CN117347796A - Intelligent gateway-based switching equipment partial discharge diagnosis system and method - Google Patents
Intelligent gateway-based switching equipment partial discharge diagnosis system and method Download PDFInfo
- Publication number
- CN117347796A CN117347796A CN202311271326.3A CN202311271326A CN117347796A CN 117347796 A CN117347796 A CN 117347796A CN 202311271326 A CN202311271326 A CN 202311271326A CN 117347796 A CN117347796 A CN 117347796A
- Authority
- CN
- China
- Prior art keywords
- partial discharge
- discharge
- network model
- tev
- sensing data
- 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
- 208000028659 discharge Diseases 0.000 title claims abstract description 329
- 238000003745 diagnosis Methods 0.000 title claims abstract description 37
- 238000000034 method Methods 0.000 title claims abstract description 30
- 238000004364 calculation method Methods 0.000 claims abstract description 24
- 238000003062 neural network model Methods 0.000 claims abstract description 23
- 238000012544 monitoring process Methods 0.000 claims abstract description 21
- 230000004927 fusion Effects 0.000 claims abstract description 10
- 238000013500 data storage Methods 0.000 claims abstract description 7
- 238000007726 management method Methods 0.000 claims abstract description 7
- 239000011159 matrix material Substances 0.000 claims description 66
- 238000012549 training Methods 0.000 claims description 25
- 238000013527 convolutional neural network Methods 0.000 claims description 21
- 238000001514 detection method Methods 0.000 claims description 19
- 238000007619 statistical method Methods 0.000 claims description 8
- QVFWZNCVPCJQOP-UHFFFAOYSA-N chloralodol Chemical compound CC(O)(C)CC(C)OC(O)C(Cl)(Cl)Cl QVFWZNCVPCJQOP-UHFFFAOYSA-N 0.000 claims description 4
- 238000004891 communication Methods 0.000 claims description 4
- 238000002405 diagnostic procedure Methods 0.000 claims 2
- 230000007547 defect Effects 0.000 abstract description 9
- 238000012360 testing method Methods 0.000 description 6
- 238000007689 inspection Methods 0.000 description 5
- 238000009826 distribution Methods 0.000 description 4
- 230000005856 abnormality Effects 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000009413 insulation Methods 0.000 description 3
- 238000012423 maintenance Methods 0.000 description 3
- 238000011176 pooling Methods 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 238000012795 verification Methods 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 230000006835 compression Effects 0.000 description 1
- 238000007906 compression Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003909 pattern recognition Methods 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 238000003860 storage Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
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
- G01R31/1227—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 of components, parts or materials
- G01R31/1263—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 of components, parts or materials of solid or fluid materials, e.g. insulation films, bulk material; of semiconductors or LV electronic components or parts; of cable, line or wire insulation
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Testing Relating To Insulation (AREA)
Abstract
The invention discloses a switch equipment partial discharge diagnosis system and method based on an intelligent gateway, wherein the system comprises: the intelligent sensing terminal, the intelligent gateway and the remote monitoring terminal; the intelligent sensing terminal acquires partial discharge sensing data of the switching equipment, wherein the partial discharge sensing data comprises characteristic parameters and a map of partial discharge; the intelligent gateway performs edge calculation according to the partial discharge sensing data to perform discharge diagnosis on the switching equipment, performs partial discharge type identification based on neural network model fusion decision, and performs data storage and uploading management on an edge calculation result; and the remote monitoring terminal carries out partial discharge alarm prompt according to the edge calculation and identification result. According to the invention, the switching equipment is subjected to discharge diagnosis, the weight coefficients are constructed based on the recognition results of the multiple neural network models on different defects, the weight coefficients output by the different models are organically combined to be used as the discharge type recognition result, and the discharge type recognition based on multiple state quantities is realized.
Description
Technical Field
The invention relates to the technical field of partial discharge diagnosis, in particular to a switching equipment partial discharge diagnosis system and method based on an intelligent gateway.
Background
The high-voltage switch equipment is a key element for controlling and protecting the power system, in particular to a distribution network switch cabinet which has large equipment quantity, multiple equipment types, low manufacturing cost and poor reliability, and meanwhile, the distribution network has weak operation and maintenance force, so that the failure rate of the distribution network equipment is high.
Partial discharge detection is an important core means for insulation monitoring and state operation and maintenance of high-voltage switch equipment, and has the remarkable advantages of high sensitivity and high timeliness. Because of cost limitation, the existing partial discharge on-line detection technology is difficult to directly popularize and apply to power distribution network switch equipment, and aiming at a large number of power grid switch equipment, power grid enterprises input a large amount of manpower and financial resources to develop operation and maintenance in a live inspection mode. Because the on-site electromagnetic interference environment of the equipment is complex, the work task of partial discharge detection is heavy, the threshold of the partial discharge live detection technology is high, the partial discharge process is greatly affected by the running environment of the equipment and the condition factors, and the level of the majority of first-line inspection personnel is limited, only the inspection equipment with simple functions can be used, so that the serious problems of low detection rate of insulation defects and poor input-output ratio of the existing live inspection are caused, and the great potential risk exists for the personal life safety of the operation inspection personnel.
Disclosure of Invention
Aiming at the defects in the prior art, the invention provides a switching equipment partial discharge diagnosis system and method based on an intelligent gateway.
In a first aspect, a switching device partial discharge diagnostic system based on an intelligent gateway includes: the intelligent sensing terminal is in wireless connection with the intelligent gateway, and the intelligent gateway is in wireless connection with the remote monitoring terminal;
the intelligent sensing terminal acquires partial discharge sensing data of the switching equipment, wherein the partial discharge sensing data comprises characteristic parameters and a map of partial discharge;
the intelligent gateway performs edge calculation according to the partial discharge sensing data to perform discharge diagnosis on the switching equipment, performs partial discharge type identification based on neural network model fusion decision, and performs data storage and uploading management on an edge calculation result;
and the remote monitoring terminal carries out partial discharge alarm prompt according to the edge calculation and identification result.
Further, the intelligent gateway comprises a lateral statistical analysis module, and the lateral statistical analysis module is specifically configured to:
acquiring real-time partial discharge sensing data, and acquiring a TEV partial discharge detection value of each switch device under the same voltage level according to the real-time partial discharge sensing data;
calculating the average level of the TEV partial discharge of all the switching equipment under the same voltage level according to the TEV partial discharge detection value;
analyzing the degree of deviation of each switching device from the TEV partial discharge average level based on the TEV partial discharge average level;
and carrying out discharge diagnosis according to the degree of deviation from the TEV partial discharge average level, and identifying the switching equipment with partial discharge based on an acousto-electric combined positioning method.
Further, the intelligent gateway further comprises a discharge type identification module, and the discharge type identification module is specifically configured to:
acquiring a history partial discharge signal, wherein the history partial discharge signal comprises a history discharge type and corresponding partial discharge sensing data;
acquiring historical ultrasonic discharge data according to the historical partial discharge sensing data, performing model training based on a BP neural network model according to the historical ultrasonic discharge data, generating a first network model, and obtaining a first weight matrix;
acquiring a historical high-frequency PRPD map according to the historical partial discharge sensing data, performing model training based on a convolutional neural network model according to the historical high-frequency PRPD map, generating a second network model, and obtaining a second weight matrix;
and acquiring a historical TEV PRPD map according to the historical partial discharge sensing data, performing model training based on a convolutional neural network model according to the historical TEV PRPD map, generating a third network model, and obtaining a third weight matrix.
Further, the discharge type identification module is further configured to:
acquiring real-time partial discharge sensing data, and acquiring switching equipment discharge parameters based on a switching equipment discharge diagnosis result according to the real-time partial discharge sensing data, wherein the discharge parameters comprise, but are not limited to, real-time ultrasonic discharge data, a real-time high-frequency PRPD map and a real-time TEV PRPD map of the switching equipment;
inputting the real-time ultrasonic discharge data into a first network model to identify a discharge type, and generating a first confidence probability matrix;
inputting the real-time high-frequency PRPD map into a second network model to identify discharge types, and generating a second confidence probability matrix;
inputting the TEV PRPD pattern into a third network model to identify discharge types and generating a third confidence probability matrix;
weighting the weight matrixes of the first network model, the second network model and the third network model with the corresponding confidence probability matrixes respectively, and generating decision matrixes of the first network model, the second network model and the third network model respectively;
and acquiring the maximum decision matrix in the decision matrixes corresponding to the first network model, the second network model and the third network model, and taking the discharge type identified by the network model corresponding to the maximum decision matrix as a final partial discharge type identification result.
Further, the intelligent sensing terminal comprises, but is not limited to, an electric wave intelligent sensing terminal, a high-frequency intelligent sensing terminal and an ultrasonic intelligent sensing terminal, and the intelligent sensing terminal uploads the partial discharge sensing data to the intelligent gateway in a loRa communication mode.
In a second aspect, a method for diagnosing partial discharge of a switching device based on an intelligent gateway, where the method is based on the switching device partial discharge diagnosing system based on the intelligent gateway according to the first aspect, and the steps include:
the intelligent sensing terminal acquires partial discharge sensing data of the switching equipment, wherein the partial discharge sensing data comprises characteristic parameters and a map of partial discharge;
the intelligent gateway performs edge calculation according to the partial discharge sensing data to perform discharge diagnosis on the switching equipment, performs partial discharge type identification based on neural network model fusion decision, and performs data storage and uploading management on an edge calculation result;
and the remote monitoring terminal carries out partial discharge alarm prompt according to the edge calculation and identification result.
Further, the diagnosing the discharge of the switching device includes:
acquiring real-time partial discharge sensing data, and acquiring a TEV partial discharge detection value of each switch device under the same voltage level according to the real-time partial discharge sensing data;
calculating the average level of the TEV partial discharge of all the switching equipment under the same voltage level according to the TEV partial discharge detection value;
analyzing the degree of deviation of each switching device from the TEV partial discharge average level based on the TEV partial discharge average level;
and carrying out discharge diagnosis according to the degree of deviation from the TEV partial discharge average level, and identifying the switching equipment with partial discharge based on an acousto-electric combined positioning method.
Further, the partial discharge type identification based on the neural network model fusion decision comprises:
acquiring a history partial discharge signal, wherein the history partial discharge signal comprises a history discharge type and corresponding partial discharge sensing data;
acquiring historical ultrasonic discharge data according to the historical partial discharge sensing data, performing model training based on a BP neural network model according to the historical ultrasonic discharge data, generating a first network model, and obtaining a first weight matrix;
acquiring a historical high-frequency PRPD map according to the historical partial discharge sensing data, performing model training based on a convolutional neural network model according to the historical high-frequency PRPD map, generating a second network model, and obtaining a second weight matrix;
and acquiring a historical TEV PRPD map according to the historical partial discharge sensing data, performing model training based on a convolutional neural network model according to the historical TEV PRPD map, generating a third network model, and obtaining a third weight matrix.
Further, the method further comprises the following steps:
acquiring real-time partial discharge sensing data, and acquiring switching equipment discharge parameters based on a switching equipment discharge diagnosis result according to the real-time partial discharge sensing data, wherein the discharge parameters comprise, but are not limited to, real-time ultrasonic discharge data, a real-time high-frequency PRPD map and a real-time TEV PRPD map of the switching equipment;
inputting the real-time ultrasonic discharge data into a first network model to identify a discharge type, and generating a first confidence probability matrix;
inputting the real-time high-frequency PRPD map into a second network model to identify discharge types, and generating a second confidence probability matrix;
inputting the TEV PRPD pattern into a third network model to identify discharge types and generating a third confidence probability matrix;
weighting the weight matrixes of the first network model, the second network model and the third network model with the corresponding confidence probability matrixes respectively, and generating decision matrixes of the first network model, the second network model and the third network model respectively;
and acquiring the maximum decision matrix in the decision matrixes corresponding to the first network model, the second network model and the third network model, and taking the discharge type identified by the network model corresponding to the maximum decision matrix as a final partial discharge type identification result.
The beneficial effects of the invention are as follows: the intelligent sensing terminal is used for collecting partial discharge sensing data of the switching equipment, the intelligent gateway is used for carrying out discharge diagnosis on the switching equipment according to the partial discharge sensing data, weight coefficients are constructed based on recognition results of a plurality of neural network models on different defects, the weight coefficients output by the different models are organically combined to be used as final discharge type recognition results, the discharge type recognition based on multiple state quantities is realized, and meanwhile, the remote monitoring terminal is used for carrying out partial discharge alarm prompt according to the recognition results, so that intelligent monitoring is realized.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below. Like elements or portions are generally identified by like reference numerals throughout the several figures. In the drawings, elements or portions thereof are not necessarily drawn to scale.
Fig. 1 is a block diagram of a switchgear partial discharge diagnosis system based on an intelligent gateway according to an embodiment of the present invention;
fig. 2 is a schematic architecture diagram of a switching device partial discharge diagnosis system based on an intelligent gateway according to a first embodiment of the present invention;
fig. 3 is a schematic diagram of a neural network model training process of a switching device partial discharge diagnosis system based on an intelligent gateway according to a first embodiment of the present invention;
fig. 4 is a flowchart of a switching device partial discharge diagnosis method based on an intelligent gateway according to a second embodiment of the present invention.
Detailed Description
Embodiments of the technical scheme of the present invention will be described in detail below with reference to the accompanying drawings. The following examples are only for more clearly illustrating the technical aspects of the present invention, and thus are merely examples, and are not intended to limit the scope of the present invention.
It is noted that unless otherwise indicated, technical or scientific terms used herein should be given the ordinary meaning as understood by one of ordinary skill in the art to which this invention pertains.
Example 1
As shown in fig. 1 and 2, a switchgear partial discharge diagnosis system based on an intelligent gateway includes: the intelligent sensing terminal is in wireless connection with the intelligent gateway, and the intelligent gateway is in wireless connection with the remote monitoring terminal;
the intelligent sensing terminal acquires partial discharge sensing data of the switching equipment, wherein the partial discharge sensing data comprises characteristic parameters and a map of partial discharge;
the intelligent gateway performs edge calculation according to the partial discharge sensing data to perform discharge diagnosis on the switching equipment, performs partial discharge type identification based on neural network model fusion decision, and performs data storage and uploading management on an edge calculation result;
and the remote monitoring terminal carries out partial discharge alarm prompt according to the edge calculation and identification result.
Further, the intelligent sensing terminals include, but are not limited to, an electric wave intelligent sensing terminal, a high-frequency intelligent sensing terminal and an ultrasonic intelligent sensing terminal, and all intelligent sensing terminals upload the partial discharge sensing data to the intelligent gateway in a loRa communication mode.
Further, the intelligent gateway comprises a transverse statistical analysis module, the transverse statistical analysis module is used for acquiring real-time partial discharge sensing data acquired by the intelligent sensing terminal, acquiring a TEV partial discharge detection value of each switch device under the same voltage level according to the real-time partial discharge sensing data, simultaneously calculating TEV partial discharge average levels of all the switch devices under the same voltage level according to the TEV partial discharge detection value, and respectively analyzing the degree of each switch deviating from the TEV partial discharge average level based on the TEV partial discharge average levels, wherein the calculation formula is as follows:
wherein N is the number of switching devices under the same voltage level, V i Is the TEV partial discharge detection value of the ith switching device,for the TEV partial discharge average level of all switching devices at the same voltage level, σ is the degree to which each switching device deviates from the TEV partial discharge average level at the same voltage level.
Conventionally, the switch indoor switch cabinets are generally arranged in a straight line shape and numbered sequentially, and as a plurality of switch indoor switch cabinets are from the same manufacturer, the operation years are not very different, the operation environment and the electromagnetic environment are basically the same, and the insulation level of the normally operated switch equipment can be considered to be not obviously different. Therefore, when the degree σ of deviation from the TEV partial discharge average level is greater than the set threshold, it is determined that the switching device has a partial discharge phenomenon. Preferably, in practical application, the electric company can reasonably select the allowable equipment abnormality probability level according to annual overhaul fund budget and personnel configuration conditions, and the state judgment threshold value corresponding to the equipment abnormality probability level can be calculated by Vtev and sigma, so that whether the switching equipment has a partial discharge phenomenon can be judged according to the equipment abnormality probability level.
When the partial discharge phenomenon of a certain switching device is diagnosed, the partial discharge is detected by combining a TEV method and an ultrasonic method to perform rough positioning. Acquiring the arrival time of the TEV signal acquired by the ground electric wave intelligent sensing terminal and the arrival time of the ultrasonic signal acquired by the ultrasonic intelligent sensing terminal, and calculating the time difference t between the arrival time of the ultrasonic signal and the arrival time of the TEV signal by taking the arrival time of the TEV signal as a starting point i The time difference t i As the propagation time of the discharge acoustic signal, only the time difference t is determined i The linear distance between the local discharge position and the intelligent sensing terminal can be determined, and then the partial discharge switching equipment is positioned according to the linear distance, so that rough positioning of the partial discharge is realized, and the calculation formula is as follows:
L＝v×t i
wherein L is the linear distance between the partial discharge position and the intelligent sensing terminal, and v is the equivalent propagation speed of the discharge acoustic signal.
Through the transverse statistical analysis module, the intelligent gateway can carry out statistical analysis and threshold comparison on the node data of a single intelligent terminal, and can also carry out transverse analysis on the node data of the same type of intelligent terminals on all the switch cabinets, and if the detection result of certain switch cabinet equipment is larger than that of other switch cabinet equipment under the same voltage level, the equipment can be judged that partial discharge is likely to occur.
Further, the intelligent gateway also comprises a discharge type identification module, which is used for carrying out pattern identification on partial discharge by utilizing a machine learning algorithm according to the real-time partial discharge sensing data acquired by the intelligent sensing terminal.
Specifically, the discharge type identification module adopts a convolutional neural network model to identify the defect type of the partial discharge map, and adopts a BP artificial neural network to identify the defect type of the ultrasonic statistical characteristic parameter, and the method comprises the following steps:
the method comprises the steps of obtaining a history partial discharge signal, wherein the history partial discharge signal comprises a history discharge type, and corresponding history partial discharge sensing data collected by an intelligent sensing terminal in a history monitoring time period, obtaining history ultrasonic discharge data, a history high-frequency PRPD (pulse-with-noise) map and a history TEV (pulse-with-noise) PRPD (pulse-with-noise) map according to the history partial discharge sensing data, and randomly dividing each type of history partial discharge sensing data into a test set and a training set respectively.
As shown in fig. 3, the ultrasonic statistical characteristic parameters are analyzed, the training set of the historical ultrasonic discharge data is utilized to train the BP neural network model, the test set is input into the BP neural network model for verification, the recognition accuracy is obtained, the recognition accuracy reflects the difference between the training result and the verification result, the weight matrix of the model is determined according to the recognition accuracy, and then the first network model is generated, so that the first weight matrix is obtained.
Analyzing the high-frequency PRPD map, carrying out image normalization processing on a training set and a test set of the historical high-frequency PRPD map, carrying out model training on the convolutional neural network model by using the normalized historical high-frequency PRPD map training set, and verifying by using the test set to determine a weight matrix of the model, so as to generate a second network model, and obtaining a second weight matrix.
Analyzing the TEV PRPD pattern, carrying out image normalization processing on a training set and a test set of the historical TEV PRPD pattern, training a convolutional neural network model by using the normalized historical TEV PRPD pattern training set, and verifying by using the test set to determine a weight matrix of the model, so as to generate a third network model, and obtaining a third weight matrix.
Further, real-time partial discharge sensing data acquired by the intelligent sensing terminal are acquired, and real-time discharge parameters corresponding to the switching equipment with partial discharge are acquired based on the switching equipment discharge diagnosis result, wherein the discharge parameters comprise, but are not limited to, real-time ultrasonic discharge data, a real-time high-frequency PRPD map and a real-time TEV PRPD map of the switching equipment.
Inputting real-time ultrasonic discharge data into a first network model to identify a discharge type based on a BP neural network according to an ultrasonic discharge signal and generate a first confidence probability matrix; inputting the real-time high-frequency PRPD pattern into a second network model to identify the discharge type based on a convolutional neural network according to the high-frequency PRPD pattern and generate a second confidence probability matrix; inputting the TEV PRPD pattern into a third network model to identify discharge types based on a convolutional neural network according to the TEV discharge pattern and generating a third confidence probability matrix.
Performing weighted operation on a first weight matrix of the first network model and a first confidence probability matrix to generate a first decision matrix; performing weighted operation on a second weight matrix of the second network model and a second confidence probability matrix to generate a second decision matrix; and carrying out weighted operation on a third weight matrix of the third network model and a third confidence probability matrix to generate a third decision matrix. And acquiring the maximum decision matrix in the first decision matrix, the second decision matrix and the third decision matrix, and taking the discharge type identified by the network model corresponding to the maximum decision matrix as a final partial discharge type result.
Preferably, the convolutional neural network model includes, but is not limited to, a CNN model, where the CNN model is formed by splicing multiple independent network strings, each layer of the CNN model is formed by multiple parallel convolutional kernels, and is responsible for performing convolution or pooling operation on a Feature Map (Feature Map) obtained by calculation of a network of a previous layer, extracting features contained in data in the process, connecting the front layer and the full-connection layer in series, and sending Feature quantities calculated and output by the front layer and the full-connection layer into a softmax layer to give a classification result. The structure of the general CNN model is as follows: an image matrix input layer, a stack of multiple convolutional layers and a pooling layer (or downsampling layer), a fully connected layer, and a softmax output layer. In this embodiment, the network takes the RGB color image of the partial discharge PRPD as input, that is, three image matrices are input, the image is converted into a pixel matrix and preprocessed, then the pixel matrix is sent to the convolution layer to extract characteristic parameters, the characteristic compression is performed through the pooling layer, and finally the classification result is calculated in the full connection layer and the softmax layer.
