CN106707060A - Method for acquiring discrete state parameters of power transformer - Google Patents
Method for acquiring discrete state parameters of power transformer Download PDFInfo
- Publication number
- CN106707060A CN106707060A CN201611166135.0A CN201611166135A CN106707060A CN 106707060 A CN106707060 A CN 106707060A CN 201611166135 A CN201611166135 A CN 201611166135A CN 106707060 A CN106707060 A CN 106707060A
- Authority
- CN
- China
- Prior art keywords
- power transformer
- parameter
- state parameter
- transformer
- designated
- 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.)
- Granted
Links
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
Abstract
The invention discloses a method for acquiring discrete state parameters of a power transformer, which solves problems of difficult power transformer state parameter recognition and difficult power transformer multi-source data mining. According to the invention, a discrete gas state parameter is acquired by adopting an improved equal width method according to the type and the content of dissolved gases in tested power transformer oil; a discrete partial discharge state parameter is acquired by adopting an information entropy decision tree method according to a discharge capacity statistical parameter, a discharge phase statistical parameter and a number of discharge statistical parameter; a discrete winding state parameter is acquired by adopting a binning method according to winding deformation and short-circuit current; and a discrete mechanical state parameter is acquired by adopting a chi-square test method according to the transformer temperature, vibration and noises. The method disclosed by the invention is applicable to the fields of power equipment detection and data mining, has the characteristics o f accurate data, high work efficiency, wide fault coverage, easy computer processing and the like, and has high economic values and broad market application prospects.
Description
Technical field
It is particularly a kind of to obtain power transformer the present invention relates to a kind of method for obtaining power equipment discrete type state parameter
The method of device discrete type state parameter, belongs to power system detection field.
Background technology
The power system energy backing powerful for the sustainable development of national economy is provided, and electrical equipment is used as power system
Composed component, its reliability is the guarantee of electric power netting safe running.With the change and progress of power load, original power transmission and transformation set
The limitation of the technical conditions such as the thermally-stabilised limit of current-carrying capacity is enjoyed, transmission capacity is substantially reduced, oneself undertakes power transmission task through being difficult to,
The need for power consumer can not being met, especially under peak times of power consumption, maintenance or when breaking down, whole transmission of electricity system
System intercommunication off-capacity will cause to press limited load, cause the economy of operation of power networks and reliability degradation.
It is technological means more effective at present to carry out detection to power equipment, and it can be carried out to Electric Power Equipment Insulation in real time
Detection, polytype detection data is collected and quick storage allows that facility information accumulates mass data.Then, extraction has
Facility information has become huge challenge.Much research shows both at home and abroad, electrical equipment online supervision information data
Related to factors, feelings are easily misrepresented deliberately, fail to report and reported by mistake to such as environment temperature, equipment service condition, equipment enlistment age
Condition.Power equipment is subject to various stress such as electricity, heat, machinery and chemistry in the process of running, and it is unavoidably destroyed
And damage, it is capable of achieving to be estimated the life-span of integral device by the analysis to one or more equipment.Power transformer is joined
Number can be various power transformer parameters such as oil dissolved gas, shelf depreciation, vibration, sound, according to power transformer with
Mapping relations between its parameter, by power equipment parameter detecting realize to one or more comprehensive characterization of equipment with comment
Valency.
Power Transformer Condition parameter is continuous data, and data are not easily separate during parameter detecting, and is easily occurred aobvious
Error is write, being difficult with computer carries out later stage supplemental characteristic mining analysis, it is therefore desirable to continuity Data Discretization.At present
There is no maturation suitable for status of electric power parameter discretization method, it is impossible to meet grid equipment state acquisition requirement, more
It is unable to reach the requirement of equipment safety reliability service.
The content of the invention
The technical problem to be solved in the present invention is to provide a kind of method of power transformer discrete type state parameter, solves existing
There is the problem that the identification of Power Transformer Condition parameter is difficult, power transformer multi-source data excavates difficulty in technology.
In order to solve the above-mentioned technical problem, the invention provides a kind of side for obtaining power transformer discrete type state parameter
Method, comprises the following steps：
The species and content of tested Detection Ssytem of Dissolved Gases in Power Transformer Oil Base are obtained by gas detection equipment；Judge every kind of molten
Whether the content for solving gas is more than predetermined threshold value：If so, the first mark is designated as, if it is not, the second mark is designated as, as power transformer
Device discrete type gaseous state parameter；
Power transformer discharge capacity statistical parameter, discharge phase statistical parameter are obtained with electric discharge by Partial discharge detector
Number of times statistical parameter, calculates the information gain maximum of these three statistical parameters and generates decision tree；Judged to unite according to decision tree
Whether meter parameter is abnormal：If so, being designated as the first mark, if it is not, being designated as the second mark, locally put as power transformer discrete type
Electricity condition parameter；
Power transformer is obtained by winding deformation detector to normalize under short circuit current and system maximum operational mode
Whether short circuit current, judge to normalize the ratio of the short circuit current under short circuit current and system maximum operational mode more than 1：If
It is to be designated as the first mark, if it is not, the second mark is designated as, as power transformer discrete type winding state parameter；
By machine performance testing equipment obtain power transformer machine performance parameter, by machine performance parameter each not
Identical numerical value is set to an interval；Whether the chi amount of each pair adjacent interval is calculated, the interval chi amount of each pair is judged
More than card side's significance value：If so, the first mark is designated as, if it is not, the second mark is designated as, as power transformer discrete type machinery
State parameter.
