基于数据驱动和多判据融合的油色谱监测传感器有效性评估方法
|
摘要
变压器是电力系统的枢纽设备之一,其运行的可靠性直接关系电力系统的安全运行。油色谱在线监测技术通过使用油色谱监测传感器监测变压器油中溶解各气体的含量能够及时发现变压器的早期故障。在现场的实际应用中,大量无效监测传感器的存在使得在线监测数据的质量降低,从而使得在线监测系统无法及时准确地监测变压器的运行状态。提出了一种基于数据驱动和多判据融合的油色谱监测传感器有效性评估方法,针对油色谱传感器采集的在线数据,首先从在线数据中选取用于评估的固定长度的特征数据集,然后将特征数据集中异常值的分布情况、连续相同值的分布情况、变异系数的变化情况以及产气率的变化情况作为判据对特征数据集进行判定,得到对应的判别值。之后,根据对每个判据的侧重情况,赋予判别值权重,获得传感器状态值,将该状态值与预先设置的容忍度进行对比,即可得到传感器的评估结果。使用现场的实例进行验证可知:该方法从多个方面对在线监测传感器的有效性进行评估,可以及时发现故障传感器,为提高在线监测数据的准确性及监测装置的可靠性提供必要的基础支撑。
Power transformer is a key electrical apparatus in power system,whose reliability is closely related to power system safety.On-line dissolved gas analysis(DGA) monitoring technique can discover early-stage faults of transformer by analyzing concentration of dissolved gases in transformer oil.However,the on-line system cannot monitor transformer operation status in time because there are many invalid sensors in actual field application,making on-line data quality relatively low.This paper proposed a validity assessment method of DGA sensor based on data driving and multiple criterion.This method was based on large amount of on-line DGA data in field,and the characteristic set used to assess validity was acquired with a sliding window.Then number and distribution of abnormal values,distribution of continual same value,changes of variation coefficients and change of gas production rate are used as criteria,and discriminant values can be obtained according to these criterion.State values of the sensors were calculated by summing the discriminant values with weighting value of every criterion.At last,validity of the sensors can be acquired by comparing the state value with tolerance value set in advance.The invalid sensors can be discovered in time and accuracy of on-line monitoring system and reliability of monitoring device can be improved with this method.
Power transformer is a key electrical apparatus in power system,whose reliability is closely related to power system safety.On-line dissolved gas analysis(DGA) monitoring technique can discover early-stage faults of transformer by analyzing concentration of dissolved gases in transformer oil.However,the on-line system cannot monitor transformer operation status in time because there are many invalid sensors in actual field application,making on-line data quality relatively low.This paper proposed a validity assessment method of DGA sensor based on data driving and multiple criterion.This method was based on large amount of on-line DGA data in field,and the characteristic set used to assess validity was acquired with a sliding window.Then number and distribution of abnormal values,distribution of continual same value,changes of variation coefficients and change of gas production rate are used as criteria,and discriminant values can be obtained according to these criterion.State values of the sensors were calculated by summing the discriminant values with weighting value of every criterion.At last,validity of the sensors can be acquired by comparing the state value with tolerance value set in advance.The invalid sensors can be discovered in time and accuracy of on-line monitoring system and reliability of monitoring device can be improved with this method.
引文
[1]Mackenzie E A,Crossey J,De Pablo A,et al.On-line monitoring and diagnostics for power transformers[C]∥2010 IEEE International Symposium on Electrical Insulation(ISEI).San Diego,USA:IEEE,2010:1-5.
[2]代杰杰,宋辉,杨祎,等.基于深度信念网络的变压器油中溶解气体浓度预测方法[J].电网技术,2017,41(8):2737-2742.Dai Jiejie,Song Hui,Yang Yi,et al.Concentration prediction of dissolved gases in transformer oil based on deep belief networks[J].Power System Technology,2017,41(8):2737-2742(in Chinese).
[3]倪鹤立,王同磊,袁炜雄,等.应用耦合电容传感器的变压器类绕组暂态电压非接触式测量方法[J].电网技术,2016,40(12):3930-3937.Ni Heli,Wang Tonglei,Yuan Weixiong,et al.Non-contact measurement of transient voltage in transformer type windings with coupling capacitive sensors[J].Power System Technology,2016,40(12):3930-3937(in Chinese).
[4]张强,李成榕,刘齐,等.变压器升高座内部套管局部放电特高频信号传播特征[J].电网技术,2017,41(4):1332-1337.Zhang Qiang,Li Chengrong,Liu Qi,et al.Research on propagation of UHF signals by partial discharges in ascending flanged base[J].Power System Technology,2017,41(4):1332-1337(in Chinese).
[5]张新伯,唐炬,潘成,等.用于局部放电模式识别的深度置信网络方法[J].电网技术,2016,40(10):3272-3278.Zhang Xinbo,Tang Ju,Pan Cheng,et al.Research of partial discharge recognition based on deep belief nets[J].Power System Technology,2016,40(10):3272-3278(in Chinese).
[6]卞建鹏,廖瑞金,杨丽君.应用弱化缓冲算子与最小二乘支持向量机的变压器油中溶解气体浓度预测[J].电网技术,2012,36(2):195-199.Bian Jianpeng,Liao Ruijin,Yang Lijun.Concentration prediction of gases dissolved in transformer oil based on weakening buffer operator and least square support vector machine[J].Power System Technology,2012,36(2):195-199(in Chinese).
[7]许坤,周建华,茹秋实,等.变压器油中溶解气体在线监测技术发展与展望[J].高电压技术,2005,31(8):30-32.Xu Kun,Zhou Jianhua,Ru Qiushi,et al.Development and prospect of transformer oil dissolved gas on-line monitoring technology[J].High Voltage Engineering,2005,31(8):30-32(in Chinese).
[8]高树国,王学磊,李庆民,等.基于MCD稳健统计分析的变压器油色谱异常值检测及分布特性[J].高电压技术,2014,40(11):3477-3482.Gao Shuguo,Wang Xuelei,Li Qingmin,et al.Outliers detection and distribution characteristics of the transformer DGA data based on MCD robust statistics[J].High Voltage Engineering,2014,40(11):3477-3482(in Chinese).
[9]邹建明.在线监测技术在电网中的应用[J].高电压技术,2007,33(8):203-206.Zou Jianming.Application of on line monitoring technology on power grid[J].High Voltage Engineering,2007,33(8):203-206(in Chinese).
[10]黄丹,徐平安,王其军,等.基于PCA神经网络和D-S决策的瓦斯传感器故障辨识[J].仪表技术与传感器,2015(3):99-103.Huang Dan,Xu Pingan,Wang Qijun,et.al.Fault identification for gas sensor based on PCA-neural network and D-S Evidence Theory[J].Instrument Technique and Sensor,2015(3):99-103(in Chinese).
[11]赵金宪,金鸿章.基于小波包和神经网络的瓦斯传感器故障诊断[J].传感器与微系统,2010,29(5):80-82.Zhao Jinxian,Jin Hongzhang.Gas sensor fault diagnosis based on wavelet packet and neural network[J].Transducer and Microsystem Technologies,2010,29(5):80-82(in Chinese).
[12]何慧娟,陈健,邹宇华.基于DPCA方法的传感器故障检测与诊断[J].传感器与微系统,2009,28(12):35-38.He Huijuan,Chen Jian,Zou Yuhua.Fault detection and diagnosis for sensors based on DPCA[J].Transducer and Microsystem Technologies,2009,28(12):35-38(in Chinese).
[13]高运广,刘顺波,张振仁.基于KPCA的HVAC系统传感器故障诊断[J].传感器与微系统,2008,27(5):37-39.Gao Yunguang,Liu Shunbo,Zhang Zhenren.Fault detection and diagnosis for sensor in HVAC system based on KPCA[J].Transducer and Microsystem Technologies,2008,27(5):37-39(in Chinese).
[14]汪辉,税爱社,宗福兴,等.基于SKPCA的卫星整流罩空调系统传感器故障检测研究[J].传感器与微系统,2015,34(2):23-27.Wang Hui,Shui Aishe,Zong Fuxing,et.al.Research on sensor fault detection for air conditioning system of satellite fairing based on SKPCA[J].Transducer and Microsystem Technologies,2015,34(2):23-27(in Chinese).
[15]王婷,李国勇,吕世轩.基于HPSO-RBF神经网络的瓦斯传感器故障诊断[J].仪表技术与传感器,2015(3):78-81.Wang Ting,Li Guoyong,LüShixuan.Gas sensor fault diagnosis based on HPSO-RBF neural network[J].Instrument Technique and Sensor,2015(3):78-81(in Chinese).
