Price prediction of power transformer materials based on CEEMD and GRU

The rapid growth of the Chinese economy has fueled the expansion of power grids.Power transformers are key equipment in power grid projects,and their price changes have a significant impact on cost control.However,the prices of power transformer materials manifest as nonsmooth and nonlinear sequences.Hence,estimating the acquisition costs of power grid projects is difficult,hindering the normal operation of power engineering construction.To more accurately predict the price of power transformer materials,this study proposes a method based on complementary ensemble empirical mode decomposition(CEEMD)and gated recurrent unit(GRU)network.First,the CEEMD decomposed the price series into multiple intrinsic mode functions(IMFs).Multiple IMFs were clustered to obtain several aggregated sequences based on the sample entropy of each IMF.Then,an empirical wavelet transform(EWT)was applied to the aggregation sequence with a large sample entropy,and the multiple subsequences obtained from the decomposition were predicted by the GRU model.The GRU model was used to directly predict the aggregation sequences with a small sample entropy.In this study,we used authentic historical pricing data for power transformer materials to validate the proposed approach.The empirical findings demonstrated the efficacy of our method across both datasets,with mean absolute percentage errors(MAPEs)of less than 1%and 3%.This approach holds a significant reference value for future research in the field of power transformer material price prediction.

考虑特征重组与改进Transformer的风电功率短期日前预测方法

短期日前风电功率预测对电力系统调度计划制定有重要意义,该文为提高风电功率预测的准确性,提出了一种基于Transformer的预测模型Powerformer。模型通过因果注意力机制挖掘序列的时序依赖;通过去平稳化模块优化因果注意力以提高数据本身的可预测性;通过设计趋势增强和周期增强模块提高模型的预测能力;通过改进解码器的多头注意力层,使模型提取周期特征和趋势特征。该文首先对风电数据进行预处理,采用完全自适应噪声集合经验模态分解(complete ensemble empirical mode decomposition with adaptive noise,CEEMDAN)将风电数据序列分解为不同频率的本征模态函数并计算其样本熵,使得风电功率序列重组为周期序列和趋势序列,然后将序列输入到Powerformer模型,实现对风电功率短期日前准确预测。结果表明,虽然训练时间长于已有预测模型,但Poweformer模型预测精度得到提升;同时,消融实验结果验证了模型各模块的必要性和有效性,具有一定的应用价值。

基于生成对抗Transformer的电力负荷数据异常检测

电力负荷异常数据将给电力系统规划、负荷预测以及用能分析等带来较大的负面影响,因此亟须对负荷数据异常进行检测与识别。首先,针对电力负荷数据异常分类、原因及其特征开展分析。其次,改进传统Transformer编码器结构,采用多头注意力层代替掩码多头注意力层,同时移除前馈网络,以提高模型对负荷时序序列的全局注意力。基于生成对抗网络(generative adversarial networks,GAN)生成器与判别器的博弈结构,提出一种改进的GAN-Transformer模型,以更好地捕捉趋势性特征并加速模型收敛。然后,引入多阶段映射与训练方法,综合焦点分数打分机制,通过分阶段负荷序列重构帮助模型更好地提取负荷数据异常特征。最后,算例分析结果表明,GAN-Transformer模型在负荷数据异常检测精确率、召回率、F_(1)值以及训练时间方面均具有更优的性能,验证了所提方法的有效性和优越性。文中研究工作为基于深度学习进一步实现电力负荷数据异常分类与数据修复提供了有益参考。

基于非平稳Transformer的超短期风电功率多步预测

针对风电预测中波动性和随机性造成的风电功率多步预测精确度不高的问题,提出一种基于非平稳Transformer的超短期风电功率多步预测模型。利用皮尔逊相关系数法(PCC)和主成分分析法(PCA)对风电功率及其影响因素的分析确定输入数据,结合可以提升非平稳时序预测效果的非平稳Transformer模型,高效充分地挖掘输入数据与输出功率的复杂关系,构建风电功率超短期预测模型。实例分析表明,所提方法对不同预测步长下的风电功率进行预测时均具有较高的预测精度,且预测结果更稳定。

ViTH:面向医学图像检索的视觉Transformer哈希改进算法

对海量的医学图像进行有效检索会给医学诊断和治疗带来极其重要的意义.哈希方法是图像检索领域中的一种主流方法,但在医学图像领域的应用相对较少.针对此,提出一种面向医学图像检索的视觉Transformer哈希改进算法.首先使用视觉Transformer模型作为基础的特征提取模块,其次在Transformer编码器的前、后端分别加入幂均值变换(Power-Mean Transformation,PMT),进一步增强模型的非线性性能,接着在Transformer编码器内部的多头注意力(Multi-Head Attention,MHA)层引入空间金字塔池化(Spatial Pyramid Pooling,SPP)形成多头空间金字塔池化注意力(Multi-Head Spatial Pyramid Pooling Attention,MHSPA)模块,该模块不仅可以提取全局的上下文特征,而且可以提取多尺度的局部上下文特征,并将不同尺度的特征进行融合.最后在输出幂均值变换层之后将提取到的特征分别通过两个多层感知机(Multi-Layer Perceptrons,MLPs),上分支的MLP用来预测图像的类别,下分支的MLP用来学习图像的哈希码.在损失函数部分,充分考虑了成对损失、量化损失、平衡损失以及分类损失来优化整个模型.在医学图像数据集ChestX-ray14和ISIC 2018上的实验结果表明,该研究所提出的算法相比于经典的哈希算法具有更好的检索效果.

Research on the longitudinal protection of a through-type cophase traction direct power supply system based on the empirical wavelet transform

This paper proposes a longitudinal protection scheme utilizing empirical wavelet transform(EWT)for a through-type cophase traction direct power supply system,where both sides of a traction network line exhibit a distinctive boundary structure.This approach capitalizes on the boundary’s capacity to attenuate the high-frequency component of fault signals,resulting in a variation in the high-frequency transient energy ratio when faults occur inside or outside the line.During internal line faults,the high-frequency transient energy at the checkpoints located at both ends surpasses that of its neighboring lines.Conversely,for faults external to the line,the energy is lower compared to adjacent lines.EWT is employed to decompose the collected fault current signals,allowing access to the high-frequency transient energy.The longitudinal protection for the traction network line is established based on disparities between both ends of the traction network line and the high-frequency transient energy on either side of the boundary.Moreover,simulation verification through experimental results demonstrates the effectiveness of the proposed protection scheme across various initial fault angles,distances to faults,and fault transition resistances.

Power Transformer Fault Diagnosis Using Random Forest and Optimized Kernel Extreme Learning Machine

Power transformer is one of the most crucial devices in power grid.It is significant to determine incipient faults of power transformers fast and accurately.Input features play critical roles in fault diagnosis accuracy.In order to further improve the fault diagnosis performance of power trans-formers,a random forest feature selection method coupled with optimized kernel extreme learning machine is presented in this study.Firstly,the random forest feature selection approach is adopted to rank 42 related input features derived from gas concentration,gas ratio and energy-weighted dissolved gas analysis.Afterwards,a kernel extreme learning machine tuned by the Aquila optimization algorithm is implemented to adjust crucial parameters and select the optimal feature subsets.The diagnosis accuracy is used to assess the fault diagnosis capability of concerned feature subsets.Finally,the optimal feature subsets are applied to establish fault diagnosis model.According to the experimental results based on two public datasets and comparison with 5 conventional approaches,it can be seen that the average accuracy of the pro-posed method is up to 94.5%,which is superior to that of other conventional approaches.Fault diagnosis performances verify that the optimum feature subset obtained by the presented method can dramatically improve power transformers fault diagnosis accuracy.