The dissolved gas include hydrogen, methane, ethane, ethene and acetylene.
The predetermined threshold value includes hydrogen threshold value, methane threshold value, ethane threshold value, ethene threshold value and acetylene threshold value；Judge every
Whether the content for planting dissolved gas is more than the predetermined threshold value of same gas.
The predetermined threshold value setting procedure includes：
The Power Transformer Faults case relevant with the dissolved gas is collected, fault data data acquisition system is set up；
Normal distribution analysis is carried out according to fault data data acquisition system, the content curve of every kind of dissolved gas is obtained；
By the numerical value of corresponding 95% distribution of the content curve of every kind of dissolved gas, as corresponding predetermined threshold value.
The discharge phase includes maximum discharge phase and averaged discharge phase.
The statistical parameter is included with Types Below：Degree of skewness, standout, the cross-correlation factor in shelf depreciation collection of illustrative plates with it is right
Title property；The method of the generation decision tree is as follows：The corresponding information gain maximum of each type statistical parameter is calculated respectively,
Using the corresponding statistical parameter symbol of information gain maximum as the splitting condition of decision tree, and decision tree is generated；
Whether exception procedure is as follows to judge statistical parameter：
From the root node of decision tree, if statistical parameter symbol is less than splitting condition, the second mark is designated as, represented just
Often；If statistical parameter symbol is more than splitting condition, into the corresponding subtree root node of current node, subtree root node pair is carried out
The splitting condition answered judges, until into abnormal terminal node is represented, being designated as the first mark.
By detecting and the ratio of short circuit in winding electric current and power transformer rated current is calculated, obtain the normalization short
Road electric current.
The machine performance parameter includes transformer temperature, transformer vibration and transformer noise；The transformer temperature
Corresponding mechanical detection equipment is infrared temperature imager, and transformer vibrates corresponding mechanical detection equipment for mechanical oscillation are detected
Instrument, the corresponding mechanical detection equipment of transformer noise is transformer noise detector.
Card side's significance value takes 0.05.
First mark takes 1, and the second mark takes 0.
The present invention is converted into discrete shape parameter by detection by by the serial number shape parameter of Power Transformer Condition, greatly
Width reduces calculating cycle, improves the disposal ability to mass data.
Discrete type state parameter is divided into Gases Dissolved in Transformer Oil state, shelf depreciation state, winding shape by the present invention
State and machine performance parameter, with failure diversity, comprehensive, improve power system detection efficiency.
The present invention is carried out discrete by the method for comentropy decision tree to local discharge condition statistical parameter, improves data
The accuracy in source.
The present invention is run by detecting that power transformer normalizes short circuit current according to winding deformation principle with system maximum
The discrete winding state parameter of method of the ratio of the short circuit current under mode, simple to operate, data are intuitive and reliable.
The present invention obtains discrete type machine performance parameter by the method for Chi-square Test, is not limited by data overall distribution
System, it is applied widely.
The present invention is classified by state-detection, the continuous type state parameter of power transformer is obtained, using corresponding method
Discrete is the numerical value of the first mark and the second mark pattern.With data acquisition is easy, simple to operate, failure broad covered area and inspection
The features such as surveying efficiency high, has dramatically speeded up data mining recognition speed, possesses wide market application foreground.
Brief description of the drawings
Fig. 1 is the inventive method schematic diagram；
Fig. 2 is Detection Ssytem of Dissolved Gases in Power Transformer Oil Base species and predetermined threshold value schematic diagram；
Fig. 3 is shelf depreciation status information entropy decision tree schematic diagram；
Fig. 4 is that winding state judges schematic diagram.
Specific embodiment
The invention will be further described with specific embodiment below in conjunction with the accompanying drawings, so that those skilled in the art can be with
More fully understand the present invention and can be practiced, but illustrated embodiment is not as a limitation of the invention.
Winding is the main building block of power transformer, and it realizes energy conversion using electromagnetic induction principle.Winding becomes
Shape refers in the case of short circuit in winding, to have very big short-circuit electric stress and produce, and be will act on winding, causes winding to occur
Certain deformation.Winding deformation testing equipment knows winding circuit state by basket vibration characteristic.Because Transformer Winding knot
Structure is different, and different short circuit currents is directly relatively difficult, therefore is compared with its equipment rated current by short circuit current, obtains
Data Discretization operation is carried out again after obtaining normalized short circuit current.
Power Transformer Condition parameter discretization is the important preparation of power transformer data mining.Data mining is walked
Suddenly mainly include：Understand data and data source, obtain relevant knowledge and technology, integrate with check data, removal it is wrong or
Inconsistent data, model and hypothesis, real data excacation, test and checking Result are set up, is explained and is applied.One
As for, data mining has involved substantial amounts of preparation and planning, and the time and efforts for having 80% is to spend in data
Pretreatment stage.The present invention be exactly by various structures and distinct transformer state supplemental characteristic, by discretization step,
Unified is discretization data as 1 or 0, has reached purification, Data Format Transform, variable conversion and the tables of data of data
The work such as link.
Embodiment 1：
Referring to accompanying drawing 1, the present embodiment is related to equipment to be put including tested power transformer system, gas detection equipment, part
Electro-detection instrument, winding deformation detector, infrared temperature imager, mechanical vibration detector and transformer noise detector, and
One computer.
The present embodiment course of work is as follows：
The species and content of tested Detection Ssytem of Dissolved Gases in Power Transformer Oil Base are obtained by gas detection equipment, and detection is tied
Fruit is sent to computer.Dissolved gas include hydrogen H2, methane CH4, ethane C2H6, ethene C2H4With acetylene C2H2, wherein methane,
Ethane, ethene and acetylene are referred to as total hydrocarbon in the lump.Predetermined threshold value is set in a computer：By power transformer in electrical engineering technology
The safe limit value of gas is used as predetermined threshold value.Referring to accompanying drawing 2, predetermined threshold value and total hydrocarbon of the predetermined threshold value including every kind of gas
Predetermined threshold value, unit is ppm ppm, is numerically equal to μ L/L.Wherein H2Predetermined threshold value be 150ppm, CH4's
Predetermined threshold value is 45ppm, C2H6Predetermined threshold value be 35ppm, C2H4Predetermined threshold value be 65ppm, C2H2Predetermined threshold value be
5ppm, the predetermined threshold value of total hydrocarbon is 150ppm.Judge the content of every kind of dissolved gas whether more than same gas by computer
Predetermined threshold value：If so, 1 is designated as, if it is not, being designated as 0；Judge the content of total hydrocarbon whether more than total hydrocarbon predetermined threshold value：If so, 1 is designated as,
If it is not, being designated as 0；Using above-mentioned boolean value as power transformer discrete type gaseous state parameter.
Power transformer discharge capacity statistical parameter, discharge phase statistical parameter are obtained with electric discharge by Partial discharge detector
Number of times statistical parameter, and send to computer.Statistical parameter includes standout Ku, the cross-correlation factor cc in shelf depreciation collection of illustrative plates
With symmetry Asym.Comentropy per class statistical parameter is calculated by computer, the maximum average value E (A) of comentropy is tried to achieve, i.e.,
Information gain maximum, A be the corresponding statistical parameter symbol of information gain maximum, and using A as decision tree splitting condition.
Referring to accompanying drawing 3, " Hqmax_cc " is the corresponding splitting conditions of cross-correlation factor cc, and " Hqn_asym " is that symmetry Asym is corresponding
Splitting condition, " Ku+, Ku- " is the corresponding splitting conditions of standout Ku, and wherein "+" represents the positive half period of standout, "-" table
Show the negative half-cycle of standout.From the root node of decision tree, if statistical parameter symbol is less than splitting condition, 0, table are designated as
Show normal；If statistical parameter symbol is more than splitting condition, into the corresponding subtree root node (right subtree) of current node, carry out
The corresponding splitting condition of subtree root node judges, until into abnormal terminal node is represented, being designated as 1；By above-mentioned boolean value
Used as power transformer discrete type shelf depreciation state parameter, the method can be described as " comentropy decision tree method ".
The short circuit in winding electric current and power transformer rated current of power transformer are obtained by winding deformation detector
Ratio, obtains normalization short circuit current Id；Short circuit current I under measuring system maximum operational modes, and send to computer.Ginseng
See accompanying drawing 4, I is judged by computerdWith IsRatio whether be more than 1：If so, being designated as 1, represent abnormal；If it is not, being designated as 0, table
Show normal；Above-mentioned boolean value as power transformer discrete type winding state parameter, the method can be described as " branch mailbox method ".
Transformer temperature is obtained by infrared temperature imager, transformer vibration is obtained by mechanical vibration detector, is led to
Cross transformer noise detector and obtain transformer noise, and send to computer.Computer shakes transformer temperature, transformer
Each of dynamic, transformer noise differs numerical value and is set to an interval；The chi amount of each pair adjacent interval is calculated, judges every
Whether more than card side significance value α (according to statistical law, α=0.05) is taken to interval chi amount：If so, being designated as 1, represent
It is abnormal；If it is not, being designated as 0, represent normal；Using above-mentioned boolean value as power transformer discrete type machine performance parameter, this side
Method can be described as " Chi-square method ".
Combined power transformer discrete type gas, shelf depreciation, winding and machine performance parameter, obtain power transformer from
Dissipate type state parameter.
The present embodiment is using the safe limit value of power transformer gas in electrical engineering technology as predetermined threshold value, and operation is simple
It is single, it is easy to which that method is transplanted.
The present embodiment judged total hydrocarbon content simultaneously, improves the accuracy of state parameter.
The present embodiment carries out Data Discretization by the method for comentropy decision tree (binary tree), with reference to computer data knot
Structure judged, improves data discrete efficiency.
The present embodiment, comprehensively will using Chi-square method by machine performance parameters such as transformer temperature, vibration and noises
Numerical discretization, with mechanical breakdown coverage rate wider.
Embodiment 2：
Referring to accompanying drawing 1, the present embodiment is related to equipment to be put including tested power transformer system, gas detection equipment, part
Electro-detection instrument, winding deformation detector, infrared temperature imager, mechanical vibration detector and transformer noise detector, and
One computer.
The present embodiment course of work is as follows：
The species and content of tested Detection Ssytem of Dissolved Gases in Power Transformer Oil Base are obtained by gas detection equipment, and detection is tied
Fruit is sent to computer.Dissolved gas include hydrogen H2, methane CH4, ethane C2H6, ethene C2H4With acetylene C2H2, wherein methane,
Ethane, ethene and acetylene are referred to as total hydrocarbon in the lump.The setting steps of predetermined threshold value are as follows：
The Power Transformer Faults case relevant with the dissolved gas is collected, fault data data is set up by computer
Set；
Normal distribution analysis is carried out according to fault data data acquisition system, the content curve of every kind of dissolved gas is obtained；
The numerical value that the content curve of every kind of dissolved gas corresponding 95% is distributed, as corresponding predetermined threshold value (electrically
It is generally acknowledged that probability of equipment failure is for 5%) in engineering field.
Predetermined threshold value includes the predetermined threshold value of every kind of gas and the predetermined threshold value of total hydrocarbon, and unit is ppm ppm,
It is numerically equal to μ L/L.Judge the content of every kind of dissolved gas whether more than same gas predetermined threshold value by computer：If
It is to be designated as 1, if it is not, being designated as 0；Judge the content of total hydrocarbon whether more than total hydrocarbon predetermined threshold value：If so, 1 is designated as, if it is not, being designated as 0；
Using above-mentioned boolean value as power transformer discrete type gaseous state parameter, the method can be described as " improving wide method ".
By Partial discharge detector obtain power transformer discharge capacity statistical parameter, maximum discharge phase statistical parameter,
Averaged discharge phase statistical parameter and discharge time statistical parameter, and send to computer.Statistical parameter includes shelf depreciation figure
Degree of skewness Sk, standout Ku, cross-correlation factor cc and symmetry Asym in spectrum.Calculated per class statistical parameter by computer
Comentropy, tries to achieve the maximum average value E (A) of comentropy, i.e. information gain maximum, and A is the corresponding system of information gain maximum
Meter pa-rameter symbols, and using A as decision tree splitting condition.From the root node of decision tree, if statistical parameter symbol is less than
Splitting condition, then be designated as 0, represents normal；If statistical parameter symbol is more than splitting condition, into the corresponding subtree of current node
Root node, carries out the corresponding splitting condition of subtree root node and judges, until into abnormal terminal node is represented, being designated as 1；Will be upper
Boolean value is stated as power transformer discrete type shelf depreciation state parameter, the method can be described as " comentropy decision tree method ".
The short circuit in winding electric current and power transformer rated current of power transformer are obtained by winding deformation detector
Ratio, obtains normalization short circuit current Id；Short circuit current I under measuring system maximum operational modes, and send to computer.Ginseng
See accompanying drawing 4, I is judged by computerdWith IsRatio whether be more than 1：If so, being designated as 1, represent abnormal；If it is not, being designated as 0, table
Show normal；Above-mentioned boolean value as power transformer discrete type winding state parameter, the method can be described as " branch mailbox method ".
Transformer temperature is obtained by infrared temperature imager, transformer vibration is obtained by mechanical vibration detector, is led to
Cross transformer noise detector and obtain transformer noise, and send to computer.Computer shakes transformer temperature, transformer
Each of dynamic, transformer noise differs numerical value and is set to an interval；The chi amount of each pair adjacent interval is calculated, judges every
Whether more than card side significance value α (according to statistical law, α=0.05) is taken to interval chi amount：If so, being designated as 1, represent
It is abnormal；If it is not, being designated as 0, represent normal；Using above-mentioned boolean value as power transformer discrete type machine performance parameter, this side
Method can be described as " Chi-square method ".
Combined power transformer discrete type gas, shelf depreciation, winding and machine performance parameter, obtain power transformer from
Dissipate type state parameter.
The present embodiment determines predetermined threshold value according to transformer fault case combination normal distribution method so that state parameter number
According to confidence level higher.
Degree of skewness Sk, standout Ku, cross-correlation factor cc and symmetry in the present embodiment synthesis shelf depreciation collection of illustrative plates
Asym improves the accuracy of data as statistical parameter.
Embodiment 3：
Referring to accompanying drawing 1, the present embodiment is related to equipment to be put including tested power transformer system, gas detection equipment, part
Electro-detection instrument, winding deformation detector, infrared temperature imager, mechanical vibration detector and transformer noise detector, and
One computer.
The present embodiment course of work is as follows：
The species and content of tested Detection Ssytem of Dissolved Gases in Power Transformer Oil Base are obtained by gas detection equipment, and detection is tied
Fruit is sent to computer.The determination step of dissolved gas is as follows：
The Power Transformer Faults case relevant with the dissolved gas is collected, fault data data is set up by computer
Set；
For each fault type, minimum support threshold is set according to artificial experience, different characteristic gas is carried out respectively
Frequency statistics, obtains all high-frequency characteristic gases, as dissolved gas.
The setting steps of predetermined threshold value are as follows：
According to the high-frequency characteristic gas (dissolved gas) for filtering out and minimum reliability, different type failure is carried out respectively
Confidence level is calculated, and the data information set set up is traveled through using Apriori algorithm, determines correlation rule；
Dissolved gas numeric type field is processed by correlation rule, dynamic partition is carried out, as corresponding
Predetermined threshold value.
Predetermined threshold value unit is ppm ppm, is numerically equal to μ L/L.Every kind of dissolving is judged by computer
Whether the content of gas is more than same gas predetermined threshold value：If so, 1 is designated as, if it is not, being designated as 0；Using above-mentioned boolean value as electricity
Power transformer discrete type gaseous state parameter.
By Partial discharge detector obtain power transformer discharge capacity statistical parameter, maximum discharge phase statistical parameter,
Averaged discharge phase statistical parameter and discharge time statistical parameter, and send to computer.Statistical parameter includes shelf depreciation figure
Degree of skewness Sk, standout Ku, cross-correlation factor cc and symmetry Asym in spectrum.Calculated per class statistical parameter by computer
Comentropy, tries to achieve the maximum average value E (A) of comentropy, i.e. information gain maximum, and A is the corresponding system of information gain maximum
Meter pa-rameter symbols, and using A as decision tree splitting condition.From the root node of decision tree, if statistical parameter symbol is less than
Splitting condition, then be designated as 0, represents normal；If statistical parameter symbol is more than splitting condition, into the corresponding subtree of current node
Root node, carries out the corresponding splitting condition of subtree root node and judges, until into abnormal terminal node is represented, being designated as 1；Will be upper
Boolean value is stated as power transformer discrete type shelf depreciation state parameter, the method can be described as " comentropy decision tree method ".
The short circuit in winding electric current and power transformer rated current of power transformer are obtained by winding deformation detector
Ratio, obtains normalization short circuit current Id；Short circuit current I under measuring system maximum operational modes, and send to computer.Ginseng
See accompanying drawing 4, I is judged by computerdWith IsRatio whether be more than 1：If so, being designated as 1, represent abnormal；If it is not, being designated as 0, table
Show normal；Above-mentioned boolean value as power transformer discrete type winding state parameter, the method can be described as " branch mailbox method ".
Transformer temperature is obtained by infrared temperature imager, transformer vibration is obtained by mechanical vibration detector, is led to
Cross transformer noise detector and obtain transformer noise, and send to computer.Computer is by each not phase of transformer temperature
An interval is set to numerical value；The chi amount of each pair adjacent interval is calculated, judges whether each pair interval chi amount is big
(according to statistical law, α=0.05) is taken in card side significance value α：If so, being designated as 1, represent abnormal；If it is not, being designated as 0, represent just
Often；Using above-mentioned boolean value as power transformer discrete type state of temperature parameter.Computer transformer is vibrated each not
Identical numerical value is set to an interval；Whether the chi amount of each pair adjacent interval is calculated, the interval chi amount of each pair is judged
(according to statistical law, α=0.05) is taken more than card side significance value α：If so, being designated as 1, represent abnormal；If it is not, being designated as 0, represent
Normally；Using above-mentioned boolean value as power transformer discrete type vibrational state parameter.Computer by transformer noise each
Differ numerical value and be set to an interval；The chi amount of each pair adjacent interval is calculated, judges that each pair interval chi amount is
It is no (according to statistical law, to take α=0.05) more than card side significance value α：If so, being designated as 1, represent abnormal；If it is not, being designated as 0, table
Show normal；Using above-mentioned boolean value as power transformer discrete type noise states parameter.
Combined power transformer discrete type gas, shelf depreciation, winding, temperature, vibration and noise state parameter, obtain electricity
Power transformer discrete type state parameter.
The present embodiment according to Power Transformer Faults case and correlation rule determine dissolved gas that data analysis needs with
Predetermined threshold value, realizes the dynamic partition of different type dissolved gas state.
The present invention is analyzed by the discretization to Power Transformer Condition parameter, realizes the total evaluation to power equipment,
Simplify Power Transformer Condition supplemental characteristic mining process.The present invention is by detection by by the consecutive numbers of Power Transformer Condition
Value shape parameter is converted into 1 and 0 discrete boolean's shape parameter, significantly reduces calculating cycle, improves the treatment energy to mass data
Power.Discrete type state parameter is divided into Gases Dissolved in Transformer Oil state, shelf depreciation state, winding state and machine performance
Parameter, with failure diversity, comprehensive, improves power system detection efficiency.According to winding deformation principle, by detecting electricity
Power transformer normalizes the discrete winding state of method of the ratio of the short circuit current under short circuit current and system maximum operational mode
Parameter, simple to operate, data are intuitive and reliable.Discrete type machine performance parameter is obtained by the method for Chi-square Test, not by data
The limitation of overall distribution, it is applied widely.The present invention is applied to Power Transformer Condition detection and data mining treatment, possesses number
According to accurate, high working efficiency, failure broad covered area, be easy to computer disposal the features such as, have economic worth higher with it is wide
Market application foreground.
It should be understood by those skilled in the art that, embodiments herein can be provided as method, system or computer program
Product.Therefore, the application can be using the reality in terms of complete hardware embodiment, complete software embodiment or combination software and hardware
Apply the form of example.And, the application can be used and wherein include the computer of computer usable program code at one or more
The computer program implemented in usable storage medium (including but not limited to magnetic disk storage, CD-ROM, optical memory etc.) is produced
The form of product.
Embodiment described above is only the preferred embodiment lifted to absolutely prove the present invention, protection model of the invention
Enclose not limited to this.Equivalent substitute or conversion that those skilled in the art are made on the basis of the present invention, in the present invention
Protection domain within.Protection scope of the present invention is defined by claims.
Claims (10)
1. it is a kind of obtain power transformer discrete type state parameter method, it is characterised in that comprise the following steps：
The species and content of tested Detection Ssytem of Dissolved Gases in Power Transformer Oil Base are obtained by gas detection equipment；Judge every kind of solution gas
Whether the content of body is more than predetermined threshold value：If so, be designated as the first mark, if it is not, be designated as the second mark, as power transformer from
Dissipate type gaseous state parameter；
Power transformer discharge capacity statistical parameter, discharge phase statistical parameter and discharge time are obtained by Partial discharge detector
Statistical parameter, calculates the information gain maximum of these three statistical parameters and generates decision tree；Statistics ginseng is judged according to decision tree
Whether number is abnormal：If so, the first mark is designated as, if it is not, the second mark is designated as, as power transformer discrete type shelf depreciation shape
State parameter；
Short circuit under power transformer normalization short circuit current and system maximum operational mode is obtained by winding deformation detector
Whether electric current, judge to normalize the ratio of the short circuit current under short circuit current and system maximum operational mode more than 1：If so, note
It is the first mark, if it is not, the second mark is designated as, as power transformer discrete type winding state parameter；
Power transformer machine performance parameter is obtained by machine performance testing equipment, each by machine performance parameter is differed
Numerical value is set to an interval；The chi amount of each pair adjacent interval is calculated, judges whether each pair interval chi amount is more than
Card side's significance value：If so, the first mark is designated as, if it is not, the second mark is designated as, as power transformer discrete type machine performance
Parameter.
2. it is according to claim 1 obtain power transformer discrete type state parameter method, it is characterised in that it is described molten
Solution gas includes hydrogen, methane, ethane, ethene and acetylene.
3. it is according to claim 2 obtain power transformer discrete type state parameter method, it is characterised in that it is described pre-
If threshold value includes hydrogen threshold value, methane threshold value, ethane threshold value, ethene threshold value and acetylene threshold value；Judge containing for every kind of dissolved gas
Whether amount is more than the predetermined threshold value of same gas.
4. according to claim 1,2 or 3 acquisition power transformer discrete type state parameter method, it is characterised in that
The predetermined threshold value setting procedure includes：
The Power Transformer Faults case relevant with the dissolved gas is collected, fault data data acquisition system is set up；
Normal distribution analysis is carried out according to fault data data acquisition system, the content curve of every kind of dissolved gas is obtained；
By the numerical value of corresponding 95% distribution of the content curve of every kind of dissolved gas, as corresponding predetermined threshold value.
5. it is according to claim 1 obtain power transformer discrete type state parameter method, it is characterised in that it is described to put
Electric phase includes maximum discharge phase and averaged discharge phase.
6. the method for obtaining power transformer discrete type state parameter according to claim 1 or 5, it is characterised in that institute
Stating statistical parameter is included with Types Below：Degree of skewness, standout, the cross-correlation factor and symmetry in shelf depreciation collection of illustrative plates；It is described
The method for generating decision tree is as follows：The corresponding information gain maximum of each type statistical parameter is calculated respectively, and information is increased
The corresponding statistical parameter symbol of beneficial maximum and generates decision tree as the splitting condition of decision tree；
Whether exception procedure is as follows to judge statistical parameter：
From the root node of decision tree, if statistical parameter symbol is less than splitting condition, the second mark is designated as, represents normal；
If statistical parameter symbol is more than splitting condition, into the corresponding subtree root node of current node, subtree root node correspondence is carried out
Splitting condition judge, until into abnormal terminal node is represented, being designated as the first mark.
7. it is according to claim 1 obtain power transformer discrete type state parameter method, it is characterised in that by inspection
The ratio of short circuit in winding electric current and power transformer rated current is surveyed and calculated, the normalization short circuit current is obtained.
8. it is according to claim 1 obtain power transformer discrete type state parameter method, it is characterised in that the machine
Tool state parameter includes transformer temperature, transformer vibration and transformer noise；The corresponding mechanical detection of the transformer temperature
Equipment is infrared temperature imager, and it is mechanical vibration detector, transformer noise that transformer vibrates corresponding mechanical detection equipment
Corresponding mechanical detection equipment is transformer noise detector.
9. according to claim 1 or 8 acquisition power transformer discrete type state parameter method, it is characterised in that institute
State card side's significance value and take 0.05.
10. it is according to claim 1 obtain power transformer discrete type state parameter method, it is characterised in that it is described
First mark takes 1, and the second mark takes 0.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611166135.0A CN106707060B (en) | 2016-12-16 | 2016-12-16 | A method of obtaining power transformer discrete type state parameter |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611166135.0A CN106707060B (en) | 2016-12-16 | 2016-12-16 | A method of obtaining power transformer discrete type state parameter |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106707060A true CN106707060A (en) | 2017-05-24 |
| CN106707060B CN106707060B (en) | 2019-09-17 |
Family
ID=58938968
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611166135.0A Active CN106707060B (en) | 2016-12-16 | 2016-12-16 | A method of obtaining power transformer discrete type state parameter |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106707060B (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107907783A (en) * | 2017-12-19 | 2018-04-13 | 西安交通大学 | Transformer fault integrated diagnostic system and diagnostic method based on fuzzy association rules |
| CN108466903A (en) * | 2018-05-28 | 2018-08-31 | 广州广日电梯工业有限公司 | A kind of elevator load bearing beam self-regulated damping device and method |
| CN108663582A (en) * | 2017-11-30 | 2018-10-16 | 全球能源互联网研究院有限公司 | A kind of fault diagnosis method and system of transformer |
| CN109408583A (en) * | 2018-09-25 | 2019-03-01 | 平安科技（深圳）有限公司 | Data processing method and device, computer readable storage medium, electronic equipment |
| CN109709938A (en) * | 2018-12-29 | 2019-05-03 | 云南电网有限责任公司电力科学研究院 | A kind of appraisal procedure of transformer state |
| CN109855740A (en) * | 2019-01-09 | 2019-06-07 | 桂林电子科技大学 | A kind of transformer fault online test method and system |
| CN111476318A (en) * | 2020-04-30 | 2020-07-31 | 常州大学 | Transformer fault diagnosis method and system based on fuzzy decision |
| CN111830439A (en) * | 2019-04-19 | 2020-10-27 | 宁波奥克斯高科技有限公司 | Transformer fault detection method and transformer |
| CN113176522A (en) * | 2021-04-16 | 2021-07-27 | 国网江苏省电力有限公司南通供电分公司 | Transformer short-circuit fault detection method |
| CN113252827A (en) * | 2021-05-12 | 2021-08-13 | 国网安徽省电力有限公司电力科学研究院 | Transformer oil chromatographic device performance evaluation method and system based on chi-square test |
| CN113933757A (en) * | 2020-06-29 | 2022-01-14 | 株洲中车时代电气股份有限公司 | Traction transformer overcurrent diagnosis protection device and method |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101551428A (en) * | 2009-05-12 | 2009-10-07 | 东北大学 | Fault diagnosis device of transformer based on rough set theory and diagnosis method thereof |
| CN102435893A (en) * | 2011-11-04 | 2012-05-02 | 国电南京自动化股份有限公司 | Oil-immersed transformer fault diagnosis method based on self-adaptive genetic algorithm |
| CN102662113A (en) * | 2012-04-17 | 2012-09-12 | 国网电力科学研究院 | Comprehensive diagnosis method of oil-immersed transformer based on fault tree |
| CN103928906A (en) * | 2014-03-26 | 2014-07-16 | 广州白云电器设备股份有限公司 | Thermal overload protection method of electric system |
| CN103954865A (en) * | 2014-05-06 | 2014-07-30 | 国家电网公司 | Mechanical-state on-line monitoring device of transformer winding |
| CN104535865A (en) * | 2014-12-30 | 2015-04-22 | 西安工程大学 | Comprehensive diagnosing method for operation troubles of power transformer based on multiple parameters |
| CN104897784A (en) * | 2015-03-15 | 2015-09-09 | 国家电网公司 | Transformer fault diagnosis method based on coupled hidden Markov model |
| CN105093039A (en) * | 2015-09-09 | 2015-11-25 | 成都比善科技开发有限公司 | On-line monitoring system for transformer station |
| CN105823960A (en) * | 2016-03-18 | 2016-08-03 | 国家电网公司 | Method and system for comprehensively diagnosing deformation of transformer winding |
-
2016
- 2016-12-16 CN CN201611166135.0A patent/CN106707060B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101551428A (en) * | 2009-05-12 | 2009-10-07 | 东北大学 | Fault diagnosis device of transformer based on rough set theory and diagnosis method thereof |
| CN102435893A (en) * | 2011-11-04 | 2012-05-02 | 国电南京自动化股份有限公司 | Oil-immersed transformer fault diagnosis method based on self-adaptive genetic algorithm |
| CN102662113A (en) * | 2012-04-17 | 2012-09-12 | 国网电力科学研究院 | Comprehensive diagnosis method of oil-immersed transformer based on fault tree |
| CN103928906A (en) * | 2014-03-26 | 2014-07-16 | 广州白云电器设备股份有限公司 | Thermal overload protection method of electric system |
| CN103954865A (en) * | 2014-05-06 | 2014-07-30 | 国家电网公司 | Mechanical-state on-line monitoring device of transformer winding |
| CN104535865A (en) * | 2014-12-30 | 2015-04-22 | 西安工程大学 | Comprehensive diagnosing method for operation troubles of power transformer based on multiple parameters |
| CN104897784A (en) * | 2015-03-15 | 2015-09-09 | 国家电网公司 | Transformer fault diagnosis method based on coupled hidden Markov model |
| CN105093039A (en) * | 2015-09-09 | 2015-11-25 | 成都比善科技开发有限公司 | On-line monitoring system for transformer station |
| CN105823960A (en) * | 2016-03-18 | 2016-08-03 | 国家电网公司 | Method and system for comprehensively diagnosing deformation of transformer winding |
Non-Patent Citations (1)
| Title |
|---|
| 赵宇明: "《模式识别》", 31 October 2013 * |
