CN111610418B - GIS partial discharge positioning method based on intelligent ultrahigh frequency sensor - Google Patents
GIS partial discharge positioning method based on intelligent ultrahigh frequency sensor Download PDFInfo
- Publication number
- CN111610418B CN111610418B CN202010487230.0A CN202010487230A CN111610418B CN 111610418 B CN111610418 B CN 111610418B CN 202010487230 A CN202010487230 A CN 202010487230A CN 111610418 B CN111610418 B CN 111610418B
- Authority
- CN
- China
- Prior art keywords
- sensor
- discharge
- sensors
- ultrahigh frequency
- abnormal
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/12—Testing dielectric strength or breakdown voltage ; Testing or monitoring effectiveness or level of insulation, e.g. of a cable or of an apparatus, for example using partial discharge measurements; Electrostatic testing
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/08—Locating faults in cables, transmission lines, or networks
- G01R31/088—Aspects of digital computing
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Theoretical Computer Science (AREA)
- Testing Relating To Insulation (AREA)
Abstract
The invention discloses a GIS partial discharge positioning method based on an intelligent ultrahigh frequency sensor, which comprises a sensor space position correlation step, an ultrahigh frequency signal abnormity judgment step, an abnormal sensor correlation analysis step, a multi-discharge source separation step, an interference signal elimination step and a partial discharge source positioning step. The GIS partial discharge positioning method based on the intelligent ultrahigh frequency sensor can solve the problem of resource consumption of an intelligent monitoring device caused by manual retest positioning, collects ultrahigh frequency abnormal signals by using the deployed intelligent ultrahigh frequency sensor, can perform defect positioning by using the positioning method at a platform end, does not need manual retest positioning, reduces resource investment and improves diagnosis efficiency.
Description
Technical Field
The invention relates to a GIS partial discharge positioning method based on an intelligent ultrahigh frequency sensor.
Background
GIS (Gas Insulation Switchgear) is more and more widely applied at home and abroad in recent decades due to the advantages of small occupied area, small influence from external environment, safe and reliable operation, simple maintenance and the like. However, in the manufacturing and assembling process of GIS, some small defects such as metal particles, insulation air gaps, etc. may be left inside GIS due to problems of process, design, etc., and these small defects may be developed into dangerous discharge channels during the operation of GIS, and finally cause insulation breakdown accidents. Therefore, the method has very important significance for monitoring partial discharge of the GIS in operation in order to prevent insulation faults of GIS equipment and guarantee safe operation of a power system.
The traditional GIS ultrahigh frequency online monitoring device is limited in application of national grids due to the problems of fixation, high cost, inconvenience in wiring deployment, complex operation and the like. In recent years, with the rapid development of technologies such as internet of things and artificial intelligence, partial discharge detection devices have gradually developed towards intellectualization. The wireless passive intelligent sensor has the advantages of convenience in deployment, convenience in installation and low cost, and has a wide application scene in the field of power transformation and distribution. The existing method for local discharge source positioning by using ultrahigh frequency technology is mostly applied to live detection and online monitoring equipment, for example, chinese patent CN201811044918.0 discloses a positioning method and system for partial discharge of electrical equipment, chinese patent CN201710380145.2 discloses a discharge source positioning method of a GIS ultrahigh frequency partial discharge online monitoring device, and chinese patent CN201510056192.2 discloses a GIS partial discharge positioning method based on ultrahigh frequency discharge online monitoring data, which require that a collection device has very high sampling frequency and bandwidth, and simultaneously, signals of a plurality of sensors must be collected synchronously. The wireless passive intelligent ultrahigh frequency sensor is limited by low power consumption, the acquisition frequency of the sensor is low, and the acquisition time of the sensor is asynchronous, so that the traditional positioning method cannot be applied. Document "partial discharge localization method based on ultrahigh frequency wireless smart sensor" (yangsen, xiong Jun, zheng Fuli, zhong Shaoquan, luo Lingen. Partial discharge localization method based on ultrahigh frequency wireless smart sensor, electric automation, 2017, 39 (2): 110-112+115] and 'partial discharge positioning method based on ultrahigh frequency wireless sensor and pattern recognition algorithm' (Huang Hui, huang Feng, li Chunlong, li Zhen. Partial discharge positioning method based on ultrahigh frequency wireless sensor and pattern recognition algorithm, scientific technology and engineering, 2018, 18 (4): 65-70 partial discharge positioning method based on ultrahigh frequency wireless sensor is researched, but no measures for eliminating space electromagnetic interference are mentioned, signal homology analysis is not carried out, a plurality of discharge sources cannot be distinguished, and positioning accuracy is affected.
Disclosure of Invention
The invention aims to overcome the defects of the prior art, provides a GIS partial discharge positioning method based on an intelligent ultrahigh frequency sensor, can solve the problem of resource consumption of an intelligent monitoring device caused by manual retest positioning, collects ultrahigh frequency abnormal signals by using the deployed intelligent ultrahigh frequency sensor, can perform defect positioning by using the positioning method at a platform end without manual retest positioning, reduces resource investment and improves diagnosis efficiency.