[16]汤琳.基于马尔科夫预测的多传感器故障检测与诊断机制[J].传感器与微系统,2014,33(11):52-55.Tang Lin.Multi-sensor fault detection and diagnosis mechanism based on Markov prediction[J].Transducer and Microsystem Technologies,2014,33(11):52-55(in Chinese).
[17]李冬辉,李晟.基于小波神经网络的空调系统传感器故障诊断[J].传感器与微系统,2007,26(4):54-57.Li Donghui,Li Sheng.Sensor fault diagnosis in air-condition system based on wavelet neural network[J].Transducer and Microsystem Technologies,2007,26(4):54-57(in Chinese).
[18]阳少军,肖登明,潘勇斌,等.变压器色谱在线监测的新型传感器[J].高电压技术,2002,28(1):30-31.Yang Shaojun,Xiao Dengming,Pan Yongbin,et al.A new kind of sensor used for the on-line measurement of transformer chromatography[J].High Voltage Engineering,2002,28(1):30-31(in Chinese).
[19]赵杰,潘勇斌,李红雷,等.新型变压器油色谱在线监测系统的研制[J].高电压技术,2000,26(6):20-21.Zhao Jie,Pan Yongbin,Li Honglei,et al.The study of new transformer online monitoring system[J].High Voltage Engineering,2000,26(6):20-21(in Chinese).
[20]王伦宾.基于分布传感技术的变压器油色谱在线监测技术应用研究[D].重庆:重庆大学,2004.
[21]苏展,郭军科,卢立秋,等.变压器油色谱在线监测装置的现场评估分析[J].宁夏电力,2013(2):59-62.Su Zhan,Guo Junke,Lu Liqiu,et al.Site evaluation analysis on online monitoring device for transformer insulating oil chromatography[J].Ningxia Power,2013(2):59-62(in Chinese).
[22]郭伟,武志峰,郝菊屏,等.变压器油在线色谱监测装置的应用现状分析[J].电力安全技术,2014,16(3):55-59.Guo Wei,Wu Zhifeng,Hao Juping,et al.Application situation analysis of transformer on-line DGA monitoring device[J].Electric Safety Technology,2014,16(3):55-59(in Chinese).
[23]陈伟根,蔚超,孙才新.变压器油纸绝缘气隙放电特性及其产气规律[J].高电压技术,2010,36(4):849-855.Chen Weigen,Wei Chao,Sun Caixin.Air-gap discharge characteristics in transformer oil-paper insulation and gas generation law[J].High Voltage Engineering,2010,36(4):849-855(in Chinese).
[24]杜威,邹先霞.基于数据流的滑动窗口机制的研究[J].计算机工程与设计,2005,26(11):64-66.Du Wei,Zou Xianxia.Research of sliding windows scheme based on data stream[J].Computer Engineering and Design,2005,26(11):64-66(in Chinese).
[25]Mosinski F,Piotrowski T.New statistical methods for evaluation of DGA data[J].IEEE Transactions on Dielectrics and Electrical Insulation,2003,10(2):260-265.
[26]Ma H,Saha T K,Ekanayake C.Statistical learning techniques and their applications for condition assessment of power transformer[J].IEEE Transactions on Dielectrics and Electrical Insulation,2012,19(2):481-489.
[27]付靖.变异系数在既有结构可靠性检验中的应用研究初探[D].西安:西安建筑科技大学,2009.
[28]陈林,李永福,司马文霞,等.采用变异系数和Allan因子的雷电时间丛集性特征分析[J].高电压技术,2011,37(7):1706-1712.Chen Lin,Li Yongfu,Sima Wenxia,et al.Time-clustering characteristics of lightning data using coefficient of variation and allan factor method[J].High Voltage Engineering,2011,37(7):1706-1712(in Chinese).
[29]GB/T 7252—2001变压器油中溶解气体分析和判断导则[S].北京:中国标准出版社,2001.
[2]代杰杰,宋辉,杨祎,等.基于深度信念网络的变压器油中溶解气体浓度预测方法[J].电网技术,2017,41(8):2737-2742.Dai Jiejie,Song Hui,Yang Yi,et al.Concentration prediction of dissolved gases in transformer oil based on deep belief networks[J].Power System Technology,2017,41(8):2737-2742(in Chinese).
[3]倪鹤立,王同磊,袁炜雄,等.应用耦合电容传感器的变压器类绕组暂态电压非接触式测量方法[J].电网技术,2016,40(12):3930-3937.Ni Heli,Wang Tonglei,Yuan Weixiong,et al.Non-contact measurement of transient voltage in transformer type windings with coupling capacitive sensors[J].Power System Technology,2016,40(12):3930-3937(in Chinese).
[4]张强,李成榕,刘齐,等.变压器升高座内部套管局部放电特高频信号传播特征[J].电网技术,2017,41(4):1332-1337.Zhang Qiang,Li Chengrong,Liu Qi,et al.Research on propagation of UHF signals by partial discharges in ascending flanged base[J].Power System Technology,2017,41(4):1332-1337(in Chinese).
[5]张新伯,唐炬,潘成,等.用于局部放电模式识别的深度置信网络方法[J].电网技术,2016,40(10):3272-3278.Zhang Xinbo,Tang Ju,Pan Cheng,et al.Research of partial discharge recognition based on deep belief nets[J].Power System Technology,2016,40(10):3272-3278(in Chinese).
[6]卞建鹏,廖瑞金,杨丽君.应用弱化缓冲算子与最小二乘支持向量机的变压器油中溶解气体浓度预测[J].电网技术,2012,36(2):195-199.Bian Jianpeng,Liao Ruijin,Yang Lijun.Concentration prediction of gases dissolved in transformer oil based on weakening buffer operator and least square support vector machine[J].Power System Technology,2012,36(2):195-199(in Chinese).
[7]许坤,周建华,茹秋实,等.变压器油中溶解气体在线监测技术发展与展望[J].高电压技术,2005,31(8):30-32.Xu Kun,Zhou Jianhua,Ru Qiushi,et al.Development and prospect of transformer oil dissolved gas on-line monitoring technology[J].High Voltage Engineering,2005,31(8):30-32(in Chinese).
[8]高树国,王学磊,李庆民,等.基于MCD稳健统计分析的变压器油色谱异常值检测及分布特性[J].高电压技术,2014,40(11):3477-3482.Gao Shuguo,Wang Xuelei,Li Qingmin,et al.Outliers detection and distribution characteristics of the transformer DGA data based on MCD robust statistics[J].High Voltage Engineering,2014,40(11):3477-3482(in Chinese).
[9]邹建明.在线监测技术在电网中的应用[J].高电压技术,2007,33(8):203-206.Zou Jianming.Application of on line monitoring technology on power grid[J].High Voltage Engineering,2007,33(8):203-206(in Chinese).
[10]黄丹,徐平安,王其军,等.基于PCA神经网络和D-S决策的瓦斯传感器故障辨识[J].仪表技术与传感器,2015(3):99-103.Huang Dan,Xu Pingan,Wang Qijun,et.al.Fault identification for gas sensor based on PCA-neural network and D-S Evidence Theory[J].Instrument Technique and Sensor,2015(3):99-103(in Chinese).
[11]赵金宪,金鸿章.基于小波包和神经网络的瓦斯传感器故障诊断[J].传感器与微系统,2010,29(5):80-82.Zhao Jinxian,Jin Hongzhang.Gas sensor fault diagnosis based on wavelet packet and neural network[J].Transducer and Microsystem Technologies,2010,29(5):80-82(in Chinese).
[12]何慧娟,陈健,邹宇华.基于DPCA方法的传感器故障检测与诊断[J].传感器与微系统,2009,28(12):35-38.He Huijuan,Chen Jian,Zou Yuhua.Fault detection and diagnosis for sensors based on DPCA[J].Transducer and Microsystem Technologies,2009,28(12):35-38(in Chinese).
[13]高运广,刘顺波,张振仁.基于KPCA的HVAC系统传感器故障诊断[J].传感器与微系统,2008,27(5):37-39.Gao Yunguang,Liu Shunbo,Zhang Zhenren.Fault detection and diagnosis for sensor in HVAC system based on KPCA[J].Transducer and Microsystem Technologies,2008,27(5):37-39(in Chinese).
[14]汪辉,税爱社,宗福兴,等.基于SKPCA的卫星整流罩空调系统传感器故障检测研究[J].传感器与微系统,2015,34(2):23-27.Wang Hui,Shui Aishe,Zong Fuxing,et.al.Research on sensor fault detection for air conditioning system of satellite fairing based on SKPCA[J].Transducer and Microsystem Technologies,2015,34(2):23-27(in Chinese).
