Vibration ultrasonic partial discharge signal analysis method based on factor analysis weighting method

By using a vibration ultrasonic partial discharge signal analysis method based on factor analysis weighting, combined with mutual information theory, the problem of fault location and maintenance of gas-insulated combined electrical appliances was solved, and efficient assessment and accurate location of equipment condition were achieved.

CN117250444BActive Publication Date: 2024-11-08STATE GRID HUBEI EXTRA HIGH VOLTAGE CO
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202310842606.9
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2023-07-10
Publication Date
2024-11-08
Estimated Expiration
2043-07-10

AI Technical Summary

Technical Problem

In the existing technology, gas-insulated switchgear is difficult to locate faults efficiently and repair quickly when a fault occurs.

Method used

A vibration-ultrasonic partial discharge signal analysis method based on factor analysis weighting is adopted. By extracting the time-domain and frequency-domain characteristic parameters of mechanical vibration signals and ultrasonic partial discharge signals, and combining factor analysis based on mutual information theory, the equipment condition of gas-insulated switchgear is evaluated.

Benefits of technology

It enables accurate location and rapid repair of faults in gas-insulated switchgear, improves the accuracy of detection results, and can evaluate the operating status of equipment through the joint detection of mechanical vibration and electrical status signals.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117250444B_ABST
    Figure CN117250444B_ABST
Patent Text Reader

Abstract

The present invention belongs to the technical field of online monitoring of equipment. The vibration ultrasonic partial discharge signal analysis method based on the factor analysis weighting method includes the following steps: A001: extracting ultrasonic partial discharge evaluation indicators of gas-insulated combination electrical appliances; A002: extracting vibration signal evaluation indicators of gas-insulated combination electrical appliances at a frequency of 0-1000Hz during normal operation and under test operation; A003: establishing an evaluation indicator set U; A004: classifying the insulation state of gas-insulated combination electrical appliances, and corresponding insulation levels according to the insulation state types, and constructing a comment set X for the evaluation model according to the insulation state types and insulation levels; A005: using the mutual information factor analysis method to weight and analyze the evaluation indicator weights in the evaluation indicator set U, calculating the information relevance of the evaluation indicators, and analyzing the equipment status of the gas-insulated combination electrical appliances through the information relevance. The technical solution can analyze the equipment status by extracting characteristic information, which is quite meaningful for positioning and finding faults.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the technical field of online monitoring of equipment, and in particular relates to a vibration ultrasonic partial discharge signal analysis method based on a factor analysis weighting method. Background Art

[0002] Gas Insulated Switchgear (GIS) is a key device in the power transmission and transformation system. Its advantage is that it seals components such as circuit breakers, disconnectors and transformers in a metal casing, making the operation of each electrical equipment less affected by the environment. However, it is difficult to locate and repair the fault when a fault occurs. Therefore, it is very important to establish an efficient and accurate status assessment system to analyze the equipment status when the equipment is in operation. Summary of the invention

[0003] The present invention provides a vibration ultrasonic partial discharge signal analysis method based on a factor analysis weighting method, which is mainly a detection method for mechanical vibration and ultrasonic partial discharge information fusion based on a factor analysis method of mutual information theory. The severity of partial discharge is described by extracting time domain and frequency domain characteristic parameters of mechanical vibration signals and ultrasonic partial discharge pulse signals, and information source evaluation indicators and insulation status levels are established. The factor analysis method of mutual information theory is used to objectively analyze the information correlation between the evaluation parameter set and the insulation status level and the evaluation indicator. The equipment status of the gas-insulated combination electrical appliance is evaluated through information correlation analysis, which solves the technical problem in the prior art that it is difficult to locate and quickly repair faults of electrical equipment when a fault occurs.

[0004] The present invention is achieved through the following technical solutions:

[0005] A vibration ultrasonic partial discharge signal analysis method based on a factor analysis weighting method, and a vibration ultrasonic partial discharge signal analysis method based on a factor analysis weighting method, comprising the following steps:

[0006] A001: Extract ultrasonic partial discharge evaluation indicators of gas-insulated switchgear;

[0007] A002: Extract the vibration signal evaluation index of the gas-insulated switchgear during normal operation and under test operation at a frequency of 0-1000Hz;

[0008] A003: establishing an evaluation index set U according to the ultrasonic partial discharge evaluation index extracted in step A001 and the vibration signal evaluation index extracted in step A002;

[0009] A004: Classify the insulation status of gas-insulated switchgear, and correspond the insulation levels according to the insulation status types, and construct the comment set X for the evaluation model according to the insulation status types and insulation levels;

[0010] A005: Through the evaluation index set U and the comment set X, the weights of the evaluation indexes in the evaluation index set U are analyzed by using the factor analysis method of mutual information, the information relevance of the evaluation indexes is calculated, and the equipment status of the gas-insulated combination electrical appliances is analyzed by the information relevance.

[0011] Optionally, in step A001, the ultrasonic partial discharge evaluation index includes a power spectrum maximum value v1, a median frequency v2 and an average power frequency v3 in the frequency domain characteristics.

[0012] Optionally, in step A002, the vibration information evaluation index includes the vibration signal amplitude v4 during normal operation and the vibration signal amplitude v5 during the operation to be detected.

[0013] Optionally, in step A004, the insulation status of the gas-insulated combination electrical appliances is classified, and the insulation levels are respectively corresponded according to the insulation status types, and the comment set X for the evaluation model is constructed according to the insulation status types and insulation levels: the insulation status of the gas-insulated combination electrical appliances is divided into four states: normal, warning, abnormal, and severe. The four insulation states correspond to four insulation status levels of H1, H2, H3, and H4, respectively.

[0014] Optionally, in step A005, the information relevance method of the evaluation index is calculated as follows:

[0015] A0051: Define each evaluation index v through fuzzy membership function i The degree of membership;

[0016]

[0017] in, Indicates that the evaluation index vi belongs to state H m The membership degree of H(u) is the corresponding membership degree of the equipment insulation level H.

[0018] Represents a collection of evaluations;

[0019] A0052: Determine the evaluation index v by fuzzy clustering method i The membership boundary of

[0020] A0053: Establishing an evaluation matrix R, and using the evaluation index set U to judge the comment set X by establishing the evaluation matrix R;

[0021] A0054: Calculate the weights of the evaluation indicators and calculate the information relevance of all variables.

[0022] Optionally, in step A0052, each evaluation index v is determined by a fuzzy clustering method. i The method of obtaining the membership boundary of the evaluation index is as follows: obtaining the cluster center and cluster radius of each evaluation index in the evaluation index set U, and constructing the values ​​of a1, a2, a3, a4, a5, a6, a7, and a8 according to the cluster center and cluster radius;

[0023] Set the evaluation metric v i After FCM function clustering, the cluster centers are o i1 , o i2 , o i3 , o i4 , the corresponding cluster radius is r i1 , r i2 , r i3 , r i4 , then the evaluation index v i The corresponding boundaries of the fuzzy membership function are a i1 , a i2 , a i3 , a i4 , a i5 , a i6 , a i7 , a i8 , can be calculated by the following formula:

[0024]

[0025] Among them, a i1 , a i3 and a i5 are the center value and two boundary values ​​of H2 level, respectively. i2 is the right boundary value of H1 level, a i6 , a i4 and a i8 are the center value and two boundary values ​​of H3 level, respectively. i7 is the left boundary value of H4 level, r i1 , r i2 , r i3 , r i4 They respectively represent the membership of the i-th evaluation index to the four insulation status levels of normal, caution, abnormal and severe.

[0026] Optionally, in step A0052, the evaluation index v i The degree of membership r belongs to four insulation status levels H1, H2, H3, H4 i1 , r i2 , r i3 , r i4 As shown below,

[0027] H1: Normal state

[0028]

[0029] H2: Attention state

[0030]

[0031] H3: Abnormal state

[0032]

[0033] H4: Severe status

[0034]

[0035] Among them, U i Represents the value of the i-th evaluation parameter in the evaluation indicators v1-v5.

[0036] Optionally, in A0053, an evaluation matrix R is established, and the evaluation index set U is used to evaluate the comment set X by establishing the evaluation matrix R = (r i1 , K, r ij , K, r in ), r ij Indicates the evaluation results of the evaluation factors in the comment set X j The possible degree on u i x j The degree of membership, among which,

[0037]

[0038] Among them, the number of rows and columns of M, K, and O are uncertain, and the row vectors and column vectors corresponding to M, K, and O are omitted respectively.

[0039] Optionally, step A0054: weight the evaluation indicators and calculate the information relevance of all variables as follows:

[0040] Evaluation parameter set V, evaluation index v i The calculation formula for the information relevance D with insulation status level H is:

[0041]

[0042] Evaluation Metrics i And the evaluation index v j The information relevance R is:

[0043]

[0044] D and R are used as the benchmark for calculating the weights of evaluation indicators;

[0045] The information relevance of all variables is calculated by the following formula;

[0046]

[0047] Among them, Φ is an operator, indicating the vector y = [y1, y2, ...] T To measure the factors, the optimization conditions of the above formula are:

[0048]

[0049] Optionally, in the formula for calculating the relevance of all variable information, the number k of evaluation indicators is set to 3 and 2 respectively, and the operator Φ is calculated by determining the constant K, and the equipment condition of the gas-insulated combination electrical appliance is analyzed and evaluated by the operator Φ.

[0050] Compared with the prior art, the present invention has the following beneficial effects:

[0051] 1) This technical solution is based on a detection method for mechanical vibration and ultrasonic partial discharge information fusion based on the factor analysis method of mutual information theory. The severity of partial discharge is described by extracting time domain and frequency domain characteristic parameters of mechanical vibration signals and ultrasonic partial discharge pulse signals, and information source evaluation indicators and insulation status levels are established. The factor analysis method of mutual information theory is used to objectively analyze the information correlation between the evaluation parameter set and the insulation status level and the evaluation indicator. The equipment condition of gas-insulated combination electrical appliances is evaluated through information correlation analysis, which solves the technical problem in the prior art that it is difficult to locate and quickly repair faults of electrical equipment when they encounter faults. It is of great significance to be able to analyze the equipment condition by extracting characteristic information.

[0052] 2) There are typical insulation defects such as free metal particles, insulation air gaps, and metal contamination on the insulation surface in gas-insulated switchgear, and the identification of insulation defects is an indispensable part of GIS insulation status assessment. At the same time, the evaluation of the severity of GIS internal discharge is the most important theoretical support for GIS insulation status assessment research. At present, the theoretical research on insulation in GIS is mainly carried out around the development law of partial discharge. The insulation degradation process inside GIS is simulated by the different severity of partial discharge inside the laboratory. According to the various physical and chemical phenomena and collected information caused by the partial discharge degradation process, the change law of degradation information is analyzed, and the insulation degradation law inside GIS is formed accordingly. Partial discharge information is the main source of insulation information monitoring for gas-insulated equipment. The insulation status assessment research of gas-insulated equipment currently carried out is mainly based on the partial discharge information in the equipment. The mechanical state in the equipment can also reflect the operating state of the equipment. For example, some mechanical vibrations can also cause the occurrence of floating potential and further partial discharge. The accuracy of the detection results will be greatly improved by jointly detecting the mechanical state and electrical state signals. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 It is a schematic diagram of membership function;

[0054] Figure 2 The flowchart of insulation condition assessment of gas-insulated equipment using vibration ultrasonic partial discharge signal analysis method based on factor analysis weighting method is presented. DETAILED DESCRIPTION

[0055] In order to make the purpose, technical scheme and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with embodiments. The illustrative implementation modes of the present invention and their description are only used to explain the present invention and are not intended to limit the present invention.

[0056] Embodiment 1:

[0057] like Figure 1-2 As shown, the vibration ultrasonic partial discharge signal analysis method based on the factor analysis weighting method includes the following steps:

[0058] A001: Extract ultrasonic partial discharge evaluation indicators of gas-insulated switchgear;

[0059] A002: Extract the vibration signal evaluation index of the gas-insulated switchgear during normal operation and under test operation at a frequency of 0-1000Hz;

[0060] A003: establishing an evaluation index set U according to the ultrasonic partial discharge evaluation index extracted in step A001 and the vibration signal evaluation index extracted in step A002;

[0061] A004: Classify the insulation status of gas-insulated switchgear, and correspond the insulation levels according to the insulation status types, and construct the comment set X for the evaluation model according to the insulation status types and insulation levels;

[0062] A005: Through the evaluation index set U and the comment set X, the weights of the evaluation indexes in the evaluation index set U are analyzed by using the factor analysis method of mutual information, the information relevance of the evaluation indexes is calculated, and the equipment status of the gas-insulated combination electrical appliances is analyzed by the information relevance.

[0063] In this embodiment, partial discharge hazard is used to confirm the insulation system status of gas insulated switchgear GIS equipment. Partial discharge hazard includes ultrasonic partial discharge assessment index and mechanical vibration partial discharge assessment index.

[0064] Wherein, in step A001, the ultrasonic partial discharge evaluation index includes the power spectrum maximum value v1, the median frequency v2 and the average power frequency v3 in the frequency domain characteristics.

[0065] Among them, in the step A002, the mechanical vibration partial discharge evaluation index includes the vibration signal amplitude v4 during normal operation and the vibration signal amplitude v5 during the operation to be detected.

[0066] In step A004, for insulation status rating, since the sources of insulation status assessment information are multi-source, the insulation status of the gas-insulated combination electrical appliances is classified, and the insulation levels are respectively corresponded according to the insulation status types. The evaluation model is constructed according to the insulation status types and insulation levels. The comment set X for is: the insulation status of the gas-insulated combination electrical appliances is divided into four states: normal, caution, abnormal, and severe. The four insulation states correspond to the four insulation status levels of H1, H2, H3, and H4, respectively.

[0067] In mutual information theory, mutual information I(x,y) represents the correlation between two random variables x and y, and in the definition, p(x), p(y), and p(x,y) represent the probability densities of the variables respectively.

[0068] Among them, in step A005, the method for calculating the information relevance of the evaluation index is:

[0069] A0051: Define each evaluation index v through fuzzy membership function i The degree of membership;

[0070]

[0071] in, Indicates that the evaluation index vi belongs to state H mThe membership degree of H(u) represents the membership degree corresponding to the equipment insulation level H, and only represents an evaluation set, including the three level states that need to be evaluated;

[0072] A0052: Determine the evaluation index v by fuzzy clustering method i The membership boundary of

[0073] A0053: Establishing an evaluation matrix R, and using the evaluation index set U to judge the comment set X by establishing the evaluation matrix R;

[0074] A0054: Calculate the weights of the evaluation indicators and calculate the information relevance of all variables.

[0075] Among them, in step A0052, each evaluation index v is determined by fuzzy clustering method. i The method of membership boundary is: obtain the cluster center and cluster radius of each evaluation indicator in the evaluation indicator set U, and construct the values ​​of a1, a2, a3, a4, a5, a6, a7, and a8 according to the cluster center and cluster radius;

[0076] Set the evaluation metric v i After FCM function clustering, the cluster centers are o i1 , o i2 , o i3 , o i4 , the corresponding cluster radius is r i1 , r i2 , r i3 , r i4 , then the evaluation index v i The corresponding boundaries of the fuzzy membership function are a i1 , a i2 , a i3 , a i4 , a i5 , a i6 , a i7 , a i8 , can be calculated by the following formula:

[0077]

[0078] Among them, a i1 , a i3 and a i5 are the center value and two boundary values ​​of H2 level, respectively. i2 is the right boundary value of H1 level, a i6 , a i4 and a i8 are the center value and two boundary values ​​of H3 level, respectively. i7 is the left boundary value of H4 level, ri1 , r i2 , r i3 , r i4 They respectively represent the membership of the i-th evaluation index to the four insulation status levels of normal, caution, abnormal and severe.

[0079] Among them, in step A0052, the evaluation index v i The degree of membership r belongs to four insulation status levels H1, H2, H3, H4 i1 , r i2 , r i3 , r i4 As shown below,

[0080] H1: Normal state

[0081]

[0082] H2: Attention state

[0083]

[0084] H3: Abnormal state

[0085]

[0086] H4: Severe status

[0087]

[0088] Among them, U i Represents the value of the i-th evaluation parameter in the evaluation indicators v1-v5.

[0089] Among them, in A0053, an evaluation matrix R is established, and the evaluation index set U is used to judge the comment set X by establishing the evaluation matrix R = (r i1 , K, r ij , K, r in ), r ij Indicates the evaluation results of the evaluation factors in the comment set X j The possible degree on u i x j The degree of membership, among which,

[0090]

[0091] The number of rows and columns of M, K, and O is uncertain, and the row vectors and column vectors corresponding to M, K, and O are omitted respectively.

[0092] Among them, step A0054: weight calculation is performed on the evaluation indicators, and the information relevance of all variables is calculated as:

[0093] Evaluation parameter set V, evaluation index v i The calculation formula for the information relevance D with insulation status level H is:

[0094]

[0095] Evaluation Metrics i And the evaluation index v j The information relevance R is:

[0096]

[0097] D and R are used as the benchmark for calculating the weights of evaluation indicators;

[0098] The information relevance of all variables is calculated by the following formula;

[0099]

[0100] Among them, Φ is an operator, indicating the vector y = [y1, y2, ...] T To measure the factor, the optimization condition of the above formula is: Solve the quadratic programming function to obtain the y value of the indicator vector corresponding to each evaluation indicator, so that the information system constructed by all evaluation indicators achieves maximum relevance and minimum redundancy. y is the required weight factor.

[0101] In the formula for calculating the relevance of all variable information, the number k of evaluation indicators is set to 3 and 2 respectively, and the operator Φ is calculated by determining the constant K, and the equipment condition of the gas-insulated combination electrical appliance is analyzed and evaluated by the operator Φ.

[0102] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A vibration ultrasonic partial discharge signal analysis method based on factor analysis weighting method, characterized by: The following steps are included: A001: Extract ultrasonic partial discharge evaluation indicators of gas-insulated switchgear; A002: Extract the vibration signal evaluation index of the gas-insulated switchgear during normal operation and under test operation at a frequency of 0-1000Hz; A003: establishing an evaluation index set U according to the ultrasonic partial discharge evaluation index extracted in step A001 and the vibration signal evaluation index extracted in step A002; A004: Classify the insulation status of gas-insulated switchgear, and correspond the insulation levels according to the insulation status types, and construct the comment set X for the evaluation model according to the insulation status types and insulation levels; A005: Through the evaluation index set U and the comment set X, the weights of the evaluation indexes in the evaluation index set U are analyzed by using the factor analysis method of mutual information, the information relevance of the evaluation indexes is calculated, and the equipment status of the gas-insulated switchgear is analyzed by the information relevance; In the step A001, the ultrasonic partial discharge evaluation index includes the power spectrum maximum value v1, the median frequency v2 and the average power frequency v3 in the frequency domain characteristics; In the step A002, the vibration information evaluation index includes the vibration signal amplitude v4 during normal operation and the vibration signal amplitude v5 during the operation to be detected; In step A004, the insulation status of the gas-insulated switchgear is classified, and the insulation status types correspond to the insulation levels respectively, and the evaluation model is constructed according to the insulation status types and insulation levels. The comment set X is as follows: the insulation status of the gas-insulated switchgear is divided into four states: normal, caution, abnormal, and severe. The four insulation states correspond to the four insulation status levels of H1, H2, H3, and H4 respectively; In step A005, the information relevance method of the evaluation index is calculated as follows: A0051: Define each evaluation index v through fuzzy membership function i The degree of membership; in, Denotes the evaluation index v i Belongs to state H m The membership degree of , H(μ) represents the evaluation set; A0052: Determine the evaluation index v by fuzzy clustering method i The membership boundary of A0053: Establishing an evaluation matrix R, and using the evaluation index set U to judge the comment set X by establishing the evaluation matrix R; A0054: Calculate the weights of the evaluation indicators and calculate the information relevance of all variables; In step A0052, the evaluation index v is determined by fuzzy clustering method. i The method of obtaining the membership boundary of the evaluation index is as follows: obtaining the cluster center and cluster radius of each evaluation index in the evaluation index set U, and constructing the values ​​of a1, a2, a3, a4, a5, a6, a7, and a8 according to the cluster center and cluster radius; Set the evaluation metric v i After FCM function clustering, the cluster centers are o i1 , o i2 , o i3 , o i4 , the corresponding cluster radius is r i1 , r i2 , r i3 , r i4 , then the evaluation index v i The corresponding boundaries of the fuzzy membership function are a i1 , a i2 , a i3 , a i4 , a i5 , a i6 , a i7 , a i8 , can be calculated by the following formula: Among them, a i1 , a i3 and a i5 are the center value and two boundary values ​​of H2 level, respectively. i2 is the right boundary value of H1 level, a i6 , a i4 and a i8 are the center value and two boundary values ​​of H3 level, respectively. i7 is the left boundary value of H4 level, r i1 , r i2 , r i3 , r i4 They represent the membership of the i-th evaluation index to the four insulation status levels of normal, caution, abnormal, and severe; In step A0052, the evaluation index v i The degree of membership r belongs to four insulation status levels H1, H2, H3, H4 i1 , r i2 , r i3 , r i4 As shown below, H1: Normal state H2: Attention state H3: Abnormal state H4: Severe status Among them, U i Represents the value of the i-th evaluation parameter in the evaluation indicators v1-v5.