CN110865260A - Method for monitoring and evaluating MOV actual state based on outlier detection - Google Patents

Method for monitoring and evaluating MOV actual state based on outlier detection Download PDF

Info

Publication number
CN110865260A
CN110865260A CN201911196768.XA CN201911196768A CN110865260A CN 110865260 A CN110865260 A CN 110865260A CN 201911196768 A CN201911196768 A CN 201911196768A CN 110865260 A CN110865260 A CN 110865260A
Authority
CN
China
Prior art keywords
distance
point
mov
data
points
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
CN201911196768.XA
Other languages
Chinese (zh)
Inventor
杨仲江
马俊彦
王昊
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.)
Nanjing University of Information Science and Technology
Original Assignee
Nanjing University of Information Science and Technology
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 Nanjing University of Information Science and Technology filed Critical Nanjing University of Information Science and Technology
Priority to CN201911196768.XA priority Critical patent/CN110865260A/en
Publication of CN110865260A publication Critical patent/CN110865260A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
    • 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
    • G01R31/003—Environmental or reliability tests
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/16—Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Computational Mathematics (AREA)
  • Mathematical Analysis (AREA)
  • Mathematical Optimization (AREA)
  • Pure & Applied Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Theoretical Computer Science (AREA)
  • Computing Systems (AREA)
  • Environmental & Geological Engineering (AREA)
  • Algebra (AREA)
  • Databases & Information Systems (AREA)
  • Software Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Testing Or Calibration Of Command Recording Devices (AREA)

Abstract

The invention discloses a method for monitoring and evaluating the actual state of an MOV (metal oxide varistor) based on outlier detection, which comprises the following steps of firstly, sampling actual parameters of the MOV at the operation moment for several times at different time points to form an original data set; secondly, performing principal component analysis on the original data set to reduce the collected characteristic dimension and realize visualization on a two-dimensional or three-dimensional layer; then, calculating the distance between two adjacent points, the kth distance neighborhood, the reachable distance and the local reachable density between data points in the data set after PCA according to the local abnormal factor; and finally, calculating local outlier factors to obtain an outlier detection evaluation result of the MOV random sampling data. The invention reduces the executable difficulty of the traditional on-line monitoring, has industrial application value, strengthens the effective way of monitoring and maintaining the lightning protection device by people, and prevents the risk conditions of damage, fire, explosion and the like of an electrical system caused by abnormal operation or degradation of an MOV device.

Description

Method for monitoring and evaluating MOV actual state based on outlier detection
Technical Field
The invention belongs to the field of monitoring design of lightning protection devices, and particularly relates to a method for monitoring and evaluating the actual state of an MOV (metal oxide varistor) based on outlier detection.
Background
A Metal Oxide Varistor (MOV) is a resistive device with nonlinear current-voltage characteristics that protects sensitive devices within the system by voltage clamping when there is a risk of lightning overvoltage generation. Therefore, the lightning surge protection device is widely applied to lightning surge protection of various electrical and communication systems and the like. In addition to the threats of lightning overvoltage, operation overvoltage and the like, the MOV bears the influence of external environmental factors such as temperature, humidity, chemical pollution, dirt and the like during operation, so that the electrical characteristics and the physical state are changed. This leads to abnormal current-voltage characteristics of the MOV, reduced thermal stability, thermal breakdown after operation, and the like, and finally, irreversible deterioration of the physical state of the MOV valve sheet occurs. This severely affects the performance of the MOV and also exposes the electrical system to the risk of lightning damage. In order to be able to evaluate the actual physical state of each MOV in an electrical system and to find devices with potentially poor physical state and risk of degradation, various researchers in various fields have made much research work with this as an objective.
For the related MOV monitoring technology, Liwai et al propose to evaluate the deterioration condition of the ZnO varistor based on a single circulation energy based on an energy model, determine the deterioration degree of the varistor by analyzing and calculating the energy change generated during the action, but the energy characteristic is obtained by establishing a linear relation based on the leakage current, and further study on more generalized non-linear relation description is lacked, Yangzhongjiang et al find that the non-linear coefficient α decreases with the deterioration of the arrester by studying the non-linear coefficient α of the varistor, so that the non-linear coefficient α is added as an index into an evaluation system of the deterioration of the arrester, but the deterioration condition of the arrester itself needs to be comprehensively and superimposed by a plurality of parameters to obtain more accurate results, only one parameter using the non-linear coefficient α generates a certain error, and simultaneously research finds that good monitoring of the deterioration problem of the device can be realized based on the indexes of the resistance, resistance current and the like can be obtained by combining the optimal calculation method of obtaining a resistance and the optimal compensation algorithm.
With the development of fire and heat of machine learning, a data-driven model is more and more emphasized by learners, so that the data of the original sampling parameters of the MOVs, which belong to outlier distribution in the data space, are monitored by a new idea based on the construction of a multi-parameterized data set and combined with a local abnormal factor algorithm in an outlier detection method, and are converted into abnormal scores, thereby providing a new monitoring and evaluation method for the actual state or physical degradation degree of each MOV in the electrical field and system.
Disclosure of Invention
The purpose of the invention is as follows: the invention provides a method for monitoring and evaluating the actual state of the MOV based on outlier detection, which reduces the executable difficulty of the traditional online monitoring and has industrial application value.
The technical scheme is as follows: the invention discloses a method for monitoring and evaluating the actual state of an MOV (metal oxide varistor) based on outlier detection, which comprises the following steps of:
(3) sampling each actual parameter of MOV operation moment for several times at different time points to form an original data set;
(4) performing principal component analysis on the original data set to reduce the collected characteristic dimension so as to improve the result precision and realize the visualization on a two-dimensional or three-dimensional layer;
(3) calculating the distance between two adjacent points, the kth distance neighborhood, the reachable distance and the local reachable density between data points in the data set after PCA according to the local abnormal factor;
(4) and calculating local outlier factors by using the parameters obtained by calculation to obtain an outlier detection evaluation result of the MOV sampling data at any time.
Further, the step (1) includes the steps of:
(11) determining MOV in the electrical system to be detected as an outlier detection target;
(12) considering characteristic parameters influencing MOV physical characteristics and actual operation states, including physical parameters and environment influence parameters, recording and acquiring original characteristic data through a detection instrument, wherein the characteristic parameters include voltage-dependent voltage, leakage current, Cp (50Hz), Cp (10kHz), Ls (10kHz), Rs (10kHz), ambient temperature, humidity and other characteristic factors;
(13) and (5) sorting and checking the characteristic data to form an original data set.
Further, the step (2) comprises the steps of:
(21) carrying out standardization processing on the raw data:
suppose P samples { X1,X2,...,XMN-dimensional features per sample }
Figure BDA0002294849480000031
Each feature xjAll have respective eigenvalues, construct a data matrix, and perform the following normalized conversion on the data matrix:
Figure BDA0002294849480000032
wherein the content of the first and second substances,
Figure BDA0002294849480000033
obtaining a standardized matrix Z;
(22) the covariance matrix R is calculated for the normalized dataset:
Figure BDA0002294849480000034
wherein the content of the first and second substances,
Figure BDA0002294849480000035
(23) calculating characteristic equation | R-lambada I of covariance matrix R corresponding to data setpObtaining p characteristic roots and determining a principal component, wherein | ═ 0;
(24) converting the normalized index variable into principal component
Figure BDA0002294849480000036
Wherein, U1Is a first main component, U2Is the second main component, UpIs the pth principal component;
(25) and weighting and summing the original standardized data set according to the principal components to obtain a new dimension-reduced optimized data set.
Further, the step (3) includes the steps of:
(31) calculating the straight-line distance d (p, o) between two adjacent data points
(32) Calculate the kth distance of data point p: the k-th distance of the point p is dk(p) ═ d (p, o), and satisfies: at least k points o 'epsilon C { x ≠ p } in the set, which do not include p, and d (p, o') is less than or equal to d (p, o); at most k-1 points o 'belonging to C { x ≠ p } excluding p in the set, and d (p, o') < d (p, o);
(33) calculate the kth distance neighborhood of data point p: k-th distance neighborhood N of point pk(p) all data points within the kth distance of p, including the kth distance; p is the number of k-th distance neighborhood points | Nk(p)|≥k；
(34) The k-th reachable distance from point o to point p is defined as:
reach-distancek(p,o)＝max{k-distance(o),d(p,o)}
the k-th reachable distance from the point o to the point p, or the k-th distance of the point o, and the real distance between the point o and the point p;
(35) the local reachable density of point p is expressed as:
Figure BDA0002294849480000041
expressed as the inverse of the average reachable distance of a point p from p within the kth distance neighborhood of point p.
Further, the local outlier factor in step (4) is represented as:
Figure BDA0002294849480000042
where this value represents the neighborhood point N of point pk(p) the average of the ratio of the local achievable density of point p to the local achievable density of point p; if the ratio is closer to 1, the density of the neighborhood points of p is similar, and p and the neighborhood belong to the same cluster; if the ratio is less than 1, the density of p is higher than that of the neighborhood points, and p is a dense point; if the ratio is greater than 1, indicating that the density of p is less than its neighborhood point density, p is more likely to be an outlier.
Has the advantages that: compared with the prior art, the invention has the beneficial effects that: 1. the advantages of the outlier detection algorithm are fully utilized, and the outlier detection is carried out by utilizing the moment multi-parameter sampling data of the lightning protection device (MOV) based on the local abnormal factor algorithm, so that the evaluation and monitoring of the actual running state and the potential degradation tendency of the device are realized; 2. the invention reduces the executable difficulty of the traditional on-line monitoring, has industrial application value, strengthens the effective way of monitoring and maintaining the lightning protection device by people, and prevents the risk conditions of damage, fire, explosion and the like of an electrical system caused by abnormal operation or degradation of an MOV device.
Drawings
FIG. 1 is a flow chart of the present invention;
FIG. 2 is a distribution plot in two-dimensional space of sampled data (assumed values) after principal component analysis according to the present invention;
FIG. 3 is a schematic diagram illustrating the calculation of the kth distance according to the present invention;
FIG. 4 is a schematic diagram illustrating the calculation of the reachable distance according to the present invention;
FIG. 5 is a graph of outlier data results after outlier detection by local outlier factors.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings.
Local anomaly factors (LOFs) are an effective method based on performing outlier detection on medium and high dimensional datasets. The main purpose of the method is to detect abnormal data or behaviors which are greatly different from normal data behaviors or characteristic attributes. The algorithm reflects the degree of abnormality of a set of sample data based on the characteristic parameter of local reachable density, and if the local reachable density of the set of data points is larger, the more likely the point is an outlier. The monitoring and evaluating work of the MOV actual state means that the method can be used for combining a large amount of MOV multi-parameter sampling data to train a model to learn the distribution of normal parameter data, and the learned distribution is used for detecting abnormal parameter data which cannot be fitted into the distribution. The greater the local achievable density of MOV data, the worse the corresponding actual performance.
The local anomaly factor (LOF) calculates parameters such as the distance between two adjacent points, the kth distance neighborhood, the reachable distance and the like to obtain the local reachable density between the sample data point sets, wherein the farther the distance between the data points is, the lower the density is, the closer the distance is, and the higher the density is. Then according to the local reachable density, further defining local abnormal factor, and judging whether the data to be detected is an outlier or not by comparing the density of the target data point p with the density of the neighborhood of the target data point p by the characteristic, and judging the local abnormal factor LOFkThe more the value of (p) is greater than 1, the more likely the data point p to be measured is an outlier.
The LOF algorithm can be used for identifying the sampling data corresponding to MOV in the deterioration and abnormal operation states, and the sampling data are described in a concise and vivid manner for qualitative analysis and system evaluation. The algorithm takes data as drive, obtains an evaluation result by analyzing the characteristics among the data, effectively reduces the difficulty of the practical application of the MOV monitoring technology in the past, and has more accurate quantitative evaluation result for monitoring the actual state of the MOV along with the increase of the accumulation of the sampling data. As shown in fig. 1, the present invention specifically includes the following steps:
1. and sampling actual parameters of the MOV at the running action moment for several times at different time points to form an original data set.
(1) And determining the MOV in the electric system to be detected as the target of the outlier detection.
(2) Considering characteristic parameters influencing MOV physical characteristics and actual operation states, including physical parameters and environment influence parameters, recording and acquiring original characteristic data including voltage-dependent voltage, leakage current, Cp (50Hz), Cp (10kHz), Ls (10kHz), Rs (10kHz), ambient temperature, humidity and other characteristic factors through a related detection instrument.
(3) And (5) sorting and checking the characteristic data to form an original data set.
2. And carrying out principal component analysis on the original data set to reduce the acquired characteristic dimension so as to improve the result precision and realize the visualization on a two-dimensional or three-dimensional layer.
(1) Data is standardized
Suppose P samples { X1,X2,...,XMEach sample has N-dimensional features
Figure BDA0002294849480000061
Each feature xjEach having a respective characteristic value. Constructing a data matrix, and carrying out standardized conversion on the data matrix as follows:
Figure BDA0002294849480000062
wherein the content of the first and second substances,
Figure BDA0002294849480000063
a normalized matrix Z is obtained.
(2) The covariance matrix R is calculated for the normalized dataset:
Figure BDA0002294849480000064
wherein the content of the first and second substances,
Figure BDA0002294849480000065
(3) calculating characteristic equation | R-lambada I of covariance matrix R corresponding to data setpAnd (5) obtaining p characteristic roots and determining the principal component. Push button
Figure BDA0002294849480000066
To determine the m value to make the information utilization rate reach more than 85%, and for each lambdajJ 1,2, m, solving the system of equations Rb λjb, obtaining unit characteristic vector
Figure BDA0002294849480000067
(4) Converting the normalized index variable into principal component
Figure BDA0002294849480000068
Wherein, U1Is a first main component, U2Is the second main component, UpIs the pth principal component.
(5) And weighting and summing the original standardized data set according to the principal components to obtain a new dimension-reduced optimized data set. As shown in fig. 2, the raw data set consisting of the sampled parameters obtained by instrumental post-recording of the MOV is visualized in two dimensions after PCA.
3. And calculating the distance between two adjacent points in the data set after PCA, the kth distance neighborhood, the reachable distance and the Local reachable density according to a Local Outlier Factor (LOF algorithm).
(1) The straight-line distance d (p, o) between two adjacent data points is calculated.
(2) The kth distance (k-distance) of the data point p is calculated. As shown in FIG. 3, the kth distance of point p is dk(p) ═ d (p, o), and satisfies: at least k points o 'epsilon C { x ≠ p } in the set, which do not include p, satisfy d (p, o') ≦ d (p)O); at most k-1 points o 'belonging to C { x ≠ p } excluding p in the set, and d (p, o') < d (p, o);
(3) computing a kth distance neighborhood N of a kth distance neighborhood (kth distance neighborhood of p) point p of a data point pk(p) all data points within the kth distance of p, including the kth distance. p is the number of k-th distance neighborhood points | Nk(p)|≥k。
(4) Reach-distance (reach-distance), as shown in fig. 4, the k-th reachable distance from point o to point p is defined as:
reach-distancek(p,o)＝max{k-distance(o),d(p,o)}
the k-th reachable distance from point o to point p, or the k-th distance of o, the true distance between o and p. This means that the k points nearest to point o, the reachable distances between o and them are considered equal and all equal to dk(o)。
(5) Local accessibility density (local accessibility density), the local accessibility density of point p is expressed as:
Figure BDA0002294849480000071
expressed as the inverse of the average reachable distance of a point p from p within the kth distance neighborhood of point p. The resulting value represents a density, the higher the density, the more likely it is considered to belong to the same cluster, and the lower the density, the more likely it is an outlier. If p and surrounding neighborhood points are in the same cluster, the more likely the reachable distance is d, which is smallerk(o), resulting in smaller sum of reachable distances and higher density values; if p and surrounding neighbor points are far apart, the reachable distance may both take a larger value of dp(o) determining the density value as a low density value and determining the density value as an outlier.
4. And calculating local outlier factors by using the parameters obtained by calculation to obtain an outlier detection evaluation result of the MOV sampling data at any time. According to the evaluation result, the actual MOV operation state corresponding to the sampling data is obtained, the actual sampling data parameters are researched, the system safety analysis suggestion is given, and the measures for preventing the further deterioration of the device are made.
From the parameters calculated above, a local outlier factor (local outlier factor) is calculated, and the local outlier factor for point p is expressed as:
Figure BDA0002294849480000081
the value representing the neighborhood point N of point pk(p) an average of a ratio of the local achievable density of (p) to the local achievable density of point p. If the ratio is closer to 1, the density of the neighborhood points of p is similar, and p may belong to the same cluster as the neighborhood; if the ratio is less than 1, the density of p is higher than that of the neighborhood points, and p is a dense point; if this ratio is greater than 1, indicating that the density of p is less than its neighborhood point density, p is more likely to be an outlier.
The evaluation result of the sampling data by using the local outlier factor algorithm is shown in fig. 5, the actual MOV operating state corresponding to the sampling data is known according to the evaluation result, the actual sampling data parameters are researched, the system safety analysis opinion is given, and the measures for preventing the further deterioration of the device are made.

Claims (5)

1. A method for monitoring and evaluating the actual state of an MOV based on outlier detection is characterized by comprising the following steps:
(1) sampling each actual parameter of MOV operation moment for several times at different time points to form an original data set;
(2) performing principal component analysis on the original data set to reduce the collected characteristic dimension so as to improve the result precision and realize the visualization on a two-dimensional or three-dimensional layer;
(3) calculating the distance between two adjacent points, the kth distance neighborhood, the reachable distance and the local reachable density between data points in the data set after PCA according to the local abnormal factor;
(4) and calculating local outlier factors by using the parameters obtained by calculation to obtain an outlier detection evaluation result of the MOV sampling data at any time.
2. A method of assessing MOV physical condition monitoring based on outlier detection as claimed in claim 1 wherein said step (1) comprises the steps of:
(11) determining MOV in the electrical system to be detected as an outlier detection target;
(12) considering characteristic parameters influencing MOV physical characteristics and actual operation states, including physical parameters and environment influence parameters, recording and acquiring original characteristic data through a detection instrument, wherein the characteristic parameters include voltage-dependent voltage, leakage current, Cp (50Hz), Cp (10kHz), Ls (10kHz), Rs (10kHz), ambient temperature, humidity and other characteristic factors;
(13) and (5) sorting and checking the characteristic data to form an original data set.
3. A method of monitoring and assessing the actual status of an MOV based on outlier detection as claimed in claim 1 wherein said step (2) comprises the steps of:
(21) carrying out standardization processing on the raw data:
suppose P samples { X1,X2,...,XMN-dimensional features per sample }
Figure FDA0002294849470000011
Each feature xjAll have respective eigenvalues, construct a data matrix, and perform the following normalized conversion on the data matrix:
Figure FDA0002294849470000012
wherein the content of the first and second substances,
Figure FDA0002294849470000013
obtaining a standardized matrix Z;
(22) the covariance matrix R is calculated for the normalized dataset:
Figure FDA0002294849470000021
wherein the content of the first and second substances,
Figure FDA0002294849470000022
(23) calculating characteristic equation | R-lambada I of covariance matrix R corresponding to data setpObtaining p characteristic roots and determining a principal component, wherein | ═ 0;
(24) converting the normalized index variable into principal component
Figure FDA0002294849470000023
Wherein, U1Is a first main component, U2Is the second main component, UpIs the pth principal component;
(25) and weighting and summing the original standardized data set according to the principal components to obtain a new dimension-reduced optimized data set.
4. A method of assessing MOV physical condition monitoring based on outlier detection as claimed in claim 1 wherein said step (3) comprises the steps of:
(31) calculating the straight-line distance d (p, o) between two adjacent data points
(32) Calculate the kth distance of data point p: the k-th distance of the point p is dk(p) ═ d (p, o), and satisfies: at least k points o 'epsilon C { x ≠ p } in the set, which do not include p, and d (p, o') is less than or equal to d (p, o); at most k-1 points o 'belonging to C { x ≠ p } excluding p in the set, and d (p, o') < d (p, o);
(33) calculate the kth distance neighborhood of data point p: k-th distance neighborhood N of point pk(p) all data points within the kth distance of p, including the kth distance; p is the number of k-th distance neighborhood points | Nk(p)|≥k；
(34) The k-th reachable distance from point o to point p is defined as:
reach-distancek(p,o)＝max{k-distance(o),d(p,o)}
the k-th reachable distance from the point o to the point p, or the k-th distance of the point o, and the real distance between the point o and the point p;
(35) the local reachable density of point p is expressed as:
Figure FDA0002294849470000031
expressed as the inverse of the average reachable distance of a point p from p within the kth distance neighborhood of point p.
5. A method for MOV physical condition monitoring assessment based on outlier detection as claimed in claim 1 wherein said local outlier factor of step (4) is expressed as:
Figure FDA0002294849470000032
where this value represents the neighborhood point N of point pk(p) the average of the ratio of the local achievable density of point p to the local achievable density of point p; if the ratio is closer to 1, the density of the neighborhood points of p is similar, and p and the neighborhood belong to the same cluster; if the ratio is less than 1, the density of p is higher than that of the neighborhood points, and p is a dense point; if the ratio is greater than 1, indicating that the density of p is less than its neighborhood point density, p is more likely to be an outlier.
CN201911196768.XA 2019-11-29 2019-11-29 Method for monitoring and evaluating MOV actual state based on outlier detection Pending CN110865260A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201911196768.XA CN110865260A (en) 2019-11-29 2019-11-29 Method for monitoring and evaluating MOV actual state based on outlier detection