The technical scheme for realizing the purpose is as follows: a GIS partial discharge positioning method based on an intelligent ultrahigh frequency sensor comprises the following steps:
s1, sensor spatial position correlation: installing a plurality of partial discharge ultrahigh frequency sensors on a GIS body, installing a plurality of space ultrahigh frequency sensors on the wall of a surrounding wall of the GIS, coding all the sensors, associating the coded sensors with the GIS space positions according to the GIS space structure, an electrical connection diagram and the sensor installation positions, and establishing a sensor space topological diagram, wherein the intelligent sensors are nodes of the topological diagram, two sensors which are directly and electrically connected are connected with each other by a straight line to form a side line of the topological diagram, and the length of the side line represents the distance of electrical connection between the two corresponding sensors;
s2, judging the abnormal condition of the ultrahigh frequency signal: judging whether the ultrahigh frequency signals of the sensors are abnormal or not by utilizing a method of combining map recognition based on deep learning and ultrahigh frequency signal intensity dynamic threshold diagnosis, and adding the sensors with abnormal signals into an abnormal sensor set;
s3, an abnormal sensor correlation analysis step: and (2) performing signal correlation analysis on each abnormal sensor in the abnormal sensor set by combining the sensor space topological graph established in the step (S1) to find a correlation sensor of the abnormal sensor on a signal propagation path, wherein the specific steps are as follows:
s31, selecting 2 abnormal sensors;
s32, calculating the shortest path among the 2 abnormal sensors by utilizing a Dijkstra algorithm;
s33, calculating the number k of sensors on the shortest path, and judging the sensors on the shortest path among the 2 abnormal sensors as related sensors when k is less than or equal to 2; when k is greater than 2, judging that no correlation sensor exists among the 2 abnormal sensors;
s34, repeating the steps S31-S33, obtaining the associated sensors among all the abnormal sensors, and adding the associated sensors into an abnormal sensor set;
s4, separating multiple discharge sources: carrying out homology analysis by comparing the correlation of characteristic sequences of monitoring data of different sensors, judging the homology of the data, separating a plurality of partial discharge sources, constructing a community network, determining a signal community according to the community network, and completing the separation of the discharge sources;
s5, interference signal elimination step: the ultrahigh frequency data signal intensity of each sensor in the signal community is compared, interference signals are eliminated, and effective identification of partial discharge signals in the electrical equipment is achieved;
s6, a partial discharge source positioning step: the approximate position of the partial discharge source in the signal community is judged by comparing the signal intensity acquired by different position sensors in the signal community, and the internal partial discharge defect is positioned.
In the above GIS partial discharge positioning method based on the intelligent ultrahigh frequency sensor, in step S1, the sensor spatial topological graph includes not only spatial position information of the sensors and the GIS, but also electrical connection information of each sensor on a partial discharge signal propagation path.
In the above method for positioning the partial discharge of the GIS based on the intelligent ultrahigh frequency sensor, step S2 specifically includes the following steps:
s21, the intelligent ultrahigh frequency sensor is started automatically to acquire data, and ultrahigh frequency map data are acquired;
s22, diagnosing ultrahigh frequency map data by using a map identification method based on deep learning, judging whether the map data is abnormal or not, and identifying a sensor acquiring abnormal data;
s23, calculating the ultrahigh frequency signal intensity collected by the sensor with abnormal data:
V i ＝N i ×F i
in the formula: v i The intensity of the ultrahigh frequency signal acquired for the ith time of the sensor; n is a radical of i And F i Respectively acquiring the number of pulses and the average discharge amplitude of data acquired by the ultrahigh frequency sensor for the ith time;
s24, calculating the accumulated average signal intensity of data acquired by the ultrahigh frequency sensor with abnormal data before the ith time:
in the formula: t is the length of the cumulative time window and represents the selected cumulative times; n is a radical of hydrogen j Acquiring the number of pulses of normal data for the jth time of the ultrahigh frequency sensor; f j Acquiring the average discharge amplitude of the normal data for the jth time of the ultrahigh frequency sensor;
s25, screening sensors with abnormal data, and when the sensors are V i >V j The sensor is added to the abnormal sensor set.
In the above method for positioning the partial discharge of the GIS based on the intelligent ultrahigh frequency sensor, in step S4, the characteristic sequence includes a daily average discharge amplitude sequence and a daily average discharge frequency sequence; the step S4 specifically includes the following steps:
s41, extracting all sensors in the abnormal sensor set by combining the sensor space topological graph in the step 1 to form an abnormal sensor topological graph;
s42, homology analysis: according to the attenuation characteristics of electromagnetic waves, the signal intensity of the electromagnetic waves emitted by the same discharge source can be attenuated to different degrees along a propagation path, but the variation trends of the signal intensity are the same, homology analysis is carried out by comparing the correlation of characteristic sequences of monitoring data of different sensors, and the specific steps of the homology analysis are as follows:
s421, acquiring historical m-day partial discharge monitoring data of the abnormal sensor centralized sensor;
s422, calculating a daily average discharge amplitude sequence and a daily average discharge frequency sequence of historical m days;
s423, carrying out normalization processing on the daily average discharge amplitude sequence and the daily average discharge frequency sequence, wherein the normalization formula is as follows:
in the formula, x * Is the value after the normalization of each vector in the feature sequence, x is the value of each vector in the feature sequence, x min Is the minimum value of the vector in the feature sequence, x max The maximum value of the vector in the characteristic sequence;
after each vector in the characteristic sequence is subjected to normalization processing, a normalized daily average discharge amplitude sequence and a normalized daily average discharge frequency sequence are obtained;
s424, calculating the daily average discharge amplitude similarity r between the sensors in the abnormal sensor topological graph c Similarity r with daily average discharge frequency d The calculation formula is as follows:
wherein r is the similarity of A, B sensor characteristic sequence, A (t) is the characteristic sequence of A sensor,
is the mean value sequence of A (t), B (t) is the characteristic sequence of B sensor,
is the mean sequence of B (t);
s425, judging the homology of signals: judging that the abnormal data monitored by the two sensors have correlation when the similarity satisfies the following formula, wherein the abnormal data come from the same discharge source:
r c ∩r d >δ
wherein delta is a similarity threshold value, and is more than or equal to 0.6 and less than or equal to 0.9;
s43, constructing a single community network: two sensors which are judged to be the same discharge source are used as nodes of a community network, and a sideline is added between the two nodes to form 1 single community network;
s44, repeating the step S43 to obtain a single community network among other nodes, and continuously combining and overlapping all the single community networks to obtain a final community network;
s45, separating a discharge source: determining a signal community according to the final community network to complete the separation of the discharge source; in the community network, nodes connected with edges form 1 signal community, nodes connected with no edges form 1 signal community independently, and each signal community corresponds to an abnormal signal source.
