CN117741517A - Capacitor equipment online monitoring method and system based on artificial intelligence - Google Patents
Capacitor equipment online monitoring method and system based on artificial intelligence Download PDFInfo
- Publication number
- CN117741517A CN117741517A CN202311758289.9A CN202311758289A CN117741517A CN 117741517 A CN117741517 A CN 117741517A CN 202311758289 A CN202311758289 A CN 202311758289A CN 117741517 A CN117741517 A CN 117741517A
- Authority
- CN
- China
- Prior art keywords
- capacitor
- coefficient
- fault
- data
- instruction
- Prior art date
- Legal status (The legal status 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 status listed.)
- Pending
Links
- 239000003990 capacitor Substances 0.000 title claims abstract description 314
- 238000000034 method Methods 0.000 title claims abstract description 93
- 238000012544 monitoring process Methods 0.000 title claims abstract description 39
- 238000013473 artificial intelligence Methods 0.000 title claims abstract description 30
- 230000007613 environmental effect Effects 0.000 claims abstract description 79
- 238000005211 surface analysis Methods 0.000 claims abstract description 63
- 238000012549 training Methods 0.000 claims abstract description 61
- 239000007789 gas Substances 0.000 claims description 60
- RAHZWNYVWXNFOC-UHFFFAOYSA-N Sulphur dioxide Chemical compound O=S=O RAHZWNYVWXNFOC-UHFFFAOYSA-N 0.000 claims description 42
- 238000004458 analytical method Methods 0.000 claims description 39
- 230000002159 abnormal effect Effects 0.000 claims description 24
- UGFAIRIUMAVXCW-UHFFFAOYSA-N Carbon monoxide Chemical compound [O+]#[C-] UGFAIRIUMAVXCW-UHFFFAOYSA-N 0.000 claims description 21
- 229910002091 carbon monoxide Inorganic materials 0.000 claims description 21
- UFHFLCQGNIYNRP-UHFFFAOYSA-N Hydrogen Chemical compound [H][H] UFHFLCQGNIYNRP-UHFFFAOYSA-N 0.000 claims description 20
- 229910052739 hydrogen Inorganic materials 0.000 claims description 20
- 239000001257 hydrogen Substances 0.000 claims description 20
- 230000008569 process Effects 0.000 claims description 20
- 238000004364 calculation method Methods 0.000 claims description 16
- 238000012360 testing method Methods 0.000 claims description 14
- 238000012545 processing Methods 0.000 claims description 9
- 239000013598 vector Substances 0.000 claims description 9
- 238000003708 edge detection Methods 0.000 claims description 8
- 238000001914 filtration Methods 0.000 claims description 8
- 238000003062 neural network model Methods 0.000 claims description 5
- 238000013527 convolutional neural network Methods 0.000 claims description 4
- 238000010183 spectrum analysis Methods 0.000 claims description 4
- 230000008030 elimination Effects 0.000 description 11
- 238000003379 elimination reaction Methods 0.000 description 11
- 230000015654 memory Effects 0.000 description 10
- 230000006870 function Effects 0.000 description 7
- 238000010586 diagram Methods 0.000 description 5
- 230000007246 mechanism Effects 0.000 description 4
- 230000008859 change Effects 0.000 description 3
- 230000005856 abnormality Effects 0.000 description 2
- 230000032683 aging Effects 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 238000003745 diagnosis Methods 0.000 description 2
- 238000010891 electric arc Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000010438 heat treatment Methods 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 1
- 238000002485 combustion reaction Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 239000003792 electrolyte Substances 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000017525 heat dissipation Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 239000011810 insulating material Substances 0.000 description 1
- 238000009413 insulation Methods 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
- 230000002028 premature Effects 0.000 description 1
- 238000001228 spectrum Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000004804 winding Methods 0.000 description 1
Landscapes
- Testing Electric Properties And Detecting Electric Faults (AREA)
Abstract
The invention belongs to the technical field of fault monitoring, and discloses an artificial intelligence-based capacitor equipment online monitoring method and system; collecting m groups of historical equipment operation data; historical equipment operational data includes harmonic data, vibration data, gas data, magnetic field data, temperature data, and capacitor images; calculating a harmonic coefficient, a vibration coefficient, a gas coefficient and a temperature coefficient corresponding to the historical equipment operation data; calculating an environmental coefficient and an operation coefficient corresponding to the historical equipment operation data; based on the capacitor image, training a surface analysis model for analyzing whether the surface of the capacitor is normal; collecting equipment operation data in real time, calculating corresponding environment coefficients and operation coefficients, inputting real-time capacitor images into a surface analysis model, and analyzing whether the surface of the capacitor is normal or not; judging whether to generate a primary fault instruction according to the environment coefficient and the operation coefficient; the monitoring capability of the running state of the capacitor is improved, and the stable running of the capacitor is ensured.
Description
Technical Field
The invention relates to the technical field of fault monitoring, in particular to an artificial intelligence-based capacitor equipment online monitoring method and system.
Background
In the power industry, capacitors are a key electrical device for improving the power factor and stability of an electrical power system; however, there is a potential risk of failure and damage during operation of the capacitor device, which may lead to instability of the power system, reduced energy efficiency and premature damage of the device; the traditional capacitor performance monitoring method mainly relies on off-line testing and periodic inspection, cannot capture instantaneous changes and implicit fault signs of the capacitor performance, and cannot monitor the state of the capacitor in real time; the evaluation and fault judgment of the capacitor mainly depend on manual experience and subjective judgment, and are easily influenced by artificial factors, so that inaccurate results are caused;
there is of course also an intelligent monitoring method, for example, chinese patent publication No. CN115128513a discloses a method for detecting abnormal capacitor based on heat and related device, including: acquiring operation related data in the operation process of the capacitor, wherein the operation related data comprise voltage, current, surface temperature and indoor temperature; calculating capacitor operation parameters according to an equivalent circuit constructed based on the operation related data, wherein the capacitor operation parameters comprise a harmonic effective value, a fundamental wave, unbalance and capacitance; carrying out heat dissipation analysis according to operation related data by adopting a preset BP neural network model to obtain a capacitor predicted temperature; performing temperature difference analysis according to the predicted temperature and the actually measured temperature of the capacitor to obtain a capacitor heating temperature difference value; if the heating temperature difference value of the capacitor exceeds a preset temperature range, carrying out anomaly analysis on the capacitor according to the operation parameters of the capacitor to obtain an anomaly detection result, and solving the technical problems of high workload and low efficiency caused by the lack of an efficient and reliable detection strategy in the prior art;
However, the operation related data collected by the technology is less in variety, so that the generalization capability of the model is limited, and the actual condition of the operation of the capacitor can not be comprehensively reflected; in the running process of the capacitor, the capacitor can work in a high-load state transiently due to load or environmental change, and the temperature rises, but the abnormal running state of the capacitor is not described, and the technology judges whether the capacitor is abnormal or not according to the difference value between the calculated and predicted temperature and the actual temperature, and the false alarm condition can occur;
in view of the above, the present invention proposes an artificial intelligence based method and system for online monitoring of capacitor devices to solve the above-mentioned problems.
Disclosure of Invention
In order to overcome the defects in the prior art, the invention provides the following technical scheme for achieving the purposes: an artificial intelligence based capacitor device on-line monitoring method comprises the following steps:
in the historical operation process of the capacitor, collecting m groups of historical equipment operation data, wherein m is an integer greater than 1; historical equipment operation data includes harmonic data, vibration data, gas data, magnetic field data, temperature data, and capacitor images;
calculating a harmonic coefficient, a vibration coefficient, a gas coefficient and a temperature coefficient corresponding to the historical equipment operation data;
Calculating an environmental coefficient and an operation coefficient corresponding to the historical equipment operation data;
based on the capacitor image, training a surface analysis model for analyzing whether the surface of the capacitor is normal;
collecting equipment operation data in real time, calculating corresponding environment coefficients and operation coefficients, inputting real-time capacitor images into a surface analysis model, and analyzing whether the surface of the capacitor is normal or not; judging whether to generate a primary fault instruction according to the environmental coefficient, the operation coefficient and the capacitor surface analysis condition;
if the primary fault instruction is generated according to the environment coefficient or the operation coefficient, analyzing the fault cause;
analyzing the real-time capacitor image, and judging whether a primary fault instruction is generated or not;
and judging whether to generate a middle-level fault instruction or a high-level fault instruction according to the generation quantity of the primary fault instructions.
