As the core equipment of electric energy collection, boost and grid connection, the transformer of new energy station may cause serious consequences such as station shutdown. Aiming at the time domain and spectrum feature significance of the fault voiceprint signal, this paper proposes a diagnosis method based on feature fusion and improved wave search algorithm. The MFCC and GFCC features are extracted by the Mel filter and the gamma-pass filter respectively. After linear combination, the improved Fisher ratio method is used to eliminate the low contribution components, and the regularization is introduced to prevent over-fitting, forming an optimized fusion feature MGCC. The MGCC features are input into the IWSA optimized support vector machine (SVM) model to achieve accurate classification of transformer fault voiceprints. Experiments show that the MGCC feature fusion extraction method and the IWSA-SVM model proposed in this paper have good recognition effects. The accuracy of fault recognition of new energy stations under small samples can reach 99.52 %, which is superior to other methods. It is an effective new energy station transformer voiceprint fault diagnosis method.

Addressing the shortcomings of the Sparrow Search Algorithm (SSA), such as low accuracy of convergence and tendency of falling into local optimum, a Multi-strategy Integrated Sparrow Search Algorithm (MISSA) is proposed. In this method, by improving the black-winged kite algorithm and applying it to the producer’s position update formula, an improved search strategy (ISS) is firstly proposed to enhance search ability. Secondly, a new strategy inspired by the Coot algorithm, called the group follow strategy (GFS), is proposed to improve the ability to jump out of the local optimum. Finally, a proposed random opposition-based learning strategy (ROBLS) is applied to the population after each iteration to enhance its diversity. To verify MISSA’s effectiveness, extensive testing is conducted on 24 benchmark functions as well as CEC 2017 functions. The experimental results, complemented by Wilcoxon rank-sum tests, conclusively demonstrate that MISSA outperforms SSA and other advanced optimization algorithms, exhibiting superior overall performance.

With the improvement in power grid simulation accuracy requirements, the existing typical load model parameters can no longer meet the accuracy requirements and become the short board that restricts the stable operation of power system. This paper mainly proposes an improved butterfly optimization algorithm based on the population optimization and dynamic strategy (PODSBOA) for commonly used synthesis load model (SLM) parameters to realize the refined and personalized identification of SLM key parameters:[pu, qu, Rs, Xs, Rr, Xr, Km, and Mif]. The results indicate that in the 2-s load data experiment, the identification error is 0.02, the identification accuracy is 4.09, and the convergence time of the PODSBOA is 12.048 s. In the 5-s load data experiment, the identification error is 0.013, the identification accuracy is 6.65, and the convergence time of the PODSBOA is 23.405 s. The identification errors in the two sets of experiments are reduced by 0.02023–0.06443 compared with other algorithms. The comparison results of different load model parameter identification algorithms indicate that the improved PODSBOA proposed in this paper has high recognition accuracy and fast convergence speed and solves the problem of low accuracy and instability of the identification results of the existing identification schemes.

In this paper, a novel T-clamp is designed to address the heat generation problem of clamp over-heating in high-voltage transmission lines. By comparing with the widely-used parallel groove clamp, a three-dimensional finite element analysis model is created in COMSOL to simulate and calculate the temperature distribution and heating generation power of the T -clamp from different current power flows, contact resistance and ambient temperature and relative humidity. The simulation results indicate that the heat generation power of the T -clamp is about one-sixth times in comparison to the parallel groove clamp under different current distribution conditions. Also, across all scenarios of current input, when the main line on the left side and the branch line each input 150A and 50A and the output is on the opposite side of the main line, in which the T -clamp is optimal. When the bolt torque increases, the contact resistance and heat generation power decrease. The heating power of T -clamp is about one-fifth times higher compared to that of parallel groove clamp when using same torque. The environmental humidity has little influence on the heat generation of the clamp. Furthermore, the temperature rise of the T - clamp is about 5°C, which is approximately 0.5°C lower than that of the parallel groove clamp at different environmental temperatures. The new designed T -clamp is experimentally verified to be simple installation and adaptable, capable of maintaining relatively lower heat generation and temperature rise even in the long-term use.

Accurate differentiation of energy consumption information of residential users is of great significance for load planning, scheduling, operation and management of power system, and is the basic premise for realizing intelligent perception of energy system and energy saving and carbon reduction. Considering that the conventional single-layer clustering method has limited clustering stability and clustering effect, this paper takes the key family feature factors as the modified feature quantity of quadratic clustering, and proposes a study of user energy characteristics based on double-layer clustering and modification. Firstly, the user’s energy consumption data is collected and pre-processed, and the user’s energy consumption curve is clustered and analyzed by using the integrated clustering algorithm based on voting and the advantages of each member algorithm. Then, the key family characteristic factors are obtained, and the results of one-layer clustering and key family characteristic factors are combined to carry out two-layer clustering of the same category of users in the form of questionnaire survey. Finally, the nonlinear mapping capability of Support Vector Machine (SVM) is used to reverse correct the results of the one-layer clustering. The actual algorithm data of the residents’ demand response experiment in a southeastern province are compared. The results show that compared with the single-layer clustering algorithm, the proposed method can accurately distinguish the energy consumption characteristics and adjustable potential of different users, and correct the wrong clustering results in the single-layer clustering. The clustering stability and clustering effect have been effectively improved.The example results show that the clustering results modified by SVM can better mine and distinguish user energy characteristics, and can be used to solve the problem of the current demand response clustering algorithm not being able to comprehensively and objectively describe the participation willingness and response-ability of residential users in the implementation process. It can also provide a basis for peak shaving and power grid frequency regulation.

The current wind power prediction scheme still has a large error in the transitional weather period. In the case of large-scale wind power integration, it will affect the safe operation of the entire power grid. In order to solve the above problems, an adaptive prediction model based on transitional weather classification is proposed. Firstly, the quartile method is used to clean and interpolate the abnormal data of the wind farm, and then the parameters of the extreme gradient lifting tree (XGB) are optimized by the improved snake swarm algorithm (CBAMSO). The scene classification model is established to divide the transitional weather, and the sensitive meteorological factors of typical transitional weather are selected to construct the input feature sequence. The convolutional neural network (CNN) fusing spatial pyramid pooling (SPP) is used to extract variable dimension features. Finally, the final wind power prediction value is obtained by using the attention mechanism (ATT) to redistribute the weight to the output of long and short memory network (LSTM). The results show that CBAMSO-XGB accurately divides all kinds of transitional weather, and the average absolute error and root mean square error of adaptive prediction model are about 42.49% to 72.91% and 65.34% to 91.2% compared with CNN-LSTM model.

In high-voltage switchgear, under the action of high electric field intensity, the insulation medium will not penetrate between the electrodes, and partial discharge may damage the insulation performance of the equipment, thus affecting the normal operation and life of the equipment. However, based on the problems of traditional manual inspection, such as low efficiency, single detection principle and insufficient safety, a four-in-one automatic on-line monitoring system for switchgear partial discharge is proposed. The system selects four detection methods: ultrasonic sensor, geoelectric wave sensor, UHF antenna sensor and temperature sensor. The inspection robot can accurately locate the detection points, SqueezeNet convolutional neural network can judge the type of partial discharge, and the correlation coefficient can be calculated by the discharge frequency. The test results show that the system can detect the status of switchgear more comprehensively and timely. Compared with traditional manual inspection, the system can realize automatic inspection, greatly improving the safety and efficiency of inspection, and the inspection benefit is about four times that of manual inspection.

Power theft has a large impact on both power supply enterprises and power users, and in view of the huge amount of data required for some existing power theft detection methods based on machine learning data analysis and the problem of low accuracy, this paper proposes a power theft identification method that integrates clustering and improved sparrow search algorithm. First, the FCM clustering algorithm is used to classify the typical daily load curves of the users and form a "portrait" of the user's electricity consumption behavior; second, by calculating the matching degree of the load curves to be tested and the user's electricity consumption behavior "portrait", the "suspected" electricity theft detection method is locked in place and the "suspect" electricity theft detection method is applied. Secondly, by calculating the matching degree between the load profile to be tested and the "portrait" of the customer's electricity consumption behavior, the "suspected" customer is locked; finally, the "suspected" customer is further detected by using the Improved Sparrow Search Algorithm (ISSA). Experimentally, the proposed method combined with FCM clustering algorithm can narrow the detection range of power theft users to a greater extent, and the improved sparrow search algorithm can accurately locate power theft users, which greatly improves the efficiency and accuracy of power theft detection.

Clean energy transmission is predominant in highaltitude areas, and in order to achieve the development and transmission of clean energy, a large number of composite insulators are needed. However, the harsh environment in highaltitude areas can accelerate the aging of composite insulators. This article uses an artificial aging test platform to conduct coupled aging tests. Through surface flashover, mechanical testing, thermogravimetric analysis, and Fourier transform infrared spectroscopy analysis of silicone rubber samples with different test cycles, the performance degradation law and failure mechanism of silicone rubber under coupled environment are studied. The results showed that the surface flashover voltage and mechanical properties of silicone rubber decreased, mass fraction first increased and then decreased, and the characteristic peak areas of the main chain and side chains decreased; The interface between the matrix and the filler is damaged, and the trihydrate aluminum filler and white carbon black filler precipitate from the silicone rubber matrix, reducing the organic components and affecting the high temperature resistance, cold resistance, and UV resistance of the silicone rubber, thereby reducing its electrical and mechanical properties. To provide useful reference for the study of aging failure mechanism of inservice composite insulators considering the influence of temperature and irradiation environment.

Residential electricity data in the collection, transmission and other aspects of the irregular missing will lead to errors in the subsequent application of the analysis, in order to ensure the integrity of the residential electricity data, for the current data missing because of smart meter sensors are subject to interference and other issues, research a similar day selection based on similar day Bayesian Gaussian tensor decomposition model of residential electricity data interpolation method. The method selects the meter data adjacent to the missing data meters in space, and in time, comprehensively considers the time series data and the similar day series data, so as to construct the spatio-temporal tensor model, fully exploits the intrinsic relationship existing between each data, and utilizes the Bayesian Gaussian CP decomposition (BGCP) model to The missing data are recovered to achieve the purpose of data restoration and improve the integrity and accuracy of the grid data. The experimental results show that: the proposed method can effectively interpolation the missing data of residential electricity consumption, and the error rate after data interpolation is below 0.4%, which can meet the needs of data application analysis.

针对江西电网用电负荷增长迅速,电网负荷呈现明显的高尖峰、短持续特征,在充分考虑供需双方收益的前提下,建立了考虑碳排放和用户满意度的需求响应激励策略优化模型,并采用强化学习的Q学习算法对模型进行迭代求解.将用户与电网进行交互的强化学习框架转换为马尔可夫决策过程(Markov decision process,MDP),并通过积累的真实需求响应历史数据辨识模型参数求解最大综合收益值,分析不同权重因子对用户满意度、电网收益、居民用户收益的影响.算例结果表明,所提出的需求响应策略优化模型能够有效平衡电网和用户双方的收益,缓解电网用电高峰时段供需不平衡问题.

Magnetic flux leakage testing is a new method in the field of testing the crimping quality of the strain clamp, but it has some disadvantages, such as weak signals and unclear characteristics. In this paper, the magnetic flux leakage signal and characteristic quantity are increased by adding magnetic flux gather structure, and three magnetic flux gather structures with different shapes are designed. A comparative study of different sizes is made by means of simulation. In this paper, a corresponding relationship between the magnetic flux leakage signal and the steel anchor structure of the strain clamp is proposed, which has certain guiding significance for the magnetic flux leakage signal to calculate the stress zone of the strain clamp.

基于帝王蝶优化算法,提出了一种新的帝王蝶-BP(Back Propagation)神经网络预测模型,以预测结果的平均绝对误差为目标函数,对BP神经网络模型的初始权重和阈值进行寻优,实现了对江西省能源供需的准确预测,并依据预测结果制定江西省低碳转型路径.通过与已有文献方法和权威公开数据的对比,验证了帝王蝶-BP神经网络优化预测模型的有效性和优越性.

With the increasing demand for reliable power supply and the widespread integration of distributed energy sources, the topology of distribution networks is subject to frequent changes. Consequently, the dynamic alterations in the connection relationships between distribution transformers and feeders occur frequently, and these changes are not accurately monitored by grid companies in real-time. In this paper, we present a data-driven machine learning approach for identifying the feeder-transformer relationship in distribution networks. Initially, we preprocess the collected three-phase voltage magnitude data of distribution transformers, addressing data quality and enhancing usability through three-phase voltage normalization. Subsequently, we derive the correlation coefficient calculations between distribution transformers, as well as between distribution transformers and feeders. To tackle the challenging task of determining the correlation coefficient threshold, we propose a multi-feature fusion approach. We extracted additional features from the feeders and combined them with the correlation coefficients to create a feature matrix. Machine learning algorithms were then applied to calculate the results. Through experimentation on a real distribution network in Jiangxi province, we demonstrated the effectiveness of the proposed method. When compared to other approaches, our method achieved outstanding results with an F1 score of 0.977, indicating high precision and recall. The precision value was 0.973 and the recall value was 0.981. Importantly, our method eliminates the need for additional measurement installations, as the required data can be obtained using existing collection devices. This significantly reduces the application cost associated with implementing our approach.

传统的空预器积灰监测系统采用的热电偶进行监测,对换热元件监测范围具有局限性.为了能全区域监测空预器换热元件的局部积灰状态,提出了一种基于红外机器视觉技术的空预器积灰监测方案.该方案利用红外热像仪实时监测空预器换热元件运行状态,并采集其高清红外热像视频上传至数据库,以进一步标识空预器积灰区域;此外,该方案对摄像头安装角度进行优化,以实现全区域监测.试验结果表明,基于该方案的监测系统能够全区域、高精度判别空预器换热元件的积灰区域,为清除积灰提供了导向作用,保证了空预器的安全稳定运行.

电晕放电严重威胁输电线路的安全运行,如何提高其放电区域识别分割准确率是一个亟待解决的问题.而因环境影响及设备性能限制,夜间型紫外成像仪常出现成像不清晰、放电区域对比度不明显等特征,导致难以有效实现电晕放电区域的分割,从而影响放电故障的判定.为此提出了基于Deeplabv3+与Otsu模型的输电线电晕放电紫外图像精确分割方法,首先构建基于Deeplabv3+语义分割模型,对放电区域进行类别分割得到大致区域;然后,利用改进Otsu算法对语义分割结果中放电目标区域方差自适应加权,使得分割阈值近似理想阈值,从而实现电晕放电区域的精确分割.实验结果表明,本文提出的分割方法在测试集中平均像素精度为93.97％,平均交并比为90.85％,分割性能良好.

Power plant enterprises lack intuitive monitoring means for primary frequency modulation (PFM) process, and in complex power environment, there may be information delay or inconsistency between the monitoring and assessment of dispatching center and the actual frequency modulation (FM) of power plants, which will seriously affect the accuracy of unit FM and the economic benefits of power plants. In view of the above problems, a monitoring and assessment system for PFM of thermal power units is designed. The system is built in the power plant to dynamically monitor the running status of grid connected units and conduct real-time assessment calculation of PFM, ensuring the stability of the unit operation and fairness of dispatching assessment.

Source storage flexible resources such as central air conditioning, electric vehicles, distributed wind power, photovoltaic, and energy storage have considerable regulation potential and are effective alternative resources for grid-side regulation capacity. However, there is no complete theoretical approach for the construction of multi-timescale optimal model and optimization strategy for multi-subject flexible loads. Considering these problems, this paper proposes a multi-timescale model and optimal scheduling strategy for flexible load aggregators based on the quadratic attack of improved seagull algorithm. Firstly, a flexible load hierarchical optimal scheduling architecture is proposed to aggregate flexible load resources; secondly, a generalized aggregation model is established for electric vehicles, temperature-controlled loads and distributed energy storage loads, and in the day-ahead phase, the start-stop schedule of conventional units, optimal charging of electric vehicles and flexible load aggregator operation constraints are considered. In the intra-day phase, the system rotation backup cost and temperature-controlled load constraints are considered to construct an intra-day optimal dispatching model. Then an economically optimal dispatching strategy is proposed to construct a model based on flexible loads such as electric vehicles, temperature-controlled loads, and distributed storage loads, and an improved seagull optimization algorithm based on population optimization and dynamic strategy is proposed to solve the problem. Finally, the correctness and effectiveness of the proposed model algorithm is verified on the improved IEEE33 node system.

With the expansion of the scale of wind power integration, the safe operation of the grid is challenged. At present, the research mainly focuses on the prediction of a single wind farm, lacking coordinated control of the cluster, and there is a large prediction error in transitional weather. In view of the above problems, this study proposes an adaptive wind farm cluster prediction model based on transitional weather classification, aiming to improve the prediction accuracy of the cluster under transitional weather conditions. First, the reference wind farm is selected, and then the improved snake algorithm is used to optimize the extreme gradient boosting tree (CBAMSO-XGB) to divide the transitional weather, and the sensitive meteorological factors under typical transitional weather conditions are optimized. A convolutional neural network (CNN) with a multi-layer spatial pyramid pooling (SPP) structure is utilized to extract variable dimensional features. Finally, the attention (ATT) mechanism is used to redistribute the weight of the long and short term memory (LSTM) network output to obtain the predicted value, and the cluster wind power prediction value is obtained by upscaling it. The results show that the classification accuracy of the CBAMSO-XGB algorithm in the transitional weather of the two test periods is 99.5833% and 95.4167%, respectively, which is higher than the snake optimization (SO) before the improvement and the other two algorithms; compared to the CNN–LSTM model, the mean absolute error (MAE) of the adaptive prediction model is decreased by approximately 42.49%–72.91% under various transitional weather conditions. The relative root mean square error (RMSE) of the cluster is lower than that of each reference wind farm and the prediction method without upscaling. The results show that the method proposed in this paper effectively improves the prediction accuracy of wind farm clusters during transitional weather.