In the above method for positioning partial discharge of a GIS based on an intelligent ultrahigh frequency sensor, step S5 specifically includes the following steps:
s51, checking whether the signal community contains a space ultrahigh frequency sensor or not, if not, judging that a discharge source of the signal community is internal partial discharge, and if so, entering the step S52;
s52, adjusting the acquisition frequency of the sensor, triggering and acquiring n groups of ultrahigh frequency abnormal data at the same time, filtering the data acquired at the ultrahigh frequency in the space, and removing noise interference;
s53, calculating the signal intensity mean value of each ultrahigh frequency sensor in the signal community, wherein the calculation formula of the signal intensity mean value is as follows:
wherein V Av The signal intensity mean value of the sensor A is obtained; n is a radical of hydrogen Ai The number of pulses of the ith group of signals of the sensor A; f Ai The average discharge amplitude of the ith group signal of the sensor A is obtained; n is the number of abnormal data groups collected by the abnormal sensor group;
s54, comparing the signal intensity mean values of the sensors, and if the signal intensity mean value of the space ultrahigh frequency sensor is the largest, judging that the discharge source of the signal community is an external interference signal; and if the average value of the signal intensity of the partial discharge ultrahigh frequency sensor is maximum, judging that the discharge source of the signal community is an internal partial discharge signal.
In the above method for positioning the partial discharge of the GIS based on the intelligent ultrahigh frequency sensor, step S6 specifically includes the following steps:
s61, obtaining a local discharge ultrahigh-frequency sensor A with the largest signal intensity mean value and a second-largest local discharge ultrahigh-frequency sensor B in the signal community according to the signal intensity mean value calculation formula in the step 5, and obtaining a suspected discharge gas chamber which is a gas chamber directly and electrically connected with the local discharge ultrahigh-frequency sensor A and the local discharge ultrahigh-frequency sensor B by combining the sensor space topological graph in the step S1; meanwhile, the air chamber only provided with 1 partial discharge ultrahigh frequency sensor A is listed as a suspected discharge air chamber;
s62, dividing the suspected discharge gas chamber by using a middle vertical line between the partial discharge ultrahigh frequency sensor A and the partial discharge ultrahigh frequency sensor B to obtain a region between the middle vertical line and the partial discharge ultrahigh frequency sensor A as a suspected discharge region; if only one partial discharge ultrahigh-frequency sensor is deployed in the suspected discharge gas chamber, the gas chamber is not required to be divided into a suspected discharge area;
s63, obtaining S suspected discharge areas through the step S62, and judging that the intersection of the S suspected discharge areas is the area where the partial discharge source is located; if the s suspected discharge areas have no intersection, judging that the union of the s suspected discharge areas is the area where the partial discharge source is located.
The GIS partial discharge positioning method based on the intelligent ultrahigh frequency sensor has the following beneficial effects:
(1) Compared with the traditional online monitoring device, the wireless passive intelligent sensor has the advantages of convenience in installation, low cost and the like, the quantity and the positions of the wireless passive intelligent sensors can be flexibly deployed according to the state of equipment, and the positioning method can realize the positioning of the partial discharge source by deploying the wireless passive intelligent sensor, thereby effectively reducing the resource investment and improving the flexibility and the working efficiency of detection;
(2) The positioning method can be applied to a monitoring system based on the Internet of things, effectively solves the problems of complexity and high requirement on professional quality of personnel in the conventional partial discharge source positioning method, can effectively improve the diagnosis efficiency and the intelligent level by applying the positioning algorithm on a platform, reduces the workload of partial discharge positioning, and has wide application scenes;
(3) The invention separates the multiple discharge sources by adopting the spatial position information of the sensor and the characteristic information of the collected data, can effectively identify the discharge sources and eliminate spatial interference under the environment of the multiple discharge sources and complex electromagnetic interference, and improves the accuracy of the positioning of the local discharge sources;
(4) The positioning method has no requirement on the synchronism of data acquisition of a plurality of sensors, and has wide application range.
Drawings
FIG. 1 is a flow chart of a GIS partial discharge positioning method based on an intelligent ultrahigh frequency sensor according to the invention;
FIG. 2 is a flow chart of the interference signal elimination step of the GIS partial discharge positioning method based on the intelligent ultrahigh frequency sensor of the invention;
FIG. 3 is a schematic diagram of sensor position information;
FIG. 4 is a sensor spatial topology;
FIG. 5 is an anomaly sensor topology;
FIG. 6 is a schematic diagram of an abnormal sensor community network;
fig. 7 is a schematic structural diagram of the gas chamber of the isolating switch.
Detailed Description
In order that those skilled in the art will better understand the technical solution of the present invention, the following detailed description is given with reference to the accompanying drawings:
referring to fig. 1 to 7, in an embodiment of the present invention, a method for positioning a GIS partial discharge based on an intelligent uhf sensor includes the following steps:
s1, sensor spatial position correlation: referring to fig. 3 and 4, the intelligent ultrahigh frequency sensor is deployed in the GIS body and the surrounding space to collect abnormal electromagnetic signals inside and outside the GIS, wherein the sensor mounted on the GIS insulating plate through a binding band is used for monitoring a partial discharge signal emitted inside the GIS, and is called as a partial discharge ultrahigh frequency sensor; the sensor is arranged on the wall around the GIS through a bracket and is used for monitoring electromagnetic interference signals in the surrounding environment of the GIS, and the sensor is called a space ultrahigh frequency sensor. Installing 8 partial discharge ultrahigh frequency sensors on the GIS body, wherein the number of the sensors is 1-8; installing 2 space ultrahigh frequency sensors on the wall of a surrounding wall of a GIS, numbering 9-10, coding all sensors, associating the coded sensors with the GIS space positions according to the GIS space structure, an electrical connection diagram and the sensor installation positions, and establishing a sensor space topological diagram, wherein intelligent sensors are nodes of the topological diagram, two sensors which are directly and electrically connected are connected by straight lines to form side lines of the topological diagram, and the length of each side line represents the distance of electrical connection between the two corresponding sensors; the sensor space topological graph not only comprises the space position information of the sensors and the GIS, but also comprises the electrical connection information of each sensor on a partial discharge signal propagation path.
