve the best economic cost compared with other state-of-the-art techniques.

微电网孤岛运行时,非线性负载和电力电子装置产生的谐波会造成损耗加大、变流器过载,危及系统安全稳定运行.针对具有LC滤波器的分布式电源,文章提出了将电压微分跟踪项和延迟补偿的有限集模型预测相结合的控制策略,用于降低输出电压谐波.首先阐述了逆变器预测模型的建模方法;然后设计了带电压微分跟踪项的有限集模型预测控制成本函数;最后通过建立仿真模型,验证了所提出的有限集模型预测控制策略具有较快的响应特性和鲁棒性,可以有效降低输出的谐波.

The accuracy and stability of wind power forecasting are very important for the operation of wind farms. However, for the newly built wind farms without sufficient historical data, it is difficult to make a more accurate prediction. Therefore, it is of great significance to explore a method to improve the wind power prediction accuracy with no sufficient historical data available. In this paper, a novel prediction model is proposed to address the few-shot learning problem of wind power prediction in new-built wind farms based on secondary evolutionary generative adversarial networks (SEGAN) and dual-dimension attention mechanism (DDAM) assisted bidirectional gate recurrent unit (BiGRU). The SEGAN first introduces the secondary evolutionary learning paradigm into learning GAN, aiming to learn the marginal distribution of real data and generate high-quality realistic data to augment the training dataset. In the prediction stage, the DDAM is attempted to obtain a new input matrix with global weight allocation and improve the sensitivity of the BiGRU model to the key information of the input data. The proposed SEGAN-DDAM-BiGRU model is validated on the data from the Galicia Wind Farm in Sotavento and the experimental results show that the proposed model is applicative for short-term prediction of new-built wind farms.

In this paper, we presented a new structure of a magnetostrictive energy harvester, which has a short-time impact piezomagnetic effect and can generate electricity during human walking. The harvester utilizes the hybrid magnetostrictive/electromagnetic effect, which helps to enhance its piezomagnetic effect in TbDyFe Alloy (magnetostrictive material). Also, an air gap is designed to improve its harvesting effect, which exhibits a larger numerical variation of magnetic field and impact stress, helping to generate more electricity. A prototype of a magnetostrictive harvester with a basic magnetic field and suitable air gap is configured, and its harvesting effect is calculated. The distribution nephogram of magnetization is analyzed, the magnetic density in the TbdyFe Alloy are calculated, and the optimized length of the air gap is determined. Then, we fabricated a prototype of the harvester, its harvesting effects are tested and calculated. The proposed structure shows a larger induced voltage for impact situations, which can generate up to 0.5-8V peak value for different impact vibrations. Also, the presented harvester can be safely used in the situation of low-frequency human walking (about 600N with 7.6MPa stress) and faster human running (about 1850N with 23.6MPa transient impact), which is suitable to be used for the application of gait monitoring systems.

在弱电网条件下,由于电网阻抗存在宽范围变化的特点,将会对并网逆变器的运行稳定性带来挑战.针对这个问题,建立三相LCL型并网逆变器电流双闭环控制的并网系统,采用基于阻抗模型的频域分析方法,通过分析并网逆变器等效输出阻抗的特点和电网阻抗对系统稳定性的影响,对并网逆变器系统输出等效阻抗进行重塑.在重塑方法上,提出一种具有宽频特性的阻抗重塑环节,通过该阻抗重塑环节可以在宽频范围内提高系统等效输出阻抗的相位,从而提高系统对弱电网的适应性.最后,通过Matlab/Simulink仿真,证明了理论分析的合理性以及所提出阻抗重塑方法的有效性.