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201911196768.XA CN110865260A (en) 2019-11-29 2019-11-29 Method for monitoring and evaluating MOV actual state based on outlier detection

Publications (1)

Publication Number Publication Date
CN110865260A true CN110865260A (en) 2020-03-06

Family

ID=69657614

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201911196768.XA Pending CN110865260A (en) 2019-11-29 2019-11-29 Method for monitoring and evaluating MOV actual state based on outlier detection

Country Status (1)

Country Link
CN (1) CN110865260A (en)

Cited By (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111443259A (en) * 2020-03-30 2020-07-24 国网山东省电力公司德州供电公司 Active power distribution network fault diagnosis method and system based on local abnormal factor detection
CN113190792A (en) * 2021-04-18 2021-07-30 宁波大学科学技术学院 Ethylene cracking furnace operation state monitoring method based on neighbor local abnormal factors
CN115238753A (en) * 2022-09-21 2022-10-25 西南交通大学 Self-adaptive SHM data cleaning method based on local outlier factor
CN115423807A (en) * 2022-11-04 2022-12-02 山东益民服饰有限公司 Cloth defect detection method based on outlier detection
CN115809435A (en) * 2023-02-06 2023-03-17 山东星科智能科技股份有限公司 Simulator-based automobile operation fault identification method
CN116879662A (en) * 2023-09-06 2023-10-13 山东华尚电气有限公司 Transformer fault detection method based on data analysis
CN117708748A (en) * 2024-02-05 2024-03-15 苏州众志新环冷却设备有限公司 Operation monitoring system and method for centrifugal fan

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102565604A (en) * 2010-12-07 2012-07-11 中国电信股份有限公司 Device and method for automatic alarming of power supply arrestor
CN105137163A (en) * 2015-07-29 2015-12-09 王富元 Metal oxide varistor(MOV)-type arrester or surge protector resistive leakage current online monitoring method
CN105279761A (en) * 2015-11-18 2016-01-27 山东大学 A background modeling method based on sample local density outlier detection
CN107294213A (en) * 2017-07-29 2017-10-24 梧州井儿铺贸易有限公司 A kind of grid equipment intelligent monitor system
CN107947666A (en) * 2017-12-26 2018-04-20 奥克斯空调股份有限公司 A kind of resistance parameter on-line identification method and device
CN108599365A (en) * 2017-12-08 2018-09-28 国网浙江平阳县供电有限责任公司 A kind of intelligent power distribution net
CN109446189A (en) * 2018-10-31 2019-03-08 成都天衡智造科技有限公司 A kind of technological parameter outlier detection system and method

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102565604A (en) * 2010-12-07 2012-07-11 中国电信股份有限公司 Device and method for automatic alarming of power supply arrestor
CN105137163A (en) * 2015-07-29 2015-12-09 王富元 Metal oxide varistor(MOV)-type arrester or surge protector resistive leakage current online monitoring method
CN105279761A (en) * 2015-11-18 2016-01-27 山东大学 A background modeling method based on sample local density outlier detection
CN107294213A (en) * 2017-07-29 2017-10-24 梧州井儿铺贸易有限公司 A kind of grid equipment intelligent monitor system
CN108599365A (en) * 2017-12-08 2018-09-28 国网浙江平阳县供电有限责任公司 A kind of intelligent power distribution net
CN107947666A (en) * 2017-12-26 2018-04-20 奥克斯空调股份有限公司 A kind of resistance parameter on-line identification method and device
CN109446189A (en) * 2018-10-31 2019-03-08 成都天衡智造科技有限公司 A kind of technological parameter outlier detection system and method

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
廖玉芳: "《湖南油茶产量气象条件分析》", 30 April 2019 *
郭丽娟: "基于离群点检测的变电站主设备异常辨识与规律分析", 《南方电网技术》 *
靳邵云: "压敏电阻劣化敏感参数分析与劣化模型构建研究", 《电瓷避雷器》 *

Cited By (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111443259A (en) * 2020-03-30 2020-07-24 国网山东省电力公司德州供电公司 Active power distribution network fault diagnosis method and system based on local abnormal factor detection
CN113190792A (en) * 2021-04-18 2021-07-30 宁波大学科学技术学院 Ethylene cracking furnace operation state monitoring method based on neighbor local abnormal factors
CN113190792B (en) * 2021-04-18 2023-10-24 宁波大学科学技术学院 Ethylene cracking furnace running state monitoring method based on neighbor local abnormal factors
CN115238753A (en) * 2022-09-21 2022-10-25 西南交通大学 Self-adaptive SHM data cleaning method based on local outlier factor
CN115423807A (en) * 2022-11-04 2022-12-02 山东益民服饰有限公司 Cloth defect detection method based on outlier detection
CN115809435A (en) * 2023-02-06 2023-03-17 山东星科智能科技股份有限公司 Simulator-based automobile operation fault identification method
CN115809435B (en) * 2023-02-06 2023-05-12 山东星科智能科技股份有限公司 Automobile operation fault identification method based on simulator
CN116879662A (en) * 2023-09-06 2023-10-13 山东华尚电气有限公司 Transformer fault detection method based on data analysis
CN116879662B (en) * 2023-09-06 2023-12-08 山东华尚电气有限公司 Transformer fault detection method based on data analysis
CN117708748A (en) * 2024-02-05 2024-03-15 苏州众志新环冷却设备有限公司 Operation monitoring system and method for centrifugal fan
CN117708748B (en) * 2024-02-05 2024-05-03 苏州众志新环冷却设备有限公司 Operation monitoring system and method for centrifugal fan

Similar Documents

Publication Publication Date Title
CN110865260A (en) Method for monitoring and evaluating MOV actual state based on outlier detection
CN110336534B (en) Fault diagnosis method based on photovoltaic array electrical parameter time series feature extraction
CN112255484B (en) Lightning arrester operation state online monitoring and assessment method and system
CN113687176B (en) Deep neural network-based power consumption abnormity detection method and system
Zuo et al. Calculation of a health index of oil‐paper transformers insulation with binary logistic regression
CN103592545A (en) Transformer temperature rise abnormity monitoring and diagnosis method based on probability statistics
CN114910756A (en) Insulation performance evaluation method and system for low-voltage bus duct
CN114004162A (en) Modeling method for smelting load harmonic emission level under multi-working-condition scene
CN114217025B (en) Analysis method for evaluating influence of meteorological data on air quality concentration prediction
Zaka et al. Performance of attribute control charts for monitoring the shape parameter of modified power function distribution in the presence of measurement error
CN110298765B (en) Power distribution network power consumption abnormality detection method based on objective correlation factors
CN117407675A (en) Lightning arrester leakage current prediction method based on multi-variable reconstruction combined dynamic weight
CN117235664A (en) Fault diagnosis method and system for power distribution communication equipment and computer equipment
CN117216956A (en) Device, method and equipment for correcting thermal-oxidative aging model of XLPE cable
CN116819382A (en) DC-DC converter state monitoring method based on digital twin
Chen et al. Parameter optimisation of support vector machine using mutant particle swarm optimisation for diagnosis of metal-oxide surge arrester conditions
CN116400266A (en) Transformer fault detection method, device and medium based on digital twin model
CN115935814A (en) Transformer fault prediction method based on ARIMA-SVM model
CN113672658B (en) Power equipment online monitoring error data identification method based on complex correlation coefficient
Khodsuz et al. Surge Arrester Monitoring under Different Operating Conditions Using Bees-ANFIS.
CN115062533A (en) Construction and prediction method and medium of multiple regression prediction model
CN113190728A (en) Oil-immersed transformer fault diagnosis method based on cluster optimization
CN112800672A (en) Boiler fouling coefficient evaluation method, system, medium and electronic equipment
CN105975441A (en) Entity value quantitative evaluation method based on relation matrix
Liu et al. Research on nonlinear compensation scheme of yarn tension sensor using SAW devices based on SSA–SVR model

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
RJ01 Rejection of invention patent application after publication
RJ01 Rejection of invention patent application after publication

Application publication date: 20200306