针对含分布式光伏的新型电力系统异常用电事件检测难度大、异常用电行为特征复杂、难以满足新型电力系统对样本数据质量要求的问题,提出一种深度循环神经网络-改进生成对抗网络（DRNN-WGAN）模型的新型电力系统负荷窃电行为对应的数据特征增强方法。首先,针对电力用户窃电导致数据存在异常值的问题,对分布式光伏的用户窃电以及纯负荷用户窃电导致的数据异常情况进行分析。其次,通过Wasserstein距离对生成对抗网络进行改进,利用生成对抗网络模型中对新型电力系统负荷异常用电数据进行生成和判别,建立新型电力系统负荷异常用电样本的生成对抗网络模型。然后,提出DRNN-WGAN网络的新型电力系统异常用电数据增强方法,并以LightGBM算法进行电网负荷数据分类与检测。最后,以辽宁省某地含高比例新能源的配电网数据进行仿真,仿真结果表明,与其他数据增强方法进行对比,该文提出的基于DRNN-WGAN网络的新型电力系统异常用电数据增强方法可加快模型的收敛速度,提升电网数据质量和检测精度。
Abstract
It is difficult to detect abnormal power consumption events in the new power system with distributed photovoltaic, and the characteristics of abnormal power consumption behavior are complex, which is difficult to meet the requirements of the new power system for sample data quality. A new data feature enhancement method for load stealing behavior of power system based on deep recurrent neural network-improved generative adversarial network （DRNN-WGAN） model is proposed. Firstly, aiming at the problem of abnormal data caused by power stealing by power users, the abnormal data caused by power stealing by distributed photovoltaic users and power stealing by pure load users are analyzed. Secondly, the generative adversarial network is improved by Wasserstein distance. The generative adversarial network model is used to generate and discriminate the abnormal power consumption data of the new power system load, and the generative adversarial network model of the new power system load abnormal power consumption sample is established. Then, a new power system abnormal electricity data enhancement method based on DRNN-WGAN network is proposed, and the LightGBM algorithm is used to classify and detect the power grid load data. Finally, the simulation is carried out with the data of a distribution network with high proportion of new energy in Liaoning Province. The simulation results show that compared with other data enhancement methods, the new power system abnormal electricity data enhancement method based on DRNN-WGAN network proposed in this paper can accelerate the convergence speed of the model and improve the data quality and detection accuracy of the power grid.
关键词
分布式光伏 /
生成对抗网络 /
异常用电 /
新型电力系统 /
Wasserstein距离 /
数据增强
Key words
distributed PV /
generative adversarial network /
abnormal electricity /
new power system /
Wasserstein distance /
data enhancement
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] SPIRIĆ J V, STANKOVIĆ S S, DOČIĆ M B, et al. Using the rough set theory to detect fraud committed by electricity customers[J]. International journal of electrical power & energy systems, 2014, 62: 727-734.
[2] 潘华, 梁作放, 肖雨涵, 等. 多场景下区域综合能源系统的优化运行[J]. 太阳能学报, 2021, 42(1): 484-492.
PAN H, LIANG Z F, XIAO Y H, et al.Optimal operation of regional integrated energy system under multiple scenes[J]. Acta energiae solaris sinica, 2021, 42(1): 484-492.
[3] BUZAU M M, TEJEDOR-AGUILERA J, CRUZ-ROMERO P, et al.Hybrid deep neural networks for detection of non-technical losses in electricity smart meters[J]. IEEE transactions on power systems, 2020, 35(2): 1254-1263.
[4] 刘科研, 盛万兴, 马晓晨, 等. 基于多种群遗传算法的分布式光伏接入配电网规划研究[J]. 太阳能学报, 2021, 42(6): 146-155.
LIU K Y, SHENG W X, MA X C, et al.Planning research of distributed photovoltaic source access distribution network based on multi-population genetic algorithm[J]. Acta energiae solaris sinica, 2021, 42(6): 146-155.
[5] ENDSLEY M R.Designing for Situation Awareness: An Approach to User-Centered Design, Second Edition[M]. Boca Raton, USA, CRC Press, 2016
[6] HAO Y, TIAN C S.A novel two-stage forecasting model based on error factor and ensemble method for multi-step wind power forecasting[J]. Applied energy, 2019, 238: 368-383.