Fault Diagnosis of Power Transformer Based on Improved ACGAN Under Imbalanced Data

The imbalance of dissolved gas analysis(DGA)data will lead to over-fitting,weak generalization and poor recognition performance for fault diagnosis models based on deep learning.To handle this problem,a novel transformer fault diagnosis method based on improved auxiliary classifier generative adversarial network(ACGAN)under imbalanced data is proposed in this paper,which meets both the requirements of balancing DGA data and supplying accurate diagnosis results.The generator combines one-dimensional convolutional neural networks(1D-CNN)and long short-term memories(LSTM),which can deeply extract the features from DGA samples and be greatly beneficial to ACGAN’s data balancing and fault diagnosis.The discriminator adopts multilayer perceptron networks(MLP),which prevents the discriminator from losing important features of DGA data when the network is too complex and the number of layers is too large.The experimental results suggest that the presented approach can effectively improve the adverse effects of DGA data imbalance on the deep learning models,enhance fault diagnosis performance and supply desirable diagnosis accuracy up to 99.46%.Furthermore,the comparison results indicate the fault diagnosis performance of the proposed approach is superior to that of other conventional methods.Therefore,the method presented in this study has excellent and reliable fault diagnosis performance for various unbalanced datasets.In addition,the proposed approach can also solve the problems of insufficient and imbalanced fault data in other practical application fields.

A Temperature Prediction Model for Oil-immersed Transformer Based on Thermal-circuit Theory

This paper proposed an improved temperature prediction model for oil-immersed transformer.The influences of the environmental temperature and heat-sinking capability changing with temperature were considered.When calculating the heat dissipation from the transformer tank to surroundings,the average oil temperature was selected as the node value in the thermal circuit.The new thermal models will be validated with the delivery experimental data of three transformers: a 220 kV-300 MV.A unit,a 110 kV40 MV.A unit and a 220 kV-75 MV.A unit.Meanwhile,the results from the proposed model were also compared with two methods recommended in the IEC loading guide.

基于多尺度时间序列块自编码Transformer神经网络模型的风电超短期功率预测

风电超短期功率预测过程中对时间依赖性的有效捕捉与建模,将直接影响风电功率时间序列预测模型的稳定性和泛化性。为此,提出一种新型时序Transformer风电功率预测模型。模型架构在逻辑上分为时间块自编码、隐空间Transformer自注意力时序自回归、随机方差缩减梯度(stochastic variance reduce gradient,SVRG)优化3个部分。首先,依稀疏约束及低秩近似规则,风电功率时空数据被半监督映射至隐空间;其次,隐空间编码经由多头自注意力网络完成时序自回归预测;最后,模型采用方差缩减SVRG优化算法降低噪声,达到更高预测效能。实验结果表明,所提新型Transformer架构能稳定有效进行超短期风电功率预测,预测结果在准确性、泛化性方面相较于传统机器学习模型都有明显提升。

基于强化学习和Transformer的输电线路缺陷智能检测方法研究

为了解决传统输电线路缺陷检测方法的不足,该文提出了一种基于强化学习和Transformer的输电线路缺陷智能识别方法。首先,采用具有较大感受野的空洞卷积网络(deterministic networking, DetNet)对输电线路巡检缺陷图像进行特征提取,继而使用深度Q网络(deepQ-network,DQN)筛选出包含前景信息的重要区域。其次,基于双线性注意力机制对背景区域特征向量进行投影压缩,使得融合特征向量聚焦于目标区域。最后,针对不确定缺陷检测结果定义可信度评测指标,构建Transformer网络编码层级的自适应调整机制,建立具有不同编码层级的Transformer模型库,以获取多模态缺陷的多层次差异化特征,采用Soft-NMS获取集成检测结果,提升识别模型的鲁棒性。通过对输电线路缺陷航拍图像进行了实验研究,该文方法检测精度平均值为89.7%,与其他算法相比具有更优的检测精度和泛化能力。

基于CATPCA的优化Transformer卫星电源消耗时序预测研究

提出一种由基于最优尺度量化的分类主成分分析数据处理模块和优化Transformer时序预测模块组成的卫星电源消耗预测方法.针对卫星工程数据的高冗余问题,建立了基于赫斯特指数分析(Hurst)、灰色关联分析以及分类主成分分析(CATPCA)的卫星高维数据处理模型,对百维度时序数据进行有效提取,重构输入数据.采用对抗学习网络架构,建立多学习Transformer的卫星电量预测模型,模型综合考虑影响卫星能源消耗的多种因素以及时序数据依赖,可以在较短的时间内完成高精度卫星电源消耗时序预测.实验部分采用卫星真实运行数据,综合考虑影响卫星能源消耗的多种因素,12 h预测拟合优度达到94%,比BP神经网络,长短期记忆网络(LSTM)精度更高.可以有效克服常规工程数据的冗余、缺失以及脏数据问题,解决了常规时序预测需要依赖长期数据的不足缺陷,有效完成卫星能源短时消耗高精度预测.这对卫星在轨任务规划、卫星在轨健康管理等后续任务提供可靠支持.

基于Transformer神经网络的变压器状态监测

电力变压器是电力系统的关键设备,为保障电力系统健康稳定运行,对电力变压器开展状态监测十分必要。提出基于Transformer神经网络的变压器状态监测方法,Transformer神经网络具有自注意力机制,能够挖掘不同特征维度之间的关联性,为变压器状态监测提供更可靠的决策能力。在进行变压器数据收集时,将采集到的数据集分为健康、亚健康、病态3个类别;之后采用原始数据、小波特征以及傅里叶特征融合的方式对数据进行预处理,增加特征维度;通过数据生成和Focal Loss的方法降低模型训练时数据不平衡带来影响,再将处理后的数据输入Transformer神经网络进行模型训练,最终利用训练好的模型预测变压器健康状态。与传统机器学习方法、卷积神经网络、长短时记忆网络相对比,所提方法预测精度有明显提升,能够准确的监测变压器设备状态,预测准确率能达到90%,是一种有效的变压器状态监测方法。

Extracting Power Transformer Vibration Features by a Time-Scale-Frequency Analysis Method

In order to take advantage of the merits of WPT and HHT in feature extraction from vibration signals of power transformer, a time-scale-frequency analysis method is developed based on the combination of these two techniques. This method consists of two steps. First, the desirable wavelet packet nodes corresponding to characteristic frequency bands of power transformer are selected through a Correlation Degree Threshold Screening (CDTS) technique for reconstructing a time-domain signal that contains useful information of power transformer. Second, the HHT is then conducted on the reconstructed signal to track the instantaneous frequencies corresponding to natural characteristics of power transformer. Experimental results are provided by analyzing a real power transformer vibration signal. Compared with the features extracted by directly using HHT, the features obtained by the proposed method reveal clearer condition pattern of the transformer, which shows the potential of this method in condition monitoring of power transformer.

基于自注意力Transformer编码器的多阶段电力系统暂态稳定评估方法

人工智能方法在电力系统暂态稳定评估研究中已经取得了一定的成果。常规深层网络普遍被视为“黑盒”模型,这限制了智能算法在实际工程应用中的可信赖性;同时,常规算法对电力系统时序信息的提取能力不足。针对以上问题,构建基于Transformer编码器的多阶段暂态稳定评估方法,其多阶段预测能够有效降低失稳漏判率。和常规算法相比,Transformer模型具有良好的可解释性,其注意力机制引导模型自适应识别并聚焦于关键特征,在一定程度上揭示深层网络内部工作决策过程。此外,采用多时刻信息构建特征空间,Transformer通过注意力机制实现全局感受野,使模型快速捕获电力系统前后时刻间的状态依赖。IEEE-39节点系统上的仿真结果表明,所提方法相比常见数据驱动模型具有更高的暂稳评估准确性,呈现出良好的可解释性,并在数据污染时依然维持较高的性能。

基于LightGBM-Transformer算法的短期电力负荷预测

高精度的短期电力负荷预测是优化电网运行策略和提高电网运行效率的可靠保障。为进一步提高短期电力负荷预测精度,提出轻量梯度提升机(light gradient boosting machine,LightGBM)-Transformer组合模型。该模型考虑时间特征和天气因素对短期电力负荷预测的影响,首先利用LightGBM算法获取特征重要性,排除无关噪声的影响,再将选择后的特征向量作为Transformer模型的输入,最后完成短期电力负荷预测。实验以澳大利亚能源市场运营商(Australian energy market operators,AEMO)检索的开放数据集为基础,并与多种类似模型进行对比。结果表明,LightGBM-Transformer组合模型的平均绝对百分比误差(mean absolute percent error,MAPE)为1.87%,误差指标显著低于其他对比模型,具有较高的预测精度,验证了该模型应用于短期电力负荷预测的可行性和有效性。

Detection of Mechanical Deformation in Old Aged Power Transformer Using Cross Correlation Co-Efficient Analysis Method

Detection of minor faults in power transformer active part is essential because minor faults may develop and lead to major faults and finally irretrievable damages occur. Sweep Frequency Response Analysis (SFRA) is an effective low-voltage, off-line diagnostic tool used for finding out any possible winding displacement or mechanical deterioration inside the Transformer, due to large electromechanical forces occurring from the fault currents or due to Transformer transportation and relocation. In this method, the frequency response of a transformer is taken both at manufacturing industry and concern site. Then both the response is compared to predict the fault taken place in active part. But in old aged transformers, the primary reference response is unavailable. So Cross Correlation Co-Efficient (CCF) measurement technique can be a vital process for fault detection in these transformers. In this paper, theoretical background of SFRA technique has been elaborated and through several case studies, the effectiveness of CCF parameter for fault detection has been represented.

Power Transformer No-Load Loss Prediction with FEM Modeling and Building Factor Optimization

Estimation of power transformer no-load loss is a critical issue in the design of distribution transformers. Any deviation in estimation of the core losses during the design stage can lead to a financial penalty for the transformer manufacturer. In this paper an effective and novel method is proposed to determine all components of the iron core losses applying a combination of the empirical and numerical techniques. In this method at the first stage all computable components of the core losses are calculated, using Finite Element Method (FEM) modeling and analysis of the transformer iron core. This method takes into account magnetic sheets anisotropy, joint losses and stacking holes. Next, a Quadratic Programming (QP) optimization technique is employed to estimate the incomputable components of the core losses. This method provides a chance for improvement of the core loss estimation over the time when more measured data become available. The optimization process handles the singular deviations caused by different manufacturing machineries and labor during the transformer manufacturing and overhaul process. Therefore, application of this method enables different companies to obtain different results for the same designs and materials employed, using their historical data. Effectiveness of this method is verified by inspection of 54 full size distribution transformer measurement data.

Research on power electronic transformer applied in AC/DC hybrid distribution networks

The AC/DC hybrid distribution network is one of the trends in distribution network development, which poses great challenges to the traditional distribution transformer. In this paper, a new topology suitable for AC/DC hybrid distribution network is put forward according to the demands of power grid, with advantages of accepting DG and DC loads, while clearing DC fault by blocking the clamping double sub-module(CDSM) of input stage. Then, this paper shows the typical structure of AC/DC distribution network that is hand in hand. Based on the new topology, this paper designs the control and modulation strategies of each stage, where the outer loop controller of input stage is emphasized for its twocontrol mode. At last, the rationality of new topology and the validity of control strategies are verified by the steady and dynamic state simulation. At the same time, the simulation results highlight the role of PET in energy regulation.

Large Power Transformer Fault Diagnosis and Prognostic Based on DBNC and D-S Evidence Theory

Power transformer is a core equipment of power system, which undertakes the important functions of power transmission and transformation, and its safe and stable operation has great significance to the normal operation of the whole power system. Due to the complex structure of the transformer, the use of single information for condition-based maintenance (CBM) has certain limitations, with the help of advanced sensor monitoring and information fusion technology, multi-source information is applied to the prognostic and health management (PHM) of power transformer, which is an important way to realize the CBM of power transformer. This paper presents a method which combine deep belief network classifier (DBNC) and D-S evidence theory, and it is applied to the PHM of the large power transformer. The experimental results show that the proposed method has a high correct rate of fault diagnosis for the power transformer with a large number of multi-source data.