Cited By (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108663582A (en) * | 2017-11-30 | 2018-10-16 | 全球能源互联网研究院有限公司 | A kind of fault diagnosis method and system of transformer |
| CN107907783B (en) * | 2017-12-19 | 2019-08-13 | 西安交通大学 | Transformer fault integrated diagnostic system and diagnostic method based on fuzzy association rules |
| CN107907783A (en) * | 2017-12-19 | 2018-04-13 | 西安交通大学 | Transformer fault integrated diagnostic system and diagnostic method based on fuzzy association rules |
| CN108466903A (en) * | 2018-05-28 | 2018-08-31 | 广州广日电梯工业有限公司 | A kind of elevator load bearing beam self-regulated damping device and method |
| CN108466903B (en) * | 2018-05-28 | 2024-01-09 | 广州广日电梯工业有限公司 | Self-adjusting damping device and method for elevator spandrel girder |
| CN109408583B (en) * | 2018-09-25 | 2023-04-07 | 平安科技（深圳）有限公司 | Data processing method and device, computer readable storage medium and electronic equipment |
| CN109408583A (en) * | 2018-09-25 | 2019-03-01 | 平安科技（深圳）有限公司 | Data processing method and device, computer readable storage medium, electronic equipment |
| CN109709938A (en) * | 2018-12-29 | 2019-05-03 | 云南电网有限责任公司电力科学研究院 | A kind of appraisal procedure of transformer state |
| CN109855740A (en) * | 2019-01-09 | 2019-06-07 | 桂林电子科技大学 | A kind of transformer fault online test method and system |
| CN111830439A (en) * | 2019-04-19 | 2020-10-27 | 宁波奥克斯高科技有限公司 | Transformer fault detection method and transformer |
| CN111830439B (en) * | 2019-04-19 | 2022-10-11 | 宁波奥克斯高科技有限公司 | Transformer fault detection method and transformer |
| CN111476318A (en) * | 2020-04-30 | 2020-07-31 | 常州大学 | Transformer fault diagnosis method and system based on fuzzy decision |
| CN113933757A (en) * | 2020-06-29 | 2022-01-14 | 株洲中车时代电气股份有限公司 | Traction transformer overcurrent diagnosis protection device and method |
| CN113933757B (en) * | 2020-06-29 | 2022-09-16 | 株洲中车时代电气股份有限公司 | Traction transformer overcurrent diagnosis protection device and method |
| CN113176522A (en) * | 2021-04-16 | 2021-07-27 | 国网江苏省电力有限公司南通供电分公司 | Transformer short-circuit fault detection method |
| CN113252827A (en) * | 2021-05-12 | 2021-08-13 | 国网安徽省电力有限公司电力科学研究院 | Transformer oil chromatographic device performance evaluation method and system based on chi-square test |
| CN113252827B (en) * | 2021-05-12 | 2022-07-22 | 国网安徽省电力有限公司电力科学研究院 | Method and system for evaluating performance of transformer oil chromatographic device based on chi-square test |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106707060B (en) | 2019-09-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106707060A (en) | Method for acquiring discrete state parameters of power transformer | |
| Gargoom et al. | Automatic classification and characterization of power quality events | |
| Baloch et al. | An intelligent data mining-based fault detection and classification strategy for microgrid | |
| Biswal et al. | Measurement and classification of simultaneous power signal patterns with an S-transform variant and fuzzy decision tree | |
| CN109830972B (en) | New energy station oscillation source rapid identification system and method | |
| CN108053095B (en) | Power quality disturbance event feature extraction method and system | |
| CN105139295A (en) | Data mining method of mass information of on-line monitoring on power equipment | |
| Koochi et al. | A synchrophasor-based decision tree approach for identification of most coherent generating units | |
| CN111092442A (en) | Hydroelectric generating set multi-dimensional vibration region fine division method based on decision tree model | |
| CN112834224A (en) | Method and system for evaluating health state of nuclear power steam turbine generator | |
| CN117332322A (en) | Boundary region importance sampling method | |
| Arvani et al. | Detection and protection against intrusions on smart grid systems | |
| Yatsugi et al. | Common diagnosis approach to three-class induction motor faults using stator current feature and support vector machine | |
| CN109901003B (en) | Inverter power fault detection method and system | |
| Streubel et al. | Detection and monitoring of supraharmonic anomalies of an electric vehicle charging station | |
| CN110119858A (en) | The Data Quality Assessment Methodology of automation system for the power network dispatching based on big data | |
| Dhumale et al. | Diagnosis of multiple open switch faults in three phase voltage source inverter | |
| Luo et al. | Short-term voltage stability assessment based on local autopattern discovery | |
| Li et al. | Improvement of Electromechanical Mode Identification from Ambient Data with Stochastic Subspace Method | |
| Arvani et al. | Cyber security of smart grid systems using intrusion detection methods | |
| CN113890018B (en) | Power distribution network weak point identification method based on data association analysis | |
| Balamurugan et al. | Bearing/incipient/open phase fault detection and diagnosis of multi-phase induction motor drives equipped by GBDTI2HO technique | |
| CN112068028B (en) | Intermittent single-phase earth fault identification method | |
| CN108876138A (en) | A kind of supervisory control of substation information leakage prison detection method based on big data analysis | |
| Dwiputranto et al. | DGA-Based Early Transformer Fault Detection using Rough Set Theory Classifier |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |