WO2023083366A1 - 基于在线监测数据的避雷器运行状态识别方法、装置 - Google Patents

基于在线监测数据的避雷器运行状态识别方法、装置 Download PDF

Info

Publication number
WO2023083366A1
WO2023083366A1 PCT/CN2022/131878 CN2022131878W WO2023083366A1 WO 2023083366 A1 WO2023083366 A1 WO 2023083366A1 CN 2022131878 W CN2022131878 W CN 2022131878W WO 2023083366 A1 WO2023083366 A1 WO 2023083366A1
Authority
WO
WIPO (PCT)
Prior art keywords
arrester
data sequence
leakage current
online
distortion rate
Prior art date
Application number
PCT/CN2022/131878
Other languages
English (en)
French (fr)
Inventor
方文田
李旭宏
杨永辉
蔡伟贤
陈晓彬
李涛
朱育钊
方逸越
Original Assignee
广东电网有限责任公司揭阳供电局
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 广东电网有限责任公司揭阳供电局 filed Critical 广东电网有限责任公司揭阳供电局
Publication of WO2023083366A1 publication Critical patent/WO2023083366A1/zh

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00Computer-aided design [CAD]
    • G06F30/20Design optimisation, verification or simulation
    • G06F30/27Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01RMEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R23/00Arrangements for measuring frequencies; Arrangements for analysing frequency spectra
    • G01R23/16Spectrum analysis; Fourier analysis
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01RMEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R31/00Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01RMEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R31/00Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
    • G01R31/50Testing of electric apparatus, lines, cables or components for short-circuits, continuity, leakage current or incorrect line connections
    • G01R31/52Testing for short-circuits, leakage current or ground faults
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06NCOMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00Computing arrangements based on biological models
    • G06N3/004Artificial life, i.e. computing arrangements simulating life
    • G06N3/006Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06NCOMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00Computing arrangements based on biological models
    • G06N3/02Neural networks
    • G06N3/04Architecture, e.g. interconnection topology
    • G06N3/045Combinations of networks
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06NCOMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00Computing arrangements based on biological models
    • G06N3/02Neural networks
    • G06N3/08Learning methods
    • YGENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y04INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
    • Y04SSYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
    • Y04S10/00Systems supporting electrical power generation, transmission or distribution
    • Y04S10/50Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications

Definitions

  • zinc oxide arresters Due to its simple structure, small size and strong flow capacity, zinc oxide arresters are widely used in the protection of lightning overvoltage and operating overvoltage in power systems, providing a reliable guarantee for the continuous and safe operation of power supply systems.
  • zinc oxide arresters In addition to being threatened by lightning overvoltage and operating overvoltage, zinc oxide arresters will also be affected by factors such as aging and moisture during operation, resulting in increased leakage current and heat generation in the arrester. In severe cases, thermal runaway may occur, seriously affecting The performance of zinc oxide arresters and the safety of power grid systems. Therefore, the operating status of zinc oxide arresters must be monitored. With the popularization of digital smart substations, online monitoring of surge arresters has gradually replaced part of live tests and operational inspections. The principle of on-line monitoring of surge arresters is consistent with the capacitive compensation method of live test of surge arresters. It can be summarized that the live test equipment is installed at the production site, and then the test data is transmitted to the monitoring terminal in real time through the communication line. The advantage of online monitoring of arresters is that it can continuously track the operating resistive current and full current of arresters, which can provide a basis for the formulation of test plans, but there are still problems of being susceptible to phase-to-phase interference and harmonic interference.
  • the present application provides a method and a device for identifying the operating state of a lightning arrester based on online monitoring data.
  • the harmonic elimination module is configured to input the online data sequence of the total distortion rate of the arrester leakage current in the online monitoring data and the online data sequence of the total harmonic distortion rate of the arrester voltage into the LSTM-RNN model that has been trained to obtain the update for eliminating the harmonics of the power grid
  • the online data sequence of the total distortion rate of the leakage current of the arrester
  • the embodiment of the present application also provides a computer-readable storage medium, on which a computer program is stored.
  • the program is executed by a processor, the online monitoring data-based A method for identifying the operating state of a lightning arrester.
  • the embodiment of the present application also provides an electronic device, including: at least one processor; a memory configured to store at least one program, and when the at least one program is executed by the at least one processor, the at least one A processor implements the above-mentioned method for identifying the operating state of the arrester based on online monitoring data.
  • FIG. 1 is a flowchart of a method for identifying the operating state of an arrester based on online monitoring data provided in Embodiment 1 of the present application;
  • FIG. 2 is a flow chart of a method for identifying the operating state of an arrester based on online monitoring data provided in Embodiment 2 of the present application;
  • Fig. 3 is the equivalent circuit diagram of the arrester under the operating voltage provided by the second embodiment of the present application.
  • Fig. 10 is a schematic structural diagram of an arrester operating state identification device based on online monitoring data provided in Embodiment 4 of the present application;
  • FIG. 11 is a schematic structural diagram of an electronic device provided in Embodiment 6 of the present application.
  • This application provides a method and device for identifying the operating state of an arrester based on online monitoring data, so as to realize real-time monitoring of the operating state of the arrester, discover hidden dangers in the early stage of arrester degradation, provide a reference for the power supply department to formulate an arrester maintenance test plan, and improve the power grid. intelligence level.
  • Figure 1 is a flow chart of a method for identifying the operating state of an arrester based on online monitoring data provided in Embodiment 1 of the present application. This embodiment is applicable to monitoring the operating state of an arrester, and the method can be executed by an operating state identification device for an arrester. , including the following steps:
  • the total harmonic distortion rate of arrester voltage can be understood as a parameter used to measure the degree of voltage waveform distortion, expressed as a percentage of the ratio of the root mean square value of each harmonic voltage to the effective value of the fundamental voltage.
  • the voltage waveform should be a periodic standard sine wave, but because there are a large number of power supply and consumption equipment with nonlinear impedance characteristics in the power system, these equipment inject harmonic current into the public grid or generate harmonics in the public grid voltage, known as a harmonic source.
  • the harmonic source makes the actual voltage waveform deviate from the sine wave, this phenomenon is called voltage sine wave distortion.
  • the current passing through the zinc oxide resistor is called the leakage current of the zinc oxide arrester, which is also considered as the full leakage current of the arrester.
  • the full current of the arrester leakage current can reflect the insulation condition of the arrester, and it is an important means to judge the quality of the arrester under the operating voltage.
  • the data can be obtained through the arrester online monitoring system applied in the production site, the data includes the online data sequence of the total distortion rate of the arrester leakage current, the online data sequence of the total harmonic distortion rate of the arrester voltage, the online data sequence of the full current of the arrester leakage current and the leakage current of the arrester Resistive current online data series.
  • long short-term memory recurrent neural network is a special neural network (RNN) model.
  • the RNN model is a deep learning model for processing sequence input and is widely used in speech recognition, state recognition and machine translation. Compared with LSTM-RNN, RNN has no memory function, and it is easy to cause gradient disappearance and gradient explosion.
  • LSTM-RNN introduces three core elements of forgetting operation, input operation and output operation, which meet the needs of long-term memory information and The need for forgotten information.
  • This method uses the existing monitoring data of the arrester as the feature quantity, and does not need to add additional sensors; it realizes the real-time monitoring of the operation status of the arrester, and can detect hidden dangers in time in the early stage of arrester degradation, which provides a reference for the power supply department to formulate an arrester maintenance test plan According to the basis, the intelligence level and power supply reliability of the power grid have been improved.
  • the online monitoring data at least include the online data sequence of the total distortion rate of the arrester leakage current, the online data sequence of the total harmonic distortion rate of the arrester voltage, the online data sequence of the full current of the arrester leakage current and the online data sequence of the arrester leakage current resistive current data sequence.
  • the equivalent circuit simulation model of the lightning arrester under the operating voltage is established, and model training is carried out according to the simulation data obtained by the equivalent circuit simulation model of the lightning arrester, and the trained LSTM-RNN model is obtained, including:
  • Zinc oxide surge arrester is a device used to protect electrical equipment, mainly composed of a cavity, a porcelain sleeve, a primary terminal and a stack of multiple zinc oxide valve plates.
  • Fig. 3 is the equivalent circuit diagram of the surge arrester under the operating voltage provided by the second embodiment of the present application.
  • the equivalent circuit simulation model of the arrester includes a voltage source and a plurality of resistance-capacitance units connected in series on one side of the voltage source, and the resistance-capacitance unit Including fixed resistors and fixed capacitors connected in series, the equivalent circuit simulation model of the surge arrester includes multiple non-linear resistors, each of which is connected in parallel with a fixed capacitor.
  • C 1 represents the equivalent capacitance of the first valve
  • C 2 represents the equivalent capacitance of the second valve
  • C n represents the equivalent capacitance of the nth valve
  • C 0 represents the equivalent capacitance of the arrester cavity, porcelain sleeve and air
  • R 0 represents the equivalent resistance of the arrester cavity, porcelain sleeve and air.
  • the resistance of the zinc oxide arrester has nonlinear characteristics, and under normal voltage, the current flowing through the arrester is extremely small. Under overvoltage, the resistance of the zinc oxide arrester decreases rapidly, and the energy is quickly released, which plays a role in protecting electrical equipment.
  • the prediction steps of leakage current harmonics are:
  • Step 1 Using the model shown in Figure 3, set the frequency of the voltage source to the grid standard frequency of 50 Hz, randomly set the nonlinear parameters (K i , a i ) of the arrester so that the resistance of the nonlinear resistor is random, and obtain the arrester at N times Leakage current harmonic total distortion rate, and record these N data outputs as the first simulation data sequence X 1 .
  • Step 3 Obtain the total distortion rate of the arrester leakage current at N times simulated in step 2, and record this N data as the second simulation data sequence X 2 ; at the same time, obtain the total distortion rate of the arrester voltage harmonics at N times , and record these N data as the simulation data sequence Y 1 of the arrester voltage harmonic total distortion rate.
  • Step 4 Use the toolkit Tensorflow of the machine learning algorithm to run on the software PYCHARM, so as to build the LSTM-RNN model, and use the second simulation data sequence X 2 of the total harmonic distortion rate of the leakage current of the arrester and the total harmonic distortion rate of the arrester voltage
  • the simulation data sequence Y 1 is used as the input data of the LSTM-RNN initial model
  • the first simulation data sequence X 1 of the leakage current harmonic distortion rate of the arrester is used as the output data of the LSTM-RNN initial model, and the model training is performed and the trained model is saved.
  • FIG. 4 is a flow chart of a method for identifying the operating state of an arrester based on online monitoring data provided by Embodiment 3 of the present application. Referring to FIG. 4, the method includes:
  • the eigenvalue sets include the full current simulation data sequence of the arrester leakage current and the arrester leakage current resistance
  • the random forest model is trained by using the training sample set, and the trained random forest model including multiple decision trees is obtained.
  • the deterioration of the arrester refers to the deterioration of the zinc oxide valve plate, and the deterioration of the zinc oxide valve plate is a slow process. In the early stage of deterioration, it only shows the increase of the harmonic content of the leakage current of the arrester, but the influence of the resistive current and the capacitive current of the leakage current of the arrester is very small.
  • the pollution on the surface of the lightning arrester porcelain sleeve has a great influence on the lightning arrester.
  • the internal insulating rod may turn brownish yellow, and the upper part of the core may even be covered with oil, and the insulation of the core is obviously aging.
  • the insulating glaze on the side of the resistor is also frosted and loses its luster. Its long-term operation at a temperature slightly higher than the normal operating temperature will cause the resistors to deteriorate.
  • the equivalent circuit simulation model of the arrester includes a voltage source and a plurality of resistance-capacitance units connected in series on one side of the voltage source, and the resistance-capacity units include fixed resistors and fixed capacitors connected in series, and the equivalent circuit simulation model of the arrester includes Multiple nonlinear resistors, each nonlinear resistor is connected in parallel with a fixed capacitor; the equivalent circuit simulation model of the arrester also includes the equivalent resistance and equivalent capacitance of the arrester cavity, porcelain sleeve and air, where the equivalent resistance and the series connection A plurality of resistance-capacitance units are connected in parallel, and the equivalent capacitance and the equivalent resistance are connected in parallel.
  • the eigenvalue sets include the full current simulation data sequence of the arrester leakage current and the arrester leakage current resistance
  • the current simulation data sequence and the third simulation data sequence of the total harmonic distortion rate of the leakage current of the arrester, and the training sample set of the random forest model of the eigenvalue set including:
  • a shunt resistor is connected in parallel at both ends of at least one non-linear resistor, and the resistance value of the shunt resistor is changed to obtain a set of output characteristic values when the state of the arrester is wet by simulation.
  • Fig. 7 is a schematic diagram of the leakage current change when the arrester provided by Embodiment 3 of the present application is affected by pollution. Referring to Fig. 7, the vertical axis in Fig. 7 represents the current, and the unit is uA, and the horizontal axis represents resistance, and the unit is G ⁇ . From Fig. 7 It can be seen that the growth rate of the resistive current of the leakage current of the arrester is obviously faster than the full current of the leakage current of the arrester.
  • the nonlinear characteristic of the arrester is represented by the following formula:
  • I is the current flowing through the valve plate of the arrester
  • U is the terminal voltage of the valve plate of the arrester
  • ki is the first parameter determining the non-linear characteristic of the valve plate of the arrester
  • a i is the second parameter determining the non-linear characteristic of the valve plate of the arrester.
  • the first parameter of the nonlinear characteristic of the arrester valve plate is related to the material of the valve plate.
  • the resistance of the valve plate of the arrester changes with the change of the terminal voltage, therefore, the function describing its nonlinear characteristics is a piecewise function.
  • Fig. 8 is a schematic diagram of the volt-ampere characteristic curve of the zinc oxide arrester provided by the third embodiment of the present application. Referring to Fig. 8, the ordinate in the figure represents the current I, and the abscissa represents the voltage U. It can be seen from the figure that the initial stage is in the linear region, and the voltage is in the nonlinear region after reaching the reference voltage Vref.
  • ki corresponds to the first parameter at the boundary point between the linear region and the nonlinear region
  • a i corresponds to the second parameter at the boundary point between the linear region and the nonlinear region.
  • Fig. 9 is a schematic diagram of the relationship between the total distortion rate of the harmonic current and the nonlinear characteristic of the arrester under the operating voltage provided by Embodiment 3 of the present application.
  • the vertical axis in the figure represents the first parameter K i , the nonlinear characteristic of the arrester valve plate
  • the horizontal axis represents the second parameter a i of the non-linear characteristic of the arrester valve plate
  • the vertical axis represents the total distortion rate of the harmonic current of the arrester.
  • the equivalent circuit of the arrester is constructed by Simulink, and the corresponding total harmonic distortion rate of the leakage current of the arrester is obtained by changing the nonlinear characteristics of the zinc oxide valve plate.
  • the total distortion rate of the arrester leakage current becomes larger as the value of K i becomes smaller, and becomes smaller as the value of a i becomes larger.
  • the change of K i value is more sensitive to the influence of the total distortion of the leakage current.
  • the training sample set is used to perform random forest model training to obtain a trained random forest model including multiple decision trees, including:
  • Step 1 Use the extraction method with replacement to extract n training samples from the N training samples in the training sample set as the root node samples of the decision tree in the random forest model to train the decision tree, where n ⁇ N.
  • the number N of training samples in the training sample set and the number n of extracted training samples can be set according to actual needs, and the embodiment of the present application does not limit the number.
  • the attribute refers to the full current simulation data sequence of the arrester leakage current, the resistive current simulation data sequence of the arrester leakage current and the third simulation data sequence of the total harmonic distortion rate of the arrester leakage current;
  • the maximum information entropy refers to the division of the decision tree using The principle of maximum information gain, information gain is detailed in the following formula (7).
  • Step 3 Repeat step 2 until all child nodes of the decision tree can no longer be split, and output a decision tree model.
  • Step 4 Repeat steps 1 to 3 to output multiple decision tree models to form a trained random forest model.
  • the selected feature sets are all continuous values. Therefore, the decision tree algorithm is used to train the data, and the division of the decision tree adopts the principle of maximum information gain, and the information gain can be defined by formula (1).
  • y is the output category of the decision tree, and the output categories are normal, damp, polluted, and degraded, and P k is the output probability corresponding to the kth category.
  • the combination strategy of the random forest adopts the voting method, and the final output result is the probability corresponding to each category as shown in formula (3):
  • P k is the output probability corresponding to the k-th category
  • M is the number of decision trees
  • m k is the number of decision trees whose output result is the k-th category.
  • Fig. 10 is a schematic structural diagram of a lightning arrester operating state identification device based on online monitoring data provided by Embodiment 4 of the present application. This embodiment corresponds to the above-mentioned method embodiment.
  • the device 400 includes: an acquisition module 401, a harmonic elimination module 402 and prediction module 403 .
  • the acquisition module 401 is set to acquire the online monitoring data of the arrester.
  • the online monitoring data at least include the online data sequence of the total distortion rate of the arrester leakage current, the online data sequence of the total harmonic distortion rate of the arrester voltage, the online data sequence of the full current of the arrester leakage current and the leakage current resistance of the arrester.
  • Sexual current online data series At least include the online data sequence of the total distortion rate of the arrester leakage current, the online data sequence of the total harmonic distortion rate of the arrester voltage, the online data sequence of the full current of the arrester leakage current and the leakage current resistance of the arrester.
  • the harmonic elimination module 402 is set to input the online data sequence of the total distortion rate of the arrester leakage current in the online monitoring data and the online data sequence of the total distortion rate of the arrester voltage harmonic into the trained LSTM-RNN model to obtain an updated arrester leakage for eliminating grid harmonics Current total distortion rate online data series.
  • the arrester operating state identification device based on online monitoring data provided by the embodiment of the present application can execute the method for identifying the operating state of the arrester based on online monitoring data provided by any embodiment of the present application, and has corresponding functional modules and beneficial effects of the execution method.
  • the included modules are only divided according to functional logic, but are not limited to the above-mentioned division, as long as the corresponding functions can be realized; in addition, The specific names of the modules are only for the convenience of distinguishing each other, and are not used to limit the protection scope of the present application.
  • Embodiment 5 of the present application also provides a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, the operating state identification of a lightning arrester based on online monitoring data provided by any embodiment of the present application is realized. method.
  • the storage medium may be a non-transitory storage medium.
  • the computer storage medium in the embodiments of the present application may use any combination of one or more computer-readable media.
  • the computer readable medium may be a computer readable signal medium or a computer readable storage medium.
  • a computer readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or device, or any combination thereof.
  • a computer readable signal medium may include a data signal carrying computer readable program code in baseband or as part of a carrier wave. Such propagated data signals may take many forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the foregoing.
  • a computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device. .
  • the remote computer can be connected to the user's computer through any kind of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or it can be connected to an external computer (for example, use an Internet service provider to connect via the Internet).
  • LAN Local Area Network
  • WAN Wide Area Network
  • FIG. 11 is a schematic structural diagram of an electronic device provided in Embodiment 6 of the present application. As shown in FIG. 11 , the electronic device includes: one or more processors 110 and memory 120 . A processor 110 is taken as an example in FIG. 11 .
  • the electronic device may further include: an input device 130 and an output device 140 .
  • the electronic device may also not include the input device 130 and the output device 140 .
  • the processor 110, the memory 120, the input device 130 and the output device 140 in the electronic device may be connected via a bus or in other ways.
  • connection via a bus is taken as an example.
  • Memory 120 may be a non-transitory computer storage medium or a transitory computer storage medium.
  • the non-transitory computer storage medium is, for example, at least one magnetic disk storage device, flash memory device or other non-volatile solid-state storage device.
  • the memory 120 may optionally include memory located remotely relative to the processor 110, and these remote memories may be connected to the electronic device through a network. Examples of the above-mentioned network may include Internet, enterprise intranet, local area network, mobile communication network and combinations thereof.
  • the input device 130 can be configured to receive input numbers or character information, and generate key signal input related to user settings and function control of the electronic device.
  • the output device 140 may include a display device such as a display screen.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Evolutionary Computation (AREA)
  • Software Systems (AREA)
  • Artificial Intelligence (AREA)
  • Mathematical Physics (AREA)
  • General Engineering & Computer Science (AREA)
  • Computing Systems (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Molecular Biology (AREA)
  • Data Mining & Analysis (AREA)
  • Computational Linguistics (AREA)
  • Biophysics (AREA)
  • Biomedical Technology (AREA)
  • General Health & Medical Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Medical Informatics (AREA)
  • Computer Hardware Design (AREA)
  • Geometry (AREA)
  • Testing Electric Properties And Detecting Electric Faults (AREA)
  • Thermistors And Varistors (AREA)

Abstract

本申请实施例公开了一种基于在线监测数据的避雷器运行状态识别方法、装置,该方法包括:获取避雷器在线监测数据;将在线监测数据中避雷器泄漏电流总畸变率在线数据序列、避雷器电压谐波总畸变率在线数据序列输入训练完成的LSTM-RNN模型,得到消除电网谐波的更新避雷器泄漏电流总畸变率在线数据序列;将更新避雷器泄漏电流总畸变率在线数据序列、避雷器泄漏电流全电流在线数据序列和避雷器泄漏电流阻性电流在线数据序列输入到训练完成的随机森林模型,得到避雷器的状态识别预测结果。

Description

基于在线监测数据的避雷器运行状态识别方法、装置
本申请要求在2021年11月15日提交中国专利局、申请号为202111344700.9的中国专利申请的优先权,以上申请的全部内容通过引用结合在本申请中。
技术领域
本申请涉及电网系统避雷器运行状态监测技术,例如涉及一种基于在线监测数据的避雷器运行状态识别方法、装置。
背景技术
氧化锌避雷器由于具备结构简单、体积小以及通流能力强等诸多优点,广泛应用于电力系统的雷电过电压和操作过电压的防护,为供电系统的持续安全运行提供了可靠保证。
氧化锌避雷器除了遭受雷电过电压和操作过电压等威胁外,在运行过程中还会受老化及受潮等因素的影响,导致泄漏电流增大及避雷器发热,严重时可能出现热失控现象,严重影响氧化锌避雷器的性能及电网系统的安全。因此,必须对氧化锌避雷器的运行状态进行监控。随着数字智能变电站的推广,避雷器在线监测逐渐取代了带电试验以及运行巡视的部分内容。避雷器在线监测的原理与避雷器带电测试容性补偿法一致,可以总结为把带电测试设备安装在了生产现场,再通过通讯线路把测试数据实时传输到监控终端。避雷器在线监测的优势在于其能不间断的跟踪避雷器的运行阻性电流与全电流,可为试验计划的制定提供依据,但也仍存在易受相间干扰影响与谐波干扰的问题。
发明内容
本申请提供一种基于在线监测数据的避雷器运行状态识别方法、装置。
第一方面,本申请实施例提供了一种基于在线监测数据的避雷器运行状态识别方法,该方法包括:
获取避雷器在线监测数据,所述在线监测数据至少包括避雷器泄漏电流总畸变率在线数据序列、避雷器电压谐波总畸变率在线数据序列、避雷器泄漏电流全电流在线数据序列和避雷器泄漏电流阻性电流在线数据序列;
将所述在线监测数据中避雷器泄漏电流总畸变率在线数据序列、避雷器电压谐波总畸变率在线数据序列输入训练完成的LSTM-RNN模型,得到消除电网谐波的更新避雷器泄漏电流总畸变率在线数据序列;
将所述更新避雷器泄漏电流总畸变率在线数据序列、所述避雷器泄漏电流全电流在线数据序列和所述避雷器泄漏电流阻性电流在线数据序列输入到训练完成的随机森林模型,得到所述避雷器的状态识别预测结果。
第二方面,本申请实施例还提供了一种基于在线监测数据的避雷器运行状态识别装置,该装置包括:
获取模块,设置为获取避雷器在线监测数据,所述在线监测数据至少包括避雷器泄漏电流总畸变率在线数据序列、避雷器电压谐波总畸变率在线数据序列、避雷器泄漏电流全电流在线数据序列和避雷器泄漏电流阻性电流在线数据序列;
谐波消除模块,设置为将所述在线监测数据中避雷器泄漏电流总畸变率在线数据序列、避雷器电压谐波总畸变率在线数据序列输入训练完成的LSTM-RNN模型,得到消除电网谐波的更新避雷器泄漏电流总畸变率在线数据序列;
预测模块,设置为将所述更新避雷器泄漏电流总畸变率在线数据序列、所述避雷器泄漏电流全电流在线数据序列和所述避雷器泄漏电流阻性电流在线数据序列输入到训练完成的随机森林模型,得到所述避雷器的状态识别预测结果。
第三方面,本申请实施例还提供了一种计算机可读存储介质,其上存储有计算机程序,该程序被处理器执行时实现如本申请实施例中任一所述的基于在线监测数据的避雷器运行状态识别方法。
第四方面,本申请实施例还提供一种电子设备,包括:至少一个处理器;存储器,设置为存储至少一个程序,当所述至少一个程序被所述至少一个处理器执行,使得所述至少一个处理器实现如上所述的基于在线监测数据的避雷器运行状态识别方法。
附图说明
图1是本申请实施例一提供的一种基于在线监测数据的避雷器运行状态识别方法的流程图;
图2是本申请实施例二提供的一种基于在线监测数据的避雷器运行状态识别方法的流程图;
图3是本申请实施例二提供的运行电压下避雷器的等效电路图;
图4是本申请实施例三提供的一种基于在线监测数据的避雷器运行状态识别方法的流程图;
图5是本申请实施例三提供的避雷器阀片部分受潮等效电路图;
图6是本申请实施例三提供的避雷器受潮时泄漏电流变化示意图;
图7是本申请实施例三提供的避雷器受污秽影响时泄漏电流变化示意图;
图8是本申请实施例三提供的氧化锌避雷器伏安特性曲线示意图;
图9是本申请实施例三提供的运行电压下避雷器谐波电流总畸变率与非线性特性关系示意图;
图10是本申请实施例四提供的一种基于在线监测数据的避雷器运行状态识别装置的结构示意图;
图11是本申请实施例六提供的一种电子设备的结构示意图。
具体实施方式
下面结合附图和实施例对本申请进行说明。
本申请提供一种基于在线监测数据的避雷器运行状态识别方法、装置,以实现对避雷器运行状态进行实时监测,在避雷器劣化前期发现隐患,为供电部门制定避雷器检修试验计划提供了参考依据,提高电网的智能化水平。
实施例一
图1是本申请实施例一提供的一种基于在线监测数据的避雷器运行状态识别方法的流程图,本实施例可适用于对避雷器运行状态进行监测,该方法可以由避雷器运行状态识别装置来执行,包括如下步骤:
S110、获取避雷器在线监测数据,在线监测数据至少包括避雷器泄漏电流总畸变率在线数据序列、避雷器电压谐波总畸变率在线数据序列、避雷器泄漏电流全电流在线数据序列和避雷器泄漏电流阻性电流在线数据序列。
其中,避雷器电压谐波总畸变率可以理解为用于衡量电压波形畸变程度的参数,以各次谐波电压的均方根值与基波电压有效值之比的百分数来表示。
在理想状态下,电压波形应该是周期性标准正弦波,但由于电力系统中存在有大量非线性阻抗特性的供用电设备,这些设备向公用电网注入谐波电流或在公用电网中产生谐波电压,称为谐波源。谐波源使得实际的电压波形偏离正弦波,这种现象称为电压正弦波形畸变。
通过氧化锌电阻片的电流叫做氧化锌避雷器的泄漏电流,也被认为成避雷器的泄漏全电流。避雷器泄漏电流全电流可以反应避雷器的绝缘情况,是运行电压下判断避雷器好坏的重要手段。
避雷器的总泄漏全电流包含阻性电流(有功分量)和容性电流(无功分量), 即避雷器泄漏电流全电流包含泄漏电流阻性电流和泄漏电流容性电流。在正常运行情况下,流过避雷器的主要电流为容性电流,阻性电流只占很小一部分,约为10%~20%左右。泄漏电流中的阻性成分对非线性金属氧化物电阻的电压-电流特性变化较为敏感,因此,避雷器泄漏电流阻性电流可以作为一个良好的诊断指示,用以指示运行中的氧化锌避雷器的健康状态。
可以通过应用在生产现场的避雷器在线监测系统获取数据,该数据包括避雷器泄漏电流总畸变率在线数据序列、避雷器电压谐波总畸变率在线数据序列、避雷器泄漏电流全电流在线数据序列和避雷器泄漏电流阻性电流在线数据序列。
S120、将在线监测数据中避雷器泄漏电流总畸变率在线数据序列、避雷器电压谐波总畸变率在线数据序列输入训练完成的LSTM-RNN模型,得到消除电网谐波的更新避雷器泄漏电流总畸变率在线数据序列。
其中,长短时记忆循环神经网络(LSTM-RNN)是一种特殊的神经网络(RNN)模型。RNN模型是一种用以处理序列输入的深度学习模型,广泛用于语音识别、状态识别与机器翻译。相较于LSTM-RNN而言,RNN没有记忆功能,容易造成梯度消失与梯度爆炸现象,而LSTM-RNN引入了遗忘操作、输入操作与输出操作三个核心要素,满足了需要长时间记忆信息和遗忘信息的需求。
在实际应用中,泄漏电流的谐波主要有两个来源,一个是来自避雷器氧化锌阀片的非线性特性造成的,另一个来源于电网谐波。而随着非线性负载及换流站的增加,电网谐波的污染会逐渐加剧。电网的谐波大致为无规律的波动,但也存在一定的周期性,比如白天非线性负载接入较多,谐波含量也较大这将导致在实际测试中,工作人员无法直接利用避雷器在线监测数据的泄漏电流谐波总畸变率作为评价避雷器健康状态的指标。避雷器在线监测系统获取的数据还包括了避雷器电压谐波序列,而避雷器电压谐波含量恰恰能反映电网谐波含量。因此,采用LSTM-RNN模型来处理避雷器泄漏电流总畸变率在线数据序列、避雷器电压谐波总畸变率在线数据序列,以消除电网谐波对后续分析的影响。
S130、将更新避雷器泄漏电流总畸变率在线数据序列、避雷器泄漏电流全电流在线数据序列和避雷器泄漏电流阻性电流在线数据序列输入到训练完成的随机森林模型,得到避雷器的状态识别预测结果。
其中,在机器学习中,随机森林是一个包含多个决策树的分类器,并且其输出的类别是由个别树输出的类别的众数而定。随机森林是集成式机器学习的其中一种算法,随机森林在某种程度上可以说是决策树的优化版本,它的基本 思想是通过合成多个决策树的输出结果,以提高模型的泛化能力。
因此,将更新避雷器泄漏电流总畸变率在线数据序列、避雷器泄漏电流全电流在线数据序列和避雷器泄漏电流阻性电流在线数据序列输入到训练完成的随机森林模型,得出避雷器健康状态评价模型。然后根据这个评价模型,利用避雷器在线监测系统获取的实时数据,实现对避雷器状态的实时跟踪,进而得到避雷器的状态识别预测结果。
本实施例提供的基于在线监测数据的避雷器运行状态识别方法,通过获取避雷器在线监测数据,在线监测数据至少包括避雷器泄漏电流总畸变率在线数据序列、避雷器电压谐波总畸变率在线数据序列、避雷器泄漏电流全电流在线数据序列和避雷器泄漏电流阻性电流在线数据序列;将在线监测数据中避雷器泄漏电流总畸变率在线数据序列、避雷器电压谐波总畸变率在线数据序列输入训练完成的LSTM-RNN模型,得到消除电网谐波的更新避雷器泄漏电流总畸变率在线数据序列;将更新避雷器泄漏电流总畸变率在线数据序列、避雷器泄漏电流全电流在线数据序列和避雷器泄漏电流阻性电流在线数据序列输入到训练完成的随机森林模型,得到避雷器的状态识别预测结果。本申请实施例利用LSTM-RNN模型对避雷器泄漏电流谐波进行处理,消除了电网谐波对后续分析的影响。利用随机森林模型,选取了避雷器泄漏电流阻性电流、避雷器泄漏电流全电流、避雷器泄漏电流谐波总畸变率作为特征量,实现对避雷器运行状态的识别。该方法利用了现有的避雷器监测数据作为特征量,不需要额外增加传感器;实现了对避雷器运行状态的实时监控,在避雷器劣化前期就可以及时发现隐患,为供电部门制定避雷器检修试验计划提供参考依据,提高了电网的智能化水平与供电可靠性。
实施例二
本实施例以上述实施例为基础进行细化,图2是本申请实施例二提供的一种基于在线监测数据的避雷器运行状态识别方法的流程图,该方法包括:
S210、获取避雷器在线监测数据,在线监测数据至少包括避雷器泄漏电流总畸变率在线数据序列、避雷器电压谐波总畸变率在线数据序列、避雷器泄漏电流全电流在线数据序列和避雷器泄漏电流阻性电流在线数据序列。
S220、建立运行电压下避雷器的等效电路仿真模型,并根据避雷器的等效电路仿真模型得到的仿真数据进行模型训练,得到训练完成的LSTM-RNN模型。
在一示例中,建立运行电压下避雷器的等效电路仿真模型,并根据避雷器的等效电路仿真模型得到的仿真数据进行模型训练,得到训练完成的 LSTM-RNN模型,包括:
建立运行电压下避雷器的等效电路仿真模型,避雷器的等效电路仿真模型包括电压源和串联在电压源一侧的多个阻容单元,阻容单元包括串联的固定电阻和固定电容,避雷器的等效电路仿真模型多个非线性电阻,每个非线性电阻与一固定电容并联连接;
设置电压源的频率为电网标准频率,并随机设置避雷器的非线性参数以使非线性电阻的阻值随机,获取避雷器泄漏电流谐波总畸变率第一仿真数据序列;
随机设置电压源的谐波幅值,获取避雷器泄漏电流谐波总畸变率第二仿真数据序列,并获取避雷器电压谐波总畸变率仿真数据序列;
搭建LSTM-RNN初始模型,将所述避雷器泄漏电流谐波总畸变率第二仿真数据序列和所述避雷器电压谐波总畸变率仿真数据序列作为所述LSTM-RNN初始模型的输入,将所述避雷器泄漏电流谐波总畸变率第一仿真数据序列作为所述LSTM-RNN初始模型的输出进行模型训练,得到训练完成的所述LSTM-RNN模型。
氧化锌避雷器是一种用来保护电力设备的装置,主要由腔体、瓷套、一次端子以及多个氧化锌阀片堆叠组成。图3是本申请实施例二提供的运行电压下避雷器的等效电路图,参考图3,避雷器的等效电路仿真模型包括电压源和串联在电压源一侧的多个阻容单元,阻容单元包括串联的固定电阻和固定电容,避雷器的等效电路仿真模型多个非线性电阻,每个非线性电阻与一个固定电容并联连接。其中R n表示第n个氧化锌阀片之间的接触电阻(其中,n>1时,第n个氧化锌阀片之间的接触电阻是第n个氧化锌阀片与第n-1个氧化锌阀片之间的接触电阻;其中n=1时,第n个氧化锌阀片之间的接触电阻是第n个氧化锌阀片与电压源之间的接触电阻),R nl1用来表示第1个阀片的非线性特性电阻,R nl2用来表示第2个阀片的非线性特性电阻,R nln用来表示第n个阀片的非线性特性电阻。C 1表示第1个阀片的等效电容,C 2表示第2个阀片的等效电容,C n表示第n个阀片的等效电容。C 0表示避雷器腔体、瓷套与空气的等效电容,R 0表示避雷器腔体、瓷套与空气的等效电阻。氧化锌避雷器电阻具有非线性特性,在正常电压下,流过避雷器的电流极小。而在过电压下,氧化锌避雷器电阻迅速减小,能量得到快速释放,起到了保护电力设备的作用。
示例性的,泄漏电流谐波的预测步骤为:
步骤一:采用图3所示模型,设置电压源频率为电网标准频率50Hz,随机设置避雷器的非线性参数(K i,a i)以使非线性电阻的阻值随机,获取N个时刻的 避雷器泄漏电流谐波总畸变率,并把这N个数据输出记为第一仿真数据序列X 1
步骤二:在步骤一的基础上,在避雷器高压端增加电压源,模拟电网所含的谐波,并随机设置电压源的谐波幅值,仿真得到N个时刻的数据。可以通过设置一个随机数发生器,使得每个时刻的电压谐波都随机波动,模拟电网谐波的波动。
步骤三:获取步骤二所仿真得到的N个时刻的避雷器泄漏电流总畸变率,并把这N个数据记为第二仿真数据序列X 2;同时获取N个时刻的避雷器电压谐波总畸变率,并把这N个数据记为避雷器电压谐波总畸变率仿真数据序列Y 1
步骤四:利用机器学习算法的工具包Tensorflow在软件PYCHARM上进行运行,从而搭建LSTM-RNN模型,把上述避雷器泄漏电流谐波总畸变率第二仿真数据序列X 2和避雷器电压谐波总畸变率仿真数据序列Y 1作为LSTM-RNN初始模型的输入数据,避雷器泄漏电流谐波总畸变率第一仿真数据序列X 1作为LSTM-RNN初始模型的输出数据,进行模型训练并保存训练好的模型。
S230、将在线监测数据中避雷器泄漏电流总畸变率在线数据序列、避雷器电压谐波总畸变率在线数据序列输入训练完成的LSTM-RNN模型,得到消除电网谐波的更新避雷器泄漏电流总畸变率在线数据序列。
S240、将更新避雷器泄漏电流总畸变率在线数据序列、避雷器泄漏电流全电流在线数据序列和避雷器泄漏电流阻性电流在线数据序列输入到训练完成的随机森林模型,得到避雷器的状态识别预测结果。
本实施例,通过建立运行电压下避雷器的等效电路仿真模型,并根据避雷器的等效电路仿真模型得到的仿真数据进行模型训练,得到训练完成的LSTM-RNN模型;利用LSTM-RNN模型来处理避雷器泄漏电流谐波序列,以消除电网谐波的影响,实现对避雷器泄漏电流谐波的预测。
实施例三
本实施例以上述实施例为基础进行细化,图4是本申请实施例三提供的一种基于在线监测数据的避雷器运行状态识别方法的流程图,参考图4,方法包括:
S310、获取避雷器在线监测数据,在线监测数据至少包括避雷器泄漏电流总畸变率在线数据序列、避雷器电压谐波总畸变率在线数据序列、避雷器泄漏电流全电流在线数据序列和避雷器泄漏电流阻性电流在线数据序列。
S320、将在线监测数据中避雷器泄漏电流总畸变率在线数据序列、避雷器电压谐波总畸变率在线数据序列输入训练完成的LSTM-RNN模型,得到消除电网谐波的更新避雷器泄漏电流总畸变率在线数据序列。
S330、将建立运行电压下避雷器的等效电路仿真模型;
改变避雷器的等效电路仿真模型的参数,分别仿真得到避雷器状态为正常、劣化、受潮、污秽时输出的特征值集合,特征值集合中包括避雷器泄漏电流全电流仿真数据序列和避雷器泄漏电流阻性电流仿真数据序列和避雷器泄漏电流谐波总畸变率第三仿真数据序列,并把特征值集合随机森林模型的训练样本集;
采用训练样本集进行随机森林模型训练,得到包括多个决策树的训练完成的随机森林模型。
长时间处于生产现场运行的避雷器难免会受到外界环境的干扰,避雷器劣化指的是氧化锌阀片的劣化,氧化锌阀片发生劣化是一个缓慢的过程。在劣化初期仅表现为避雷器泄漏电流谐波含量的增加,而对避雷器泄漏电流阻性电流,避雷器泄漏电流容性电流的影响很微小。
避雷器密封不良是造成受潮的主要原因,氧化锌阀片烘干不彻底,密封垫圈安放位置不当甚至没有安装,使用材料不合格也会造成水分渗入。阀片受潮后,导致避雷器泄漏电流全电流以及避雷器泄漏电流阻性电流明显增大。
避雷器瓷套表面污秽对避雷器的影响较大。避雷器在运行过程中,外套污秽则可能导致内部绝缘杆已变为棕黄色,芯体上部甚至附着油状物,芯体绝缘出现明显老化电阻片侧面绝缘釉也呈磨砂状,失去光泽。其长期运行在比正常运行温度稍高的温度下,导致电阻片劣化。
首先建立运行电压下避雷器的等效电路仿真模型,改变避雷器的等效电路仿真模型的参数,分别仿真得到避雷器状态为正常、劣化、受潮、污秽时输出的特征值集合。
在一实施例中,避雷器的等效电路仿真模型包括电压源和串联在电压源一侧的多个阻容单元,阻容单元包括串联的固定电阻和固定电容,避雷器的等效电路仿真模型包括多个非线性电阻,每个非线性电阻与一固定电容并联连接;避雷器的等效电路仿真模型还包括避雷器腔体、瓷套与空气的等效电阻与等效电容,其中等效电阻与串联的多个阻容单元并联连接,等效电容与等效电阻并联连接。
改变避雷器的等效电路仿真模型的参数,分别仿真得到避雷器状态为正常、劣化、受潮、污秽时输出的特征值集合,特征值集合中包括避雷器泄漏电流全电流仿真数据序列和避雷器泄漏电流阻性电流仿真数据序列和避雷器泄漏电流谐波总畸变率第三仿真数据序列,并把特征值集合随机森林模型的训练样本集,包括:
(1)改变避雷器的非线性参数,以仿真得到避雷器状态为劣化时的输出特征值集合。
(2)在至少一个非线性电阻的两端并联旁路电阻,改变旁路电阻的阻值,以仿真得到避雷器状态为受潮时的输出特征值集合。
(3)改变等效电阻的阻值,以仿真得到避雷器状态为污秽时的输出特征值结合。
图5是本申请实施例三提供的避雷器阀片部分受潮等效电路图,参考图5,避雷器的等效电路仿真模型包括电压源和串联在电压源一侧的多个阻容单元,阻容单元包括串联的固定电阻和固定电容,避雷器的等效电路仿真模型包括多个非线性电阻,每个非线性电阻与一固定电容并联连接;避雷器的等效电路仿真模型还包括避雷器腔体、瓷套与空气的等效电阻与等效电容,其中等效电阻与串联的多个阻容单元并联连接,等效电容与等效电阻并联连接。其中R n表示第n个氧化锌阀片之间的接触电阻(其中,n>1时,第n个氧化锌阀片之间的接触电阻是第n个氧化锌阀片与第n-1个氧化锌阀片之间的接触电阻;其中n=1时,第n个氧化锌阀片之间的接触电阻是第n个氧化锌阀片与电压源之间的接触电阻),R nl1用来表示第1个阀片的非线性特性电阻,R nl2用来表示第2个阀片的非线性特性电阻,R nln用来表示第n个阀片的非线性特性电阻。C 1表示第1个阀片的等效电容,C 2表示第2个阀片的等效电容,C n表示第n个阀片的等效电容。C 0表示避雷器腔体、瓷套与空气的等效电容,R 0表示避雷器腔体、瓷套与空气的等效电阻。
避雷器阀片受潮将导致阀片间形成泄漏旁路,增大泄漏电流,但对避雷器非线性特性的影响不大。当避雷器阀片部分受潮时,图3所示的等效模型变为图5,具体改变是在避雷器阀片等效电路上并连一个旁路电阻R s,模拟水汽的泄漏旁路。通过改变旁路电阻R s的值,获取运行电压下氧化锌阀片受潮时的避雷器泄漏电流全电流与避雷器泄漏电流阻性电流。
图6是本申请实施例三提供的避雷器受潮时泄漏电流变化示意图,参考图6,图6的纵轴表示电流,单位是uA,横轴表示电阻,单位是MΩ,从图6中可以看出:随着受潮程度的增加,避雷器泄漏电流全电流逐渐增加,并且避雷器泄漏电流阻性电流的增速明显比避雷器泄漏电流全电流快。
当避雷器瓷套表面污秽严重时,图5所示避雷器的等效电路仿真模型中的等效电阻R 0将减小,导致避雷器泄漏电流阻性电流增大。图7是本申请实施例三提供的避雷器受污秽影响时泄漏电流变化示意图,参考图7,图7中的纵轴表 示电流,单位是uA,横轴表示电阻,单位是GΩ,从图7中可以看出:避雷器泄漏电流阻性电流的增速明显快于避雷器泄漏电流全电流。
通过改变避雷器的非线性参数,以仿真得到避雷器状态为劣化时的输出特征值集合;在至少一个非线性电阻的两端并联旁路电阻R s,改变旁路电阻R s的阻值,以仿真得到避雷器状态为受潮时的输出特征值集合;改变等效电阻R 0的阻值,以仿真得到避雷器状态为污秽时的输出特征值结合。
在一实施例中,避雷器的非线性特性由以下公式表示:
Figure PCTCN2022131878-appb-000001
I为流过避雷器阀片电流,U为避雷器阀片端电压,k i为决定避雷器阀片非线性特性的第一参数,a i为决定避雷器阀片非线性特性第二参数。
避雷器阀片非线性特性的第一参数与阀片材料有关。避雷器阀片的电阻随端电压的变化而变化,因此,描述其非线性特性的函数为分段函数。图8是本申请实施例三提供的氧化锌避雷器伏安特性曲线示意图,参考图8,图中的纵坐标表示电流I、横坐标表示电压U。从图中可以看出,初始阶段处于线性区,电压达到参考电压Vref以后处于非线性区。其中k i对应于线性区与非线性区分界点处的第一参数,a i对应于线性区与非线性区分界点处的第二参数。氧化锌阀片发生劣化是一个缓慢的过程,在劣化初期仅表现为泄漏电流谐波含量的增加,而对阻性电流基波,容性电容的影响很微小。因此,可通过获取氧化锌避雷器运行电压下的泄漏电流,并对其成分加以分析实现对阀片劣化的识别。
图9是本申请实施例三提供的运行电压下避雷器谐波电流总畸变率与非线性特性关系示意图,参考图9,图中的纵轴表示避雷器阀片非线性特性的第一参数K i、横轴表示避雷器阀片非线性特性第二参数a i、竖轴表示避雷器谐波电流总畸变率。通过Simulink构建避雷器等效电路,通过改变氧化锌阀片非线性特性,获得与之相对应的避雷器泄漏电流谐波总畸变率。根据避雷器非线性特性公式
Figure PCTCN2022131878-appb-000002
的计算结果得出结论:避雷器泄漏电流总畸率随K i值的变小而变大,随a i值的变大而变小。其中,K i值的变化对泄漏电流总畸率的影响较敏感。
在一实施例中,采用训练样本集进行随机森林模型训练,得到包括多个决策树的训练完成的随机森林模型,包括:
步骤一:采用有放回的抽取方式在训练样本集中的N个训练样本中抽取n个训练样本作为随机森林模型中决策树的根节点样本来训练决策树,其中n<N。
其中,训练样本集中训练样本个数N以及抽取的训练样本个数n可以根据 实际需要进行设置,本申请实施例对个数并不进行限制。
步骤二:进行在每个训练样本的M个属性中随机选择m个属性,根据最大信息熵选择选取决策树的子节点的分裂属性,其中m<M。
需要说明的是属性是指避雷器泄漏电流全电流仿真数据序列和避雷器泄漏电流阻性电流仿真数据序列和避雷器泄漏电流谐波总畸变率第三仿真数据序列;最大信息熵是指决策树的划分采用最大信息增益原则,信息增益详见下述公式(7)。
步骤三:重复进行步骤二直至决策树的所有子节点不能再分裂,输出一个决策树模型。
步骤四:重复进行步骤一至步骤三,输出多个决策树模型构成训练完成的随机森林模型。
本实施例中,选取的特征集合均为连续值,因此,采用决策树算法训练数据,决策树的划分采用最大信息增益原则,信息增益可用公式(1)定义。
Figure PCTCN2022131878-appb-000003
式中,D表示所有训练样本,a表示特征集合,特征集合可以为避雷器泄漏电流全电流仿真数据序列和避雷器泄漏电流阻性电流仿真数据序列和避雷器泄漏电流谐波总畸变率第三仿真数据序列,V表示特征a可能的取值集合。Ent(D)为信息熵的定义,如公式(2)所示:
Figure PCTCN2022131878-appb-000004
式中,y为决策树的输出类别,输出类别分别为正常、受潮、污秽、劣化四类,P k为第k个类别所对应的输出概率。
S340、将更新避雷器泄漏电流总畸变率在线数据序列、避雷器泄漏电流全电流在线数据序列和避雷器泄漏电流阻性电流在线数据序列输入到训练完成的随机森林模型,得到避雷器的状态识别预测结果。
在一实施例中,根据随机森林模型中输出相同状态的决策树模型的个数和随机森林模型所包括的决策树模型的总个数确定避雷器每个状态的概率。
随机森林的结合策略采用投票法,最后的输出结果为每个类别所对应的概率如公式(3)所示:
Figure PCTCN2022131878-appb-000005
Figure PCTCN2022131878-appb-000006
式中P k为第k个类别所对应的输出概率,M为决策树个数,m k为输出结果为第k个类别的决策树个数。
其中,决策数模型的个数作为参数可以根据实际需要进行设置,完成决策树模型个数的设置后,对数据集进行训练,训练好之后,结合随机森林模型的总个数,根据每个决策树相对应的预测结果进行统计,可确定避雷器每个状态的概率。示例性的,可以设置100个决策树,100个决策树都会得到100个预测结果,比如预测结果是受潮的有20个,正常的40个,劣化的有10个、污秽的有30个、总数是100个,这样可以求得避雷器受潮的概率为20%,避雷器正常的概率为40%,避雷器劣化的概率为10%,避雷器污秽的概率为30%。
本实施例通过仿真获取避雷器劣化、受潮或表面污秽时各特征量的变化情况,运用随机森林模型对数据进行训练,得出避雷器健康状态评价模型。根据该评价模型,利用避雷器在线监测的实时数据,实现对避雷器状态的实时跟踪。
实施例四
图10是本申请实施例四提供的一种基于在线监测数据的避雷器运行状态识别装置的结构示意图,本实施例对应上述方法实施例,该装置400包括:获取模块401、谐波消除模块402和预测模块403。
获取模块401设置为获取避雷器在线监测数据,在线监测数据至少包括避雷器泄漏电流总畸变率在线数据序列、避雷器电压谐波总畸变率在线数据序列、避雷器泄漏电流全电流在线数据序列和避雷器泄漏电流阻性电流在线数据序列。
谐波消除模块402设置为将在线监测数据中避雷器泄漏电流总畸变率在线数据序列、避雷器电压谐波总畸变率在线数据序列输入训练完成的LSTM-RNN模型,得到消除电网谐波的更新避雷器泄漏电流总畸变率在线数据序列。
预测模块403设置为将更新避雷器泄漏电流总畸变率在线数据序列、避雷器泄漏电流全电流在线数据序列和避雷器泄漏电流阻性电流在线数据序列输入到训练完成的随机森林模型,得到避雷器的状态识别预测结果。
本申请实施例所提供的基于在线监测数据的避雷器运行状态识别装置可执行本申请任意实施例所提供的基于在线监测数据的避雷器运行状态识别方法,具备执行方法相应的功能模块和有益效果。
值得注意的是,上述避雷器运行状态识别确定装置的实施例中,所包括的各个模块只是按照功能逻辑进行划分的,但并不局限于上述的划分,只要能够实现相应的功能即可;另外,各模块的具体名称也只是为了便于相互区分,并 不用于限制本申请的保护范围。
实施例五
本申请实施例五还提供了一种计算机可读存储介质,其上存储有计算机程序,该程序被处理器执行时实现本申请任意实施例所提供的一种基于在线监测数据的避雷器运行状态识别方法。
存储介质可以是非暂态(non-transitory)存储介质。
本申请实施例的计算机存储介质,可以采用一个或多个计算机可读的介质的任意组合。计算机可读介质可以是计算机可读信号介质或者计算机可读存储介质。计算机可读存储介质例如可以是——但不限于——电、磁、光、电磁、红外线、或半导体的系统、装置或器件,或者任意以上的组合。计算机可读存储介质的例子(非穷举的列表)包括:具有一个或多个导线的电连接、便携式计算机磁盘、硬盘、随机存取存储器(RAM)、只读存储器(Read-Only Memory,ROM)、可擦式可编程只读存储器(Erasable Programmable Read-Only Memory,EPROM)或闪存、光纤、便携式紧凑磁盘只读存储器(Compact Disc Read-Only Memory,CD-ROM)、光存储器件、磁存储器件、或者上述的任意合适的组合。在本文件中,计算机可读存储介质可以是任何包含或存储程序的有形介质,该程序可以被指令执行系统、装置或者器件使用或者与其结合使用。
计算机可读的信号介质可以包括在基带中或者作为载波一部分传播的数据信号,其中承载了计算机可读的程序代码。这种传播的数据信号可以采用多种形式,包括但不限于电磁信号、光信号或上述的任意合适的组合。计算机可读的信号介质还可以是计算机可读存储介质以外的任何计算机可读介质,该计算机可读介质可以发送、传播或者传输用于由指令执行系统、装置或者器件使用或者与其结合使用的程序。
计算机可读介质上包含的程序代码可以用任何适当的介质传输,包括——但不限于无线、电线、光缆、RF等等,或者上述的任意合适的组合。
可以以一种或多种程序设计语言或其组合来编写用于执行本申请操作的计算机程序代码,所述程序设计语言包括面向对象的程序设计语言—诸如Java、Smalltalk、C++、Ruby、Go,还包括常规的过程式程序设计语言—诸如“C”语言或类似的程序设计语言,以及AI算法的计算机语言。程序代码可以完全地在用户计算机上执行、部分地在用户计算机上执行、作为一个独立的软件包执行、部分在用户计算机上部分在远程计算机上执行、或者完全在远程计算机或服务器上执行。在涉及远程计算机的情形中,远程计算机可以通过任意种类的网络 ——包括局域网(Local Area Network,LAN)或广域网(Wide Area Network,WAN)—连接到用户计算机,或者,可以连接到外部计算机(例如利用因特网服务提供商来通过因特网连接)。
实施例六
图11是本申请实施例六提供的一种电子设备的结构示意图,如图11所示,该电子设备包括:一个或多个处理器110和存储器120。图11中以一个处理器110为例。
所述电子设备还可以包括:输入装置130和输出装置140。
可以理解地,所述电子设备也可以不包括,输入装置130和输出装置140。
所述电子设备中的处理器110、存储器120、输入装置130和输出装置140可以通过总线或者其他方式连接,图11中以通过总线连接为例。
存储器120作为一种计算机可读存储介质,可设置为存储软件程序、计算机可执行程序以及模块。处理器110通过运行存储在存储器120中的软件程序、指令以及模块,从而执行多种功能应用以及数据处理,以实现上述实施例中的任意一种方法。
存储器120可以包括存储程序区和存储数据区,其中,存储程序区可存储操作系统、至少一个功能所需要的应用程序;存储数据区可存储根据电子设备的使用所创建的数据等。此外,存储器可以包括随机存取存储器(Random Access Memory,RAM)等易失性存储器,还可以包括非易失性存储器,例如至少一个磁盘存储器件、闪存器件或者其他非暂态固态存储器件。
存储器120可以是非暂态计算机存储介质或暂态计算机存储介质。该非暂态计算机存储介质,例如至少一个磁盘存储器件、闪存器件或其他非易失性固态存储器件。在一些实施例中,存储器120可选包括相对于处理器110远程设置的存储器,这些远程存储器可以通过网络连接至电子设备。上述网络的实例可以包括互联网、企业内部网、局域网、移动通信网及其组合。
输入装置130可设置为接收输入的数字或字符信息,以及产生与电子设备的用户设置以及功能控制有关的键信号输入。输出装置140可包括显示屏等显示设备。
本领域技术人员会理解,本申请不限于这里所述的特定实施例,对本领域技术人员来说能够进行各种明显的变化、重新调整和替代而不会脱离本申请的保护范围。

Claims (11)

  1. 一种基于在线监测数据的避雷器运行状态识别方法,包括:
    获取避雷器在线监测数据,所述在线监测数据至少包括避雷器泄漏电流总畸变率在线数据序列、避雷器电压谐波总畸变率在线数据序列、避雷器泄漏电流全电流在线数据序列和避雷器泄漏电流阻性电流在线数据序列;
    将所述在线监测数据中避雷器泄漏电流总畸变率在线数据序列、避雷器电压谐波总畸变率在线数据序列输入训练完成的LSTM-RNN模型,得到消除电网谐波的更新避雷器泄漏电流总畸变率在线数据序列;
    将所述更新避雷器泄漏电流总畸变率在线数据序列、所述避雷器泄漏电流全电流在线数据序列和所述避雷器泄漏电流阻性电流在线数据序列输入到训练完成的随机森林模型,得到所述避雷器的状态识别预测结果。
  2. 根据权利要求1所述的基于在线监测数据的避雷器运行状态识别方法,在将所述在线监测数据中避雷器泄漏电流总畸变率在线数据序列、避雷器电压谐波总畸变率在线数据序列输入训练完成的LSTM-RNN模型,得到消除电网谐波的更新避雷器泄漏电流总畸变率在线数据序列之前,还包括:
    建立运行电压下所述避雷器的等效电路仿真模型,并根据所述避雷器的等效电路仿真模型得到的仿真数据进行模型训练,得到训练完成的LSTM-RNN模型。
  3. 根据权利要求2所述的基于在线监测数据的避雷器运行状态识别方法,其中,所述建立运行电压下所述避雷器的等效电路仿真模型,并根据所述避雷器的等效电路仿真模型得到的仿真数据进行模型训练,得到训练完成的LSTM-RNN模型,包括:
    建立运行电压下所述避雷器的等效电路仿真模型,所述避雷器的等效电路仿真模型包括电压源和串联在所述电压源一侧的多个阻容单元,所述阻容单元包括串联的固定电阻和固定电容,所述避雷器的等效电路仿真模型多个非线性电阻,每个所述非线性电阻与一所述固定电容并联连接;
    设置电压源的频率为电网标准频率,并随机设置避雷器的非线性参数以使所述非线性电阻的阻值随机,获取避雷器泄漏电流谐波总畸变率第一仿真数据序列;
    随机设置所述电压源的谐波幅值,获取避雷器泄漏电流谐波总畸变率第二仿真数据序列,并获取避雷器电压谐波总畸变率仿真数据序列;
    搭建LSTM-RNN初始模型,将所述避雷器泄漏电流谐波总畸变率第二仿真数据序列和所述避雷器电压谐波总畸变率仿真数据序列作为所述LSTM-RNN初 始模型的输入,将所述避雷器泄漏电流谐波总畸变率第一仿真数据序列作为所述LSTM-RNN初始模型的输出进行模型训练,得到训练完成的所述LSTM-RNN模型。
  4. 根据权利要求1所述的基于在线监测数据的避雷器运行状态识别方法,在所述将所述更新避雷器泄漏电流总畸变率在线数据序列、所述避雷器泄漏电流全电流在线数据序列和所述避雷器泄漏电流阻性电流在线数据序列输入到训练完成的随机森林模型,得到所述避雷器的状态识别预测结果之前,还包括:
    建立运行电压下所述避雷器的等效电路仿真模型;
    改变所述避雷器的等效电路仿真模型的参数,分别仿真得到避雷器状态为正常、劣化、受潮、污秽时输出的特征值集合,所述特征值集合中包括避雷器泄漏电流全电流仿真数据序列和所述避雷器泄漏电流阻性电流仿真数据序列和避雷器泄漏电流谐波总畸变率第三仿真数据序列,并把所述特征值集合所述随机森林模型的训练样本集;
    采用所述训练样本集进行随机森林模型训练,得到包括多个决策树的训练完成的随机森林模型。
  5. 根据权利要求4所述的基于在线监测数据的避雷器运行状态识别方法,其中,所述采用所述训练样本集进行随机森林模型训练,得到包括多个决策树的训练完成的随机森林模型,包括:
    步骤一:采用有放回的抽取方式在所述训练样本集中的N个训练样本中抽取n个训练样本作为所述随机森林模型中决策树的根节点样本来训练决策树,其中n<N;
    步骤二:进行在每个所述训练样本的M个属性中随机选择m个属性,根据最大信息熵选择选取所述决策树的子节点的分裂属性,其中m<M;
    步骤三:重复进行所述步骤二直至所述决策树的所有子节点不能再分裂,输出一个决策树模型;
    步骤四:重复进行所述步骤一至所述步骤三,输出多个决策树模型构成训练完成的随机森林模型。
  6. 根据权利要求4所述的基于在线监测数据的避雷器运行状态识别方法,其中,所述避雷器的等效电路仿真模型包括电压源和串联在所述电压源一侧的多个阻容单元,所述阻容单元包括串联的固定电阻和固定电容,所述避雷器的等效电路仿真模型包括多个非线性电阻,每个所述非线性电阻与一所述固定电容并联连接;所述避雷器的等效电路仿真模型还包括避雷器腔体、瓷套与空气 的等效电阻与等效电容,其中所述等效电阻与串联的多个阻容单元并联连接,所述等效电容与所述等效电阻并联连接;
    所述改变所述避雷器的等效电路仿真模型的参数,分别仿真得到避雷器状态为正常、劣化、受潮、污秽时输出的特征值集合,所述特征值集合中包括避雷器泄漏电流全电流仿真数据序列和所述避雷器泄漏电流阻性电流仿真数据序列和避雷器泄漏电流谐波总畸变率第三仿真数据序列,并把所述特征值集合所述随机森林模型的训练样本集,包括:
    改变所述避雷器的非线性参数,以仿真得到避雷器状态为劣化时的输出特征值集合;
    在至少一个所述非线性电阻的两端并联旁路电阻,改变所述旁路电阻的阻值,以仿真得到避雷器状态为受潮时的输出特征值集合;
    改变所述等效电阻的阻值,以仿真得到所述避雷器状态为污秽时的输出特征值结合。
  7. 根据权利要求3或6所述的基于在线监测数据的避雷器运行状态识别方法,其中,所述避雷器的非线性特性由以下公式表示:
    Figure PCTCN2022131878-appb-100001
    I为流过避雷器阀片电流,U为避雷器阀片端电压,k i为决定避雷器阀片非线性特性的第一参数,a i为决定避雷器阀片非线性特性第二参数。
  8. 根据权利要求4所述的基于在线监测数据的避雷器运行状态识别方法,其中,将所述更新避雷器泄漏电流总畸变率在线数据序列、所述避雷器泄漏电流全电流在线数据序列和所述避雷器泄漏电流阻性电流在线数据序列输入到训练完成的随机森林模型,得到所述避雷器的状态识别预测结果,包括:
    根据所述随机森林模型中输出相同状态的决策树模型的个数和随机森林模型所包括的决策树模型的总个数确定所述避雷器每个状态的概率。
  9. 一种基于在线监测数据的避雷器运行状态识别装置,包括:
    获取模块,设置为获取避雷器在线监测数据,所述在线监测数据至少包括避雷器泄漏电流总畸变率在线数据序列、避雷器电压谐波总畸变率在线数据序列、避雷器泄漏电流全电流在线数据序列和避雷器泄漏电流阻性电流在线数据序列;
    谐波消除模块,设置为将所述在线监测数据中避雷器泄漏电流总畸变率在线数据序列、避雷器电压谐波总畸变率在线数据序列输入训练完成的LSTM-RNN模型,得到消除电网谐波的更新避雷器泄漏电流总畸变率在线数据 序列;
    预测模块,设置为将所述更新避雷器泄漏电流总畸变率在线数据序列、所述避雷器泄漏电流全电流在线数据序列和所述避雷器泄漏电流阻性电流在线数据序列输入到训练完成的随机森林模型,得到所述避雷器的状态识别预测结果。
  10. 一种计算机可读存储介质,所述计算机可读存储介质上存储有计算机程序,所述计算机程序被处理器执行时实现如权利要求1-8中任一所述的基于在线监测数据的避雷器运行状态识别方法。
  11. 一种电子设备,包括:
    至少一个处理器;
    存储器,设置为存储至少一个程序,
    当所述至少一个程序被所述至少一个处理器执行,使得所述至少一个处理器实现如权利要求1-8中任一所述的基于在线监测数据的避雷器运行状态识别方法。
PCT/CN2022/131878 2021-11-15 2022-11-15 基于在线监测数据的避雷器运行状态识别方法、装置 WO2023083366A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202111344700.9A CN113792495B (zh) 2021-11-15 2021-11-15 一种基于在线监测数据的避雷器运行状态识别方法、装置
CN202111344700.9 2021-11-15

Publications (1)

Publication Number Publication Date
WO2023083366A1 true WO2023083366A1 (zh) 2023-05-19

Family

ID=78955192

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/131878 WO2023083366A1 (zh) 2021-11-15 2022-11-15 基于在线监测数据的避雷器运行状态识别方法、装置

Country Status (2)

Country Link
CN (1) CN113792495B (zh)
WO (1) WO2023083366A1 (zh)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116399405A (zh) * 2023-06-07 2023-07-07 国网上海市电力公司 一种基于多模态融合感知的绝缘子串状态诊断方法和系统
CN116930667A (zh) * 2023-09-15 2023-10-24 广东电网有限责任公司 一种台区电网边缘测试方法、装置、设备及存储介质
CN117332410A (zh) * 2023-10-25 2024-01-02 北京航空航天大学 一种基于信息熵特征的电磁泄漏红黑信号辨识方法
CN117407675A (zh) * 2023-10-26 2024-01-16 国网青海省电力公司海北供电公司 基于多变量重构联合动态权重的避雷器泄漏电流预测方法
CN117435947A (zh) * 2023-12-20 2024-01-23 山东和兑智能科技有限公司 一种避雷器状态监测系统和方法

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113792495B (zh) * 2021-11-15 2022-04-01 广东电网有限责任公司揭阳供电局 一种基于在线监测数据的避雷器运行状态识别方法、装置
CN116910668B (zh) * 2023-09-11 2024-04-02 国网浙江省电力有限公司余姚市供电公司 一种避雷器故障预警方法、装置、设备及存储介质

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103207336A (zh) * 2013-03-22 2013-07-17 中国电力科学研究院 一种避雷器运行模拟仿真系统
US20150331967A1 (en) * 2014-05-13 2015-11-19 Lsis Co., Ltd. Apparatus and method for verifying lightning arrester of high voltage direct current transmission system
CN111707971A (zh) * 2020-06-30 2020-09-25 国网冀北电力有限公司唐山供电公司 一种避雷器绝缘状态检测方法
CN112016246A (zh) * 2020-08-26 2020-12-01 国网湖南省电力有限公司长沙市望城区供电分公司 基于防雷效率预测的电磁泄流型避雷器分布优化方法
CN113792495A (zh) * 2021-11-15 2021-12-14 广东电网有限责任公司揭阳供电局 一种基于在线监测数据的避雷器运行状态识别方法、装置

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106650037A (zh) * 2016-11-30 2017-05-10 国网江苏省电力公司盐城供电公司 一种基于支持向量机回归的避雷器状态诊断方法
CN107367673B (zh) * 2017-08-16 2019-06-04 上海电力学院 一种避雷器阀片电阻运行状态诊断方法
CN110826750B (zh) * 2018-08-08 2023-09-26 阿里巴巴集团控股有限公司 一种电力负荷预测方法、装置、设备及系统
CN109444637A (zh) * 2018-11-20 2019-03-08 武汉大学 一种考虑谐波影响的避雷器阻性电流计算方法
US11777303B2 (en) * 2019-03-11 2023-10-03 Hitachi Energy Switzerland Ag Leakage current based remote monitoring device and method for disconnector devices
CN113255147B (zh) * 2021-06-04 2021-11-02 广东电网有限责任公司 Cvt电容量在线监测方法、装置、终端设备及存储介质

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103207336A (zh) * 2013-03-22 2013-07-17 中国电力科学研究院 一种避雷器运行模拟仿真系统
US20150331967A1 (en) * 2014-05-13 2015-11-19 Lsis Co., Ltd. Apparatus and method for verifying lightning arrester of high voltage direct current transmission system
CN111707971A (zh) * 2020-06-30 2020-09-25 国网冀北电力有限公司唐山供电公司 一种避雷器绝缘状态检测方法
CN112016246A (zh) * 2020-08-26 2020-12-01 国网湖南省电力有限公司长沙市望城区供电分公司 基于防雷效率预测的电磁泄流型避雷器分布优化方法
CN113792495A (zh) * 2021-11-15 2021-12-14 广东电网有限责任公司揭阳供电局 一种基于在线监测数据的避雷器运行状态识别方法、装置

Cited By (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116399405A (zh) * 2023-06-07 2023-07-07 国网上海市电力公司 一种基于多模态融合感知的绝缘子串状态诊断方法和系统
CN116399405B (zh) * 2023-06-07 2023-09-15 国网上海市电力公司 一种基于多模态融合感知的绝缘子串状态诊断方法和系统
CN116930667A (zh) * 2023-09-15 2023-10-24 广东电网有限责任公司 一种台区电网边缘测试方法、装置、设备及存储介质
CN117332410A (zh) * 2023-10-25 2024-01-02 北京航空航天大学 一种基于信息熵特征的电磁泄漏红黑信号辨识方法
CN117332410B (zh) * 2023-10-25 2024-04-12 北京航空航天大学 一种基于信息熵特征的电磁泄漏红黑信号辨识方法
CN117407675A (zh) * 2023-10-26 2024-01-16 国网青海省电力公司海北供电公司 基于多变量重构联合动态权重的避雷器泄漏电流预测方法
CN117435947A (zh) * 2023-12-20 2024-01-23 山东和兑智能科技有限公司 一种避雷器状态监测系统和方法

Also Published As

Publication number Publication date
CN113792495A (zh) 2021-12-14
CN113792495B (zh) 2022-04-01

Similar Documents

Publication Publication Date Title
WO2023083366A1 (zh) 基于在线监测数据的避雷器运行状态识别方法、装置
Frizzo Stefenon et al. Fault diagnosis of insulators from ultrasound detection using neural networks
CN112255484B (zh) 一种避雷器运行状态在线监测及评估方法和系统
CN111123048A (zh) 一种基于卷积神经网络的串联故障电弧检测装置及方法
CN106650037A (zh) 一种基于支持向量机回归的避雷器状态诊断方法
CN102221651A (zh) 一种矿用隔爆型干式变压器故障在线诊断及预警方法
Bahador et al. Modelling and detection of live tree‐related high impedance fault in distribution systems
Wang et al. Application of extension neural network algorithm and chaos synchronization detection method to partial discharge diagnosis of power capacitor
CN116008731B (zh) 配电网高阻故障辨识方法、装置和电子设备
CN117748507B (zh) 基于高斯回归模型的配网谐波接入不确定性评估方法
Izadi et al. A synchronized lissajous-based method to detect and classify events in synchro-waveform measurements in power distribution networks
CN113850330A (zh) 一种基于短时傅里叶变换和卷积神经网络的配电网故障原因检测方法
CN107153915A (zh) 一种变电站不变化遥测数据自动检测的方法和系统
Zhang et al. Research on the application of Decision Tree and Random Forest Algorithm in the main transformer fault evaluation
Palangar et al. Detection of critical conditions in ceramic insulators based on harmonic analysis of leakage current
CN115902700A (zh) 输电线路绝缘子的污闪预警方法、装置、设备及介质
CN113255203B (zh) 基于anfis的电气线路老化程度在线辨识系统及方法
Dey et al. Leakage current monitoring of suspension insulator for effective determination of ESDD
CN117332330A (zh) 基于多源数据图融合模型的变压器故障评估方法及系统
Xu et al. Improved method for extracting resistive leakage current of surge protectors
Khodsuz et al. Surge Arrester Monitoring under Different Operating Conditions Using Bees-ANFIS.
Bian et al. A Novel Dynamic Weight Allocation Method for Assessing the Health Status of Remote Terminal Unit in Distribution Automation System Based on AHM and GRA
Palangar et al. Detecting of unnormal conditions of polluted insulators based on analysis phase angle of leakage current
CN114595746A (zh) 配电网故障原因的分类方法、装置、设备及存储介质
Muniraj et al. Adaptive Neurofuzzy Inference System‐Based Pollution Severity Prediction of Polymeric Insulators in Power Transmission Lines

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22892158

Country of ref document: EP

Kind code of ref document: A1