Further, the harmonic data comprises a distortion rate and total harmonic distortion, wherein the distortion rate is the waveform distortion degree corresponding to the voltage signal, and the total harmonic distortion is the total distortion degree of the fundamental wave and the harmonic corresponding to the voltage signal; the vibration data comprise vibration frequency and vibration amplitude, the gas data comprise hydrogen concentration, carbon monoxide concentration and sulfur dioxide concentration, the magnetic field data are magnetic field intensity around the capacitor, the temperature data comprise internal temperature and environment temperature, the internal temperature is the temperature inside the capacitor, and the environment temperature is the temperature of the environment where the capacitor is located;
The distortion rate and the total harmonic distortion are obtained through frequency spectrum analysis of the voltage signal; firstly, obtaining a voltage signal of an output end of a capacitor, performing fast Fourier transform on the voltage signal, decomposing the voltage signal into a fundamental wave and each harmonic wave, obtaining amplitudes corresponding to the fundamental wave and each harmonic wave in the voltage signal, and calculating a distortion rate and total harmonic distortion corresponding to the capacitor; the calculation formula of the distortion rate SZL is as follows:
XB j the amplitude corresponding to the harmonic wave is JB, N is the total number of harmonic waves, j is the j-th harmonic wave, j is [1, N ]]The method comprises the steps of carrying out a first treatment on the surface of the The calculation formula of the total harmonic distortion ZJB is as follows:
further, the method for calculating the harmonic coefficients comprises the following steps:
XBX＝ω 1 ×(ZJB 2 +SZL)；
wherein XBX is harmonic coefficient omega 1 Is of preset weight and omega 1 ＞0；
The method for calculating the vibration coefficient comprises the following steps:
wherein ZDX is the vibration coefficient, ZP is the vibration frequency, ZF is the vibration amplitude, ω 2 Is of preset weight and omega 2 ＞0；
The method for calculating the gas coefficient comprises the following steps:
presetting a hydrogen concentration threshold, a carbon monoxide concentration threshold and a sulfur dioxide concentration threshold;
QTX＝β×(QN-QNY)×(YN-YNY)×(EN-ENY)；
wherein QTX is a gas coefficient, QN is a hydrogen concentration, QNY is a hydrogen concentration threshold, YN is a carbon monoxide concentration, YNY is a carbon monoxide concentration threshold, EN is a sulfur dioxide concentration, ENY is a sulfur dioxide concentration threshold, β is a preset proportionality coefficient, and β > 0;
The temperature coefficient calculating method comprises the following steps:
WDX＝ω 3 ×NW+ω 4 ×e (NW+HW) ；
wherein WDX is a temperature coefficient, NW is an internal temperature, HW is an ambient temperature, ω 3 、ω 4 Is of preset weight and omega 3 ＞0、ω 4 And > 0, e is a natural constant.
Further, the method for calculating the environmental coefficient comprises the following steps:
HJX＝α 1 ×(WDX 3 -A)+QTX；
wherein HJX is an environmental coefficient, QTX is a gas coefficient, alpha 1 Is a preset proportionality coefficient and alpha 1 > 0, A is a preset constant; if the gas coefficient QTX is a negative number, the QTX value is 0, and if the gas coefficient QTX is a positive number, the QTX value is QTX;
the calculation method of the operation coefficient comprises the following steps:
YXX＝ln[α 2 ×(XBX+ZDX)×CC]+α 3 ×(XBX+ZDX)；
wherein YXX is an operation coefficient, CC is magnetic field data, and alpha 2 、α 3 Is a preset proportionality coefficient and alpha 2 ＞0、α 3 ＞0。
Further, the specific training process of the surface analysis model comprises:
marking the capacitor image as a training image, marking m training images, wherein marking comprises normal and broken marks, and respectively converting the normal and broken marks into digital marks; dividing the marked training image into a training set and a testing set; training the surface analysis model by using a training set, and testing the surface analysis model by using a testing set; presetting an error threshold, and outputting a surface analysis model when the prediction error is smaller than the error threshold; the surface analysis model is a convolutional neural network model.
Further, the method for judging whether to generate the primary fault instruction according to the environmental coefficient, the operation coefficient and the capacitor surface analysis condition comprises the following steps:
a. judging whether a primary fault instruction is generated according to the operation coefficient;
comparing the operation coefficient with a preset operation threshold;
if the operation coefficient is smaller than or equal to the operation threshold value, a primary fault instruction is not generated;
if the operation coefficient is greater than the operation threshold value, generating a primary fault instruction;
b. judging whether to generate a primary fault instruction according to the analysis condition of the capacitor surface;
if the digital label output by the surface analysis model corresponds to normal, a primary fault instruction is not generated;
if the digital label output by the surface analysis model corresponds to a crack, generating a primary fault instruction;
c. judging whether to generate a primary fault instruction according to the environmental coefficient;
comparing the environmental coefficient with a preset environmental threshold;
if the environmental coefficient is smaller than or equal to the environmental threshold value, a primary fault instruction is not generated;
if the environmental coefficient is larger than the environmental threshold value and a primary fault instruction is not generated according to the operation coefficient and the capacitor surface analysis condition, a suspected fault instruction is generated;
if the environmental coefficient is greater than the environmental threshold, and a primary fault instruction is generated according to the operation coefficient or the capacitor surface analysis condition, generating the primary fault instruction;
Marking a time point corresponding to the suspected fault instruction as a fault suspected point, presetting a duration threshold, acquiring temperature data and gas data of a plurality of time points after the fault suspected point according to the duration threshold, and calculating a corresponding environment coefficient, wherein the plurality of time points are the number of the time points corresponding to the duration threshold;
judging whether the multiple time points are fault suspected points or not;
if the multiple time points are fault suspected points, generating a primary fault instruction;
if the time points in the multiple time points are not fault suspected points, a primary fault instruction is not generated.
Further, the method for analyzing the fault cause comprises the following steps:
inputting an environmental coefficient or an operation coefficient corresponding to the primary fault instruction into a trained cause analysis model to obtain a fault cause of the capacitor;
the training process of the cause analysis model comprises the following steps:
taking the environment coefficient and the operation coefficient as judgment coefficients, and acquiring fault reasons corresponding to abnormal judgment coefficients in advance;
converting the judgment coefficient and the fault cause into a corresponding group of feature vectors;
taking each group of feature vectors as the input of a cause analysis model, wherein the cause analysis model takes a group of predicted fault causes corresponding to each group of judgment coefficients as the output, and takes an actual fault cause corresponding to each group of judgment coefficients as a prediction target, and the actual fault cause is the pre-acquired fault cause corresponding to the judgment coefficients; taking the sum of prediction errors of the minimum judgment coefficients as a training target; training the cause analysis model until the sum of the prediction errors reaches convergence, and stopping training; the cause analysis model is a deep neural network model.
Further, the method for analyzing the real-time capacitor image and judging whether to generate the primary fault instruction comprises the following steps:
s1: reading a real-time capacitor image;
s2: carrying out graying treatment on the real-time capacitor image, and converting the color image into a gray image;
s3: applying gaussian filtering to the gray scale image to reduce noise and detail;
s4: performing edge detection on the gray level image after Gaussian filtering is applied, and detecting the edge of a capacitor in the gray level image;
s5: performing binarization operation on the gray level image after edge detection, and converting the gray level image into a binary image only comprising edges and a background;
s6: searching the outline in the binary image by using a findContours function;
s7: drawing the searched outline on the capacitor image by using a drawContours function so as to visualize the identified outline;
s8: counting the number of pixel points in the outline of the capacitor image;
s9: comparing the counted number of the pixel points with a number threshold range, and if the number of the pixel points is in the number threshold range or equal to the number threshold range, not generating a primary fault instruction; and if the number of the pixel points is out of the number threshold range, generating a primary fault instruction.
Further, the method for judging whether to generate the middle-level fault instruction or the high-level fault instruction comprises the following steps:
preset constant d 1 And d 2 ，d 2 ＞d 1 ＞1；
If the generation number of the primary fault instructions is greater than or equal to d 1 When the fault instruction is generated, a medium-level fault instruction is generated;
if the generation number of the primary fault instructions is greater than or equal toAt d 2 Generating an advanced fault instruction;
if the primary fault instruction is generated, disconnecting the capacitor from the related circuit, and transmitting the corresponding fault instruction, the fault reason and real-time equipment operation data corresponding to the fault instruction to a central control screen of the power system; the fault instructions include primary fault instructions, intermediate fault instructions, and advanced fault instructions.
The capacitor equipment on-line monitoring system based on the artificial intelligence implements the capacitor equipment on-line monitoring method based on the artificial intelligence, and comprises the following steps:
the data acquisition module is used for acquiring m groups of historical equipment operation data in the historical operation process of the capacitor, wherein m is an integer greater than 1; historical equipment operation data includes harmonic data, vibration data, gas data, magnetic field data, temperature data, and capacitor images;
the first data processing module is used for calculating harmonic coefficients, vibration coefficients, gas coefficients and temperature coefficients corresponding to the historical equipment operation data;
The second data processing module is used for calculating the environment coefficient and the operation coefficient corresponding to the operation data of the historical equipment;
the model training module is used for training a surface analysis model for analyzing whether the surface of the capacitor is normal or not based on the capacitor image;
the first fault judging module is used for collecting equipment operation data in real time, calculating corresponding environment coefficients and operation coefficients, inputting real-time capacitor images into the surface analysis model and analyzing whether the surface of the capacitor is normal or not; judging whether to generate a primary fault instruction according to the environmental coefficient, the operation coefficient and the capacitor surface analysis condition;
the fault analysis module is used for analyzing the fault reason if the primary fault instruction is generated according to the environment coefficient or the operation coefficient;
the second fault judging module is used for analyzing the real-time capacitor image and judging whether a primary fault instruction is generated or not;
and the third fault judging module is used for judging whether to generate a middle-level fault instruction or a high-level fault instruction according to the generation quantity of the primary fault instructions.
The capacitor equipment on-line monitoring method and system based on artificial intelligence have the technical effects and advantages that:
1. based on historical equipment operation data and real-time data acquisition, calculating harmonic coefficients, vibration coefficients, gas coefficients and temperature coefficients by adopting multidimensional data such as harmonic waves, vibration, gas, magnetic fields and temperature, and carrying out real-time monitoring and analysis by combining capacitor images; further analyzing fault reasons according to conditions, and improving the monitoring capability of the running state of the capacitor; a multi-level fault judging mechanism is established, and a medium-level or high-level fault instruction is generated according to the number of the primary fault instructions, so that the safety and reliability of equipment are improved, the fault risk is reduced, and the stable operation of the capacitor is ensured.
2. Under the condition that a plurality of capacitors are closely installed, through comprehensive analysis of the internal temperature of the capacitors and the ambient temperature of a working area, other normal capacitors are prevented from being misjudged as faults due to single capacitor faults; the intelligent judging mechanism can distinguish faults caused by abnormal ambient temperature and abnormal internal temperature of the capacitor, so that the number of erroneous judgment is reduced, and the system stability is improved; and can generate fault elimination instruction in time, help the accurate diagnosis capacitor unusual, ensure the safety and the stability of equipment operation.
Drawings
FIG. 1 is a schematic diagram of an on-line monitoring system of an artificial intelligence based capacitor device according to embodiment 1 of the present invention;
FIG. 2 is a schematic diagram of an on-line monitoring system of an artificial intelligence based capacitor device according to embodiment 2 of the present invention;
FIG. 3 is a schematic diagram of an on-line monitoring method of an artificial intelligence based capacitor device according to embodiment 3 of the present invention;
fig. 4 is a schematic diagram of an electronic device according to embodiment 4 of the present invention;
fig. 5 is a schematic diagram of a storage medium according to embodiment 5 of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Example 1
Referring to fig. 1, the capacitor device online monitoring system based on artificial intelligence according to the present embodiment includes a data acquisition module, a first data processing module, a second data processing module, a model training module, a first fault judgment module, a fault analysis module, a second fault judgment module, and a third fault judgment module; each module is connected in a wired and/or wireless mode, so that data transmission among the modules is realized;
the data acquisition module is used for acquiring m groups of historical equipment operation data in the historical operation process of the capacitor, wherein m is an integer greater than 1; historical equipment operation data includes harmonic data, vibration data, gas data, magnetic field data, temperature data, and capacitor images;
the harmonic data includes distortion ratio and total harmonic distortion; the distortion ratio is the waveform distortion degree corresponding to the voltage signal, and the total harmonic distortion is the total distortion degree of the fundamental wave and the harmonic wave corresponding to the voltage signal; the distortion rate and the total harmonic distortion are obtained through frequency spectrum analysis of the voltage signal; firstly, a voltage sensor arranged at the output end of a capacitor is used for acquiring a voltage signal, performing fast Fourier transform on the voltage signal, converting the voltage signal in a time domain into an energy spectrum in a frequency domain, namely decomposing the voltage signal into a fundamental wave and each harmonic wave, acquiring amplitudes corresponding to the fundamental wave and each harmonic wave in the voltage signal, and calculating the distortion rate and the total harmonic distortion corresponding to the capacitor; the calculation formula of the distortion rate SZL is as follows:
XB j The amplitude corresponding to the harmonic is JB, the amplitude corresponding to the fundamental wave is N, the total number of the harmonic is j, the j is the j-th harmonic,j∈[1,N]the method comprises the steps of carrying out a first treatment on the surface of the The calculation formula of the total harmonic distortion ZJB is:
when the capacitor fails, the waveform of the voltage signal is affected, so that new harmonic waves appear and the amplitude of each harmonic wave is increased, and the corresponding distortion rate and the total harmonic distortion are changed;
the vibration data comprise vibration frequency and vibration amplitude, and the vibration data reflect the vibration condition of the capacitor; vibration data are acquired through a piezoelectric sensor arranged on the capacitor, and the abnormal vibration data indicate that the capacitor fails;
the gas data comprises hydrogen concentration, carbon monoxide concentration and sulfur dioxide concentration, the gas data is obtained by an infrared gas sensor arranged in a gas escape area, the gas escape area is an area where generated gas escapes when a capacitor breaks down, a person skilled in the art collects gas around the capacitor by using the infrared sensor when the capacitor breaks down, and the area where the gas generated when the capacitor breaks down can be collected is marked as the gas escape area; when the capacitor fails, gas is generated, so that the concentration of the corresponding gas in the environment where the capacitor is located is increased, for example, hydrogen is generated due to electrolyte aging in the capacitor, carbon monoxide is generated due to combustion or arc discharge of internal elements or insulation of the capacitor, and media (such as oil) in the capacitor can be damaged or aged to generate sulfur dioxide and the like;
The magnetic field data is the intensity of the magnetic field around the capacitor, and the magnetic field data is acquired by a magnetic field sensor arranged in the working area of the capacitor; when the capacitor fails, such as arc discharge, winding short circuit and the like, the magnetic field of the surrounding environment can be changed;
the temperature data comprise an internal temperature and an ambient temperature, wherein the internal temperature is the temperature of the capacitor, the ambient temperature is the temperature of the environment where the capacitor is positioned, the internal temperature is obtained by a thermocouple sensor arranged in the capacitor, and the ambient temperature is obtained by a temperature sensor arranged in a working area of the capacitor; the ambient temperature affects the internal temperature of the capacitor, and the higher the ambient temperature is, the higher the internal temperature of the capacitor is, and vice versa; if the internal temperature is too high, the capacitor is indicated to have a fault condition, and the capacitor is also indicated to have normal change caused by factors such as increase of working load or increase of ambient temperature, if the internal temperature is continuously too high, the capacitor is indicated to have a fault condition, and the capacitor needs to be overhauled in time;
the capacitor image is acquired by a CCD camera arranged in the working area of the capacitor, and the CCD camera faces the capacitor so as to acquire the image of the capacitor; when the capacitor is in fault, the surface of the capacitor may be bulged, cracked and the like, so that the capacitor image is collected and analyzed to judge whether the capacitor is in fault or not;
The first data processing module is used for calculating harmonic coefficients, vibration coefficients, gas coefficients and temperature coefficients corresponding to the historical equipment operation data;
the method for calculating the harmonic coefficients comprises the following steps:
XBX＝ω 1 ×(ZJB 2 +SZL)；
wherein XBX is harmonic coefficient omega 1 Is of preset weight and omega 1 ＞0；
The method for calculating the vibration coefficient comprises the following steps:
wherein ZDX is the vibration coefficient, ZP is the vibration frequency, ZF is the vibration amplitude, ω 2 Is of preset weight and omega 2 ＞0；
In the calculation formula of the harmonic coefficient and the vibration coefficient, a preset weight is used for acquiring a plurality of groups of comprehensive parameters by a person skilled in the art, a corresponding weight is set for each group of comprehensive parameters, the preset weight and the acquired comprehensive parameters are substituted into the formula, and the calculated weight is filtered and averaged to obtain omega 1 、ω 2 Is a value of (2);
the method for calculating the gas coefficient comprises the following steps:
presetting a hydrogen concentration threshold, a carbon monoxide concentration threshold and a sulfur dioxide concentration threshold;
QTX＝β×(QN-QNY)×(YN-YNY)×(EN-ENY)；
wherein QTX is a gas coefficient, QN is a hydrogen concentration, QNY is a hydrogen concentration threshold, YN is a carbon monoxide concentration, YNY is a carbon monoxide concentration threshold, EN is a sulfur dioxide concentration, ENY is a sulfur dioxide concentration threshold, β is a preset proportionality coefficient, and β > 0;
wherein the preset proportionality coefficient is obtained by a person skilled in the art, a plurality of groups of comprehensive parameters are collected, corresponding proportionality coefficients are set for each group of comprehensive parameters, the preset proportionality coefficient and the collected comprehensive parameters are substituted into a formula, and the calculated proportionality coefficients are filtered and averaged to obtain the value of beta;
The temperature coefficient calculating method comprises the following steps:
WDX＝ω 3 ×NW+ω 4 ×e (NW+HW) ；
wherein WDX is a temperature coefficient, NW is an internal temperature, HW is an ambient temperature, ω 3 、ω 4 Is of preset weight and omega 3 ＞0、ω 4 > 0, e is a natural constant;
wherein the preset weight is acquired by a person skilled in the art, corresponding weights are set for each group of comprehensive parameters, the preset weight and the acquired comprehensive parameters are substituted into a formula, any two formulas form a binary one-time equation set, the calculated weights are filtered and averaged to obtain omega 3 、ω 4 Is a value of (2);
it should be noted that, when the capacitor is aged by a person skilled in the art, collecting the hydrogen concentration of the multiple groups of gas dissipation areas, and taking the average value of the multiple groups of hydrogen concentrations as the hydrogen concentration threshold; the carbon monoxide concentration threshold is obtained by a person skilled in the art when the internal element of the capacitor burns, the carbon monoxide concentration of a plurality of groups of gas dissipation areas is collected, and the average value of the carbon monoxide concentrations is taken as the carbon monoxide concentration threshold; when a medium in the capacitor is damaged, the sulfur dioxide concentration threshold value is obtained by a person skilled in the art, the sulfur dioxide concentration of a plurality of groups of gas dissipation areas, and the average value of the sulfur dioxide concentrations is taken as the sulfur dioxide concentration threshold value;
The second data processing module is used for calculating the environment coefficient and the operation coefficient corresponding to the operation data of the historical equipment;
the calculating method of the environment coefficient comprises the following steps:
HJX＝α 1 ×(WDX 3 -A)+QTX；
wherein HJX is an environmental coefficient, QTX is a gas coefficient, alpha 1 Is a preset proportionality coefficient and alpha 1 > 0, A is a preset constant; if the gas coefficient QTX is a negative number, the QTX value is 0, and if the gas coefficient QTX is a positive number, the QTX value is QTX;
wherein the preset proportionality coefficient and the preset constant are collected by a person skilled in the art, corresponding proportionality coefficient and constant are set for each group of comprehensive parameters, the preset proportionality coefficient, constant and the collected comprehensive parameters are substituted into a formula, any two formulas form a binary one-time equation system, the calculated proportionality coefficient and constant are screened and averaged to obtain alpha 1 Values of A;
the operation coefficient calculating method comprises the following steps:
YXX＝ln[α 2 ×(XBX+ZDX)×CC]+α 3 ×(XBX+ZDX)；
wherein YXX is an operation coefficient, CC is magnetic field data, and alpha 2 、α 3 Is a preset proportionality coefficient and alpha 2 ＞0、α 3 ＞0；
Wherein the preset proportionality coefficient is obtained by the person skilled in the art, a plurality of groups of comprehensive parameters are collected, the corresponding proportionality coefficient is set for each group of comprehensive parameters, the preset proportionality coefficient and the collected comprehensive parameters are substituted into a formula, any two formulas form a binary once equation set, the calculated proportionality coefficient is screened and averaged to obtain alpha 2 、α 3 Is a value of (2);
it should be noted that, the operation coefficient and the environmental coefficient are only relevant parameters for judging whether the capacitor is faulty or not, and have no other practical significance, so the calculation of the operation coefficient and the environmental coefficient is dimensionality removal calculation;
the model training module is used for training a surface analysis model for analyzing whether the surface of the capacitor is normal or not based on the capacitor image;
the specific training process of the surface analysis model comprises the following steps:
marking the capacitor image as a training image, marking m training images, wherein the marking comprises normal and broken marks, respectively converting the normal and broken marks into digital marks, and converting the normal mark into 1 and the broken mark into 2 by way of example; dividing the marked training images into a training set and a testing set, taking 70% of the training images as the training set and 30% of the training images as the testing set; training the surface analysis model by using a training set, and testing the surface analysis model by using a testing set; presetting an error threshold, and outputting a surface analysis model when the prediction error is smaller than the error threshold; wherein, the calculation formula of the preset error is Z P ＝(α P -μ P ) 2 Wherein Z is P For prediction error, P is the number of the training image, α P For the prediction label corresponding to the P group training image, mu P The actual labels corresponding to the P group training images are provided; the error threshold value is preset according to the precision required by the surface analysis model;
the surface analysis model is specifically a convolutional neural network model;
the first fault judging module is used for collecting equipment operation data in real time, calculating corresponding environment coefficients and operation coefficients, inputting real-time capacitor images into the surface analysis model and analyzing whether the surface of the capacitor is normal or not; judging whether to generate a primary fault instruction according to the environmental coefficient, the operation coefficient and the capacitor surface analysis condition;
the method for judging whether to generate the primary fault instruction according to the environment coefficient, the operation coefficient and the capacitor surface analysis condition comprises the following steps:
a. judging whether a primary fault instruction is generated according to the operation coefficient;
comparing the operation coefficient with a preset operation threshold;
if the operation coefficient is smaller than or equal to the operation threshold value, a primary fault instruction is not generated; the harmonic data, vibration data and magnetic field data corresponding to the capacitor are all normal at the moment;
if the operation coefficient is greater than the operation threshold value, generating a primary fault instruction; at least one of harmonic data, vibration data and magnetic field data corresponding to the capacitor is abnormal data, so that the capacitor is indicated to have a fault condition;
b. Judging whether to generate a primary fault instruction according to the analysis condition of the capacitor surface;
if the digital label output by the surface analysis model corresponds to normal, a primary fault instruction is not generated;
indicating that no crack exists on the surface of the capacitor;
if the digital label output by the surface analysis model corresponds to a crack, generating a primary fault instruction;
indicating that cracks exist on the surface of the capacitor, so that the fault condition of the capacitor is indicated;
c. judging whether to generate a primary fault instruction according to the environmental coefficient;
comparing the environmental coefficient with a preset environmental threshold;
if the environmental coefficient is smaller than or equal to the environmental threshold value, a primary fault instruction is not generated; indicating that the temperature data and the gas data corresponding to the capacitor are normal at the moment;
if the environmental coefficient is larger than the environmental threshold value and a primary fault instruction is not generated according to the operation coefficient and the capacitor surface analysis condition, a suspected fault instruction is generated; at least one of the temperature data and the gas data corresponding to the capacitor is abnormal data, but whether the capacitor fails or not cannot be confirmed;
if the environmental coefficient is greater than the environmental threshold, and a primary fault instruction is generated according to the operation coefficient or the capacitor surface analysis condition, generating the primary fault instruction; indicating that various data anomalies exist in the capacitor at the moment, and the capacitor is in a fault state;
Whether a primary fault instruction is generated according to the operation coefficient or the capacitor surface analysis condition is judged to generate a suspected fault instruction or the primary fault instruction is because if the primary fault instruction is generated according to the operation coefficient or the capacitor surface analysis condition, the capacitor is in a fault state at the moment, the capacitor is required to be disconnected from a related circuit, and the environmental coefficient cannot be further analyzed, so that the primary fault instruction is generated, and the maintenance is further judged by staff; if a primary fault instruction is not generated according to the operation coefficient and the analysis condition of the capacitor surface, the fact that the capacitor has no other data anomalies except the environment coefficient anomalies at the moment is indicated, whether the capacitor fails or not cannot be confirmed, and the environment coefficient needs to be further analyzed, so that a suspected fault instruction is generated;
it should be noted that, the above operation threshold and the environmental threshold are all those skilled in the art, in the historical operation process of the capacitor, when the capacitor is in a fault state, corresponding historical equipment operation data is collected multiple times, multiple sets of historical equipment operation data are collected at one time, environmental coefficients and operation coefficients corresponding to the multiple sets of historical equipment operation data are calculated, mean values corresponding to the multiple environmental coefficients and mean values corresponding to the multiple operation coefficients are calculated, and as the corresponding historical equipment operation data are collected multiple times, multiple environmental coefficient mean values and multiple operation coefficient mean values are calculated, mean values corresponding to the multiple environmental coefficient mean values are calculated as the environmental threshold, and mean values corresponding to the multiple operation coefficient mean values are calculated as the operation threshold;
Marking a time point corresponding to a suspected fault instruction as a fault suspected point, presetting a duration threshold, collecting temperature data and gas data of a plurality of time points after the fault suspected point, and calculating corresponding environment coefficients, wherein the plurality of time points are the number of the time points corresponding to the duration threshold, and if the duration threshold is 5, namely the number of the time points corresponding to the duration threshold is 5, collecting the temperature data and the gas data of the 5 time points, and 1 time point is 1 second;
judging whether the multiple time points are fault suspected points or not;
if the multiple time points are fault suspected points, generating a primary fault instruction; indicating that a capacitor has a fault condition;
if the time points in the multiple time points are not fault suspected points, a primary fault instruction is not generated, and the reason for the abnormality of the environment coefficient is that the capacitor is temporarily operated in a high-load state due to load or environment change, so that the environment coefficient is temporarily abnormal;
it should be noted that, the preset value of the duration threshold is specifically determined according to the user manual and the technical specification of the capacitor;
the fault analysis module is used for analyzing the fault reason if the primary fault instruction is generated according to the environment coefficient or the operation coefficient;
The method for analyzing the fault cause comprises the following steps:
inputting an environmental coefficient or an operation coefficient corresponding to the primary fault instruction into a trained cause analysis model to obtain a fault cause of the capacitor;
the training process of the cause analysis model comprises the following steps:
taking the environment coefficient and the operation coefficient as judgment coefficients, and acquiring fault reasons corresponding to abnormal judgment coefficients in advance; in the historical operation process of the capacitor, when the capacitor is in a fault state, corresponding historical equipment operation data are acquired for a plurality of times, environmental coefficients and operation coefficients corresponding to the historical equipment operation data are calculated, the calculated environmental coefficients are abnormal environmental coefficients, and the calculated operation coefficients are abnormal operation coefficients; the technical staff in the field analyzes different abnormal environment coefficients and abnormal operation coefficients according to experience to obtain corresponding abnormal reasons, wherein the abnormal reasons are fault reasons of the capacitor; the fault reasons of the capacitor are such as mechanical part faults, loosening of internal components, internal arcing, aging of internal media, damage of insulating materials and the like;
converting the judgment coefficient and the fault cause into a corresponding group of feature vectors;
taking each group of characteristic vectors as input of a cause analysis model, wherein the cause analysis model takes a group of predicted fault causes corresponding to each group of judgment coefficients as output, takes an actual fault cause corresponding to each group of judgment coefficients as a prediction target, and takes an actual fault source The failure cause corresponding to the judgment coefficient is obtained in advance; taking the sum of prediction errors of the minimum judgment coefficients as a training target; wherein, the calculation formula of the prediction error is Z K ＝(α K -μ K ) 2 Wherein Z is K For prediction error, K is the group number of the feature vector corresponding to the judgment coefficient, alpha K For the predicted fault reason corresponding to the K-th group judgment coefficient, mu K Judging the actual fault reason corresponding to the coefficient for the K group; training the cause analysis model until the sum of the prediction errors reaches convergence, and stopping training;
the reason analysis model is specifically a deep neural network model;
the second fault judging module is used for analyzing the real-time capacitor image and judging whether a primary fault instruction is generated or not;
the method for analyzing the real-time capacitor image and judging whether to generate the primary fault instruction comprises the following steps:
s1: reading a real-time capacitor image;
image＝cv2.imread('your_image.jpg',cv2.IMREAD_COLOR)
s2: carrying out graying treatment on the real-time capacitor image, and converting the color image into a gray image;
gray＝cv2.cvtColor(image,cv2.COLOR_BGR2GRAY)
s3: applying gaussian filtering to the gray scale image to reduce noise and detail; noise is random and irregular pixel points which are not expected to exist in the gray level image;
blurred＝cv2.GaussianBlur(gray,(5,5),0)
s4: performing edge detection on the gray level image after Gaussian filtering is applied, and detecting the edge of a capacitor in the gray level image;
edges＝cv2.Canny(blurred,50,150)
S5: performing binarization operation on the gray level image after edge detection, and converting the gray level image into a binary image only comprising edges and a background;
_,thresh＝cv2.threshold(edges,0,255,cv2.THRESH_BINARY)
s6: searching the outline in the binary image by using a findContours function;
contours,_＝cv2.findContours(thresh,cv2.RETR_EXTERNAL,cv2.CHAIN_APPROX_SIMPLE)
s7: drawing the searched outline on the capacitor image by using a drawContours function so as to visualize the identified outline;
contour_image＝image.copy()
cv2.drawContours(contour_image,contours,-1,(0,255,0),2)
s8: counting the number of pixel points in the outline of the capacitor image;
for contour in contours:
area＝cv2.contourArea(contour)
print("ContourArea:",area)
s9: comparing the counted number of the pixel points with a number threshold range, and if the number of the pixel points is in the number threshold range or equal to the number threshold range, not generating a primary fault instruction, so that the surface of the capacitor is not changed; if the number of pixel points is out of the range of the number threshold, generating a primary fault instruction, wherein the primary fault instruction indicates that the surface of the capacitor is changed, such as bulge or damage, and the capacitor is in a fault state;
when the capacitor is in a normal running state, the number threshold range is that the capacitor images are acquired for a plurality of times, the pixel point number of each capacitor image is acquired each time, the pixel point number corresponding to the plurality of capacitor images acquired at one time is added and divided by the capacitor image number acquired at one time to acquire the pixel point average value, the pixel point average values are sequenced from large to small, the pixel point average value arranged at the first is taken as the maximum value of the number threshold range, and the pixel point average value arranged at the last is taken as the minimum value of the number threshold range;
The third fault judging module is used for judging whether to generate a middle-level fault instruction or a high-level fault instruction according to the generation quantity of the primary fault instructions;
the method for judging whether to generate the middle-level fault instruction or the high-level fault instruction comprises the following steps:
preset constant d 1 And d 2 ，d 2 ＞d 1 ＞1；
If the generation number of the primary fault instructions is greater than or equal to d 1 When the capacitor is abnormal, a medium-level fault instruction is generated, and the condition that various data are abnormal in the capacitor at the moment is indicated;
if the generation number of the primary fault instructions is greater than or equal to d 2 When the capacitor is in a high-level fault state, generating an advanced fault instruction, and indicating that the fault degree of the capacitor is higher at the moment, and the capacitor needs to be overhauled in time;
d is the same as 1 、d 2 According to the actual situation, the embodiment preferably selects d 1 Is 2, d 2 4;
if the primary fault instruction is generated, disconnecting the capacitor from the related circuit, and transmitting the corresponding fault instruction, fault reason and real-time equipment operation data corresponding to the fault instruction to a central control screen of the power system, so that a worker can conveniently and timely overhaul, and the capacitor is prevented from further deepening the fault degree; the fault instructions comprise primary fault instructions, intermediate fault instructions and advanced fault instructions;
the embodiment is based on historical equipment operation data and real-time data acquisition, and adopts multi-dimensional data such as harmonic wave, vibration, gas, magnetic field, temperature and the like to calculate harmonic wave coefficients, vibration coefficients, gas coefficients and temperature coefficients, and combines capacitor images to monitor and analyze in real time; further analyzing fault reasons according to conditions, and improving the monitoring capability of the running state of the capacitor; a multi-level fault judging mechanism is established, and a medium-level or high-level fault instruction is generated according to the number of the primary fault instructions, so that the safety and reliability of equipment are improved, the fault risk is reduced, and the stable operation of the capacitor is ensured.
Example 2
Referring to fig. 2, the present embodiment further improves the design based on embodiment 1, when the environmental coefficient is calculated, the environmental temperature needs to be considered, if the installation positions of the plurality of capacitors are relatively close, when one or more capacitors fail to cause the internal temperature to rise, the environmental temperature of the working area where the failed capacitor is located rises, so that the calculated environmental coefficient is affected when the other normal capacitors perform fault monitoring, and the situation that the normal capacitor is misdiagnosed as the failed capacitor occurs; therefore, the embodiment provides an on-line monitoring system for capacitor equipment based on artificial intelligence, which further comprises a fault elimination module, wherein if a plurality of capacitors exist in a working area and are operated simultaneously, when the capacitors exist in the plurality of capacitors and are monitored as faults due to overhigh internal temperature, the capacitors are marked as high-temperature capacitors, and if the fault condition of the rest capacitors is also monitored, the rest capacitors are marked as suspected fault capacitors; judging whether to eliminate the fault condition according to the internal temperature of the high-temperature capacitor and the internal temperature of the fault capacitor;
the fault elimination module judges whether to generate a fault elimination instruction according to a fault instruction corresponding to the suspected fault capacitor;
The method for judging whether to generate the fault elimination instruction comprises the following steps:
judging whether a fault instruction corresponding to the suspected fault capacitor is a primary fault instruction or not;
if the fault instruction is not the primary fault instruction, a fault elimination instruction is not generated, and the capacitor generates a plurality of primary fault instructions, so that a plurality of data anomalies exist;
if the primary fault instruction is the primary fault instruction, judging whether the primary fault instruction is generated according to the environmental coefficient;
if the primary fault instruction is not generated according to the environment coefficient, a fault elimination instruction is not generated; indicating that the capacitor is not failed due to an abnormality in the internal temperature or the ambient temperature;
if the primary fault instruction is generated according to the environmental coefficient, subtracting the internal temperature of the suspected fault capacitor from the internal temperature of the high-temperature capacitor to obtain a difference value, and comparing the absolute value of the difference value with a difference value threshold;
if the absolute value of the difference value is smaller than the difference value threshold value, a fault elimination instruction is not generated; the internal temperature difference between the suspected fault capacitor and the internal temperature of the high-temperature capacitor is smaller, and the internal temperature of the suspected fault capacitor is abnormal, so that the capacitor is in a fault state;
if the absolute value of the difference value is greater than or equal to the difference value threshold value and the difference value is not a negative number, generating a fault elimination instruction; the method includes the steps that the difference between the internal temperature of a suspected fault capacitor and the internal temperature of a high-temperature capacitor is large, the internal temperature of the suspected fault capacitor is normal, and the suspected fault capacitor is monitored to be in a fault state because the internal temperature of the high-temperature capacitor is too high, so that the ambient temperature of a working area where the suspected fault capacitor is located is increased;
If a fault elimination instruction is generated, eliminating a primary fault instruction generated by the corresponding suspected fault capacitor; the corresponding suspected fault capacitor is a normal capacitor;
it should be noted that, the difference threshold is obtained by acquiring the internal temperature of one capacitor when the capacitor is in fault due to the over-high internal temperature when the positions where a plurality of capacitors exist in one working area are relatively close to each other in the historical operation process of the capacitor, and simultaneously acquiring the internal temperatures of the other capacitors with normal internal temperature, subtracting the internal temperatures corresponding to the capacitors with the over-high internal temperature from the internal temperatures corresponding to the other capacitors with normal internal temperature respectively to obtain a plurality of differences, and calculating the average value of the plurality of differences; and similarly, obtaining average values of a plurality of differences, calculating average values corresponding to the average values of the plurality of differences, and taking the average values as difference threshold values;
in the embodiment, under the condition that a plurality of capacitors are closely installed, through comprehensive analysis of the internal temperature of the capacitors and the ambient temperature of a working area, other normal capacitors are prevented from being misjudged as faults due to single capacitor faults; the intelligent judging mechanism can distinguish faults caused by abnormal ambient temperature and abnormal internal temperature of the capacitor, so that the number of erroneous judgment is reduced, and the system stability is improved; and can generate fault elimination instruction in time, help the accurate diagnosis capacitor unusual, ensure the safety and the stability of equipment operation.
Example 3
Referring to fig. 3, this embodiment, which is not described in detail in embodiments 1 and 2, provides an artificial intelligence based on-line monitoring method for capacitor devices, comprising:
in the historical operation process of the capacitor, collecting m groups of historical equipment operation data, wherein m is an integer greater than 1; historical equipment operation data includes harmonic data, vibration data, gas data, magnetic field data, temperature data, and capacitor images;
calculating a harmonic coefficient, a vibration coefficient, a gas coefficient and a temperature coefficient corresponding to the historical equipment operation data;
calculating an environmental coefficient and an operation coefficient corresponding to the historical equipment operation data;
based on the capacitor image, training a surface analysis model for analyzing whether the surface of the capacitor is normal;
collecting equipment operation data in real time, calculating corresponding environment coefficients and operation coefficients, inputting real-time capacitor images into a surface analysis model, and analyzing whether the surface of the capacitor is normal or not; judging whether to generate a primary fault instruction according to the environmental coefficient, the operation coefficient and the capacitor surface analysis condition;
if the primary fault instruction is generated according to the environment coefficient or the operation coefficient, analyzing the fault cause;
Analyzing the real-time capacitor image, and judging whether a primary fault instruction is generated or not;
and judging whether to generate a middle-level fault instruction or a high-level fault instruction according to the generation quantity of the primary fault instructions.
Further, the harmonic data comprises a distortion rate and total harmonic distortion, wherein the distortion rate is the waveform distortion degree corresponding to the voltage signal, and the total harmonic distortion is the total distortion degree of the fundamental wave and the harmonic corresponding to the voltage signal; the vibration data comprise vibration frequency and vibration amplitude, the gas data comprise hydrogen concentration, carbon monoxide concentration and sulfur dioxide concentration, the magnetic field data are magnetic field intensity around the capacitor, the temperature data comprise internal temperature and environment temperature, the internal temperature is the temperature inside the capacitor, and the environment temperature is the temperature of the environment where the capacitor is located;
the distortion rate and the total harmonic distortion are obtained through frequency spectrum analysis of the voltage signal; firstly, obtaining a voltage signal of an output end of a capacitor, performing fast Fourier transform on the voltage signal, decomposing the voltage signal into a fundamental wave and each harmonic wave, obtaining amplitudes corresponding to the fundamental wave and each harmonic wave in the voltage signal, and calculating a distortion rate and total harmonic distortion corresponding to the capacitor; the calculation formula of the distortion rate SZL is as follows:
XB j The amplitude corresponding to the harmonic wave is JB, N is the total number of harmonic waves, j is the j-th harmonic wave, j is [1, N ]]The method comprises the steps of carrying out a first treatment on the surface of the The calculation formula of the total harmonic distortion ZJB is as follows:
further, the method for calculating the harmonic coefficients comprises the following steps:
XBX＝ω 1 ×(ZJB 2 +SZL)；
wherein XBX is harmonic coefficient omega 1 Is of preset weight and omega 1 ＞0；
The method for calculating the vibration coefficient comprises the following steps:
wherein ZDX is the vibration coefficient, ZP is the vibration frequency, ZF is the vibration amplitude, ω 2 Is of preset weight and omega 2 ＞0；
The method for calculating the gas coefficient comprises the following steps:
presetting a hydrogen concentration threshold, a carbon monoxide concentration threshold and a sulfur dioxide concentration threshold;
QTX＝β×(QN-QNY)×(YN-YNY)×(EN-ENY)；
wherein QTX is a gas coefficient, QN is a hydrogen concentration, QNY is a hydrogen concentration threshold, YN is a carbon monoxide concentration, YNY is a carbon monoxide concentration threshold, EN is a sulfur dioxide concentration, ENY is a sulfur dioxide concentration threshold, β is a preset proportionality coefficient, and β > 0;
the temperature coefficient calculating method comprises the following steps:
WDX＝ω 3 ×NW+ω 4 ×e (NW+HW) ；
wherein WDX is a temperature coefficient, NW is an internal temperature, HW is an ambient temperature, ω 3 、ω 4 Is of preset weight and omega 3 ＞0、ω 4 And > 0, e is a natural constant.
Further, the method for calculating the environmental coefficient comprises the following steps:
HJX＝α 1 ×(WDX 3 -A)+QTX；
wherein HJX is an environmental coefficient, QTX is a gas coefficient, alpha 1 Is a preset proportionality coefficient and alpha 1 > 0, A is a preset constant; if the gas coefficient QTX is a negative number, the QTX value is 0, and if the gas coefficient QTX is a positive number, the QTX value is QTX;
the calculation method of the operation coefficient comprises the following steps:
YXX＝ln[α 2 ×(XBX+ZDX)×CC]+α 3 ×(XBX+ZDX)；
wherein YXX is an operation coefficient, CC is magnetic field data, and alpha 2 、α 3 Is a preset proportionality coefficient and alpha 2 ＞0、α 3 ＞0。
Further, the specific training process of the surface analysis model comprises:
marking the capacitor image as a training image, marking m training images, wherein marking comprises normal and broken marks, and respectively converting the normal and broken marks into digital marks; dividing the marked training image into a training set and a testing set; training the surface analysis model by using a training set, and testing the surface analysis model by using a testing set; presetting an error threshold, and outputting a surface analysis model when the prediction error is smaller than the error threshold; the surface analysis model is a convolutional neural network model.
Further, the method for judging whether to generate the primary fault instruction according to the environmental coefficient, the operation coefficient and the capacitor surface analysis condition comprises the following steps:
a. judging whether a primary fault instruction is generated according to the operation coefficient;
comparing the operation coefficient with a preset operation threshold;
If the operation coefficient is smaller than or equal to the operation threshold value, a primary fault instruction is not generated;
if the operation coefficient is greater than the operation threshold value, generating a primary fault instruction;
b. judging whether to generate a primary fault instruction according to the analysis condition of the capacitor surface;
if the digital label output by the surface analysis model corresponds to normal, a primary fault instruction is not generated;
if the digital label output by the surface analysis model corresponds to a crack, generating a primary fault instruction;
c. judging whether to generate a primary fault instruction according to the environmental coefficient;
comparing the environmental coefficient with a preset environmental threshold;
if the environmental coefficient is smaller than or equal to the environmental threshold value, a primary fault instruction is not generated;
if the environmental coefficient is larger than the environmental threshold value and a primary fault instruction is not generated according to the operation coefficient and the capacitor surface analysis condition, a suspected fault instruction is generated;
if the environmental coefficient is greater than the environmental threshold, and a primary fault instruction is generated according to the operation coefficient or the capacitor surface analysis condition, generating the primary fault instruction;
marking a time point corresponding to the suspected fault instruction as a fault suspected point, presetting a duration threshold, acquiring temperature data and gas data of a plurality of time points after the fault suspected point according to the duration threshold, and calculating a corresponding environment coefficient, wherein the plurality of time points are the number of the time points corresponding to the duration threshold;
Judging whether the multiple time points are fault suspected points or not;
if the multiple time points are fault suspected points, generating a primary fault instruction;
if the time points in the multiple time points are not fault suspected points, a primary fault instruction is not generated.
Further, the method for analyzing the fault cause comprises the following steps:
inputting an environmental coefficient or an operation coefficient corresponding to the primary fault instruction into a trained cause analysis model to obtain a fault cause of the capacitor;
the training process of the cause analysis model comprises the following steps:
taking the environment coefficient and the operation coefficient as judgment coefficients, and acquiring fault reasons corresponding to abnormal judgment coefficients in advance;
converting the judgment coefficient and the fault cause into a corresponding group of feature vectors;
taking each group of feature vectors as the input of a cause analysis model, wherein the cause analysis model takes a group of predicted fault causes corresponding to each group of judgment coefficients as the output, and takes an actual fault cause corresponding to each group of judgment coefficients as a prediction target, and the actual fault cause is the pre-acquired fault cause corresponding to the judgment coefficients; taking the sum of prediction errors of the minimum judgment coefficients as a training target; training the cause analysis model until the sum of the prediction errors reaches convergence, and stopping training; the cause analysis model is a deep neural network model.
Further, the method for analyzing the real-time capacitor image and judging whether to generate the primary fault instruction comprises the following steps:
s1: reading a real-time capacitor image;
s2: carrying out graying treatment on the real-time capacitor image, and converting the color image into a gray image;
s3: applying gaussian filtering to the gray scale image to reduce noise and detail;
s4: performing edge detection on the gray level image after Gaussian filtering is applied, and detecting the edge of a capacitor in the gray level image;
s5: performing binarization operation on the gray level image after edge detection, and converting the gray level image into a binary image only comprising edges and a background;
s6: searching the outline in the binary image by using a findContours function;
s7: drawing the searched outline on the capacitor image by using a drawContours function so as to visualize the identified outline;
s8: counting the number of pixel points in the outline of the capacitor image;
s9: comparing the counted number of the pixel points with a number threshold range, and if the number of the pixel points is in the number threshold range or equal to the number threshold range, not generating a primary fault instruction; and if the number of the pixel points is out of the number threshold range, generating a primary fault instruction.
Further, the method for judging whether to generate the middle-level fault instruction or the high-level fault instruction comprises the following steps:
preset constant d 1 And d 2 ，d 2 ＞d 1 ＞1；
If the generation number of the primary fault instructions is greater than or equal to d 1 When the fault instruction is generated, a medium-level fault instruction is generated;
if the generation number of the primary fault instructions is greater than or equal to d 2 Generating an advanced fault instruction;
if the primary fault instruction is generated, disconnecting the capacitor from the related circuit, and transmitting the corresponding fault instruction, the fault reason and real-time equipment operation data corresponding to the fault instruction to a central control screen of the power system; the fault instructions include primary fault instructions, intermediate fault instructions, and advanced fault instructions.
Example 4
Referring to fig. 4, an electronic device 500 is also provided according to yet another aspect of the present application. The electronic device 500 may include one or more processors and one or more memories. Wherein the memory has stored therein computer readable code which, when executed by the one or more processors, can perform the artificial intelligence based capacitor device online monitoring method as described above.
The method or system according to embodiments of the present application may also be implemented by means of the architecture of the electronic device shown in fig. 4. As shown in fig. 4, the electronic device 500 may include a bus 501, one or more CPUs 502, a Read Only Memory (ROM) 503, a Random Access Memory (RAM) 504, a communication port 505 connected to a network, an input/output 506, a hard disk 507, and the like. A storage device in electronic device 500, such as ROM503 or hard disk 507, may store the artificial intelligence based capacitor device online monitoring methods provided herein. Further, the electronic device 500 may also include a user interface 508. Of course, the architecture shown in fig. 4 is merely exemplary, and one or more components of the electronic device shown in fig. 4 may be omitted as may be desired in implementing different devices.
Example 5
Referring to FIG. 5, a computer readable storage medium 600 according to one embodiment of the present application is shown. Computer readable storage medium 600 has stored thereon computer readable instructions. The artificial intelligence based capacitor device online monitoring method according to embodiments of the present application described with reference to the above figures may be performed when computer readable instructions are executed by a processor. Storage medium 600 includes, but is not limited to, for example, volatile memory and/or nonvolatile memory. Volatile memory can include, for example, random Access Memory (RAM), cache memory (cache), and the like. The non-volatile memory may include, for example, read Only Memory (ROM), hard disk, flash memory, and the like.
In addition, according to embodiments of the present application, the processes described above with reference to flowcharts may be implemented as computer software programs. For example, the present application provides a non-transitory machine-readable storage medium storing machine-readable instructions executable by a processor to perform instructions corresponding to the method steps provided herein, such as: an artificial intelligence based capacitor equipment on-line monitoring method. The above-described functions defined in the method of the present application are performed when the computer program is executed by a Central Processing Unit (CPU).
The foregoing is merely illustrative of the present invention, and the present invention is not limited thereto, and any person skilled in the art will readily recognize that variations or substitutions are within the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.
Finally: the foregoing description of the preferred embodiments of the invention is not intended to limit the invention to the precise form disclosed, and any such modifications, equivalents, and alternatives falling within the spirit and principles of the invention are intended to be included within the scope of the invention.
Claims (10)
1. The capacitor equipment on-line monitoring method based on artificial intelligence is characterized by comprising the following steps of:
in the historical operation process of the capacitor, collecting m groups of historical equipment operation data, wherein m is an integer greater than 1; historical equipment operation data includes harmonic data, vibration data, gas data, magnetic field data, temperature data, and capacitor images;
calculating a harmonic coefficient, a vibration coefficient, a gas coefficient and a temperature coefficient corresponding to the historical equipment operation data;
calculating an environmental coefficient and an operation coefficient corresponding to the historical equipment operation data;
Based on the capacitor image, training a surface analysis model for analyzing whether the surface of the capacitor is normal;
collecting equipment operation data in real time, calculating corresponding environment coefficients and operation coefficients, inputting real-time capacitor images into a surface analysis model, and analyzing whether the surface of the capacitor is normal or not; judging whether to generate a primary fault instruction according to the environmental coefficient, the operation coefficient and the capacitor surface analysis condition;
if the primary fault instruction is generated according to the environment coefficient or the operation coefficient, analyzing the fault cause;
analyzing the real-time capacitor image, and judging whether a primary fault instruction is generated or not;
and judging whether to generate a middle-level fault instruction or a high-level fault instruction according to the generation quantity of the primary fault instructions.
2. The online monitoring method of capacitor equipment based on artificial intelligence according to claim 1, wherein the harmonic data comprises a distortion rate and a total harmonic distortion, the distortion rate is a waveform distortion degree corresponding to a voltage signal, and the total harmonic distortion is a total distortion degree corresponding to a fundamental wave and a harmonic wave of the voltage signal; the vibration data comprise vibration frequency and vibration amplitude, the gas data comprise hydrogen concentration, carbon monoxide concentration and sulfur dioxide concentration, the magnetic field data are magnetic field intensity around the capacitor, the temperature data comprise internal temperature and environment temperature, the internal temperature is the temperature inside the capacitor, and the environment temperature is the temperature of the environment where the capacitor is located;
The distortion rate and the total harmonic distortion are obtained through frequency spectrum analysis of the voltage signal; firstly, obtaining a voltage signal of an output end of a capacitor, performing fast Fourier transform on the voltage signal, decomposing the voltage signal into a fundamental wave and each harmonic wave, obtaining amplitudes corresponding to the fundamental wave and each harmonic wave in the voltage signal, and calculating a distortion rate and total harmonic distortion corresponding to the capacitor; the calculation formula of the distortion rate SZL is as follows:
XB j the amplitude corresponding to the harmonic wave is JB, N is the total number of harmonic waves, j is the j-th harmonic wave, j is [1, N ]]The method comprises the steps of carrying out a first treatment on the surface of the The calculation formula of the total harmonic distortion ZJB is as follows:
3. the method for online monitoring of capacitor equipment based on artificial intelligence according to claim 2, wherein the method for calculating the harmonic coefficients comprises:
XBX＝ω 1 ×(ZJB 2 +SZL)；
wherein XBX is harmonic coefficient omega 1 Is of preset weight and omega 1 ＞0；
The method for calculating the vibration coefficient comprises the following steps:
wherein ZDX is the vibration coefficient, ZP is the vibration frequency, ZF is the vibration amplitude, ω 2 Is of preset weight and omega 2 ＞0；
The method for calculating the gas coefficient comprises the following steps:
presetting a hydrogen concentration threshold, a carbon monoxide concentration threshold and a sulfur dioxide concentration threshold;
QTX＝β×(QN-QNY)×(YN-YNY)×(EN-ENY)；
wherein QTX is a gas coefficient, QN is a hydrogen concentration, QNY is a hydrogen concentration threshold, YN is a carbon monoxide concentration, YNY is a carbon monoxide concentration threshold, EN is a sulfur dioxide concentration, ENY is a sulfur dioxide concentration threshold, β is a preset proportionality coefficient, and β > 0;
The temperature coefficient calculating method comprises the following steps:
WDX＝ω 3 ×NW+ω 4 ×e( NW+HW )；
wherein WDX is a temperature coefficient, NW is an internal temperature, HW is an ambient temperature, ω 3 、ω 4 Is of preset weight and omega 3 ＞0、ω 4 And > 0, e is a natural constant.
4. The artificial intelligence based capacitor device on-line monitoring method of claim 3, wherein the environmental factor calculating method comprises:
HJX＝α 1 ×(WDX 3 -A)+QTX；
wherein HJX is an environmental coefficient, QTX is a gas coefficient, alpha 1 Is a preset proportionality coefficient and alpha 1 > 0, A is a preset constant; if the gas coefficient QTX is a negative number, the QTX value is 0, and if the gas coefficient QTX is a positive number, the QTX value is QTX;
the calculation method of the operation coefficient comprises the following steps:
YXX＝ln[α 2 ×(XBX+ZDX)×CC]+α 3 ×(XBX+ZDX)；
wherein YXX is an operation coefficient, CC is magnetic field data, and alpha 2 、α 3 Is a preset proportionality coefficient and alpha 2 ＞0、α 3 ＞0。
5. The artificial intelligence based capacitor device on-line monitoring method of claim 4, wherein the training process of the surface analysis model comprises:
marking the capacitor image as a training image, marking m training images, wherein marking comprises normal and broken marks, and respectively converting the normal and broken marks into digital marks; dividing the marked training image into a training set and a testing set; training the surface analysis model by using a training set, and testing the surface analysis model by using a testing set; presetting an error threshold, and outputting a surface analysis model when the prediction error is smaller than the error threshold; the surface analysis model is a convolutional neural network model.
6. The method for online monitoring of capacitor equipment based on artificial intelligence according to claim 5, wherein the method for judging whether to generate the primary fault instruction according to the environmental coefficient, the operation coefficient and the capacitor surface analysis condition comprises the following steps:
a. judging whether a primary fault instruction is generated according to the operation coefficient;
comparing the operation coefficient with a preset operation threshold;
if the operation coefficient is smaller than or equal to the operation threshold value, a primary fault instruction is not generated;
if the operation coefficient is greater than the operation threshold value, generating a primary fault instruction;
b. judging whether to generate a primary fault instruction according to the analysis condition of the capacitor surface;
if the digital label output by the surface analysis model corresponds to normal, a primary fault instruction is not generated;
if the digital label output by the surface analysis model corresponds to a crack, generating a primary fault instruction;
c. judging whether to generate a primary fault instruction according to the environmental coefficient;
comparing the environmental coefficient with a preset environmental threshold;
if the environmental coefficient is smaller than or equal to the environmental threshold value, a primary fault instruction is not generated;
if the environmental coefficient is larger than the environmental threshold value and a primary fault instruction is not generated according to the operation coefficient and the capacitor surface analysis condition, a suspected fault instruction is generated;
If the environmental coefficient is greater than the environmental threshold, and a primary fault instruction is generated according to the operation coefficient or the capacitor surface analysis condition, generating the primary fault instruction;
marking a time point corresponding to the suspected fault instruction as a fault suspected point, presetting a duration threshold, acquiring temperature data and gas data of a plurality of time points after the fault suspected point according to the duration threshold, and calculating a corresponding environment coefficient, wherein the plurality of time points are the number of the time points corresponding to the duration threshold;
judging whether the multiple time points are fault suspected points or not;
if the multiple time points are fault suspected points, generating a primary fault instruction;
if the time points in the multiple time points are not fault suspected points, a primary fault instruction is not generated.
7. The method for online monitoring of capacitor equipment based on artificial intelligence according to claim 6, wherein the method for analyzing the cause of the fault comprises:
inputting an environmental coefficient or an operation coefficient corresponding to the primary fault instruction into a trained cause analysis model to obtain a fault cause of the capacitor;
the training process of the cause analysis model comprises the following steps:
taking the environment coefficient and the operation coefficient as judgment coefficients, and acquiring fault reasons corresponding to abnormal judgment coefficients in advance;
Converting the judgment coefficient and the fault cause into a corresponding group of feature vectors;
taking each group of feature vectors as the input of a cause analysis model, wherein the cause analysis model takes a group of predicted fault causes corresponding to each group of judgment coefficients as the output, and takes an actual fault cause corresponding to each group of judgment coefficients as a prediction target, and the actual fault cause is the pre-acquired fault cause corresponding to the judgment coefficients; taking the sum of prediction errors of the minimum judgment coefficients as a training target; training the cause analysis model until the sum of the prediction errors reaches convergence, and stopping training; the cause analysis model is a deep neural network model.
8. The method for online monitoring of capacitor equipment based on artificial intelligence according to claim 7, wherein the method for analyzing the real-time capacitor image to determine whether to generate the primary fault instruction comprises:
s1: reading a real-time capacitor image;
s2: carrying out graying treatment on the real-time capacitor image, and converting the color image into a gray image;
s3: applying gaussian filtering to the gray scale image to reduce noise and detail;
s4: performing edge detection on the gray level image after Gaussian filtering is applied, and detecting the edge of a capacitor in the gray level image;
S5: performing binarization operation on the gray level image after edge detection, and converting the gray level image into a binary image only comprising edges and a background;
s6: searching the outline in the binary image by using a findContours function;
s7: drawing the searched outline on the capacitor image by using a drawContours function so as to visualize the identified outline;
s8: counting the number of pixel points in the outline of the capacitor image;
s9: comparing the counted number of the pixel points with a number threshold range, and if the number of the pixel points is in the number threshold range or equal to the number threshold range, not generating a primary fault instruction; and if the number of the pixel points is out of the number threshold range, generating a primary fault instruction.
9. The method for online monitoring of capacitor equipment based on artificial intelligence according to claim 8, wherein the method for judging whether to generate the medium-level fault instruction or the high-level fault instruction comprises:
preset constant d 1 And d 2 ，d 2 ＞d 1 ＞1；
If the generation number of the primary fault instructions is greater than or equal to d 1 When the fault instruction is generated, a medium-level fault instruction is generated;
if the generation number of the primary fault instructions is greater than or equal to d 2 Generating an advanced fault instruction;
if the primary fault instruction is generated, disconnecting the capacitor from the related circuit, and transmitting the corresponding fault instruction, the fault reason and real-time equipment operation data corresponding to the fault instruction to a central control screen of the power system; the fault instructions include primary fault instructions, intermediate fault instructions, and advanced fault instructions.
10. An artificial intelligence based capacitor device on-line monitoring system for implementing the artificial intelligence based capacitor device on-line monitoring method according to any one of claims 1 to 9, comprising:
the data acquisition module is used for acquiring m groups of historical equipment operation data in the historical operation process of the capacitor, wherein m is an integer greater than 1; historical equipment operation data includes harmonic data, vibration data, gas data, magnetic field data, temperature data, and capacitor images;
the first data processing module is used for calculating harmonic coefficients, vibration coefficients, gas coefficients and temperature coefficients corresponding to the historical equipment operation data;
the second data processing module is used for calculating the environment coefficient and the operation coefficient corresponding to the operation data of the historical equipment;
the model training module is used for training a surface analysis model for analyzing whether the surface of the capacitor is normal or not based on the capacitor image;
the first fault judging module is used for collecting equipment operation data in real time, calculating corresponding environment coefficients and operation coefficients, inputting real-time capacitor images into the surface analysis model and analyzing whether the surface of the capacitor is normal or not; judging whether to generate a primary fault instruction according to the environmental coefficient, the operation coefficient and the capacitor surface analysis condition;
The fault analysis module is used for analyzing the fault reason if the primary fault instruction is generated according to the environment coefficient or the operation coefficient;
the second fault judging module is used for analyzing the real-time capacitor image and judging whether a primary fault instruction is generated or not;
and the third fault judging module is used for judging whether to generate a middle-level fault instruction or a high-level fault instruction according to the generation quantity of the primary fault instructions.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311758289.9A CN117741517A (en) | 2023-12-20 | 2023-12-20 | Capacitor equipment online monitoring method and system based on artificial intelligence |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311758289.9A CN117741517A (en) | 2023-12-20 | 2023-12-20 | Capacitor equipment online monitoring method and system based on artificial intelligence |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117741517A true CN117741517A (en) | 2024-03-22 |
Family
ID=90278882
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311758289.9A Pending CN117741517A (en) | 2023-12-20 | 2023-12-20 | Capacitor equipment online monitoring method and system based on artificial intelligence |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117741517A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118244166A (en) * | 2024-05-28 | 2024-06-25 | 上海兴兴泰港机技术发展有限公司 | Transformer fault detection system |
| CN118443886A (en) * | 2024-05-17 | 2024-08-06 | 技整科技(广州)有限公司 | Plasma equipment health monitoring method and system |
-
2023
- 2023-12-20 CN CN202311758289.9A patent/CN117741517A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118443886A (en) * | 2024-05-17 | 2024-08-06 | 技整科技(广州)有限公司 | Plasma equipment health monitoring method and system |
| CN118244166A (en) * | 2024-05-28 | 2024-06-25 | 上海兴兴泰港机技术发展有限公司 | Transformer fault detection system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN117741517A (en) | Capacitor equipment online monitoring method and system based on artificial intelligence | |
| CN105324900B (en) | The method and apparatus of the early warning of the defects of for power equipment | |
| CN117808456B (en) | Equipment fault early warning method and device based on intelligent operation management | |
| KR102136141B1 (en) | Emergency generator including function for fault prevention using artificial intelligence algorithm | |
| AU2010308514A1 (en) | System and method for selecting a maintenance operation | |
| CN117783769B (en) | Power distribution network fault positioning method, system, equipment and storage medium based on visual platform | |
| CN118503636A (en) | Power distribution equipment health state assessment method and system | |
| KR20210109136A (en) | Equipment and system including function for fault prevention using artificial intelligence algorithm | |
| CN117992776B (en) | Real-time prediction method for health state of power grid equipment based on artificial intelligence | |
| CN117706943B (en) | Self-adaptive control method and system for converter transformer valve side sleeve end screen voltage divider | |
| CN117074852A (en) | Power distribution network electric energy monitoring and early warning management method and system | |
| Vidhya et al. | Transformer breather thermal image decomposition for fault diagnosis | |
| CN114019422A (en) | Transformer fault monitoring system based on ATT-BilSTM | |
| CN117783689A (en) | Intelligent defect assessment method for basin-type insulator of extra-high voltage GIS equipment | |
| CN111832653B (en) | XGboost algorithm-based arrester defect diagnosis method and system | |
| CN111314110B (en) | Fault early warning method for distributed system | |
| CN113721086A (en) | Method for monitoring the electrical insulation condition of a piece of equipment of an MV or HV electrical system | |
| CN116961226A (en) | Monitoring method and system for transformer in transformer substation | |
| CN116338391A (en) | Insulator state detection method and device, electronic equipment and medium | |
| CN206656749U (en) | Heat-engine plant sensor fault diagnosis system | |
| CN115758066A (en) | Method for counting carbon emission in whole life cycle of transformer | |
| CN113406537B (en) | Quantitative evaluation method for fault degree of power equipment | |
| Jacome et al. | Real-Time Fault Identification of Photovoltaic Systems Based on Remote Monitoring with IoT | |
| RU2791597C1 (en) | System for monitoring, diagnostics and management of the technical condition of power transformers | |
| CN118503727B (en) | Efficient energy-saving intelligent regulation transformer and operation and maintenance method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication |