Diagnosis of Disc Space Variation Fault Degree of Transformer Winding Based on K-Nearest Neighbor Algorithm

Winding is one of themost important components in power transformers.Ensuring the health state of the winding is of great importance to the stable operation of the power system.To efficiently and accurately diagnose the disc space variation(DSV)fault degree of transformer winding,this paper presents a diagnostic method of winding fault based on the K-Nearest Neighbor(KNN)algorithmand the frequency response analysis(FRA)method.First,a laboratory winding model is used,and DSV faults with four different degrees are achieved by changing disc space of the discs in the winding.Then,a series of FRA tests are conducted to obtain the FRA results and set up the FRA dataset.Second,ten different numerical indices are utilized to obtain features of FRA curves of faulted winding.Third,the 10-fold cross-validation method is employed to determine the optimal k-value of KNN.In addition,to improve the accuracy of the KNN model,a comparative analysis is made between the accuracy of the KNN algorithm and k-value under four distance functions.After getting the most appropriate distance metric and kvalue,the fault classificationmodel based on theKNN and FRA is constructed and it is used to classify the degrees of DSV faults.The identification accuracy rate of the proposed model is up to 98.30%.Finally,the performance of the model is presented by comparing with the support vector machine(SVM),SVM optimized by the particle swarmoptimization(PSO-SVM)method,and randomforest(RF).The results show that the diagnosis accuracy of the proposed model is the highest and the model can be used to accurately diagnose the DSV fault degrees of the winding.

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MTF-Swin Transformer的风机齿轮箱故障诊断

针对风机齿轮箱实际工况复杂多变及含有强噪声,传统故障诊断方法对风机齿轮箱故障诊断识别准确率较低的问题,文章提出了MTF-Swin Transformer风机齿轮箱故障诊断模型。首先,采用马尔科夫变迁场(MTF)图形编码方法将原始一维振动时序信号转化为具有关联时间信息的二维特征图谱;然后,将特征图谱作为Swin Transformer模型的输入,基于自注意力机制进行自动特征提取;最后,实现对不同故障类型的分类。仿真结果表明,该方法对齿轮箱故障诊断准确率达到了99.48%,证明了该方法的有效性和优越性。

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。

基于CWT和优化Swin Transformer的风电齿轮箱故障诊断方法

针对传统故障诊断方法在风电齿轮箱运行故障诊断应用上的不足,提出一种基于小波变换(continuous wavelet transform, CWT)和优化Swin Transformer的风电齿轮箱故障诊断方法。该方法利用小波变换将风电齿轮箱振动信号转换为时频图;使用SuperMix数据增强算法对样本进行扩充;利用迁移学习技术将模型预训练参数用于训练和优化Swin Transformer模型;将训练完成的优化Swin Transformer模型应用于风场实际运维数据进行对比验证,分类准确率达到99.67%。验证结果表明该方法能够有效地实现风电齿轮箱故障诊断,并提高模型的识别准确率。

基于对抗训练与Transformer的风力发电机故障分类方法

风力发电机故障分类的复杂性和多样性严重影响风能发电效率,传统的人工方法效率低下,准确率较低,已有的深度学习模型在真实环境中易受数据噪声干扰而表现不佳。为提升风力发电机故障分类模型在真实环境下的分类性能与鲁棒性,提出一种基于对抗训练与Transformer的故障分类方法。首先通过引入一维卷积与门控线性单元(GLU)增强注意力机制对局部特征的学习,保留易被忽略的局部信息,提升模型对于局部特征的敏感度。其次结合限制因子约束对抗样本,提高对抗样本产生的准确性。最后在消除错误样本的同时反馈生成过程,使其具备更好的抗干扰能力。实验结果表明,与5种常用的分类模型相比,所提模型分类性能平均提升7.76%,与真实结果之间的误差最小。局部增强的注意力机制和所提的对抗训练方法分别使模型的分类性能平均提升4.51%、4.95%。所提模型在10%~20%噪声环境中仍保持较好性能,增强了其在真实环境中的稳定性。该方法在提高分类准确率的同时使模型具备更强的泛化能力,对于提升风力发电机故障分类性能与鲁棒性具有重要意义。

基于轻量化Transformer模型的多变量风电功率预测

风电功率预测对电力调度和新能源管理极为重要。为准确高效预测多变量风电功率,提出一种基于轻量化Transformer(Light Transformer)的风电功率预测方法。首先,采用滚动序列建模方法,确定输入数据,然后采用Transformer预测模型,改进和精简原始结构,在前馈网络模块中使用GeLU激活函数来代替传统的ReLU激活函数,提升模型的质量,轻量化网络结构,利用多头注意力机制,加快模型训练速度,提升预测模型精度。

Dynamic stability of inner windings of large capacity transformers under short circuit conditions

Presents the study on the dynamic buckling of the inner windings of large capacity transformers under short circuit conditions by a finite element method and the findings as follows: 1) No radial dynamic buckling of inner windings occurs under short circuit conditions when the windings are well supported at all the radial supports; 2) The windings buckle under short circuit conditions when the windings are not supported at one radial support (for instance, 25°) but well supported at all other radial supports; 3) When the windings are not well supported at one radial support, the ability of the windings to resist radial dynamic buckling can be greatly enhanced provided some measures are taken to provide necessary radial supporting.

Bacterial Foraging Algorithm based Parameter Estimation of Three WINDING Transformer

Transformers are one of the main components of any power system. An accurate estimation of system be-haviour, including load flow studies, protection, and safe control of the system calls for an accurate equiva-lent circuit parameters of all system components such as generators, transformers, etc. This paper presents a methodology to estimate the equivalent circuit parameters of the Three Winding Transformer (TWT) using Bacterial Foraging Algorithm (BFA). The estimation procedure based on load test data at one particular op-erating point namely supply voltage, load currents, input power. The performance characteristics, such as efficiency and voltage regulation are considered along with the name plate data in order to minimize the er-ror between the estimated and measured data. The estimation procedure is demonstrated with a sample three winding transformer and the results are compared against the directly measured performance of TWT and genetic algorithm optimization results. The simulation results show the ability of the proposed technique to capture the true values of the machine parameters and the superiority of the results obtained using the bacte-rial foraging algorithm.

Influence of Capacitive Effects on Transformer Windings

For years, capacitive effects have been the subject of research [1] and [2]. The capacitive effects are discrete capacitors that appear between active conductors of power lines and between them with the ground plane, generating capacitive reactive power to the network [1] and [2]. Indeed, it must be noted that these effects affect the windings of the transformer when the coupling is in star or triangle. This study is conducted to show that capacitive effects affect transformer windings differently when coupling is in stars or triangles. The results obtained are interesting and can be exploited in electrical transmission networks to ensure a long lifespan of transformers.

The Effect of Windings on ADSL Transformer Insertion Losses

Insertion loss （IL） is one of the important parameters of asymmetrical digital subscriber loop （ADSL） transformers. In different frequency bands, the factors that affect insertion loss are different. Windings mainly affect insertion loss in mid and high frequency bands. The effects of winding ways, winding wire diameter and winding turns on insertion loss were discussed. The presented experiment shows that the insertion loss of an ADSL transformer could be under 0.4 dB in mid frequency band when the winding is 30 turns, in which the ADSL transformer satisfies the requirement of total harmonic distortion （THD）. Our experiments also show that the sandwich winding structure is better than the side by side winding structure and the twisted-pair winding structure, and the increase of winding diameter is one means to reduce insertion losses of an ADSL transformer in mid frequency band.

Analytical Solution for Evaluation of Voltage Surge Distribution along Transformer Windings through the Method of Residues

Computer programs have definitely become indispensable for designing power transformer. Among several applications, computer programs are mostly used for electric field calculation and thus electrical insulation concerns. In consequence, studies based on analytical approach to basic studies of correlated problems have become even more important because they form the very basis of knowledge that is necessary to every transformer designer in view of taking all the advantages of computational analyses. On the other hand, one of the most important basic studies consists in the evaluation of voltage surge distribution along transformer windings for which the method of separation of variables has been extensively used thanks to some simplifying assumptions. With this aim, authors have developed and previously published works that show the applicability of an alternative and useful analytical method that is the method of the residues, which requires no simplification to be assumed. In this work, another important step is taken towards proofing the total applicability of this promising method that is through a practical problem. A comparison to the numerical method TLM （transmission line method） is also performed and concordance with TLM and experimental data confirms the proposal of the method of residues can be also applicable to several others problems of electromagnetism.

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。

Research on Numerical Simulation of 3D Leakage Magnetic Field and Short-circuit Impedance of Axial Dual-low-voltage Split-winding Transformer

It is difficult to accurately calculate the short-circuit impedance, due to the complexity of axial dual-low-voltage split-winding transformer winding structure. In this paper, firstly, the leakage magnetic field and short-circuit impedance model of axial dual-low-voltage split-winding transformer is established, and then the 2D and 3D leakage magnetic field are analyzed. Secondly, the short-circuit impedance and split parallel branch current distribution in different working conditions are calculated, which is based on field-circuit coupled method. At last, effectiveness and feasibility of the proposed model is verified by comparison between experiment, analysis and simulation. The results showed that the 3D analysis method is a better approach to calculate the short-circuit impedance, since its analytical value is more closer to the experimental value compared with the 2D analysis results, the finite element method calculation error is less than 2%, while the leakage flux method maximum error is 7.2%.

Wind Turbine Planetary Gearbox Fault Diagnosis via Proportion-Extracting Synchrosqueezing Chirplet Transform

Wind turbine planetary gearboxes usually work under time-varying conditions,leading to nonstationary vibration signals.These signals often consist of multiple time-varying components with close instantaneous frequencies.Therefore,high-quality time-frequency analysis(TFA)is needed to extract the time-frequency feature from such nonstationary signals for fault diagnosis.However,it is difficult to obtain high-quality timefrequency representations(TFRs)through conventional TFA methods due to low resolution and time-frequency blurs.To address this issue,we propose a new TFA method termed the proportion-extracting synchrosqueezing chirplet transform(PESCT).Firstly,the proportion-extracting chirplet transform is employed to generate highresolution underlying TFRs.Then,the energy concentration of the underlying TFRs is enhanced via the synchrosqueezing transform.Finally,wind turbine planetary gearbox fault can be diagnosed by analysis of the dominant time-varying components revealed by the concentrated TFRs with high resolution.The proposed PESCT is suitable for achieving high-quality TFRs for complicated nonstationary signals.Numerical and experimental analyses validate the effectiveness of the PESCT in characterizing the nonstationary signals from wind turbine planetary gearboxes.

基于DA-Transformer的风机叶片覆冰检测

恶劣的气候条件会增加风机叶片结冰的风险.随着传感器技术在风电系统中的广泛应用,数据驱动的叶片覆冰检测方法引起了广泛关注.与传统方法相比,数据驱动方法可以避免专业知识的限制,降低安装检测设备引起的额外成本.然而,传统数据驱动模型挖掘数据信息的能力有限.同时,深度学习方法存在难以调整超参数的问题.为了解决上述问题,本文提出了蜻蜓算法(DA)和Transformer的风机叶片覆冰检测混合模型.其中,Transformer中的自注意力机制可以挖掘时间序列的局部和全局特征信息,蜻蜓算法可以智能优化Transformer的超参数.实验结果表明,相比已有的模型及Transformer而言,提出的混合模型具有更好的覆冰检测效果.

Transformer real-time reliability model based on operating conditions

Operational reliability evaluation theory reflects real-time reliability level of power system. The component failure rate varies with operating conditions. The impact of real-time operating conditions such as ambient temperature and transformer MVA (megavolt-ampere) loading on transformer insulation life is studied in this paper. The formula of transformer failure rate based on the winding hottest-spot temperature (HST) is given. Thus the real-time reliability model of transformer based on oper- ating conditions is presented. The work is illustrated using the 1979 IEEE Reliability Test System. The changes of operating conditions are simulated by using hourly load curve and temperature curve, so the curves of real-time reliability indices are ob- tained by using operational reliability evaluation.

考虑多绕组耦合的Sen Transformer电磁解析模型

为剖析Sen Transformer(ST)的内部特性,提出了一种适用于三相三柱式变压器结构的考虑多绕组耦合的ST电磁解析模型。一方面,从ST的电磁耦合关系出发,基于统一电磁等值电路推导了由自感系数和互感系数构成的ST电磁耦合模型。另一方面,从ST与外接等值系统的电气连接关系出发,推导了ST的内部电压和电流的电气解析模型。进而形成了由上述电磁耦合模型与电气解析模型构成的该类考虑多绕组耦合的ST电磁解析模型。最后,借助MATLAB软件,以一个三相三柱式ST为例,通过比较其电磁解析计算结果和现有ST串联电压补偿时域仿真结果发现,仅有个别绕组电压和支路电流有较明显的差异,但幅值差异≤6%,相角差异≤7°。该结果表明,多绕组磁耦合对于此类ST影响不大,但不宜忽视;且所提模型通过其磁路的不对称性表征,能更精确地反映其内部电磁特性。

Monitoring of Wind Turbine Blades Based on Dual-Tree Complex Wavelet Transform

Structural health monitoring(SHM)in-service is very important for wind turbine system.Because the central wavelength of a fiber Bragg grating(FBG)sensor changes linearly with strain or temperature,FBG-based sensors are easily applied to structural tests.Therefore,the monitoring of wind turbine blades by FBG sensors is proposed.The method is experimentally proved to be feasible.Five FBG sensors were set along the blade length in order to measure distributed strain.However,environmental or measurement noise may cover the structural signals.Dual-tree complex wavelet transform(DT-CWT)is suggested to wipe off the noise.The experimental studies indicate that the tested strain fluctuate distinctly as one of the blades is broken.The rotation period is about 1 s at the given working condition.However,the period is about 0.3 s if all the wind blades are in good conditions.Therefore,strain monitoring by FBG sensors could predict damage of a wind turbine blade system.Moreover,the studies indicate that monitoring of one blade is adequate to diagnose the status of a wind generator.