基于云模型的变压器状态评估与故障诊断的研究
|
摘要
变压器作为电力系统中的核心设备,它的安全性和可靠性直接关系到电力系统的安全性和可靠性。由于各种内部因素和外部因素的影响,变压器在长期的运行中出现故障或事故的情况是难以避免的。作为变压器的状态检修基础,电力变压器的状态评估显得尤为重要,而作为状态评估的特例,故障诊断技术的研究同样具有非常重要的意义。因此,研究电力变压器的状态评估方法和故障诊断技术,以便及时、准确地检测出变压器的潜伏性故障,在成为电力行业普遍关注的课题的同时,也受到了学术界的广泛关注。
在变压器的状态评估方面,鉴于传统的评估方法只考虑了指标的模糊性而忽略了随机性,利用云模型理论在考虑模糊性的同时兼顾随机性的特点,将云模型理论应用其中,建立了新的状态评估模型。同时,结合层次分析法获取各指标的权重信息,利用变权公式对个别关键指标的权重进行均衡性调整,并采用云函数计算出各指标的评判矩阵,最终得到更接近实际的评判向量,从而判断出变压器的状态。通过实例分析,将新的评估模型与传统的评估方法进行对比,得出运用云模型理论对变压器进行状态评估的结果更接近于实际状态。
在变压器的故障诊断方面,鉴于物元理论在构建诊断模型时,忽略了分界值的不确定性,使诊断结果偏离了实际情况。利用云模型能合理地解决边界不确定问题,将物元理论与云模型相结合,建立了新的故障诊断模型。同时,结合油中溶解气体分析技术,通过对油中溶解气体的浓度、产气率、总烃含量以及气体间比值的分析,客观、准确地对变压器进行故障诊断。通过与其他诊断技术相比较,并结合案例分析,得出基于物元理论和云模型的变压器故障诊断技术具有更高的正确率。
In the power system, the transformer is seen to be the core equipment, so the safety and reliability of the transformer is directly related to the security and reliability of the power system. Due to various internal factors and external factors it is difficult to avoid failure or accident occurred in the transformer’s long run. As the basis of the state of transformer’s maintenance, power transformer condition assessment is particularly important, and as a special case of condition assessment, fault diagnosis technology also has a great significance. Therefore, the study of power transformer condition assessment and fault diagnosis technology for finding incipient faults as early as possible not only become common concern of power industry issues but also received widespread attention in academic.
In the study of transformer condition assessment, in view of the traditional fuzzy comprehensive evaluation model only considers fuzziness and ignored randomness of uncertainty, while the cloud model considers fuzziness and randomness of uncertainty at the same time, we establish a new condition assessment model based on cloud model. Besides, we will use sub-indicators of the evaluation object to calculate the evaluation results of objective indicators and get weight information by AHP, using variable weight formula to balance adjustment individual key indicators’weights, and get the actual evaluation vector using cloud formula, then determine the status of the transformer. Through case analysis and compared with the traditional condition assessment, we find that the result based on the new assessment model is closer to the actual state.
In the study of transformer fault diagnosis, in view of matter-element theory solving fault diagnosis problems in a simple and effective method, ignoring the uncertainty of boundary values, the result is deviated from the actual situation. Using cloud model theory to consider the uncertainty of the boundary, we combine cloud model theory with matter-element to propose a new model which is based on matter-element theory and cloud model theory. Meanwhile, we combine oil dissolved gas analysis, by analyzing the concentration of dissolved gases, gas production rate, the total hydrocarbon content and the ratio between gas in oil, and establish transformer fault diagnosis model, then determine the transformer fault accurately and objectively. By compared with other techniques, combined with case analysis, we reach the conclusion that the new technique which is based on matter-element theory and cloud model has a higher accuracy.
在变压器的状态评估方面,鉴于传统的评估方法只考虑了指标的模糊性而忽略了随机性,利用云模型理论在考虑模糊性的同时兼顾随机性的特点,将云模型理论应用其中,建立了新的状态评估模型。同时,结合层次分析法获取各指标的权重信息,利用变权公式对个别关键指标的权重进行均衡性调整,并采用云函数计算出各指标的评判矩阵,最终得到更接近实际的评判向量,从而判断出变压器的状态。通过实例分析,将新的评估模型与传统的评估方法进行对比,得出运用云模型理论对变压器进行状态评估的结果更接近于实际状态。
在变压器的故障诊断方面,鉴于物元理论在构建诊断模型时,忽略了分界值的不确定性,使诊断结果偏离了实际情况。利用云模型能合理地解决边界不确定问题,将物元理论与云模型相结合,建立了新的故障诊断模型。同时,结合油中溶解气体分析技术,通过对油中溶解气体的浓度、产气率、总烃含量以及气体间比值的分析,客观、准确地对变压器进行故障诊断。通过与其他诊断技术相比较,并结合案例分析,得出基于物元理论和云模型的变压器故障诊断技术具有更高的正确率。
In the power system, the transformer is seen to be the core equipment, so the safety and reliability of the transformer is directly related to the security and reliability of the power system. Due to various internal factors and external factors it is difficult to avoid failure or accident occurred in the transformer’s long run. As the basis of the state of transformer’s maintenance, power transformer condition assessment is particularly important, and as a special case of condition assessment, fault diagnosis technology also has a great significance. Therefore, the study of power transformer condition assessment and fault diagnosis technology for finding incipient faults as early as possible not only become common concern of power industry issues but also received widespread attention in academic.
In the study of transformer condition assessment, in view of the traditional fuzzy comprehensive evaluation model only considers fuzziness and ignored randomness of uncertainty, while the cloud model considers fuzziness and randomness of uncertainty at the same time, we establish a new condition assessment model based on cloud model. Besides, we will use sub-indicators of the evaluation object to calculate the evaluation results of objective indicators and get weight information by AHP, using variable weight formula to balance adjustment individual key indicators’weights, and get the actual evaluation vector using cloud formula, then determine the status of the transformer. Through case analysis and compared with the traditional condition assessment, we find that the result based on the new assessment model is closer to the actual state.
In the study of transformer fault diagnosis, in view of matter-element theory solving fault diagnosis problems in a simple and effective method, ignoring the uncertainty of boundary values, the result is deviated from the actual situation. Using cloud model theory to consider the uncertainty of the boundary, we combine cloud model theory with matter-element to propose a new model which is based on matter-element theory and cloud model theory. Meanwhile, we combine oil dissolved gas analysis, by analyzing the concentration of dissolved gases, gas production rate, the total hydrocarbon content and the ratio between gas in oil, and establish transformer fault diagnosis model, then determine the transformer fault accurately and objectively. By compared with other techniques, combined with case analysis, we reach the conclusion that the new technique which is based on matter-element theory and cloud model has a higher accuracy.
引文
[1]吴立增.变压器状态评估方法的研究[D].保定:华北电力大学, 2005
[2]刘有为,李光范,高克利,等.制定《电气设备状态维修导则》的原则框架[J].电网技术, 2003, 27(6):64-67
[3]中国电力科学研究院.基于状态检修基础的电气设备试验规程[S].北京:国家电网公司, 2007
[4]赵文清,朱永利.电力变压器状态评估综述[J].变压器, 2007, 44(11): 9-12
[5] Arshad M. Islam, S.M Khaliq. A Power Transformer Aging and Life Extension[A]. Probabilistic, 2004 International Conference on Methods Applied to Power System, 2004 :498-501
[6] Hsu Chih-Wei, Liu Chih-Jen. A Comparison of Mentods for Multicalss Support Vector Machines[J]. IEEE Transactions on Neural Networks, 2002, V13(2):415-425
[7] Fei SW, Zhang XB. Fault Diagnosis of Power Transformer Based on Support Vector Machine with Genetic Algorithm[J]. Expert Systems with Applications, 2009, V36(8):11352-11357
[8]刘敦楠,陈雪青,何光宇,等.电力市场评价指标体系的原理和构建方法[J].电力系统自动化, 2005, 29(23):2-7
[9]袁志坚,孙才新,袁张渝,等.变压器健康状态评估的灰色聚类决策方法[J].重庆大学学报(自然科学版), 2005, 28(3):22-25
[10]纪航,朱永利,郭伟.基于模糊综合评判的变压器状态评分方法研究[J].继电器, 2006, 34(5):29-33
[11]张青.多指标综合评价失效因素分析及模型[J].统计与决策, 2005, 18(5):6-8
[12]谢红玲,律方成.基于信息融合的变压器状态评估方法研究[J].华北电力大学学报, 2006, 33(2):8-11
[13]孙才新,陈伟根,李俭,廖瑞金.电气设备油中气体在线监测与故障诊断技术[M].北京:科学技术出版社, 2003
[14]陈新岗,李太福.基于DGA特征量的变压器绝缘故障诊断专家系统的研究[J].变压器, 2005, 42(1):33-36
[15]谢可夫,罗安.遗传算法在变压器故障诊断中的应用[J].电力自动化设备, 2005, 25:55-58
[16] Georgilakis PS, Katsigiannis JA, Valavanis KP, et al. A Systematic Stochastic Petri Net Based Methodology for Transformer Fault Diagnosis and Repair Actions[J]. Journal of Intelligent and RoboticSystems, 2006, V45(2):181-201
[17]李俭,孙才新.基于灰色聚类分析的充油电力变压器绝缘故障诊断的研究[J].电工技术学报, 2002, 17(4):80-83
[18] Yan-Chang Huang. A New Data Mining Approach to Dissolved Gas Analysis of Oil-insulated Power Apparatus. IEEE Transactions on Power Delivery, 2003, 18(4):1257-1261
[19] Q Su, C Mi. A Fuzzy Dissolved Gas Analysis Method for the Diagnosis of Multiple Incipient Faults in a Transformer. IEEE Transactions on Power Systems. 2000, 15(2):593-598
[20]范定国,贺硕,段富,等.一种基于云模型的综合评判模型[J].科技情报开发与经济, 2003, 13(12):157-159
[21]何雷,李丽平.基于云模型的电力变压器状态综合评判[J].华北电力大学学报, 2009, 36(3): 98-103
[22] Li Deyi. Uncertainty Reasoning Based on Cloud Models[J]. Computer Science and Mathematics with Application, 1998, V35(3):99-123
[23] Marley MS, Saumon D, Goldblatt C. A Patchy Cloud Model for the L to T Dwarf Transition[J]. Astrophysical Journal Letters, 2010, V723(1):117-121
[24] R. Wang, W.G. Wan, X.L.Ma, et al. Cloud Model-based Control Strategy on Cluster Communication Coverage for Wireless Sensor Networks[A].2010 2nd International Asia Conference on Informatics in Control, Automation and Robotics, 2010:307-310
[25] Gao ST. The Scheduling Algorithm of Grid Task Based on Cloud Model[A]. International Conference on Advanced Measurement and Test (AMT 2010)[C]. Switzerland Trans Tech Publications Lid, 2010:1177-1183
[26] Islam S M, Wu T, Ledwich G. A Novel Fuzzy Logic Approach to Transformer Fault Diagnosis[J]. IEEE Transactions on Dielectrics and Electrical Insulation, 2000, 7(2):177-186
[27] Cahyono B, Arifianto I. Thermal Condition Assessment of Power Transformer[A]. Proceedings of the 9th International Conference on Properties and Applications of Dielectric Materials[C], 2009:60-62
[28] Moradi M, Gholami A. Transformer Condition Assessment Via Oil Quality Parameters and DGA[A]. Proceedings of 2008 International Conference On Condition Monitoring and Diagnosis[C], 2007:993-999
[29] Teller A, Levin Z. The Effects of Aerosols on Precipitation and Dimensions of Subtropical Clouds: A Sensitivity Study using A Numerical Cloud Model[J]. Atmospheric Chemistry and Physics, 2006, V6:67-80
[30]国家电网公司. 110(66)KV~500KV油浸式变压器(电抗器)运行规范[M]. 2005, 3
[31]廖瑞金,王谦,骆思佳,等.基于模糊综合评判的电力变压器运行状态评估模型[J].电力系统自动化, 2008, 32(3):70-75
[32] Kaibo Duan, Keerthi S. Which is the Best Multiclass SVM Methods? An Empirical Study[A]. Multiple Classifier Systems, The 6th International Workshop, Seaside, CA, USA:2005
[33] Pedro Vicente Jover, Rodríguez, Antero Arkkio. Detection of Stator Winding Fault in Induction Motor Using Fuzzy Logic[J]. Applied Soft Computing, 2008, V8(2): 1112-1120
[34] Shintemirov A, Tang WH, Wu QH. Transformer Winding Condition Assessment Using Frequency Response Analysis and Evidential Reasoning[J]. IET Electric Power Applications, 2010, V4(3):198-212
[35] Su HS, Li QZ. Fuzzy Neural Classifier for Transformer Fault Diagnosis Based on EM Learning[A]. Computational Interlligence[C], 2006:222-229
[36] Tang W H, Spurgeon K, Wu Q H, et al. An Evidential Reasoning Approach to Transformer Condition Assessments[J]. IEEE Transactions on Power Delivery, 2004, 19 (4):1696-1703
[37] DL/596-1996,电力设备预防性试验规程[S].北京, 1996
[38] Wu Lizeng, Zhu Yongli, Li Xueyu. Application of Multi-agent and Data Mining Techniques in Condition Assessment of Transformers[A]. Power Conference, 2004:823-827
[39] Xu Tao, He Renmu, Wang Peng. Applications of Datamining Technique for Power System Transient Stability Prediction, Electric Utility Deregulation[A]. Restructuring and Power Technologies. Proceedings of the 2004 IEEE International Conference, 2004:389-392
[40]骆思佳,廖瑞金,王有元,等.带变权的电力变压器状态模糊综合评判[J].高电压技术, 2007, 33(8):106-110
[41] Zheng N, Lv FC. Transformer Condition Assessment Based on Matter-element Theory[A]. Proceedings of the 2007 Conference on Systems Science Dynamics[C], 2007:3101-3105
[42]熊浩,孙才新,杜鹏等.基于物元理论的电力变压器状态综合评估[J].重庆大学学报(自然科学版), 2006, 29(10):24-28
[2]刘有为,李光范,高克利,等.制定《电气设备状态维修导则》的原则框架[J].电网技术, 2003, 27(6):64-67
[3]中国电力科学研究院.基于状态检修基础的电气设备试验规程[S].北京:国家电网公司, 2007
[4]赵文清,朱永利.电力变压器状态评估综述[J].变压器, 2007, 44(11): 9-12
[5] Arshad M. Islam, S.M Khaliq. A Power Transformer Aging and Life Extension[A]. Probabilistic, 2004 International Conference on Methods Applied to Power System, 2004 :498-501
[6] Hsu Chih-Wei, Liu Chih-Jen. A Comparison of Mentods for Multicalss Support Vector Machines[J]. IEEE Transactions on Neural Networks, 2002, V13(2):415-425
[7] Fei SW, Zhang XB. Fault Diagnosis of Power Transformer Based on Support Vector Machine with Genetic Algorithm[J]. Expert Systems with Applications, 2009, V36(8):11352-11357
[8]刘敦楠,陈雪青,何光宇,等.电力市场评价指标体系的原理和构建方法[J].电力系统自动化, 2005, 29(23):2-7
[9]袁志坚,孙才新,袁张渝,等.变压器健康状态评估的灰色聚类决策方法[J].重庆大学学报(自然科学版), 2005, 28(3):22-25
[10]纪航,朱永利,郭伟.基于模糊综合评判的变压器状态评分方法研究[J].继电器, 2006, 34(5):29-33
[11]张青.多指标综合评价失效因素分析及模型[J].统计与决策, 2005, 18(5):6-8
[12]谢红玲,律方成.基于信息融合的变压器状态评估方法研究[J].华北电力大学学报, 2006, 33(2):8-11
[13]孙才新,陈伟根,李俭,廖瑞金.电气设备油中气体在线监测与故障诊断技术[M].北京:科学技术出版社, 2003
[14]陈新岗,李太福.基于DGA特征量的变压器绝缘故障诊断专家系统的研究[J].变压器, 2005, 42(1):33-36
[15]谢可夫,罗安.遗传算法在变压器故障诊断中的应用[J].电力自动化设备, 2005, 25:55-58
[16] Georgilakis PS, Katsigiannis JA, Valavanis KP, et al. A Systematic Stochastic Petri Net Based Methodology for Transformer Fault Diagnosis and Repair Actions[J]. Journal of Intelligent and RoboticSystems, 2006, V45(2):181-201
[17]李俭,孙才新.基于灰色聚类分析的充油电力变压器绝缘故障诊断的研究[J].电工技术学报, 2002, 17(4):80-83
[18] Yan-Chang Huang. A New Data Mining Approach to Dissolved Gas Analysis of Oil-insulated Power Apparatus. IEEE Transactions on Power Delivery, 2003, 18(4):1257-1261
[19] Q Su, C Mi. A Fuzzy Dissolved Gas Analysis Method for the Diagnosis of Multiple Incipient Faults in a Transformer. IEEE Transactions on Power Systems. 2000, 15(2):593-598
[20]范定国,贺硕,段富,等.一种基于云模型的综合评判模型[J].科技情报开发与经济, 2003, 13(12):157-159
[21]何雷,李丽平.基于云模型的电力变压器状态综合评判[J].华北电力大学学报, 2009, 36(3): 98-103
[22] Li Deyi. Uncertainty Reasoning Based on Cloud Models[J]. Computer Science and Mathematics with Application, 1998, V35(3):99-123
[23] Marley MS, Saumon D, Goldblatt C. A Patchy Cloud Model for the L to T Dwarf Transition[J]. Astrophysical Journal Letters, 2010, V723(1):117-121
[24] R. Wang, W.G. Wan, X.L.Ma, et al. Cloud Model-based Control Strategy on Cluster Communication Coverage for Wireless Sensor Networks[A].2010 2nd International Asia Conference on Informatics in Control, Automation and Robotics, 2010:307-310
[25] Gao ST. The Scheduling Algorithm of Grid Task Based on Cloud Model[A]. International Conference on Advanced Measurement and Test (AMT 2010)[C]. Switzerland Trans Tech Publications Lid, 2010:1177-1183
[26] Islam S M, Wu T, Ledwich G. A Novel Fuzzy Logic Approach to Transformer Fault Diagnosis[J]. IEEE Transactions on Dielectrics and Electrical Insulation, 2000, 7(2):177-186
[27] Cahyono B, Arifianto I. Thermal Condition Assessment of Power Transformer[A]. Proceedings of the 9th International Conference on Properties and Applications of Dielectric Materials[C], 2009:60-62
[28] Moradi M, Gholami A. Transformer Condition Assessment Via Oil Quality Parameters and DGA[A]. Proceedings of 2008 International Conference On Condition Monitoring and Diagnosis[C], 2007:993-999
[29] Teller A, Levin Z. The Effects of Aerosols on Precipitation and Dimensions of Subtropical Clouds: A Sensitivity Study using A Numerical Cloud Model[J]. Atmospheric Chemistry and Physics, 2006, V6:67-80
[30]国家电网公司. 110(66)KV~500KV油浸式变压器(电抗器)运行规范[M]. 2005, 3
[31]廖瑞金,王谦,骆思佳,等.基于模糊综合评判的电力变压器运行状态评估模型[J].电力系统自动化, 2008, 32(3):70-75
[32] Kaibo Duan, Keerthi S. Which is the Best Multiclass SVM Methods? An Empirical Study[A]. Multiple Classifier Systems, The 6th International Workshop, Seaside, CA, USA:2005
[33] Pedro Vicente Jover, Rodríguez, Antero Arkkio. Detection of Stator Winding Fault in Induction Motor Using Fuzzy Logic[J]. Applied Soft Computing, 2008, V8(2): 1112-1120
[34] Shintemirov A, Tang WH, Wu QH. Transformer Winding Condition Assessment Using Frequency Response Analysis and Evidential Reasoning[J]. IET Electric Power Applications, 2010, V4(3):198-212
[35] Su HS, Li QZ. Fuzzy Neural Classifier for Transformer Fault Diagnosis Based on EM Learning[A]. Computational Interlligence[C], 2006:222-229
[36] Tang W H, Spurgeon K, Wu Q H, et al. An Evidential Reasoning Approach to Transformer Condition Assessments[J]. IEEE Transactions on Power Delivery, 2004, 19 (4):1696-1703
[37] DL/596-1996,电力设备预防性试验规程[S].北京, 1996
[38] Wu Lizeng, Zhu Yongli, Li Xueyu. Application of Multi-agent and Data Mining Techniques in Condition Assessment of Transformers[A]. Power Conference, 2004:823-827
[39] Xu Tao, He Renmu, Wang Peng. Applications of Datamining Technique for Power System Transient Stability Prediction, Electric Utility Deregulation[A]. Restructuring and Power Technologies. Proceedings of the 2004 IEEE International Conference, 2004:389-392
[40]骆思佳,廖瑞金,王有元,等.带变权的电力变压器状态模糊综合评判[J].高电压技术, 2007, 33(8):106-110
[41] Zheng N, Lv FC. Transformer Condition Assessment Based on Matter-element Theory[A]. Proceedings of the 2007 Conference on Systems Science Dynamics[C], 2007:3101-3105
[42]熊浩,孙才新,杜鹏等.基于物元理论的电力变压器状态综合评估[J].重庆大学学报(自然科学版), 2006, 29(10):24-28