[15]王婷,李国勇,吕世轩.基于HPSO-RBF神经网络的瓦斯传感器故障诊断[J].仪表技术与传感器,2015(3):78-81.Wang Ting,Li Guoyong,LüShixuan.Gas sensor fault diagnosis based on HPSO-RBF neural network[J].Instrument Technique and Sensor,2015(3):78-81(in Chinese).
[16]汤琳.基于马尔科夫预测的多传感器故障检测与诊断机制[J].传感器与微系统,2014,33(11):52-55.Tang Lin.Multi-sensor fault detection and diagnosis mechanism based on Markov prediction[J].Transducer and Microsystem Technologies,2014,33(11):52-55(in Chinese).
[17]李冬辉,李晟.基于小波神经网络的空调系统传感器故障诊断[J].传感器与微系统,2007,26(4):54-57.Li Donghui,Li Sheng.Sensor fault diagnosis in air-condition system based on wavelet neural network[J].Transducer and Microsystem Technologies,2007,26(4):54-57(in Chinese).
[18]阳少军,肖登明,潘勇斌,等.变压器色谱在线监测的新型传感器[J].高电压技术,2002,28(1):30-31.Yang Shaojun,Xiao Dengming,Pan Yongbin,et al.A new kind of sensor used for the on-line measurement of transformer chromatography[J].High Voltage Engineering,2002,28(1):30-31(in Chinese).
[19]赵杰,潘勇斌,李红雷,等.新型变压器油色谱在线监测系统的研制[J].高电压技术,2000,26(6):20-21.Zhao Jie,Pan Yongbin,Li Honglei,et al.The study of new transformer online monitoring system[J].High Voltage Engineering,2000,26(6):20-21(in Chinese).
[20]王伦宾.基于分布传感技术的变压器油色谱在线监测技术应用研究[D].重庆:重庆大学,2004.
[21]苏展,郭军科,卢立秋,等.变压器油色谱在线监测装置的现场评估分析[J].宁夏电力,2013(2):59-62.Su Zhan,Guo Junke,Lu Liqiu,et al.Site evaluation analysis on online monitoring device for transformer insulating oil chromatography[J].Ningxia Power,2013(2):59-62(in Chinese).
[22]郭伟,武志峰,郝菊屏,等.变压器油在线色谱监测装置的应用现状分析[J].电力安全技术,2014,16(3):55-59.Guo Wei,Wu Zhifeng,Hao Juping,et al.Application situation analysis of transformer on-line DGA monitoring device[J].Electric Safety Technology,2014,16(3):55-59(in Chinese).
[23]陈伟根,蔚超,孙才新.变压器油纸绝缘气隙放电特性及其产气规律[J].高电压技术,2010,36(4):849-855.Chen Weigen,Wei Chao,Sun Caixin.Air-gap discharge characteristics in transformer oil-paper insulation and gas generation law[J].High Voltage Engineering,2010,36(4):849-855(in Chinese).
[24]杜威,邹先霞.基于数据流的滑动窗口机制的研究[J].计算机工程与设计,2005,26(11):64-66.Du Wei,Zou Xianxia.Research of sliding windows scheme based on data stream[J].Computer Engineering and Design,2005,26(11):64-66(in Chinese).
[25]Mosinski F,Piotrowski T.New statistical methods for evaluation of DGA data[J].IEEE Transactions on Dielectrics and Electrical Insulation,2003,10(2):260-265.
[26]Ma H,Saha T K,Ekanayake C.Statistical learning techniques and their applications for condition assessment of power transformer[J].IEEE Transactions on Dielectrics and Electrical Insulation,2012,19(2):481-489.
[27]付靖.变异系数在既有结构可靠性检验中的应用研究初探[D].西安:西安建筑科技大学,2009.
[28]陈林,李永福,司马文霞,等.采用变异系数和Allan因子的雷电时间丛集性特征分析[J].高电压技术,2011,37(7):1706-1712.Chen Lin,Li Yongfu,Sima Wenxia,et al.Time-clustering characteristics of lightning data using coefficient of variation and allan factor method[J].High Voltage Engineering,2011,37(7):1706-1712(in Chinese).
[29]GB/T 7252—2001变压器油中溶解气体分析和判断导则[S].北京:中国标准出版社,2001.