基于故障信息系统的电网智能决策系统的研究
摘要
近年来,伴随国民经济的高速增长和社会快速进步,电力作为国家支柱产业迅速发展壮大;数字化、信息化技术在电网中大量应用,使电网发展呈现出规模化、智能化的特点。继电保护作为保障电网安全稳定运行的第一道防线,担负保障电网和电力设备安全的重要职责。故障录波器是电力系统记录动态过程重要的精密设备,为评价继电保护动作行为及分析设备故障性质和原因提供了重要依据。
故障信息系统实现了将保护和故障录波器以及开关动作信息快速传送至调度端,为科学分析故障各种信息,实现对电网故障智能诊断和辅助决策提供了支撑平台,吉林省故障信信息管理系统已经投入运行,实现了故障信息的自动采集和Web发布。
通过信号分析和处理技术的应用,实现保护动作行为的子站端分析,为故障诊断提供依据。通过专家系统和信息处理技术的综合运用,实现主站端智能诊断和辅助决策,大大提高调度人员事故处理的快速性和准确性,为保证电网的安全稳定运行发挥至关重要作用。
In recent years, following the high speed growth of national economy and the social fast advancement, the electric power has been developed quickly as pillar industry of the country. Because the digitization and information technology are widely applied, the electric network development shows the scale and intellectualized characteristics. On the one hand, the function of resource optimizing configuration is improved by the large electrical network interconnection, on the other hand the affecting scope of electric network fault is expanded. The national economy and social stability are influenced very much by the fault. Since the 1960s, many cataclysm power system faults have taken place around the world, caused large economic losses and huge social impact.
The fault settlement of electric power system is the extremely complex real-time decision-making process. There are many factors including many uncertainty ones to affect the correct process. Formerly the local decision mainly depends on dispatcher's experience. When the electrical network breaks down, the dispatcher must rapidly and accurately diagnose the fault, give out the corresponding processing measure, restore the normal operation in time, according to the fault information. It is the precondition and foundation that the fault is rapidly and correctly diagnosed. But the modern electrical network is more and more complex; it is very difficult to satisfy the requirements of fault settlement that only depend on the dispatcher’s theory knowledge and experience.
As the first defense line guaranteeing the electrical network safe and stable run, the microcomputer relay protection takes on the important responsibility to ensure the electrical network and the electric equipment safe. The alternate current and voltage information are turned into the low level digital signal by the A/D transformation. Then the CPUs in the protection device carry out a series of functions according to the protection operating principle, such as data processing, value operation and logical judgment and so on, provide the hard contact information reflecting protection basic survey and the data of protection action.
The fault recorder is the important precision equipment to record the electric power system dynamic process. It gathers the current, voltage information and so on, simultaneously cuts in the circuit breaker phase splitting trip information, protection information and pilot protection channel information. The fault recorder has provided the important reference for evaluating relay protection action and analyzing the nature and reason of equipment fault.
The fault information management system has established based on the interconnection of microcomputer relay protection and digital fault recorder. It has realized that the protection, fault recorder and switch act information are quickly sent to the dispatch. And it has provided a supporting platform for the scientific analysis of all kinds of fault information, realized fault intelligent diagnosis and the assistant decision-making.
The relay protection fault information management system design of Jilin Province 220kV electric network adopts tri-level frame including master station, centralized control station and substation, uses ATM as the data transmission channel. The master station is located at Jilin dispatch center. The centralized control stations are set up at Siping, Tonghua , Yanbian, Changchun, Jilin, Liaoyuan, Baicheng and Baishan local dispatch. The substations are situated at 54 220kV substations managed by Jilin Province dispatch center.
The communication system is made up by two ATM virtual special net, each 220kV substation has two 10M Ethernet interfaces, each centralized control station has two 10M Ethernet interfaces, the master station has four l00M Ethernet interfaces, separately composing the protection and fault record net, they are isolated each other, The separate transmission between the protection and fault record information has been realized. Two kinds of information converge through exchange in master station at last. The information of the relay protection action, run, setting value, operating characteristic and channel, the circuit breaker action, the fault recorder, and so on, which concern the electrical network safe operation in substation, are transmitted to the master station by the information transmission equipments. The comprehensive analysis and use to the substation information has been realized in the master station through the information preservation and treatment.
At the information collection and management function, the categoricalness of raw information acquisition and the selectivity of report information are emphasized in the part of substation. Using the ability which wavelet transform can detect the fault signal irregularity, the data of the fault recorder has been analyzed. It has realized the fault time detection, the identification of line fault nature and the protection action, the selective upload information.
The fault diagnosis is the most important function in the electric network fault information system. It analyzes the various action signals of circuit breakers and protection devices, the electrical quantities measurement of voltage, current and so on in the substation. According to the open circuit breaker and the action logic of protection, the position and type of the fault can be concluded. In order to enhance the reliability and accuracy of the electric network fault information management system and deal with a various of electric network fault correctly, a kind of layering fault information processing and diagnosis method is proposed and has been realized in the system combinating the facts of Jilin electric network.
To the online fault diagnosis of the complex electric network fault, the protection and circuit breaker don’t correctly act at the same time, the paper presented that apply the information theory to deal with the fault messages. It is the information movement process that the protection and circuit breaker act when the electric power system breaks down. An essence of information movement is uncertainty, lying lots of random information. Therefore, the intelligent electric network fault diagnosis based on information theory can be a good reaction of the uncertainty between fault and indication, improve the reality and accuracy of fault diagnosis. The power system online fault diagnosis methods based on information theory, can give the probability distribution of many possible solutions in the decision result. The master station intelligent diagnosis and aid decision-making has been realized by the way of synthesis application of the expert systems and information processing technology. And the rapidity and accuracy which the dispatchers deal with the fault has been greatly enhanced. This system plays a vital role to ensure the electric network safe and stable operation.
故障信息系统实现了将保护和故障录波器以及开关动作信息快速传送至调度端,为科学分析故障各种信息,实现对电网故障智能诊断和辅助决策提供了支撑平台,吉林省故障信信息管理系统已经投入运行,实现了故障信息的自动采集和Web发布。
通过信号分析和处理技术的应用,实现保护动作行为的子站端分析,为故障诊断提供依据。通过专家系统和信息处理技术的综合运用,实现主站端智能诊断和辅助决策,大大提高调度人员事故处理的快速性和准确性,为保证电网的安全稳定运行发挥至关重要作用。
In recent years, following the high speed growth of national economy and the social fast advancement, the electric power has been developed quickly as pillar industry of the country. Because the digitization and information technology are widely applied, the electric network development shows the scale and intellectualized characteristics. On the one hand, the function of resource optimizing configuration is improved by the large electrical network interconnection, on the other hand the affecting scope of electric network fault is expanded. The national economy and social stability are influenced very much by the fault. Since the 1960s, many cataclysm power system faults have taken place around the world, caused large economic losses and huge social impact.
The fault settlement of electric power system is the extremely complex real-time decision-making process. There are many factors including many uncertainty ones to affect the correct process. Formerly the local decision mainly depends on dispatcher's experience. When the electrical network breaks down, the dispatcher must rapidly and accurately diagnose the fault, give out the corresponding processing measure, restore the normal operation in time, according to the fault information. It is the precondition and foundation that the fault is rapidly and correctly diagnosed. But the modern electrical network is more and more complex; it is very difficult to satisfy the requirements of fault settlement that only depend on the dispatcher’s theory knowledge and experience.
As the first defense line guaranteeing the electrical network safe and stable run, the microcomputer relay protection takes on the important responsibility to ensure the electrical network and the electric equipment safe. The alternate current and voltage information are turned into the low level digital signal by the A/D transformation. Then the CPUs in the protection device carry out a series of functions according to the protection operating principle, such as data processing, value operation and logical judgment and so on, provide the hard contact information reflecting protection basic survey and the data of protection action.
The fault recorder is the important precision equipment to record the electric power system dynamic process. It gathers the current, voltage information and so on, simultaneously cuts in the circuit breaker phase splitting trip information, protection information and pilot protection channel information. The fault recorder has provided the important reference for evaluating relay protection action and analyzing the nature and reason of equipment fault.
The fault information management system has established based on the interconnection of microcomputer relay protection and digital fault recorder. It has realized that the protection, fault recorder and switch act information are quickly sent to the dispatch. And it has provided a supporting platform for the scientific analysis of all kinds of fault information, realized fault intelligent diagnosis and the assistant decision-making.
The relay protection fault information management system design of Jilin Province 220kV electric network adopts tri-level frame including master station, centralized control station and substation, uses ATM as the data transmission channel. The master station is located at Jilin dispatch center. The centralized control stations are set up at Siping, Tonghua , Yanbian, Changchun, Jilin, Liaoyuan, Baicheng and Baishan local dispatch. The substations are situated at 54 220kV substations managed by Jilin Province dispatch center.
The communication system is made up by two ATM virtual special net, each 220kV substation has two 10M Ethernet interfaces, each centralized control station has two 10M Ethernet interfaces, the master station has four l00M Ethernet interfaces, separately composing the protection and fault record net, they are isolated each other, The separate transmission between the protection and fault record information has been realized. Two kinds of information converge through exchange in master station at last. The information of the relay protection action, run, setting value, operating characteristic and channel, the circuit breaker action, the fault recorder, and so on, which concern the electrical network safe operation in substation, are transmitted to the master station by the information transmission equipments. The comprehensive analysis and use to the substation information has been realized in the master station through the information preservation and treatment.
At the information collection and management function, the categoricalness of raw information acquisition and the selectivity of report information are emphasized in the part of substation. Using the ability which wavelet transform can detect the fault signal irregularity, the data of the fault recorder has been analyzed. It has realized the fault time detection, the identification of line fault nature and the protection action, the selective upload information.
The fault diagnosis is the most important function in the electric network fault information system. It analyzes the various action signals of circuit breakers and protection devices, the electrical quantities measurement of voltage, current and so on in the substation. According to the open circuit breaker and the action logic of protection, the position and type of the fault can be concluded. In order to enhance the reliability and accuracy of the electric network fault information management system and deal with a various of electric network fault correctly, a kind of layering fault information processing and diagnosis method is proposed and has been realized in the system combinating the facts of Jilin electric network.
To the online fault diagnosis of the complex electric network fault, the protection and circuit breaker don’t correctly act at the same time, the paper presented that apply the information theory to deal with the fault messages. It is the information movement process that the protection and circuit breaker act when the electric power system breaks down. An essence of information movement is uncertainty, lying lots of random information. Therefore, the intelligent electric network fault diagnosis based on information theory can be a good reaction of the uncertainty between fault and indication, improve the reality and accuracy of fault diagnosis. The power system online fault diagnosis methods based on information theory, can give the probability distribution of many possible solutions in the decision result. The master station intelligent diagnosis and aid decision-making has been realized by the way of synthesis application of the expert systems and information processing technology. And the rapidity and accuracy which the dispatchers deal with the fault has been greatly enhanced. This system plays a vital role to ensure the electric network safe and stable operation.
引文
【1】 唐涛,发电厂与变电站自动化技术及其应用,中国电力出版社,2005,第一版,355～580
【2】高翔,电网故障信息系统应用技术 中国电力出版社,2007.06,第一版,3～70
【3】彭玉华,小波变换与工程应用,科学出版社,1999.9,第一版,110～155
【4】张淑娥、孔英会、高强,电力系统通信技术,2005.7,第一版,195～205
【5】王世祯,电网调度运行技术,沈阳东北大学出版社,1997,第一版240～255
【6】屠强、周迎秋、肖荣国,变电所自动化实用技术及应用指南,2004.9,第一版,175～205
【7】葛耀中,新型继电保护与故障测距原理与技术,西安交通大学出版社,1993,第一版,65～95
【8】陈玉林,陈允平,孙金莉,电网故障诊断方法综述,中国电力,2006.05,第 39 卷第 5 期,27~30
【9】杨光亮、乐全明,基于小波神经网络和故障录波数据的电网故障类型识别,中国电机工程学报,2006.05,第 26 卷第 10 期,99～103
【10】任震、黄群古、黄雯莹,小波理论在电力系统微机保护中的应用研究,继电器,2001.01,第29卷第1期,1～4
【11】史志鸿、、刘伟、廖泽友、柳凤夙,继电保护故障信息系统的通信协义探讨,继电器,2004.05,第32卷第9期,40～44
【12】王高丰、杲秀芳、唐博,基于小波变换的电力系统故障暂态信号分析,电气开关,2007.03,第 28 卷第 10 期,32～38
【13】汤磊,基于信息理论的电力系统在线故障诊断,中国电机工程学报,2003,23期,5～11
【14】周岗、乐全明、郑华珍,基于单端故障电流录波数据的小波故障选线诊断,华东电力,2006.01,第34卷 第1期,34~37
【15】高署,江北,电力系统事故判断专家系统的研究,电力系统及其自动化学报,1992,51~58
【16】刘华,白义传,河南电网继电保护及故障信息管理系统建设,电力系统自动化, 2006.05,第30卷第9期,201～214
【17】刘涤尘,张琳,齐晓曼,电力系统继电保护信息分析管理专家系统的实现,2005.02,第30卷第4期,63～66
【18】Sang-Won Min,Jin-Man Sohn,Jong-Keun Park,etc.Adaptive Fault Section Estimation Using Matrix Representation With Fuzzy Relations. IEEE Transactions on Power System,2004,V19(2)
【19】章坚民、朱炳铨、蒋月良,继电保护故障信息处理主站系统设计的核心问题,电力系统自动化,2003.07,第23卷第14期,72～74
【20】刘涤尘、张琳、齐晓曼,电力系统继电保护信息分析管理专家系统的实现,继电器,2005.02,第 33 卷第 4 期,63～74
【21】石俊杰、孟碧波、顾锦汶,电网调度自动化专业综述,2004,第28卷4期,32～36
【22】张劲、章坚民、朱炳铨,220kV继电保护故障信息系统的设计,电力系统自动化,2003,第27卷第11期,26～30
【23】徐子立,变电站自动化系统IEC60870-5-103和IEC60870-5-103协义的分析和实施,电网技术,2002.06,第26卷第4期,18～22
【24】王益民,辛耀中、向力,调度自动化系统及数据网络的安全防护,电力系统自动化,2000,第25卷第21期,26～30
【25】刘青松、夏道止,基于正反向推理的电力系统故障诊断专家系统,, 电网技术,1999,第23卷第9期,8～14
【26】周亚玲、张 举,基于故障录波器信息的故障诊断系统,电网技术,2001.11,第25卷第11期,42~45
【27】杜一,张沛超、郁惟鏞 采用事例和规则混合护理的电网故障诊断专家系统,电网技术,2004.1,第28卷第1期,12～16
【28】Hong-Tzer Yang, “A New Neural Networks Approach To On-line Fault Section Estimation Using Information of Protective Relays And Circuit Breakers”, IEEE Transactions on Power Delivery, Vol.9, No. 1, January 1994: 220~229.
【29】Tanaka H., “Design and evaluation or neural network for fault diagnosis”, Proc. Of ESAP’89, USA, 1989.160~170
【30】周森,郑玉平,刘晗,张毅,IEC60870—5—103规约在微机型继电保护测试系统中的应用,继电器,200.12,第33卷,第23期,16～19
【2】高翔,电网故障信息系统应用技术 中国电力出版社,2007.06,第一版,3～70
【3】彭玉华,小波变换与工程应用,科学出版社,1999.9,第一版,110～155
【4】张淑娥、孔英会、高强,电力系统通信技术,2005.7,第一版,195～205
【5】王世祯,电网调度运行技术,沈阳东北大学出版社,1997,第一版240～255
【6】屠强、周迎秋、肖荣国,变电所自动化实用技术及应用指南,2004.9,第一版,175～205
【7】葛耀中,新型继电保护与故障测距原理与技术,西安交通大学出版社,1993,第一版,65～95
【8】陈玉林,陈允平,孙金莉,电网故障诊断方法综述,中国电力,2006.05,第 39 卷第 5 期,27~30
【9】杨光亮、乐全明,基于小波神经网络和故障录波数据的电网故障类型识别,中国电机工程学报,2006.05,第 26 卷第 10 期,99～103
【10】任震、黄群古、黄雯莹,小波理论在电力系统微机保护中的应用研究,继电器,2001.01,第29卷第1期,1～4
【11】史志鸿、、刘伟、廖泽友、柳凤夙,继电保护故障信息系统的通信协义探讨,继电器,2004.05,第32卷第9期,40～44
【12】王高丰、杲秀芳、唐博,基于小波变换的电力系统故障暂态信号分析,电气开关,2007.03,第 28 卷第 10 期,32～38
【13】汤磊,基于信息理论的电力系统在线故障诊断,中国电机工程学报,2003,23期,5～11
【14】周岗、乐全明、郑华珍,基于单端故障电流录波数据的小波故障选线诊断,华东电力,2006.01,第34卷 第1期,34~37
【15】高署,江北,电力系统事故判断专家系统的研究,电力系统及其自动化学报,1992,51~58
【16】刘华,白义传,河南电网继电保护及故障信息管理系统建设,电力系统自动化, 2006.05,第30卷第9期,201～214
【17】刘涤尘,张琳,齐晓曼,电力系统继电保护信息分析管理专家系统的实现,2005.02,第30卷第4期,63～66
【18】Sang-Won Min,Jin-Man Sohn,Jong-Keun Park,etc.Adaptive Fault Section Estimation Using Matrix Representation With Fuzzy Relations. IEEE Transactions on Power System,2004,V19(2)
【19】章坚民、朱炳铨、蒋月良,继电保护故障信息处理主站系统设计的核心问题,电力系统自动化,2003.07,第23卷第14期,72～74
【20】刘涤尘、张琳、齐晓曼,电力系统继电保护信息分析管理专家系统的实现,继电器,2005.02,第 33 卷第 4 期,63～74
【21】石俊杰、孟碧波、顾锦汶,电网调度自动化专业综述,2004,第28卷4期,32～36
【22】张劲、章坚民、朱炳铨,220kV继电保护故障信息系统的设计,电力系统自动化,2003,第27卷第11期,26～30
【23】徐子立,变电站自动化系统IEC60870-5-103和IEC60870-5-103协义的分析和实施,电网技术,2002.06,第26卷第4期,18～22
【24】王益民,辛耀中、向力,调度自动化系统及数据网络的安全防护,电力系统自动化,2000,第25卷第21期,26～30
【25】刘青松、夏道止,基于正反向推理的电力系统故障诊断专家系统,, 电网技术,1999,第23卷第9期,8～14
【26】周亚玲、张 举,基于故障录波器信息的故障诊断系统,电网技术,2001.11,第25卷第11期,42~45
【27】杜一,张沛超、郁惟鏞 采用事例和规则混合护理的电网故障诊断专家系统,电网技术,2004.1,第28卷第1期,12～16
【28】Hong-Tzer Yang, “A New Neural Networks Approach To On-line Fault Section Estimation Using Information of Protective Relays And Circuit Breakers”, IEEE Transactions on Power Delivery, Vol.9, No. 1, January 1994: 220~229.
【29】Tanaka H., “Design and evaluation or neural network for fault diagnosis”, Proc. Of ESAP’89, USA, 1989.160~170
【30】周森,郑玉平,刘晗,张毅,IEC60870—5—103规约在微机型继电保护测试系统中的应用,继电器,200.12,第33卷,第23期,16～19