S2, judging the abnormal condition of the ultrahigh frequency signal: when partial discharge or external interference occurs, map data acquired by the ultrahigh frequency sensors can present partial discharge characteristics, the electromagnetic signal intensity can be higher than that in normal conditions, a method of combining map recognition based on deep learning and ultrahigh frequency signal intensity dynamic threshold diagnosis is utilized to judge whether ultrahigh frequency signals of the sensors are abnormal, and the sensors with abnormal signals are added into an abnormal sensor set; the step S2 specifically includes the following steps:
s21, the intelligent ultrahigh frequency sensor is started automatically to acquire data, and ultrahigh frequency map data are acquired;
s22, diagnosing ultrahigh frequency map data by using a map identification method based on deep learning, judging whether the map data is abnormal or not, and identifying a sensor acquiring abnormal data; judging that the map data of the sensors 1-6 and 8 are abnormal;
s23, calculating the ultrahigh frequency signal intensity collected by the sensor with abnormal data:
V i ＝N i ×F i (1)
in the formula: v i The intensity of the ultrahigh frequency signal acquired by the sensor for the ith time; n is a radical of i And F i Respectively acquiring the number of pulses and the average discharge amplitude of data acquired by the ultrahigh frequency sensor for the ith time;
s24, calculating the cumulative average signal intensity of data acquired by the ultrahigh frequency sensor with abnormal data before the ith time:
in the formula: t is the length of the cumulative time window and represents the selected cumulative times; n is a radical of j Acquiring the number of pulses of normal data for the jth time of the ultrahigh frequency sensor; f j Acquiring the average discharge amplitude of the normal data for the jth time of the ultrahigh frequency sensor;
the signal intensity of the sensor and the average signal intensity accumulated for 30 days were obtained, and the signal intensity values of each sensor are shown in the following table:
| numbering | Signal strength | Cumulative |
| 1 | 28.4 | 5.4 |
| 2 | 44.7 | 5.7 |
| 3 | 146.1 | 9.8 |
| 4 | 121.5 | 8.1 |
| 5 | 5.3 | 6.2 |
| 6 | 68.6 | 4.5 |
| 8 | 10.4 | 4.1 |
TABLE 1
S25, screening sensors with abnormal data, and when the sensors are V i >V j The sensor is added to the abnormal sensor set. Comparing the signal intensities of the 7 sensors in table 1 with the cumulative average signal intensity, sensors 1, 2, 3, 4, 6, 8 were added to the abnormal sensor set.
S3, an abnormal sensor correlation analysis step: and (2) performing signal correlation analysis on each abnormal sensor in the abnormal sensor set by combining the sensor space topological graph (see fig. 4) established in the step (S1) to find out a correlation sensor of the abnormal sensor on a signal propagation path, wherein the specific steps are as follows:
s31, selecting 2 abnormal sensors;
s32, calculating the shortest path among the 2 abnormal sensors by utilizing a Dijkstra algorithm, wherein the following table is shown:
| sensor with a |
1 | 2 | 3 | 4 | 6 | 8 |
| 1 | - | 1-3-2 | 1-3 | 1-4 | 1-3-6 | 1-10-8 |
| 2 | - | - | 2-3 | 2-3-4 | 2-6 | 2-10-8 |
| 3 | - | - | - | 3-4 | 3-6 | 3-10-8 |
| 4 | - | - | - | - | 4-3-6 | 4-5-7-8 |
| 6 | - | - | - | - | 6-10-8 |
TABLE 2
S33, calculating the number k of sensors on the shortest path, and judging the sensors on the shortest path among the 2 abnormal sensors as related sensors when k is less than or equal to 2; when k is greater than 2, judging that no correlation sensor exists among the 2 abnormal sensors;
and S34, repeating the steps S31 to S33, obtaining the related sensors among all the abnormal sensors as the sensor 10, and adding the sensors 1, 2, 3, 4, 6, 8 and 10 into the abnormal sensor set.
S4, a multi-discharge source separation step: carrying out homology analysis by comparing the correlation of characteristic sequences of monitoring data of different sensors, judging the homology of the data, separating a plurality of partial discharge sources, constructing a community network, determining a signal community according to the community network, and completing the separation of the discharge sources; the step S4 specifically includes the following steps:
s41, referring to fig. 5, extracting each sensor in the abnormal sensor set in combination with the sensor space topological graph (see fig. 4) in step 1 to form an abnormal sensor topological graph (see fig. 5);
s42, homology analysis: according to the attenuation characteristics of electromagnetic waves, the intensity of electromagnetic wave signals emitted by the same discharge source can be attenuated to different degrees along a propagation path, but the variation trends of the electromagnetic wave signals are the same, homology analysis is carried out by comparing the correlation of characteristic sequences of monitoring data of sensors 1, 2, 3, 4, 6, 8 and 10, the characteristic sequences comprise a daily average discharge amplitude sequence and a daily average discharge frequency sequence, and the specific steps of the homology analysis are as follows:
s421, acquiring historical m-day partial discharge monitoring data of the abnormal sensor centralized sensor;
s422, calculating a daily average discharge amplitude sequence and a daily average discharge frequency sequence of 30 days in history;
s423, normalizing the daily average discharge amplitude sequence and the daily average discharge frequency sequence, wherein the normalization formula is as follows:
in the formula, x * Is the value after the normalization of each vector in the feature sequence, x is the value of each vector in the feature sequence, x min Is the minimum value of the vector in the feature sequence, x max The maximum value of the vector in the characteristic sequence;
normalizing the quantities in the characteristic sequence to obtain a normalized daily average discharge amplitude sequence and a normalized daily average discharge frequency sequence;
s424, calculating the daily average discharge amplitude similarity r between the sensors in the abnormal sensor topological graph c Similarity to the average daily discharge number r d The calculation formula is as follows:
wherein r is the similarity of A, B sensor characteristic sequence, A (t) is the characteristic sequence of A sensor,
is the mean value sequence of A (t), B (t) is the characteristic sequence of the B sensor,
is the mean sequence of B (t);
r between the sensors at two ends of each edge in the abnormal sensor topological graph is calculated and obtained through a formula (3) and a formula (4) c And r d R between the sensors c As shown in table 3:
| sensor with a |
1 | 2 | 3 | 4 | 6 | 8 | 10 |
| 1 | - | 0.7309 | 0.8438 | 0.8785 | 0.7317 | 0.3215 | 0.3441 |
| 2 | - | - | 0.8938 | 0.7486 | 0.8871 | 0.2957 | 0.3586 |
| 3 | - | - | - | 0.9034 | 0.9164 | 0.3167 | 0.3479 |
| 4 | - | - | - | - | 0.7682 | 0.3074 | 0.3085 |
| 6 | - | - | - | - | - | 0.2851 | 0.2964 |
| 8 | - | - | - | - | - | - | 0.6932 |
TABLE 3
Between each sensor r d As shown in table 4:
| sensor with a |
1 | 2 | 3 | 4 | 6 | 8 | 10 |
| 1 | - | 0.7814 | 0.8945 | 0.9013 | 0.7476 | 0.3135 | 0.5879 |
| 2 | - | - | 0.8913 | 0.7501 | 0.8937 | 0.3079 | 0.3028 |
| 3 | - | - | - | 0.9241 | 0.9187 | 0.2986 | 0.2812 |
| 4 | - | - | - | - | 0.7537 | 0.2945 | 0.2887 |
| 6 | - | - | - | - | - | 0.2894 | 0.2985 |
| 8 | - | - | - | - | - | - | 0.7259 |
TABLE 4
S425, judging the homology of signals: judging that the abnormal data monitored by the two sensors have correlation when the similarity satisfies the following formula, wherein the abnormal data come from the same discharge source:
r c ∩r d >δ (5)
in the formula, δ is a similarity threshold, δ is greater than or equal to 0.6 and less than or equal to 0.9, and δ =0.65 is taken in the example.
Comparing the similarity of tables 3 and 4 by using formula 5 to obtain that the sensors 1, 2, 3, 4 and 6 have homology, and the sensors 8 and 10 have homology;
s43, constructing a single community network: two sensors which are judged to be the same discharge source are used as nodes of a community network, and a sideline is added between the two nodes to form 1 single community network;
s44, please refer to fig. 6, repeat step S43 to obtain a single community network among other nodes, and continuously merge and overlap all the single community networks to obtain a final community network (see fig. 6);
s45, separating a discharge source: determining a signal community according to the final community network to complete the separation of the discharge source; in the community network, nodes connected with edges form 1 signal community, nodes connected with no edges form 1 signal community independently, and each signal community corresponds to an abnormal signal source.
Thus, 2 signal communities, namely 2 abnormal discharge sources, namely a signal community R composed of sensors 1, 2, 3, 4 and 6 and a signal community S composed of sensors 8 and 10 can be separated in the example.
S5, interference signal elimination: referring to fig. 2, the ultrahigh frequency data signal intensities of the sensors in the signal community are compared to eliminate interference signals, so as to effectively identify partial discharge signals inside the electrical equipment; the step S5 specifically includes the following steps:
s51, checking whether the signal community contains a space ultrahigh frequency sensor or not, if not, judging that a discharge source of the signal community is internal partial discharge, and if so, entering the step S52;
s52, adjusting the acquisition frequency of the sensor, triggering and acquiring 5 groups of ultrahigh frequency abnormal data at the same time, filtering the data acquired at the ultrahigh frequency in the space, and removing noise interference;
s53, calculating the signal intensity mean value of each ultrahigh frequency sensor in the signal community, wherein the calculation formula of the signal intensity mean value is as follows:
wherein V Av The signal intensity mean value of the sensor A is obtained; n is a radical of Ai The number of pulses of the ith group of signals of the sensor A; f Ai The average discharge amplitude of the ith group signal of the sensor A is obtained; n is the number of abnormal data groups collected by the abnormal sensor group;
s54, comparing the signal intensity mean values of the sensors, and if the signal intensity mean value of the space ultrahigh frequency sensor is the largest, judging that the discharge source of the signal community is an external interference signal; and if the average value of the signal intensity of the partial discharge ultrahigh frequency sensor is maximum, judging that the discharge source of the signal community is an internal partial discharge signal.
In the embodiment, as the signal community R does not contain a space ultrahigh frequency sensor, the discharge source of the signal community is judged to be internal partial discharge; the signal community S comprises a space ultrahigh frequency sensor, the signal intensity mean value of the sensors 8 and 10 is calculated, the signal intensity mean value of the sensor 10 is obtained through comparison and is larger than the signal intensity mean value of the sensor 8, and the discharging source of the signal community is judged to be an external interference signal.
S6, a partial discharge source positioning step: the method comprises the following steps of judging the approximate position of a partial discharge source in a signal community by comparing the signal intensity acquired by sensors at different positions in the signal community R, and positioning the internal partial discharge defect, wherein the step S6 specifically comprises the following steps:
s61, according to a signal intensity mean value calculation formula (namely formula (6)) in the step 5, obtaining a local discharge ultrahigh-frequency sensor A with the maximum signal intensity mean value in a signal community R as a sensor 3, obtaining a local discharge ultrahigh-frequency sensor B with the second-order signal intensity mean value as a sensor 4, and combining a sensor space topological graph in the step 1 to obtain a suspected discharge gas chamber which is an isolation switch gas chamber on the upper side of the sensor 3 and on the left side of the sensor 4, wherein the isolation switch gas chamber is shown in FIG. 7;
s62, dividing the suspected discharge gas chamber by using a perpendicular bisector between the sensor 3 and the sensor 4 to obtain a region between the perpendicular bisector and the sensor 3 as a suspected discharge region, such as a dotted line filling part in FIG. 7;
s63, obtaining S suspected discharge areas through the step S62, and judging that the intersection of the S suspected discharge areas is the area where the partial discharge source is located; if the s suspected discharge areas have no intersection, judging that the union of the s suspected discharge areas is the area where the partial discharge source is located.
In this embodiment, the positioning method provided by the present invention is applied to analyze data acquired by the wireless passive intelligent ultrahigh frequency sensor, and finally, it is determined that the partial discharge source is located in the isolator air chamber on the upper side of the sensor 3 and on the left side of the sensor 4, and the position is more biased toward the sensor 3.
The invention discloses a GIS (geographic information system) partial discharge positioning method based on an intelligent ultrahigh frequency sensor, which provides a spatial positioning method suitable for a wireless passive intelligent ultrahigh frequency sensor with low power consumption by utilizing signal amplitude values and spatial orientation information of a plurality of sensors according to the characteristic that the attenuation degree from a signal generated by the same discharge source to different position sensors is different. The GIS partial discharge positioning method based on the intelligent ultrahigh frequency sensor can effectively identify partial discharge signals generated in the GIS and determine the approximate discharge position under the conditions of a plurality of discharge sources and external interference.
In summary, according to the GIS local discharge positioning method based on the intelligent ultrahigh frequency sensor, the wireless passive ultrahigh frequency sensor is deployed in the GIS body and the surrounding space to sense abnormal electromagnetic signals, the spatial topological relation between the sensor and the GIS is combined, the strength and the characteristic similarity of the ultrahigh frequency signals are utilized to eliminate spatial interference signals, local discharge signals inside the GIS are identified, the position of a local discharge source is judged, the deployed intelligent ultrahigh frequency sensor is utilized to collect the ultrahigh frequency abnormal signals, the positioning method can be utilized on a platform end to perform defect positioning, manual retest positioning is not needed, resource investment is reduced, and diagnosis efficiency is improved.
It should be understood by those skilled in the art that the above embodiments are only for illustrating the present invention and are not to be used as a limitation of the present invention, and that the changes and modifications of the above embodiments are within the scope of the appended claims as long as they are within the true spirit of the present invention.
Claims (6)
1. A GIS partial discharge positioning method based on an intelligent ultrahigh frequency sensor is characterized by comprising the following steps:
s1, a sensor spatial position correlation step: installing a plurality of partial discharge ultrahigh frequency sensors on a GIS body, installing a plurality of space ultrahigh frequency sensors on the wall of a surrounding wall of the GIS, coding all the sensors, associating the coded sensors with the GIS space positions according to the GIS space structure, an electrical connection diagram and the sensor installation positions, and establishing a sensor space topological diagram, wherein intelligent sensors comprising the partial discharge ultrahigh frequency sensors and the space ultrahigh frequency sensors are nodes of the topological diagram, two sensors which are directly and electrically connected form side lines of the topological diagram by straight line connection, and the length of each side line represents the distance of electrical connection between the corresponding two sensors;
s2, judging the abnormal condition of the ultrahigh frequency signal: judging whether the ultrahigh frequency signals of the sensors are abnormal or not by utilizing a method of combining map recognition based on deep learning and ultrahigh frequency signal intensity dynamic threshold diagnosis, and adding the sensors with abnormal signals into an abnormal sensor set;
s3, an abnormal sensor correlation analysis step: and (2) performing signal correlation analysis on each abnormal sensor in the abnormal sensor set by combining the sensor space topological graph established in the step (S1) to find out a correlation sensor of the abnormal sensor on a signal propagation path, wherein the specific steps are as follows:
s31, selecting 2 abnormal sensors;
s32, calculating the shortest path among the 2 abnormal sensors by utilizing a Dijkstra algorithm;
s33, calculating the number k of sensors on the shortest path, and judging the sensors on the shortest path among the 2 abnormal sensors as related sensors when k is less than or equal to 2; when k is greater than 2, judging that no correlation sensor exists among the 2 abnormal sensors;
s34, repeating the steps S31-S33, obtaining the associated sensors among all the abnormal sensors, and adding the associated sensors into an abnormal sensor set;
s4, a multi-discharge source separation step: carrying out homology analysis by comparing the correlation of characteristic sequences of monitoring data of different sensors, judging the homology of the data, separating a plurality of partial discharge sources, constructing a community network, determining a signal community according to the community network, and completing the separation of the discharge sources;
s5, interference signal elimination: the ultrahigh frequency data signal intensity of each sensor in the signal community is compared, interference signals are eliminated, and effective identification of partial discharge signals in the electrical equipment is achieved;
s6, a partial discharge source positioning step: the approximate position of the partial discharge source in the signal community is judged by comparing the signal intensity acquired by different position sensors in the signal community, and the internal partial discharge defect is positioned.
2. The GIS local discharge positioning method based on the intelligent ultrahigh frequency sensor according to claim 1, wherein in step S1, the sensor spatial topological graph not only includes the spatial position information of the sensor and the GIS, but also includes the electrical connection information of each sensor on the local discharge signal propagation path.
3. The GIS partial discharge positioning method based on the intelligent ultrahigh frequency sensor according to claim 1, wherein the step S2 specifically comprises the following steps:
s21, the intelligent ultrahigh frequency sensor is started automatically to acquire data, and ultrahigh frequency map data are acquired;
s22, diagnosing ultrahigh frequency map data by using a map identification method based on deep learning, judging whether the map data is abnormal or not, and identifying a sensor acquiring abnormal data;
s23, calculating the ultrahigh frequency signal intensity collected by the sensor with abnormal data:
V i ＝N i ×F i
in the formula: v i The intensity of the ultrahigh frequency signal acquired by the sensor for the ith time; n is a radical of hydrogen i And F i Respectively acquiring the pulse number and the average discharge amplitude of data acquired for the ith time by the ultrahigh frequency sensor;
s24, calculating the cumulative average signal intensity of data acquired by the ultrahigh frequency sensor with abnormal data before the ith time:
in the formula: t is the length of the cumulative time window and represents the selected cumulative times; n is a radical of j Acquiring the number of pulses of normal data for the jth time of the ultrahigh frequency sensor; f j Acquiring the average discharge amplitude of the normal data for the jth time of the ultrahigh frequency sensor;
s25, screening sensors with abnormal data when V i >V j The sensor is added to the abnormal sensor set.
4. The GIS local discharge positioning method based on the intelligent ultrahigh frequency sensor according to claim 1, wherein in step S4, the characteristic sequence comprises a daily average discharge amplitude sequence and a daily average discharge frequency sequence; step S4 specifically includes the following steps:
s41, extracting all sensors in the abnormal sensor set by combining the sensor space topological graph in the step 1 to form an abnormal sensor topological graph;
s42, homology analysis: according to the attenuation characteristics of electromagnetic waves, the intensity of electromagnetic wave signals emitted by the same discharge source can be attenuated to different degrees along a propagation path, but the variation trends of the electromagnetic wave signals are the same, homology analysis is carried out by comparing the correlation of characteristic sequences of monitoring data of different sensors, and the specific steps of the homology analysis are as follows:
s421, acquiring local discharge monitoring data of the abnormal sensor set sensor in historical m days;
s422, calculating a daily average discharge amplitude sequence and a daily average discharge frequency sequence of historical m days;
s423, carrying out normalization processing on the daily average discharge amplitude sequence and the daily average discharge frequency sequence, wherein the normalization formula is as follows:
in the formula, x * Is the value after the normalization of each vector in the feature sequence, x is the value of each vector in the feature sequence, x min Is the minimum value of the vector in the feature sequence, x max The maximum value of the vector in the characteristic sequence is taken;
normalizing the quantities in the characteristic sequence to obtain a normalized daily average discharge amplitude sequence and a normalized daily average discharge frequency sequence;
s424, calculating the daily average discharge amplitude similarity r between the sensors in the abnormal sensor topological graph c Similarity r with daily average discharge frequency d The calculation formula is as follows:
wherein r is the similarity of A, B sensor characteristic sequence, A (t) is the characteristic sequence of A sensor,
is the mean value sequence of A (t), B (t) is the characteristic sequence of B sensor,
is the mean sequence of B (t);
s425, judging the signal homology: judging that the abnormal data monitored by the two sensors have correlation when the similarity satisfies the following formula, wherein the abnormal data come from the same discharge source:
r c ∩r d >δ
wherein delta is a similarity threshold value, and is more than or equal to 0.6 and less than or equal to 0.9;
s43, constructing a single community network: two sensors which are judged to be the same discharge source are used as nodes of a community network, and a sideline is added between the two nodes to form 1 single community network;
s44, repeating the step S43 to obtain a single community network among other nodes, and continuously combining and overlapping all the single community networks to obtain a final community network;
s45, separating a discharge source: determining a signal community according to the final community network to complete the separation of the discharge source; in the community network, nodes connected with edges form 1 signal community, nodes connected with no edges form 1 signal community independently, and each signal community corresponds to an abnormal signal source.
5. The GIS partial discharge positioning method based on intelligent ultrahigh frequency sensor according to claim 1, wherein the step S5 specifically comprises the following steps:
s51, checking whether the signal community contains a space ultrahigh frequency sensor, if not, judging that a discharge source of the signal community is internal partial discharge, and if so, entering the step S52;
s52, adjusting the acquisition frequency of the sensor, triggering and acquiring n groups of ultrahigh frequency abnormal data at the same time, filtering the data acquired at the ultrahigh frequency in the space, and removing noise interference;
s53, calculating the signal intensity mean value of each ultrahigh frequency sensor in the signal community, wherein the calculation formula of the signal intensity mean value is as follows:
wherein V Av The signal intensity mean value of the sensor A is obtained; n is a radical of Ai The number of pulses of the ith group of signals of the sensor A; f Ai The average discharge amplitude of the ith group signal of the sensor A is obtained; n is the number of abnormal data groups collected by the abnormal sensor group;
s54, comparing the signal intensity mean values of the sensors, and if the signal intensity mean value of the space ultrahigh frequency sensor is the largest, judging that the discharge source of the signal community is an external interference signal; and if the average value of the signal intensity of the partial discharge ultrahigh-frequency sensor is maximum, judging that the discharge source of the signal community is an internal partial discharge signal.
6. The GIS partial discharge positioning method based on the intelligent ultrahigh frequency sensor according to claim 5, wherein the step S6 specifically comprises the following steps:
s61, obtaining a local discharge ultrahigh-frequency sensor A with the largest signal intensity mean value and a second-largest local discharge ultrahigh-frequency sensor B in the signal community according to the signal intensity mean value calculation formula in the step 5, and obtaining a suspected discharge gas chamber which is a gas chamber directly and electrically connected with the local discharge ultrahigh-frequency sensor A and the local discharge ultrahigh-frequency sensor B by combining the sensor space topological graph in the step S1; meanwhile, the air chamber only provided with 1 partial discharge ultrahigh frequency sensor A is listed as a suspected discharge air chamber;
s62, dividing a suspected discharge gas chamber by using a perpendicular bisector between the partial discharge ultrahigh frequency sensor A and the partial discharge ultrahigh frequency sensor B to obtain a region between the perpendicular bisector and the partial discharge ultrahigh frequency sensor A as a suspected discharge region; if only one partial discharge ultrahigh-frequency sensor is deployed in the suspected discharge gas chamber, the gas chamber is not required to be divided into a suspected discharge area;
s63, obtaining S suspected discharge areas through the step S62, and judging that the intersection of the S suspected discharge areas is the area where the partial discharge source is located; if the s suspected discharge areas have no intersection, judging that the union of the s suspected discharge areas is the area where the partial discharge source is located.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010487230.0A CN111610418B (en) | 2020-05-28 | 2020-05-28 | GIS partial discharge positioning method based on intelligent ultrahigh frequency sensor |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010487230.0A CN111610418B (en) | 2020-05-28 | 2020-05-28 | GIS partial discharge positioning method based on intelligent ultrahigh frequency sensor |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111610418A CN111610418A (en) | 2020-09-01 |
| CN111610418B true CN111610418B (en) | 2022-11-22 |
Family
ID=72195369
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010487230.0A Active CN111610418B (en) | 2020-05-28 | 2020-05-28 | GIS partial discharge positioning method based on intelligent ultrahigh frequency sensor |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111610418B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112710928B (en) * | 2020-12-10 | 2023-02-21 | 国网宁夏电力有限公司电力科学研究院 | Direct-current partial discharge waveform interference removing method and system based on autocorrelation analysis |
| CN114487741A (en) * | 2022-04-02 | 2022-05-13 | 国网山东省电力公司东营供电公司 | Electrical equipment live detection method and system |
| CN115616363A (en) * | 2022-12-06 | 2023-01-17 | 北京鼎诚鸿安科技发展有限公司 | Intelligent diagnosis device and diagnosis method for partial discharge data |
| CN116626454B (en) * | 2023-05-26 | 2024-01-02 | 国网江苏省电力有限公司电力科学研究院 | Oil paper insulation UHF partial discharge signal anti-interference identification and positioning method and device based on correction time-frequency clustering |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR101402887B1 (en) * | 2013-01-30 | 2014-06-27 | (주)에이피엠테크놀러지스 | System and method of monitoring GIS partial discharge having analyzing function linked time and frequency domain, local unit therefor |
| CN104155585A (en) * | 2014-08-12 | 2014-11-19 | 国家电网公司 | GIS partial discharge type identification method based on GK fuzzy clustering |
| CN205246818U (en) * | 2015-12-16 | 2016-05-18 | 国家电网公司 | Device is taked care of with three frequency centers to separation of many partial discharge source |
| CN105606975A (en) * | 2016-03-09 | 2016-05-25 | 武汉华威众科电力有限公司 | Localizable ultrahigh-frequency cable partial discharge detection method and device |
| CN106771935A (en) * | 2017-02-23 | 2017-05-31 | 全球能源互联网研究院 | A kind of extra-high video sensor drawn game portion discharge on-line monitoring system |
| CN107831422A (en) * | 2017-11-02 | 2018-03-23 | 广东电网有限责任公司电力科学研究院 | A kind of GIS equipment partial discharge diagnostic method and system |
| CN108896891A (en) * | 2018-08-26 | 2018-11-27 | 国网天津市电力公司 | A kind of GIS multi-source shelf depreciation separation recognition methods based on superfrequency method |
| CN110514970A (en) * | 2019-09-03 | 2019-11-29 | 国网湖南省电力有限公司 | A kind of GIS partial discharge source localization method, system and medium based on data-driven |
-
2020
- 2020-05-28 CN CN202010487230.0A patent/CN111610418B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR101402887B1 (en) * | 2013-01-30 | 2014-06-27 | (주)에이피엠테크놀러지스 | System and method of monitoring GIS partial discharge having analyzing function linked time and frequency domain, local unit therefor |
| CN104155585A (en) * | 2014-08-12 | 2014-11-19 | 国家电网公司 | GIS partial discharge type identification method based on GK fuzzy clustering |
| CN205246818U (en) * | 2015-12-16 | 2016-05-18 | 国家电网公司 | Device is taked care of with three frequency centers to separation of many partial discharge source |
| CN105606975A (en) * | 2016-03-09 | 2016-05-25 | 武汉华威众科电力有限公司 | Localizable ultrahigh-frequency cable partial discharge detection method and device |
| CN106771935A (en) * | 2017-02-23 | 2017-05-31 | 全球能源互联网研究院 | A kind of extra-high video sensor drawn game portion discharge on-line monitoring system |
| CN107831422A (en) * | 2017-11-02 | 2018-03-23 | 广东电网有限责任公司电力科学研究院 | A kind of GIS equipment partial discharge diagnostic method and system |
| CN108896891A (en) * | 2018-08-26 | 2018-11-27 | 国网天津市电力公司 | A kind of GIS multi-source shelf depreciation separation recognition methods based on superfrequency method |
| CN110514970A (en) * | 2019-09-03 | 2019-11-29 | 国网湖南省电力有限公司 | A kind of GIS partial discharge source localization method, system and medium based on data-driven |
Non-Patent Citations (1)
| Title |
|---|
| 邵先军 等." 利用累积能量函数特征参量优化提取的多源局部放电信号分离技术".《中国电机工程学报》.2017,第37卷(第11期), * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111610418A (en) | 2020-09-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111610418B (en) | GIS partial discharge positioning method based on intelligent ultrahigh frequency sensor | |
| KR101574615B1 (en) | A partial discharge monitoring and diagnosis system for power devices by using signal detection based on statistical phase-angle patterns | |
| WO2019015311A1 (en) | Vibration signal support vector machine-based gil fault online monitoring system | |
| US10422715B2 (en) | GIS leak monitoring method based on vibration signal | |
| AU2006326931B2 (en) | Method and apparatus for detecting an event | |
| CN107390097A (en) | A kind of acoustoelectric combined shelf depreciation simulation detection system of GIS and its detection method | |
| EP2579055A1 (en) | METHOD FOR THE CONTINUOUS MONITORING AND DIAGNOSIS OF SOURCES OF PARTIAL DISCHARGES (PDs) IN HIGH-VOLTAGE CABLES DURING CONNECTION TO, AND OPERATION IN THE POWER GRID, AND PHYSICAL SYSTEM FOR CARRYING OUT SAME | |
| CN109799434A (en) | Partial discharge pattern recognition system and method | |
| CN110703078A (en) | GIS fault diagnosis method based on spectral energy analysis and self-organizing competition algorithm | |
| CN112307969B (en) | Pulse signal classification identification method and device and computer equipment | |
| CN102230902A (en) | Method for visually and intelligently identifying internal defects of GIS (Geographic Information System) equipment | |
| CN112069930A (en) | Vibration signal processing method and device for improving GIS equipment fault diagnosis accuracy | |
| El-Hag | Application of machine learning in outdoor insulators condition monitoring and diagnostics | |
| CN111999620A (en) | Multi-channel joint positioning method for partial discharge of power equipment | |
| KR102377939B1 (en) | Partial discharge monitoring and diagnosis system for distribution board using ultra frequency and high frequency current transformer signal | |
| Chan et al. | Automatic blind equalization and thresholding for partial discharge measurement in power transformer | |
| CN208076016U (en) | A kind of GIL On-line Faults monitoring system based on vibration signal support vector machines | |
| CN115954291B (en) | Crack monitoring system of TOPCON structure photovoltaic cell panel | |
| CN116910470A (en) | GIS combined electrical apparatus partial discharge fault mode identification method | |
| KR102243313B1 (en) | PARTIAL DISCHARGE JUDGING METHOD and DIAGNOSTIC SYSTEM | |
| CN114660424A (en) | GIS equipment partial discharge detection diagnostic system | |
| CN109917245B (en) | Ultrasonic detection partial discharge signal mode identification method considering phase difference | |
| CN114167240A (en) | Built-in wireless self-energy-taking ultrahigh frequency partial discharge detection method | |
| CN112526404A (en) | Power transformer fault on-line positioning system based on ultrahigh frequency external sensor | |
| CN112883639A (en) | GIS equipment service life prediction device and method based on machine learning |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |