CN115600136A - High-voltage bushing fault diagnosis method, system and medium based on multiple sensors - Google Patents

High-voltage bushing fault diagnosis method, system and medium based on multiple sensors Download PDF

Info

Publication number
CN115600136A
CN115600136A CN202211333972.3A CN202211333972A CN115600136A CN 115600136 A CN115600136 A CN 115600136A CN 202211333972 A CN202211333972 A CN 202211333972A CN 115600136 A CN115600136 A CN 115600136A
Authority
CN
China
Prior art keywords
data information
fault
voltage bushing
neural network
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
Application number
CN202211333972.3A
Other languages
Chinese (zh)
Inventor
谢茜
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Electric Power Research Institute of State Grid Sichuan Electric Power Co Ltd
Original Assignee
Electric Power Research Institute of State Grid Sichuan Electric Power Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Electric Power Research Institute of State Grid Sichuan Electric Power Co Ltd filed Critical Electric Power Research Institute of State Grid Sichuan Electric Power Co Ltd
Priority to CN202211333972.3A priority Critical patent/CN115600136A/en
Publication of CN115600136A publication Critical patent/CN115600136A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01D—MEASURING NOT SPECIALLY ADAPTED FOR A SPECIFIC VARIABLE; ARRANGEMENTS FOR MEASURING TWO OR MORE VARIABLES NOT COVERED IN A SINGLE OTHER SUBCLASS; TARIFF METERING APPARATUS; MEASURING OR TESTING NOT OTHERWISE PROVIDED FOR
    • G01D21/00—Measuring or testing not otherwise provided for
    • G01D21/02—Measuring two or more variables by means not covered by a single other subclass
    • 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
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G06N3/084—Backpropagation, e.g. using gradient descent

Landscapes

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

Abstract

The invention discloses a multi-sensor-based high-voltage bushing fault diagnosis method, a multi-sensor-based high-voltage bushing fault diagnosis system and a multi-sensor-based high-voltage bushing fault diagnosis medium, wherein first data information is acquired from signal data of each sensor in a high-voltage bushing; preprocessing the first data information to obtain second data information; extracting feature data information of the second data information by adopting an optimized BP neural network model, wherein the feature data information comprises feature data on a plurality of different sensors; calculating weight values of different fault types under the same characteristic data by adopting a D-S evidence theory algorithm to obtain a plurality of weight values; dividing a plurality of weighted values according to different fault types, and judging faults of the monitoring system according to the weighted values under the same fault type; the method has the advantages of improving the accuracy of judging the fault type of the high-voltage bushing and reducing the efficiency of judging the fault of the high-voltage bushing.