[7] 汪露露, 吴红斌, 周亦尧. 基于供能可靠性的综合能源系统优化配置[J]. 太阳能学报, 2021, 42(12): 395-400.
WANG L L, WU H B, ZHOU Y Y.Optimal configuration of integrated energy system based on energy supply reliability[J]. Acta energiae solaris sinica, 2021, 42(12): 395-400.
[8] ZHANG Y C, LE J, LIAO X B, et al.A novel combination forecasting model for wind power integrating least square support vector machine, deep belief network, singular spectrum analysis and locality-sensitive hashing[J]. Energy, 2019, 168: 558-572.
[9] BAGAVATHIAPPAN S, LAHIRI B B, SARAVANAN T, et al.Infrared thermography for condition monitoring-A review[J]. Infrared physics & technology, 2013, 60: 35-55.
[10] OKUMUS I, DINLER A.Current status of wind energy forecasting and a hybrid method for hourly predictions[J]. Energy conversion and management, 2016, 123: 362-371.
[11] 张永会, 张亚谱, 潘超, 等. 考虑源-储特性的分布式电源并网阶段优化与评估方法[J]. 太阳能学报, 2020, 41(6): 226-233.
ZHANG Y H, ZHANG Y P, PAN C, et al.Optimization and evaluation method of grid-connected distributed power supply considering source-storage characteristics[J]. Acta energiae solaris sinica, 2020, 41(6): 226-233.
[12] LAHOUAR A, BEN HADJ SLAMA J. Hour-ahead wind power forecast based on random forests[J]. Renewable energy, 2017, 109: 529-541.
[13] AMBACH D, SCHMID W.A new high-dimensional time series approach for wind speed, wind direction and air pressure forecasting[J]. Energy, 2017, 135: 833-850.
[14] WANG Y, XIE Z X, HU Q H, et al.Correlation aware multi-step ahead wind speed forecasting with heteroscedastic multi-kernel learning[J]. Energy conversion and management, 2018, 163: 384-406.
[15] DO NASCIMENTO CAMELO H, LUCIO P S, LEAL J B V Jr, et al. Innovative hybrid models for forecasting time series applied in wind generation based on the combination of time series models with artificial neural networks[J]. Energy, 2018, 151: 347-357.
[16] HU Y L, CHEN L.A nonlinear hybrid wind speed forecasting model using LSTM network, hysteretic ELM and Differential Evolution algorithm[J]. Energy conversion and management, 2018, 173: 123-142.
[17] EZZAT A A, JUN M, DING Y.Spatio-temporal asymmetry of local wind fields and its impact on short-term wind forecasting[J]. IEEE transactions on sustainable energy, 2018, 9(3): 1437-1447.
[18] ZIEL F, CROONENBROECK C, AMBACH D.Forecasting wind power-Modeling periodic and non-linear effects under conditional heteroscedasticity[J]. Applied energy, 2016, 177: 285-297.
[19] MANOBEL B, SEHNKE F, LAZZÚS J A, et al. Wind turbine power curve modeling based on Gaussian Processes and artificial neural networks[J]. Renewable energy, 2018, 125: 1015-1020.
[20] LAMSAL D, SREERAM V, MISHRA Y, et al.Achieving a minimum power fluctuation rate in wind and photovoltaic output power using discrete Kalman filter based on weighted average approach[J]. IET renewable power generation, 2018, 12(6): 633-638.
[21] KOO J, HAN G D, CHOI H J, et al.Wind-speed prediction and analysis based on geological and distance variables using an artificial neural network: a case study in South Korea[J]. Energy, 2015, 93: 1296-1302.
[22] ZHANG C, ZHOU J Z, LI C S, et al.A compound structure of ELM based on feature selection and parameter optimization using hybrid backtracking search algorithm for wind speed forecasting[J]. Energy conversion and management, 2017, 143: 360-376.
基金
国网辽宁省电力有限公司管理科技项目（2023YF-157）
PDF(1154 KB)