Compared with the traditional method, the BP neural network method adopted by the embodiment has the following obvious advantages: the input mode with noise or deformation can be identified; the self-adaptive learning ability is very strong; the identification processing and a plurality of preprocessing can be integrated; and a parallel working mode is adopted, so that the recognition speed is high.
In the embodiment, a large amount of data measured by the history of the intelligent perception terminal is used as a data set, a convolutional neural network model is trained by utilizing the atlas of high-frequency TEV signals, and a BP neural network model is trained by utilizing the statistical characteristic parameters of ultrasonic signals. When the intelligent gateway receives the real-time partial discharge characteristic parameters and the patterns uploaded by the intelligent sensing terminal, the pattern recognition of a single signal type is realized by utilizing the neural network model corresponding to various signals, and the recognition accuracy of various signals under different defects is considered to be different, and even the same partial discharge is performed, the recognition results of different signals are possibly different, so that the recognition results of various signals are required to be comprehensively considered. And constructing weight coefficients according to the identification results of the network models on different defects, selecting the corresponding weight coefficients according to the output during identification, organically combining the weight coefficients output by different models to serve as a final discharge type identification result, and realizing discharge type identification based on multiple state quantities.
After the discharge type recognition module completes the discharge type recognition based on the multiple state quantities, the intelligent gateway carries out data local storage on the number of the switching equipment with partial discharge and the corresponding discharge type, and meanwhile, the number and the corresponding discharge type are sent to the remote monitoring terminal through LoRa communication to be stored to the cloud for subsequent inquiry. The remote monitoring terminal carries out partial discharge alarm prompt in real time according to the number of the switching equipment with partial discharge and the corresponding discharge type based on a preset alarm rule so as to remind monitoring personnel to check and overhaul in time and realize intelligent monitoring of the partial discharge of the switching equipment.
Example two
As shown in fig. 4, a method for diagnosing partial discharge of a switching device based on an intelligent gateway, the method is based on the switching device partial discharge diagnosing system based on the intelligent gateway according to the first embodiment, and the steps include:
the intelligent sensing terminal acquires partial discharge sensing data of the switching equipment, wherein the partial discharge sensing data comprises characteristic parameters and a map of partial discharge;
the intelligent gateway performs edge calculation according to the partial discharge sensing data to perform discharge diagnosis on the switching equipment, performs partial discharge type identification based on neural network model fusion decision, and performs data storage and uploading management on an edge calculation result;
and the remote monitoring terminal carries out partial discharge alarm prompt according to the edge calculation and identification result.
Further, the diagnosing the discharge of the switching device includes:
acquiring real-time partial discharge sensing data, and acquiring a TEV partial discharge detection value of each switch device under the same voltage level according to the real-time partial discharge sensing data;
calculating the average level of the TEV partial discharge of all the switching equipment under the same voltage level according to the TEV partial discharge detection value;
analyzing the degree of deviation of each switching device from the TEV partial discharge average level based on the TEV partial discharge average level;
and carrying out discharge diagnosis according to the degree of deviation from the TEV partial discharge average level, and identifying the switching equipment with partial discharge based on an acousto-electric combined positioning method.
Further, the partial discharge type identification based on the neural network model fusion decision comprises:
acquiring a history partial discharge signal, wherein the history partial discharge signal comprises a history discharge type and corresponding partial discharge sensing data;
acquiring historical ultrasonic discharge data according to the historical partial discharge sensing data, performing model training based on a BP neural network model according to the historical ultrasonic discharge data, generating a first network model, and obtaining a first weight matrix;
acquiring a historical high-frequency PRPD map according to the historical partial discharge sensing data, performing model training based on a convolutional neural network model according to the historical high-frequency PRPD map, generating a second network model, and obtaining a second weight matrix;
and acquiring a historical TEV PRPD map according to the historical partial discharge sensing data, performing model training based on a convolutional neural network model according to the historical TEV PRPD map, generating a third network model, and obtaining a third weight matrix.
Further, the method further comprises the following steps:
acquiring real-time partial discharge sensing data, and acquiring switching equipment discharge parameters based on a switching equipment discharge diagnosis result according to the real-time partial discharge sensing data, wherein the discharge parameters comprise, but are not limited to, real-time ultrasonic discharge data, a real-time high-frequency PRPD map and a real-time TEV PRPD map of the switching equipment;
inputting the real-time ultrasonic discharge data into a first network model to identify a discharge type, and generating a first confidence probability matrix;
inputting the real-time high-frequency PRPD map into a second network model to identify discharge types, and generating a second confidence probability matrix;
inputting the TEV PRPD pattern into a third network model to identify discharge types and generating a third confidence probability matrix;
weighting the weight matrixes of the first network model, the second network model and the third network model with the corresponding confidence probability matrixes respectively, and generating decision matrixes of the first network model, the second network model and the third network model respectively;
and acquiring the maximum decision matrix in the decision matrixes corresponding to the first network model, the second network model and the third network model, and taking the discharge type identified by the network model corresponding to the maximum decision matrix as a final partial discharge type identification result.
It should be noted that, regarding a more specific workflow of the switch device partial discharge diagnosis method based on the intelligent gateway, please refer to the foregoing system embodiment part, and the description is omitted herein.
According to the invention, the intelligent sensing terminal is used for collecting partial discharge sensing data of the switching equipment, the intelligent gateway is used for carrying out discharge diagnosis on the switching equipment according to the partial discharge sensing data, the weight coefficients are constructed based on the identification results of a plurality of neural network models on different defects, the weight coefficients output by the different models are organically combined to be used as a final discharge type identification result, so that the discharge type identification based on multiple state quantities is realized, and meanwhile, the remote monitoring terminal is used for carrying out partial discharge alarm prompt according to the identification results, so that intelligent monitoring is realized.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present invention, and not for limiting the same; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some or all of the technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit of the invention, and are intended to be included within the scope of the appended claims and description.
Claims (9)
1. An intelligent gateway-based switchgear partial discharge diagnostic system, comprising: the intelligent sensing terminal is in wireless connection with the intelligent gateway, and the intelligent gateway is in wireless connection with the remote monitoring terminal;
the intelligent sensing terminal acquires partial discharge sensing data of the switching equipment, wherein the partial discharge sensing data comprises characteristic parameters and a map of partial discharge;
the intelligent gateway performs edge calculation according to the partial discharge sensing data to perform discharge diagnosis on the switching equipment, performs partial discharge type identification based on neural network model fusion decision, and performs data storage and uploading management on an edge calculation result;
and the remote monitoring terminal carries out partial discharge alarm prompt according to the edge calculation and identification result.
2. The intelligent gateway-based switchgear partial discharge diagnostic system of claim 1, wherein the intelligent gateway comprises a lateral statistical analysis module, the lateral statistical analysis module being specifically configured to:
acquiring real-time partial discharge sensing data, and acquiring a TEV partial discharge detection value of each switch device under the same voltage level according to the real-time partial discharge sensing data;
calculating the average level of the TEV partial discharge of all the switching equipment under the same voltage level according to the TEV partial discharge detection value;
analyzing the degree of deviation of each switching device from the TEV partial discharge average level based on the TEV partial discharge average level;
and carrying out discharge diagnosis according to the degree of deviation from the TEV partial discharge average level, and identifying the switching equipment with partial discharge based on an acousto-electric combined positioning method.
3. The intelligent gateway-based switching device partial discharge diagnostic system of claim 1, wherein the intelligent gateway further comprises a discharge type identification module, the discharge type identification module being specifically configured to:
acquiring a history partial discharge signal, wherein the history partial discharge signal comprises a history discharge type and corresponding partial discharge sensing data;
acquiring historical ultrasonic discharge data according to the historical partial discharge sensing data, performing model training based on a BP neural network model according to the historical ultrasonic discharge data, generating a first network model, and obtaining a first weight matrix;
acquiring a historical high-frequency PRPD map according to the historical partial discharge sensing data, performing model training based on a convolutional neural network model according to the historical high-frequency PRPD map, generating a second network model, and obtaining a second weight matrix;
and acquiring a historical TEV PRPD map according to the historical partial discharge sensing data, performing model training based on a convolutional neural network model according to the historical TEV PRPD map, generating a third network model, and obtaining a third weight matrix.
4. A switchgear partial discharge diagnostic system based on intelligent gateways according to claim 3, characterized in that the discharge type identification module is also adapted to:
acquiring real-time partial discharge sensing data, and acquiring switching equipment discharge parameters based on a switching equipment discharge diagnosis result according to the real-time partial discharge sensing data, wherein the discharge parameters comprise, but are not limited to, real-time ultrasonic discharge data, a real-time high-frequency PRPD map and a real-time TEV PRPD map of the switching equipment;
inputting the real-time ultrasonic discharge data into a first network model to identify a discharge type, and generating a first confidence probability matrix;
inputting the real-time high-frequency PRPD map into a second network model to identify discharge types, and generating a second confidence probability matrix;
inputting the TEV PRPD pattern into a third network model to identify discharge types and generating a third confidence probability matrix;
weighting the weight matrixes of the first network model, the second network model and the third network model with the corresponding confidence probability matrixes respectively, and generating decision matrixes of the first network model, the second network model and the third network model respectively;
and acquiring the maximum decision matrix in the decision matrixes corresponding to the first network model, the second network model and the third network model, and taking the discharge type identified by the network model corresponding to the maximum decision matrix as a final partial discharge type identification result.
5. The intelligent gateway-based switching device partial discharge diagnosis system according to claim 1, wherein the intelligent sensing terminals include, but are not limited to, an electric wave intelligent sensing terminal, a high-frequency intelligent sensing terminal and an ultrasonic intelligent sensing terminal, and the intelligent sensing terminal uploads partial discharge sensing data to the intelligent gateway in a LoRa communication mode.
6. A method for diagnosing partial discharge of a switching device based on an intelligent gateway, the method being based on the switching device partial discharge diagnosis system based on an intelligent gateway according to any one of claims 1 to 5, comprising the steps of:
the intelligent sensing terminal acquires partial discharge sensing data of the switching equipment, wherein the partial discharge sensing data comprises characteristic parameters and a map of partial discharge;
the intelligent gateway performs edge calculation according to the partial discharge sensing data to perform discharge diagnosis on the switching equipment, performs partial discharge type identification based on neural network model fusion decision, and performs data storage and uploading management on an edge calculation result;
and the remote monitoring terminal carries out partial discharge alarm prompt according to the edge calculation and identification result.
7. The intelligent gateway-based switching device partial discharge diagnostic method of claim 6, wherein the performing the discharge diagnosis on the switching device comprises:
acquiring real-time partial discharge sensing data, and acquiring a TEV partial discharge detection value of each switch device under the same voltage level according to the real-time partial discharge sensing data;
calculating the average level of the TEV partial discharge of all the switching equipment under the same voltage level according to the TEV partial discharge detection value;
analyzing the degree of deviation of each switching device from the TEV partial discharge average level based on the TEV partial discharge average level;
and carrying out discharge diagnosis according to the degree of deviation from the TEV partial discharge average level, and identifying the switching equipment with partial discharge based on an acousto-electric combined positioning method.
8. The intelligent gateway-based switching device partial discharge diagnosis method according to claim 6, wherein the neural network model fusion decision-based partial discharge type identification comprises:
acquiring a history partial discharge signal, wherein the history partial discharge signal comprises a history discharge type and corresponding partial discharge sensing data;
acquiring historical ultrasonic discharge data according to the historical partial discharge sensing data, performing model training based on a BP neural network model according to the historical ultrasonic discharge data, generating a first network model, and obtaining a first weight matrix;
acquiring a historical high-frequency PRPD map according to the historical partial discharge sensing data, performing model training based on a convolutional neural network model according to the historical high-frequency PRPD map, generating a second network model, and obtaining a second weight matrix;
and acquiring a historical TEV PRPD map according to the historical partial discharge sensing data, performing model training based on a convolutional neural network model according to the historical TEV PRPD map, generating a third network model, and obtaining a third weight matrix.
9. The intelligent gateway-based switching device partial discharge diagnostic method of claim 8, further comprising:
acquiring real-time partial discharge sensing data, and acquiring switching equipment discharge parameters based on a switching equipment discharge diagnosis result according to the real-time partial discharge sensing data, wherein the discharge parameters comprise, but are not limited to, real-time ultrasonic discharge data, a real-time high-frequency PRPD map and a real-time TEV PRPD map of the switching equipment;
inputting the real-time ultrasonic discharge data into a first network model to identify a discharge type, and generating a first confidence probability matrix;
inputting the real-time high-frequency PRPD map into a second network model to identify discharge types, and generating a second confidence probability matrix;
inputting the TEV PRPD pattern into a third network model to identify discharge types and generating a third confidence probability matrix;
weighting the weight matrixes of the first network model, the second network model and the third network model with the corresponding confidence probability matrixes respectively, and generating decision matrixes of the first network model, the second network model and the third network model respectively;
and acquiring the maximum decision matrix in the decision matrixes corresponding to the first network model, the second network model and the third network model, and taking the discharge type identified by the network model corresponding to the maximum decision matrix as a final partial discharge type identification result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311271326.3A CN117347796A (en) | 2023-09-28 | 2023-09-28 | Intelligent gateway-based switching equipment partial discharge diagnosis system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311271326.3A CN117347796A (en) | 2023-09-28 | 2023-09-28 | Intelligent gateway-based switching equipment partial discharge diagnosis system and method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117347796A true CN117347796A (en) | 2024-01-05 |
Family
ID=89355080
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311271326.3A Pending CN117347796A (en) | 2023-09-28 | 2023-09-28 | Intelligent gateway-based switching equipment partial discharge diagnosis system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117347796A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117783797A (en) * | 2024-02-28 | 2024-03-29 | 国网山东省电力公司莒县供电公司 | Power distribution automatic switch partial discharge detection early warning system |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR100853725B1 (en) * | 2007-06-20 | 2008-08-22 | (주) 피에스디테크 | Gis partial discharge diagnostic method and system thereof with prps algorithm |
| KR20120000228A (en) * | 2010-06-25 | 2012-01-02 | 주식회사 케이디파워 | Partial discharging detector and method for detecting the partial discharging |
| CN105676085A (en) * | 2016-01-31 | 2016-06-15 | 国家电网公司 | Extra-high voltage GIS partial discharge detection method based on multi-sensor information fusion |
| CN106324459A (en) * | 2016-09-30 | 2017-01-11 | 国家电网公司 | GIS local discharging detecting system and method based on sound-electricity integration sensor |
| US20200104650A1 (en) * | 2018-09-27 | 2020-04-02 | Industrial Technology Research Institute | Fusion-based classifier, classification method, and classification system |
| CN112014691A (en) * | 2020-07-10 | 2020-12-01 | 国网安徽省电力有限公司电力科学研究院 | Multi-information fusion partial discharge detection terminal and method under power internet of things |
| CN112203249A (en) * | 2020-09-29 | 2021-01-08 | 国网四川省电力公司电力科学研究院 | Intelligent gateway system suitable for partial discharge of switch cabinet |
| CN113568353A (en) * | 2021-07-30 | 2021-10-29 | 国网四川省电力公司电力科学研究院 | Switch cabinet partial discharge monitoring and early warning method based on reduced state feature set |
| CN113610180A (en) * | 2021-08-17 | 2021-11-05 | 湖南工学院 | Visible light image and infrared image fusion ship classification method and device based on deep learning |
| CN114414955A (en) * | 2021-12-24 | 2022-04-29 | 贵州电网有限责任公司 | Method and system for monitoring and evaluating state of switch cabinet |
| CN116452542A (en) * | 2023-04-19 | 2023-07-18 | 华北电力大学（保定） | GIS partial discharge defect diagnosis method based on nerve supervision decision tree |
| CN116503710A (en) * | 2023-04-17 | 2023-07-28 | 吉林省电力科学研究院有限公司 | GIS partial discharge type identification method based on self-adaptive convolutional neural network |
-
2023
- 2023-09-28 CN CN202311271326.3A patent/CN117347796A/en active Pending
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR100853725B1 (en) * | 2007-06-20 | 2008-08-22 | (주) 피에스디테크 | Gis partial discharge diagnostic method and system thereof with prps algorithm |
| KR20120000228A (en) * | 2010-06-25 | 2012-01-02 | 주식회사 케이디파워 | Partial discharging detector and method for detecting the partial discharging |
| CN105676085A (en) * | 2016-01-31 | 2016-06-15 | 国家电网公司 | Extra-high voltage GIS partial discharge detection method based on multi-sensor information fusion |
| CN106324459A (en) * | 2016-09-30 | 2017-01-11 | 国家电网公司 | GIS local discharging detecting system and method based on sound-electricity integration sensor |
| US20200104650A1 (en) * | 2018-09-27 | 2020-04-02 | Industrial Technology Research Institute | Fusion-based classifier, classification method, and classification system |
| CN112014691A (en) * | 2020-07-10 | 2020-12-01 | 国网安徽省电力有限公司电力科学研究院 | Multi-information fusion partial discharge detection terminal and method under power internet of things |
| CN112203249A (en) * | 2020-09-29 | 2021-01-08 | 国网四川省电力公司电力科学研究院 | Intelligent gateway system suitable for partial discharge of switch cabinet |
| CN113568353A (en) * | 2021-07-30 | 2021-10-29 | 国网四川省电力公司电力科学研究院 | Switch cabinet partial discharge monitoring and early warning method based on reduced state feature set |
| CN113610180A (en) * | 2021-08-17 | 2021-11-05 | 湖南工学院 | Visible light image and infrared image fusion ship classification method and device based on deep learning |
| CN114414955A (en) * | 2021-12-24 | 2022-04-29 | 贵州电网有限责任公司 | Method and system for monitoring and evaluating state of switch cabinet |
| CN116503710A (en) * | 2023-04-17 | 2023-07-28 | 吉林省电力科学研究院有限公司 | GIS partial discharge type identification method based on self-adaptive convolutional neural network |
| CN116452542A (en) * | 2023-04-19 | 2023-07-18 | 华北电力大学（保定） | GIS partial discharge defect diagnosis method based on nerve supervision decision tree |
Non-Patent Citations (2)
| Title |
|---|
| 唐炬 等: "气体绝缘电器局部放电模式的多特征信息融合决策诊断", 《高电压技术》, 30 November 2013 (2013-11-30), pages 2581 - 2588 * |
| 王磊 等: "基于CNN+D-S证据理论的多维信息源局部放电故障识别", 《电力工程技术》, 22 September 2022 (2022-09-22), pages 172 - 179 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117783797A (en) * | 2024-02-28 | 2024-03-29 | 国网山东省电力公司莒县供电公司 | Power distribution automatic switch partial discharge detection early warning system |
| CN117783797B (en) * | 2024-02-28 | 2024-05-10 | 国网山东省电力公司莒县供电公司 | Power distribution automatic switch partial discharge detection early warning system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11474163B2 (en) | Power transformer winding fault positioning method based on deep convolutional neural network integrated with visual identification | |
| CN109765883B (en) | Power distribution automation terminal operation state evaluation and fault diagnosis method | |
| CN117347796A (en) | Intelligent gateway-based switching equipment partial discharge diagnosis system and method | |
| CN110163075A (en) | A kind of multi-information fusion method for diagnosing faults based on Weight Training | |
| CN111273125A (en) | RST-CNN-based power cable channel fault diagnosis method | |
| CN115112978A (en) | Fault detection method for insulating ring main unit of power system | |
| CN117032165A (en) | Industrial equipment fault diagnosis method | |
| CN108093210A (en) | A kind of transformer oil level warning system and its alarm method | |
| CN117390403B (en) | Power grid fault detection method and system for new energy lighthouse power station | |
| CN111143835B (en) | Non-invasive protection method for business logic of electric power metering system based on machine learning | |
| CN116244600A (en) | Method, system and equipment for constructing GIS intermittent discharge mode identification model | |
| CN116578922A (en) | Valve cooling system fault diagnosis method and device based on multichannel convolutional neural network | |
| CN114167837B (en) | Intelligent fault diagnosis method and system for railway signal system | |
| CN114400776B (en) | Digital mirror image-based substation automation equipment state diagnosis method and system | |
| CN113392773A (en) | Transformer fault diagnosis system of convolutional neural network based on map information fusion | |
| CN112183745A (en) | High-voltage cable partial discharge mode identification method based on particle swarm algorithm and DBN | |
| Du et al. | Fault diagnosis method of automation equipment in independent and controllable substation based on deep reinforcement learning | |
| CN116433034B (en) | Virtual power plant operation risk detection method based on machine learning | |
| CN117056814B (en) | Transformer voiceprint vibration fault diagnosis method | |
| CN117559502B (en) | New energy storage, variable flow and boosting integrated cabin | |
| CN110942401B (en) | Intelligent communication method for electric power Internet of things | |
| CN118158565A (en) | Communication fault self-adaptive repairing method for intelligent electric energy meter | |
| CN117708667A (en) | Main transformer operation state identification method based on ant colony optimization BP neural network | |
| CN116975389A (en) | Transformer substation fault analysis method and device | |
| CN117373144A (en) | Intelligent equipment inspection method and system based on fault prediction and image recognition |
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 |