Description

High-voltage bushing fault diagnosis method, system and medium based on multiple sensors
Technical Field
The invention relates to the technical field of bushing fault diagnosis, in particular to a multi-sensor-based high-voltage bushing fault diagnosis method, system and medium.
Background
High voltage bushings are a key component of equipment such as transformers, current transformers, etc. in power systems. In recent years, various on-line monitoring devices have been developed for high voltage bushings, using different sensors to monitor the condition of the bushing. Different sensors have been developed for temperature, pressure, partial discharge, insulation, hydrogen content, etc. of the bushing. The acquisition or calculation of signals of temperature, pressure, UHF, HFCT, dielectric loss, capacitance, hydrogen content and the like of the sleeve can be realized through different sensors.
Due to the complexity of the structure of the casing, when the casing runs, a plurality of structures move inside the casing and have different state quantities, so that when different sensors are used for online monitoring and fault diagnosis of the casing, faults and the state quantities are not in one-to-one correspondence, a certain fault may correspond to a plurality of state quantities, and a certain state quantity may also be caused by a plurality of faults.
However, in the prior art, when diagnosing a bushing fault, usually, the data information of a single sensor is used for judgment, and when the information of the single sensor is used for judgment, the collected information is fuzzy or contradictory, so that the accuracy of judging the fault type of the high-voltage bushing is reduced, and the efficiency of judging the high-voltage bushing fault is increased.
In view of this, the present application is specifically proposed.
Disclosure of Invention
The invention aims to solve the technical problems that in the prior art, the fault of the high-voltage bushing is judged by collecting information data of a single sensor, the collected information is fuzzy or contradictory, the accuracy of judging the fault type of the high-voltage bushing is reduced, and the efficiency of judging the fault of the high-voltage bushing is increased.
The invention is realized by the following technical scheme:
a multi-sensor based high voltage bushing fault diagnosis method, the method steps comprising:
acquiring first data information, wherein the first data information is acquired signal data of each sensor in the high-voltage bushing;
preprocessing the first data information to obtain second data information;
extracting feature data information of the second data information by adopting an optimized BP neural network model, wherein the feature data information comprises feature data on a plurality of different sensors;
calculating weight values of different fault types under the same characteristic data by adopting a D-S evidence theory algorithm to obtain a plurality of weight values;
dividing a plurality of weighted values according to different fault types, and judging the fault of the monitoring system according to the weighted values under the same fault type.
In the conventional fault judgment of the high-voltage bushing on-line monitoring system, the diagnosis of the fault of the high-voltage bushing on-line monitoring system is usually realized by acquiring information data of a single sensor and analyzing the information data of the single sensor, but when the method is adopted, the acquired information data of the single sensor is fuzzy and inaccurate or the acquired data information is inconsistent sometimes, so that the accuracy of judging the fault type of the high-voltage bushing is reduced, and the efficiency of judging the fault of the high-voltage bushing is increased; the invention provides a multi-sensor-based high-voltage bushing fault diagnosis method, which is characterized in that various sensor information data are collected and fused, and data information is processed in a mode of combining a BP neural network with a D-S evidence theory algorithm to judge the fault of an online monitoring system, so that the accuracy of judging the type of the fault of a high-voltage bushing can be improved, and the efficiency of judging the fault of the high-voltage bushing is reduced.
Preferably, the preprocessing is to perform denoising or interference removing processing on the first data information.
Preferably, the building steps of the optimized BP neural network model are as follows:
acquiring historical data information, wherein the historical data information is signal data of faults of all sensors in the high-voltage bushing and corresponding fault types;
denoising and interference removing processing are carried out on the historical data information to obtain sub historical data information;
and constructing a BP neural network model, training the BP neural network model through the sub-historical data information, and performing optimization iteration by adopting an error function of a standard neural network to obtain an optimized BP neural network model.
Preferably, the specific steps of obtaining the weight value include:
selecting any one characteristic data from the characteristic data information, and calculating the conflict difference of the characteristic data under different fault types by adopting a BPA decision method in a D-S evidence theory to obtain a plurality of conflict differences;
processing the conflict differences by adopting a normalization method and combining a cosine theorem in a trigonometric function to obtain a weight value corresponding to the characteristic data;
and traversing the characteristic data information to obtain a plurality of weight values.
Preferably, the fault type is an overheat fault or a discharge fault.
Preferably, the specific expression of the error function is:
Figure BDA0003914658710000021
e is an error function, d k Is a target output value, o k Is the actual output value.
Preferably, the specific expression of the conflict difference is as follows:
Figure BDA0003914658710000031
Figure BDA0003914658710000032
Figure BDA0003914658710000033
as a degree of conflict, x t Is the t-th group of focal elements,
Figure BDA0003914658710000034
to conflict, x ti Is i focal element in t group, x tj Is i focal element and j focal element in t group, m (x) ti ) And distributing the credibility.
Preferably, the specific expression of the weight value is:
Figure BDA0003914658710000035
Figure BDA0003914658710000036
Figure BDA0003914658710000037
Figure BDA0003914658710000038
Figure BDA0003914658710000039
for a collision difference, H t Is an exponential calculation value, beta t As a weight value, ω t Is a weight value.
The invention also provides a multi-sensor-based high-voltage bushing fault diagnosis system, which comprises a data acquisition module, a preprocessing module, a characteristic data extraction module, a weight value calculation module and a judgment module,
the data acquisition module is used for acquiring first data information, wherein the first data information is acquired signal data of each sensor in the high-voltage bushing;
the preprocessing module is used for preprocessing the first data information to obtain second data information;
the characteristic data extraction module is used for extracting characteristic data information of the second data information by adopting an optimized BP neural network model, and the characteristic data information comprises characteristic data on a plurality of different sensors;
the weight calculation module is used for calculating weight values of different fault types under the same characteristic data by adopting a D-S evidence theory algorithm to obtain a plurality of weight values;
the judging module is used for dividing the weighted values according to different fault types and judging the fault of the monitoring system according to the weighted values under the same fault type.
The present invention also provides a computer storage medium having stored thereon a computer program which, when executed by a processor, implements the fault diagnosis method as described above.
Compared with the prior art, the invention has the following advantages and beneficial effects:
according to the multi-sensor-based high-voltage bushing fault diagnosis method, system and medium, provided by the embodiment of the invention, the information data of various sensors are collected and fused, and the data information is processed in a mode of combining a BP neural network and a D-S evidence theory algorithm to judge the fault of the online monitoring system, so that the accuracy of judging the fault type of the high-voltage bushing can be improved, and the efficiency of judging the fault of the high-voltage bushing is reduced.
Drawings
In order to more clearly illustrate the technical solutions of the exemplary embodiments of the present invention, the drawings that are required in the embodiments will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present invention and therefore should not be considered as limiting the scope, and that those skilled in the art may also derive other related drawings based on these drawings without inventive effort.
FIG. 1 is a high voltage bushing fault diagnosis model;
FIG. 2 is a structural model of a parallel multi-source information fusion system;
FIG. 3 is a feed-forward neural network architecture;
FIG. 4 is a process of multi-source information fusion.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail below with reference to examples and accompanying drawings, and the exemplary embodiments and descriptions thereof are only used for explaining the present invention and are not meant to limit the present invention.
In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention. However, it will be apparent to one of ordinary skill in the art that: it is not necessary to employ these specific details to practice the present invention. In other instances, well-known structures, circuits, materials, or methods have not been described in detail so as not to obscure the present invention.
Throughout the specification, reference to "one embodiment," "an embodiment," "one example," or "an example" means: the particular features, structures, or characteristics described in connection with the embodiment or example are included in at least one embodiment of the invention. Thus, the appearances of the phrases "one embodiment," "an embodiment," "one example" or "an example" in various places throughout this specification are not necessarily all referring to the same embodiment or example. Furthermore, the particular features, structures, or characteristics may be combined in any suitable combination and/or sub-combination in one or more embodiments or examples. Further, those of ordinary skill in the art will appreciate that the illustrations provided herein are for illustrative purposes and are not necessarily drawn to scale. As used herein, the term "and/or" includes any and all combinations of one or more of the associated listed items.
In the description of the present invention, the terms "front", "rear", "left", "right", "upper", "lower", "vertical", "horizontal", "upper", "lower", "inner", "outer", etc. indicate orientations or positional relationships based on those shown in the drawings, merely for convenience of description and simplification of description, and do not indicate or imply that the device or element referred to must have a particular orientation, be constructed in a particular orientation, and be operated, and therefore, should not be construed as limiting the scope of the invention.
Example one
In the conventional fault judgment of the high-voltage bushing online monitoring system, the fault diagnosis of the high-voltage bushing online monitoring system is usually realized by acquiring information data of a single sensor and analyzing the information data of the single sensor, but when the method is adopted, the acquired information data of the single sensor is fuzzy and inaccurate or the acquired data information is inconsistent sometimes, so that the accuracy of judging the fault type of the high-voltage bushing is reduced, and the efficiency of judging the fault of the high-voltage bushing is increased.
The embodiment discloses a multi-sensor-based high-voltage bushing fault diagnosis method, which is characterized in that various sensor information data are collected and fused, data information is processed in a mode of combining a BP neural network with a D-S evidence theory algorithm to judge the fault of an online monitoring system, the accuracy of judging the type of the fault of a high-voltage bushing can be improved, and the efficiency of judging the fault of the high-voltage bushing is reduced. The specific diagnostic method in this embodiment is schematically illustrated in fig. 1 to 4, and the method includes the steps of:
s1: acquiring first data information, wherein the first data information is acquired signal data of each sensor in the high-voltage bushing;
in step S1, the high-voltage bushing is a wire outlet device for leading the high-voltage wire inside the transformer to the outside of the oil tank, not only serves as ground insulation of the lead wire, but also serves as a fixed lead wire, and is one of important accessories of the transformer; therefore, when the transformer operates, the bushing is usually monitored on line through various sensors, and therefore, in this embodiment, the data of each sensor for monitoring the bushing on line is obtained, and the data is fused, so that the accuracy of judging the fault type can be increased.
S2: preprocessing the first data information to obtain second data information; the preprocessing is to perform denoising or interference removing processing on the first data information.
In step S2, there may be certain noise or interference information in the acquired first data information, and the noise and interference information need to be removed by filtering.
S3: extracting feature data information of the second data information by adopting an optimized BP neural network model, wherein the feature data information comprises feature data on a plurality of different sensors;
the optimized BP neural network model is constructed by the following steps:
acquiring historical data information, wherein the historical data information is signal data of faults of all sensors in the high-voltage bushing and corresponding fault types; denoising and interference removing processing are carried out on the historical data information to obtain sub-historical data information; and constructing a BP neural network model, training the BP neural network model through the sub-historical data information, and performing optimization iteration by adopting an error function of a standard neural network to obtain an optimized BP neural network model.
The BP neural network model is continuously optimized and iteratively updated through the collected historical data information, and the accuracy of the model in processing the relevant data information can be improved.
The specific BP neural network model is as follows:
the input layer, the intermediate layer and the output layer of the standard BP network are respectively called N i i、N j And N k A neuron. The input to the intermediate jth neuron is:
Figure BDA0003914658710000061
in the formula, ω ij The weights of the ith neuron to the jth neuron in the middle layer in the input layer are calculated; o i Is the output of the ith neuron in the input layer. The input to the kth neuron of the output layer is:
Figure BDA0003914658710000062
in the formula, ω jk The weight from the jth neuron in the middle layer to the kth neuron in the output layer is calculated; o j Is the output of the Kth neuron in the middle layer.
The outputs of the input layer, the intermediate layer and the output layer are respectively:
o i ＝net j ＝x i
Figure BDA0003914658710000063
Figure BDA0003914658710000064
θ j and theta k Threshold values for the intermediate layer jth neuron and the output layer kth neuron, respectively. x is the number of i For the signals acquired by each sensor.
The training of the BP network adopts a gamma learning law based on a gradient method, and the aim is to minimize the mean square error of the network output and the training sample. In the standard BP neural network, a training sample is set as P, wherein an input vector is x1, x2.. Xp; the output vector is y1, y2... Yp; the corresponding teacher value (sample) vector is t1, t2.. Tp; then the mean square error of the P samples is:
Figure BDA0003914658710000065
in the formula, tpk and ypk are the teacher value and the actual output value, respectively, of the pth sample of the kth output neuron.
At this time, the weight of the middle layer is adjusted as follows:
Δω ij (n+1)＝ηδ jp o jp +αΔw ij (n)
Figure BDA0003914658710000071
at this time, the weight of the output layer is adjusted as follows:
Δω jk (n+1)＝ηδ jp o jp +αΔw jk (n)
δ kp ＝(t kp -y kp )f k '(net kp )
in the formula, η is the learning rate, and α is the momentum factor.
The specific expression of the error function is as follows:
Figure BDA0003914658710000072
e is an error function, d k Is a target output value, o k Is the actual output value.
The error function reflects that k And o k On a different "distance" scale between the two quantities. The purpose of BP network training is to make the actual output o k Maximum possible near target output d k This proximity is measured. The error function can effectively improve the problem of low convergence speed when the BP neural network is trained, and improve the training speed of the BP neural network.
After the error function is used, the weight of the middle layer of the neural network is adjusted as follows:
Figure BDA0003914658710000073
at this time, the weight of the output layer is adjusted as follows:
Figure BDA0003914658710000074
wherein, ω is ij The weights of the ith neuron to the jth neuron in the middle layer in the input layer are calculated; omega jk The weight from the jth neuron in the middle layer to the kth neuron in the output layer is calculated; eta is learning rateAnd adjusting the factor. O is i Is the output of the ith neuron in the input layer. O is j Is the output of the Kth neuron in the middle layer.
S4: calculating weight values of different fault types under the same characteristic data by adopting a D-S evidence theory algorithm to obtain a plurality of weight values;
the specific steps of obtaining the weight value include:
selecting any one characteristic data from the characteristic data information, and calculating the conflict differences of the characteristic data under different fault types by adopting a BPA (Business Process analysis) decision method in a D-S evidence theory to obtain a plurality of conflict differences; the fault type is an overheating fault or a discharging fault.
Processing the conflict differences by adopting a normalization method and combining a cosine theorem in a trigonometric function to obtain a weight value corresponding to the characteristic data;
and traversing the characteristic data information to obtain a plurality of weight values.
In the embodiment, a method for redistributing evidence BPA is provided for a D-S evidence theory algorithm, so that the influence of evidence conflict on information fusion is effectively overcome, and the certainty factor and the hesitation degree in judgment are described by numerical values in a reasonable evidence reasoning mode.
Defining m as a BPA function on an identification space theta, and A and B as focal elements on the identification space, and redefining the BPA on the identification space by using a formula.
Figure BDA0003914658710000081
SetPm (a) indicates the degree of support of the basic probability distribution to each subset.
Figure BDA0003914658710000082
The degree of collision Diff for the same target is expressed as:
Diff(A)＝|SetP mi (A)-SetP mj (A)|
the embodiment improves the D-S evidence theory algorithm, and weights are given to the evidence again according to the reliability of the evidence, so that a better fusion result is obtained.
Let θ be the sample space, xi be the focal elements in the sample space θ, i =1,2, … N, and there are N sets of evidence functions. Note mi = { xt1, xt2, xt3.. Xti } where i =1,2, … N, t =1,2, … N. The weights are represented by β i, where i =1,2,3.
First, the evidence BPA is redistributed separately.
Figure BDA0003914658710000083
Second, the collision difference of the same focal element under different evidences is calculated.
Figure BDA0003914658710000084
Figure BDA0003914658710000085
As a degree of conflict, x t Is the t-th group of focal elements,
Figure BDA0003914658710000086
for conflict, x ti Is i focal element in t group, x tj Is i focal element and j focal element in t group, m (x) ti ) And distributing the credibility.
And thirdly, normalizing the collision difference.
Figure BDA0003914658710000087
The requirements are met,
Figure BDA0003914658710000091
Figure BDA0003914658710000092
for a collision difference, H t Is an exponential calculation value, beta t As a weight value, the weight value,ω t is a weight value.
The normalized value is then calculated.
Figure BDA0003914658710000093
The weight may be expressed as:
Figure BDA0003914658710000094
the cosine theorem in the trigonometric function is introduced, the weight is fixed between [0,1], the weight is given by utilizing the curve characteristic of the cosine function, and the obtained weight is relatively smooth. Namely:
Figure BDA0003914658710000095
thus, a weight vector consisting of the weight coefficients of the evidences is determined.
S5: dividing a plurality of weighted values according to different fault types, and judging the fault of the monitoring system according to the weighted values under the same fault type.
And combining the BP neural network and the D-S evidence theory to fuse the characteristic value and the characteristic vector. The uncertain factor fi is set as a training error of the BP neural network, the output is normalized and used as a basic probability value of each focus element, and the calculation formula is as follows:
Figure BDA0003914658710000096
in the formula: fi represents a failure mode, i = (0,1,2,3,4,5,6), and y (fi) represents an output result of the BP neural network.
Figure BDA0003914658710000097
En is the network error of the sample, tnj, ynj are pairs respectivelyThe expected and actual values for the jth neuron.
Figure BDA0003914658710000098
Firstly, processing input values of each sensor through a BP neural network by calculation to obtain mass function values confirmed by corresponding different fault types, and finally obtaining a fusion result by applying a D-S evidence theory rule.
The specific implementation process comprises the following steps:
and constructing a BP diagnostic network, wherein the number of neurons of an input layer of the BP network is 9 characteristic parameters, the number of neurons of an output layer is the number of faults (representing different fault types), namely, an output objective function F = { F1, F2, F3, F4, F5, F6}, and the expected output of the neural network is represented by [0,1], wherein 1 represents the existence of the fault and 0 represents the nonexistence of the fault. Table 1 shows input samples of the neural network, which are 9 characteristic parameters, such as a square root mean value x1, a standard deviation x2, a temperature index x3, a pressure index x4, an oil level index x5, a dielectric loss index x6, a capacitance index x7, a leakage current x8, and an angle difference x 9.
Input samples for a table-neural network
Figure BDA0003914658710000101
According to the measuring point data of the table I, the BP neural network is firstly diagnosed locally, the BP network input layer is set as 6 nodes, the hidden layer is set as 17 nodes, the output layer is set as 6 nodes and represents six faults respectively, the three-layer BP network is trained by a training set, and then part of samples are extracted from all fault samples for training. The training results are shown in table 2:
training results of table two BP neural network
Figure BDA0003914658710000102
Figure BDA0003914658710000111
By adopting the invention, the BP neural network is combined with the D-S evidence theory, and the fusion result of the samples can be obtained by applying the D-S evidence theory rule, which is shown in the third table.
Fusion results of Table III D-S evidence theory
Figure BDA0003914658710000112
As can be seen by comparing table 2 and table 3: through information fusion comprehensive diagnosis, the diagnosis precision is greatly improved. The method has the advantages that each sample has a non-ideal result, if single measuring point data is used for judgment, misjudgment is easy, a multi-sensor information fusion method is utilized, multiple parameters and multiple variables are comprehensively considered, a BP neural network is utilized to perform characteristic layer fusion, a D-S synthesis rule is utilized to perform decision layer fusion, the finally obtained result is still ideal, all the confidence degrees are above 0.95, and therefore the effectiveness of the intelligent diagnosis method for the faults of the information fusion casing pipe combining the BP neural network and the D-S evidence theory is proved.
According to the fault diagnosis method for the high-voltage bushing online monitoring system, the data of various sensor information is collected and fused, and the data information is processed in a mode that a BP neural network is combined with a D-S evidence theory algorithm to judge the fault of the online monitoring system, so that the accuracy of judging the type of the fault of the high-voltage bushing can be improved, and the efficiency of judging the fault of the high-voltage bushing is reduced.
Example two
The embodiment discloses a multi-sensor-based high-voltage bushing fault diagnosis system, which is used for realizing the fault diagnosis method in the first embodiment and comprises a data acquisition module, a preprocessing module, a characteristic data extraction module, a weight value calculation module and a judgment module,
the data acquisition module is used for acquiring first data information, wherein the first data information is acquired signal data of each sensor in the high-voltage bushing;
the preprocessing module is used for preprocessing the first data information to obtain second data information;
the characteristic data extraction module is used for extracting characteristic data information of the second data information by adopting an optimized BP neural network model, and the characteristic data information comprises characteristic data on a plurality of different sensors;
the weight calculation module is used for calculating weight values of different fault types under the same characteristic data by adopting a D-S evidence theory algorithm to obtain a plurality of weight values;
the judging module is used for dividing the weighted values according to different fault types and judging the fault of the monitoring system according to the weighted values under the same fault type.
EXAMPLE III
The present embodiment discloses a computer storage medium on which a computer program is stored, which, when executed by a processor, implements a fault diagnosis method as described in the first embodiment.
As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer programs may be provided to issue instructions to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program issue instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the issue instructions stored in the computer-readable memory produce an article of manufacture including issue instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program issue instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the issue instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
The above-mentioned embodiments are intended to illustrate the objects, technical solutions and advantages of the present invention in further detail, and it should be understood that the above-mentioned embodiments are merely exemplary embodiments of the present invention, and are not intended to limit the scope of the present invention, and any modifications, equivalent substitutions, improvements and the like made within the spirit and principle of the present invention should be included in the scope of the present invention.

Claims (10)

1. A multi-sensor based high voltage bushing fault diagnosis method, characterized in that the method steps comprise:
acquiring first data information, wherein the first data information is acquired signal data of each sensor in the high-voltage bushing;
preprocessing the first data information to obtain second data information;
extracting feature data information of the second data information by adopting an optimized BP neural network model, wherein the feature data information comprises feature data on a plurality of different sensors;
calculating weight values of different fault types under the same characteristic data by adopting a D-S evidence theory algorithm to obtain a plurality of weight values;
dividing the weighted values according to different fault types, and judging the fault of the monitoring system according to the weighted values under the same fault type.
2. A multi-sensor based high voltage bushing fault diagnosis method according to claim 1, wherein said preprocessing is de-noising or de-disturbing said first data information.
3. The method for diagnosing the fault of the high-voltage bushing based on the multiple sensors as claimed in claim 1, wherein the step of constructing the optimized BP neural network model comprises the following steps:
acquiring historical data information, wherein the historical data information is signal data of faults of all sensors in the high-voltage bushing and corresponding fault types;
denoising and interference removing processing are carried out on the historical data information to obtain sub-historical data information;
and constructing a BP neural network model, training the BP neural network model through the sub-historical data information, and performing optimization iteration by adopting an improved error function of the neural network to obtain an optimized BP neural network model.
4. The method according to claim 1, wherein the step of obtaining the weight value comprises:
selecting any one characteristic data from the characteristic data information, and calculating the conflict difference of the characteristic data under different fault types by adopting a BPA decision method in a D-S evidence theory to obtain a plurality of conflict differences;
processing the conflict differences by adopting a normalization method and combining a cosine theorem in a trigonometric function to obtain a weight value corresponding to the characteristic data;
and traversing the characteristic data information to obtain a plurality of weight values.
5. A multi-sensor based high voltage bushing fault diagnosis method according to claim 4, wherein said fault type is an over-temperature fault or a discharge fault.
6. A multi-sensor based high voltage bushing fault diagnosis method according to claim 3, characterized in that said error function is specifically expressed as:
Figure FDA0003914658700000011
e is an error function, d k Is a target output value, o k Is the actual output value.
7. A multi-sensor based high voltage bushing fault diagnosis method according to claim 4, characterized in that said conflict difference is expressed by the following specific expression:
Figure FDA0003914658700000021
Figure FDA0003914658700000022
Figure FDA0003914658700000023
as a degree of conflict, x t Is the t-th group of focal elements,
Figure FDA0003914658700000024
to conflict, x ti Is i focal element in t group, x tj Is i focal element and j focal element in t group, m (x) ti ) And distributing the credibility.
8. The method according to claim 7, wherein the specific expression of the weight value is as follows:
Figure FDA0003914658700000025
Figure FDA0003914658700000026
Figure FDA0003914658700000027
Figure FDA0003914658700000028
α xt for a collision difference, H t Is an exponential calculation value, beta t As a weight value, ω t Is a weight value.
9. A high-voltage bushing fault diagnosis system based on multiple sensors is characterized by comprising a data acquisition module, a preprocessing module, a characteristic data extraction module, a weight value calculation module and a judgment module,
the data acquisition module is used for acquiring first data information, wherein the first data information is acquired signal data of each sensor in the high-voltage bushing;
the preprocessing module is used for preprocessing the first data information to obtain second data information;
the characteristic data extraction module is used for extracting characteristic data information of the second data information by adopting an optimized BP neural network model, and the characteristic data information comprises characteristic data on a plurality of different sensors;
the weight calculation module is used for calculating weight values of different fault types under the same characteristic data by adopting a D-S evidence theory algorithm to obtain a plurality of weight values;
the judging module is used for dividing the weighted values according to different fault types and judging the fault of the monitoring system according to the weighted values under the same fault type.
10. A computer storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the fault diagnosis method according to any one of claims 1 to 8.
CN202211333972.3A 2022-10-28 2022-10-28 High-voltage bushing fault diagnosis method, system and medium based on multiple sensors Pending CN115600136A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202211333972.3A CN115600136A (en) 2022-10-28 2022-10-28 High-voltage bushing fault diagnosis method, system and medium based on multiple sensors

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202211333972.3A CN115600136A (en) 2022-10-28 2022-10-28 High-voltage bushing fault diagnosis method, system and medium based on multiple sensors

Publications (1)

Publication Number Publication Date
CN115600136A true CN115600136A (en) 2023-01-13

Family

ID=84849981

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202211333972.3A Pending CN115600136A (en) 2022-10-28 2022-10-28 High-voltage bushing fault diagnosis method, system and medium based on multiple sensors

Country Status (1)

Country Link
CN (1) CN115600136A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117076870A (en) * 2023-10-16 2023-11-17 广东石油化工学院 Rotary machine fault diagnosis method, device, equipment and storage medium
CN117572295A (en) * 2024-01-12 2024-02-20 山东和兑智能科技有限公司 Multi-mode on-line monitoring and early warning method for high-voltage sleeve

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117076870A (en) * 2023-10-16 2023-11-17 广东石油化工学院 Rotary machine fault diagnosis method, device, equipment and storage medium
CN117076870B (en) * 2023-10-16 2024-01-12 广东石油化工学院 Rotary machine fault diagnosis method, device, equipment and storage medium
CN117572295A (en) * 2024-01-12 2024-02-20 山东和兑智能科技有限公司 Multi-mode on-line monitoring and early warning method for high-voltage sleeve
CN117572295B (en) * 2024-01-12 2024-04-12 山东和兑智能科技有限公司 Multi-mode on-line monitoring and early warning method for high-voltage sleeve

Similar Documents

Publication Publication Date Title
US11586913B2 (en) Power equipment fault detecting and positioning method of artificial intelligence inference fusion
CN115600136A (en) High-voltage bushing fault diagnosis method, system and medium based on multiple sensors
CN112418277B (en) Method, system, medium and equipment for predicting residual life of rotating machine parts
CN111914883B (en) Spindle bearing state evaluation method and device based on deep fusion network
CN110766059A (en) Transformer fault prediction method, device and equipment
CN110231156B (en) Service robot motion system fault diagnosis method and device based on time sequence characteristics
CN112508105B (en) Fault detection and retrieval method for oil extraction machine
CN109829236A (en) A kind of Compressor Fault Diagnosis method based on XGBoost feature extraction
CN109145516B (en) Analog circuit fault identification method based on improved extreme learning machine
CN105825271B (en) Satellite failure diagnosis and prediction method based on evidential reasoning
Fu et al. A re-optimized deep auto-encoder for gas turbine unsupervised anomaly detection
CN105975744A (en) D-S evidence theory-based textile process data fusion system
CN115859077A (en) Multi-feature fusion motor small sample fault diagnosis method under variable working conditions
CN110991471A (en) Fault diagnosis method for high-speed train traction system
CN115587309A (en) Method, device and equipment for extracting key features of short-circuit resistance of transformer
GB2616996A (en) Method for recognizing type of vortex signal of evaporator of nuclear power plant on basis of LSTM-CNN
CN115759352A (en) Variable working condition mechanical fault diagnosis method based on multi-source domain class separation countermeasure network
CN115793590A (en) Data processing method and platform suitable for system safety operation and maintenance
Zhao et al. Bearing fault diagnosis based on mel frequency cepstrum coefficient and deformable space-frequency attention network
CN113627268A (en) Model training method, and method and device for detecting fault of speed reducer for mine hoist
CN116578922A (en) Valve cooling system fault diagnosis method and device based on multichannel convolutional neural network
CN116595465A (en) High-dimensional sparse data outlier detection method and system based on self-encoder and data enhancement
CN114252266A (en) Rolling bearing performance degradation evaluation method based on DBN-SVDD model
CN115389881A (en) Insulation state evaluation method and device for cable intermediate joint
CN111047185B (en) Method and device for determining influence of storage environment factors on ammunition storage reliability

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