CN112347695A - Method and system for predicting top-layer oil temperature of transformer - Google Patents
Method and system for predicting top-layer oil temperature of transformer Download PDFInfo
- Publication number
- CN112347695A CN112347695A CN202011178764.1A CN202011178764A CN112347695A CN 112347695 A CN112347695 A CN 112347695A CN 202011178764 A CN202011178764 A CN 202011178764A CN 112347695 A CN112347695 A CN 112347695A
- Authority
- CN
- China
- Prior art keywords
- transformer
- time
- predicted
- oil temperature
- moment
- 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
- 238000000034 method Methods 0.000 title claims abstract description 46
- 238000012706 support-vector machine Methods 0.000 claims abstract description 60
- 238000011160 research Methods 0.000 claims abstract description 29
- 238000012795 verification Methods 0.000 claims description 40
- 238000012549 training Methods 0.000 claims description 29
- 238000004364 calculation method Methods 0.000 claims description 15
- 238000004422 calculation algorithm Methods 0.000 claims description 10
- 238000011156 evaluation Methods 0.000 claims description 9
- 238000012545 processing Methods 0.000 claims description 7
- NAWXUBYGYWOOIX-SFHVURJKSA-N (2s)-2-[[4-[2-(2,4-diaminoquinazolin-6-yl)ethyl]benzoyl]amino]-4-methylidenepentanedioic acid Chemical compound C1=CC2=NC(N)=NC(N)=C2C=C1CCC1=CC=C(C(=O)N[C@@H](CC(=C)C(O)=O)C(O)=O)C=C1 NAWXUBYGYWOOIX-SFHVURJKSA-N 0.000 claims description 3
- 230000002159 abnormal effect Effects 0.000 abstract description 2
- 230000006870 function Effects 0.000 description 18
- 238000010586 diagram Methods 0.000 description 8
- 238000004590 computer program Methods 0.000 description 7
- 238000005286 illumination Methods 0.000 description 3
- 238000003860 storage Methods 0.000 description 3
- 238000013473 artificial intelligence Methods 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 238000004804 winding Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 230000017525 heat dissipation Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
- G06N20/10—Machine learning using kernel methods, e.g. support vector machines [SVM]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/10—Numerical modelling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/08—Thermal analysis or thermal optimisation
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Evolutionary Computation (AREA)
- Software Systems (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- General Engineering & Computer Science (AREA)
- Geometry (AREA)
- Computer Hardware Design (AREA)
- Data Mining & Analysis (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Housings And Mounting Of Transformers (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention discloses a method and a system for predicting top layer oil temperature of a transformer, wherein the method comprises the following steps: calculating the comprehensive correlation degree of each moment and the moment to be predicted in the research sample to obtain a similar moment set with the maximum comprehensive correlation degree with the moment to be predicted; determining optimal parameters of a support vector machine based on a research sample; and predicting the top oil temperature of the transformer according to the size of the similar time set and the optimal parameters of the support vector machine. The method effectively improves the prediction accuracy of the top oil temperature of the transformer, can find potential hidden dangers in the transformer in time, ensures the service life of the transformer, and can provide a basis for judging abnormal sounds of the transformer by predicting the top oil temperature of the transformer.
Description
Technical Field
The invention relates to a method and a system for predicting top-layer oil temperature of a transformer, and belongs to the technical field of transformers.
Background
As a core device of a power grid, the power transformer has a wide application prospect in the power grid, and the working state of the power transformer influences the safe and stable operation of the power grid. Theoretically, the winding hot spot temperature of the power transformer is an important parameter for measuring the internal thermal state, but in view of the defects of difficult measurement, high cost and the like of the winding hot spot temperature, the thermal state of the power transformer is generally evaluated by adopting the top oil temperature at present. If the top-layer oil temperature of the transformer can be predicted, the thermal state of the transformer can be estimated in advance, potential thermal faults inside the transformer can be found in time, the service life of the transformer is guaranteed, meanwhile, the power failure time is reduced, the power supply reliability of the whole society is guaranteed, and therefore the top-layer oil temperature prediction of the transformer has practical significance.
At present, a great deal of research is carried out at home and abroad aiming at the calculation and prediction of the hot spot temperature and the oil temperature of the transformer, and the research is mainly divided into three types including an empirical formula, a hot circuit model and numerical calculation in summary, wherein the empirical formula is used for roughly estimating the hot spot temperature by utilizing the existing calculation formula and manual experience, and the general error is large; the thermal circuit model is a thermoelectric temperature calculation model established based on a thermal principle and a thermoelectric analogy method, and the calculation accuracy of the method is greatly influenced by model parameters; the numerical calculation is to establish the internal temperature distribution of multi-physical fusion calculation through the research of the internal structure and the heat dissipation medium of the transformer so as to estimate the hot spot temperature, and the calculation result of the method basically conforms to the real thermal state of the transformer along with the improvement of the numerical simulation level, but the method is complex. In addition, with the continuous deepening of the artificial intelligence algorithm in the field of the transformer, some researchers apply the artificial intelligence algorithm to the prediction of the hot spot temperature and the oil temperature of the transformer step by step, but the algorithm is difficult to realize. In order to solve the above problems, it is necessary to develop a method for predicting the top-layer oil temperature of the transformer, which has a simple algorithm and high prediction accuracy.
Disclosure of Invention
The invention aims to overcome the defects in the prior art and provides a method and a system for predicting the top oil temperature of a transformer, which can effectively improve the prediction precision of the top oil temperature of the transformer. In order to achieve the purpose, the invention is realized by adopting the following technical scheme:
in a first aspect, the present invention provides a method for predicting a top-layer oil temperature of a transformer, where the method includes the following steps:
calculating the comprehensive correlation degree of each moment and the moment to be predicted in the research sample to obtain a similar moment set with the maximum comprehensive correlation degree with the moment to be predicted;
determining optimal parameters of a support vector machine based on a research sample;
and predicting the top oil temperature of the transformer according to the size of the similar time set and the optimal parameters of the support vector machine.
With reference to the first aspect, further, the calculating a comprehensive correlation between each time in the research sample and the time to be predicted includes the following steps:
calculating direct correlation coefficients of various influencing factors of the top oil temperature of the transformer;
calculating the weather correlation, the time correlation and the load correlation of each moment and the moment to be predicted in the research sample according to the calculated direct correlation coefficient;
and carrying out linear weighting based on the meteorological correlation, the time correlation and the load correlation to obtain the comprehensive correlation between each moment and the moment to be predicted in the research sample.
With reference to the first aspect, further, the influencing factors of the oil temperature at the top layer of the transformer include three factors of weather, time and load, the weather factors include temperature, humidity, wind speed, air pressure, rainfall and illumination intensity, and the time factors include horizontal date and vertical time.
With reference to the first aspect, further, the values of the influencing factors are normalized to obtain values convenient for subsequent calculation:
in the formula (1), XgijRepresenting the g-th influencing factor, C, at the j-th time on the ith day before the time to be predictedxThe normalized coefficient is expressed, and the value range is 0.01-0.45.
With reference to the first aspect, further, a direct correlation coefficient of each influence factor of the top oil temperature of the transformer is calculated, where the direct correlation coefficient is:
in the formula (2), RgRepresenting the direct correlation coefficient, RH, of the g-th influencing factor with the oil temperature at the top layer of the transformergThe transverse direct correlation coefficient of the g influencing factor and the top layer oil temperature of the transformer is shown as follows:
in the formula (3), bhgRepresents the lateral partial regression coefficient, M represents the number of samples, XgijDenotes the g-th influencing factor, Y, at the j-th time on the ith day before the time to be predictedijRepresenting the top-layer oil temperature of the transformer at the jth moment of the ith day before the moment to be predicted;
in the formula (2), RZgThe longitudinal direct correlation coefficient of the g influencing factor and the top layer oil temperature of the transformer is shown as follows:
in the formula (4), bzgRepresenting longitudinal partial regression coefficients;
in the formula (2), R7Direct correlation coefficient, RH, representing the lateral date and the top oil temperature of the transformer7A transverse direct correlation coefficient representing a transverse date and the top layer oil temperature of the transformer; r8Express verticalDirect correlation coefficient, RH, between time and top layer oil temperature of transformer8And the longitudinal direct correlation coefficient of the longitudinal time and the oil temperature of the top layer of the transformer is represented.
With reference to the first aspect, further, the comprehensive correlation degree is:
in the formula (5), DlijRepresenting the comprehensive degree of correlation, R, between the time l to be predicted and the time j on the ith day before the time to be predictedgRepresenting the direct correlation coefficient of the g influencing factor and the oil temperature of the top layer of the transformer, R9Representing the direct correlation coefficient of the load factor with the top oil temperature of the transformer, AlijRepresenting the weather correlation degree of the moment l to be predicted and the ith and jth moment before the moment to be predicted:
in formula (6), MDlijRepresenting Euclidean distance between a time l to be predicted and a j-th day before the time to be predicted:
in formula (7), Xg0lIndicating the g-th influencing factor, Xg, corresponding to the moment l to be predictedijRepresenting the g type influence factor of the j time on the ith day before the time to be predicted;
in formula (6), MDmaxMaximum euclidean distance representing meteorological factors:
in the formula (8), CxExpressing the normalized coefficient, the value range is 0.01-0.45, and MDTgIn the Euclidean distance corresponding to the g-th influence factorIntermediate quantity:
in the formula (5), BlijRepresenting the time correlation between the time l to be predicted and the ith and jth time before the time to be predicted:
in the formula (10), R7Direct correlation coefficient, R, representing transverse date and top layer oil temperature of transformer8Representing the direct correlation coefficient of the longitudinal time with the oil temperature at the top of the transformer, X70lRepresents the horizontal date corresponding to the time l to be predicted,
X7ilindicating the transverse date, MDT, of the ith and jth time before the time to be predicted7Intermediate Euclidean distance, X8, representing the correspondence of the horizontal date0lIndicating the longitudinal time corresponding to the time l to be predicted, X8ilIndicating the longitudinal time, MDT, of the ith and jth time of day before the time to be predicted8Representing the intermediate Euclidean distance corresponding to the longitudinal time;
in the formula (5), ClijRepresenting the load correlation degree of the moment l to be predicted and the ith and jth moment before the moment to be predicted:
in formula (11), X90lIndicating the load factor corresponding to the moment l to be predicted, X9ijRepresenting the load factor, MDT, at the ith and jth time before the time to be predicted9And representing the intermediate Euclidean distance corresponding to the load factors.
With reference to the first aspect, further, the obtaining a similar time set with a maximum integrated correlation with a time to be predicted from a research sample includes the following steps:
setting a comprehensive correlation degree threshold value D0；
Judging comprehensive correlation degree threshold value D0Degree of correlation withlijThe numerical value of (A):
when D is presentlij≥D0Then, the ith time before the time to be predicted is the potential similar time of the time to be predicted; when D is presentlij<D0Then, the ith time before the time to be predicted is not the potential similar time of the time to be predicted;
counting the number E of the potential similar moments;
selecting a similar time set of the time to be predicted:
when E is more than or equal to 20, selecting the comprehensive degree of correlation DlijThe largest 20 potential similar moments are used as a similar moment set of the moments l to be predicted; when E is<20 and E is more than or equal to 10, selecting the comprehensive correlation degree DlijThe maximum 10 potential similar moments are used as a similar moment set of the moment l to be predicted; e<And 10, selecting all the potential similar moments as a similar moment set of the moment l to be predicted.
With reference to the first aspect, further, the method further includes determining main influencing factors of the top-layer oil temperature of the transformer, including:
selecting the minimum value R of direct correlation coefficients of all influencing factors and the top oil temperature of the transformer0；
Judging direct correlation coefficient R of each influencing factorgAnd R0The size of (2):
in the formula (12), RgRepresenting the direct correlation coefficient of the g influencing factor and the oil temperature of the top layer of the transformer, R0And (4) showing.
With reference to the first aspect, further, the determining the optimal parameters of the support vector machine according to the similar time instant set includes the following steps:
determining main influence factors of the top oil temperature of the transformer;
dividing a research sample into a training sample and a verification sample;
calculating the similar time of each time of the verification sample from the training sample to obtain a verification sample similar time set;
the method comprises the steps that main influence factors of all moments in a verification sample similar moment set are used as training input variables of a support vector machine, and transformer top layer oil temperatures corresponding to all moments in the verification sample similar moment set are used as training output variables of the support vector machine;
initializing and setting relevant parameters of the support vector machine, including a kernel function ker, kernel parameters p1 and p2, a parameter C and a Loss function Loss;
leading the constructed input training vector and the constructed output training vector into a support vector machine, and training the support vector machine to obtain a trained support vector machine;
the main influence factors at each moment in the verification sample are used as verification input variables and input into a trained support vector machine to obtain a verification value of the top-layer oil temperature of the transformer at each moment of the verification sample;
establishing a parameter evaluation function based on the sum of squared errors:
in formula (13), Y1jRepresenting the verification value at time j obtained by the support vector machine, Y0jRepresents the actual value at time j;
when the parameter evaluation function obtains the minimum value, the related parameters of the support vector machine are the optimal parameters of the support vector machine, otherwise, the kernel function ker, the kernel parameters p1 and p2, the parameter C and the Loss function Loss of the support vector machine are updated, and the support vector machine is retrained until the parameter evaluation function obtains the minimum value;
and outputting the optimal parameters of the support vector machine to obtain a prediction support vector machine which can be used for predicting the top oil temperature of the transformer.
Specifically, the research samples are divided into training samples and verification samples according to the time sequence, wherein the training samples are arranged in the front of the time sequence, and the verification samples are arranged in the back of the time sequence.
Specifically, the training samples comprise main influencing factors of the similar time sets of the verification samples in the training samples at corresponding moments and corresponding transformer top oil temperatures, and the verification samples comprise the main influencing factors of the verification samples at the moments and the corresponding transformer top oil temperatures.
With reference to the first aspect, further, the predicting the transformer top layer oil temperature includes:
the method for predicting the top-layer oil temperature of the transformer is selected according to the number of similar moments in a similar moment set:
when the number of the similar moments in the similar moment set is less than or equal to 10, processing the top-layer oil temperature of the transformer corresponding to each moment in the similar moment set by adopting a linear weighting algorithm to obtain a predicted value of the top-layer oil temperature of the transformer;
and when the number of similar moments in the similar moment set is 20, predicting the top oil temperature of the transformer by using a prediction support vector machine to obtain a predicted value of the top oil temperature of the transformer.
With reference to the first aspect, further, the linear weighting algorithm is:
configuring different weight coefficients p for the similar moments according to the comprehensive correlation degree of the similar moments in the similar moment set and the moments to be predictedxAnd 0 < px＜1；
Calculating a predicted value of the top oil temperature of the transformer:
in the formula (14), Y represents a predicted value of the top-layer oil temperature of the transformer; y0xAnd representing the actual value of the top-layer oil temperature of the transformer at the x-th similar moment.
In a second aspect, the present invention provides a system for predicting top-layer oil temperature of a transformer, including:
a first calculation module: the method comprises the steps of calculating the comprehensive correlation degree of each moment and the moment to be predicted in a research sample, and obtaining a similar moment set with the maximum comprehensive correlation degree with the moment to be predicted;
an optimal parameter determination module: the optimal parameters for determining the support vector machine;
the transformer top oil temperature prediction module: and predicting the top oil temperature of the transformer according to the size of the similar time set and the optimal parameters of the support vector machine.
Compared with the prior art, the method and the system for predicting the top-layer oil temperature of the transformer provided by the embodiment of the invention have the following beneficial effects:
according to the method, the comprehensive correlation degree of each moment and the moment to be predicted in the research sample is calculated, the similar moment set with the maximum comprehensive correlation degree with the moment to be predicted is obtained, the effectiveness of similar moment selection can be guaranteed, and the precision of top-layer oil temperature prediction of the transformer can be improved;
according to the invention, the direct correlation coefficient of each influence factor of the top oil temperature of the transformer is calculated from the transverse direct correlation coefficient and the longitudinal direct correlation coefficient, so that the interaction among the influence factors of the top oil temperature of the transformer is eliminated;
the method can accurately predict the top oil temperature of the transformer, can find potential hidden dangers in the transformer in time, ensures the service life of the transformer, and can provide a basis for judging abnormal sounds of the transformer by predicting the top oil temperature of the transformer.
Drawings
Fig. 1 is a flowchart of a method for predicting top-layer oil temperature of a transformer according to the present invention.
Detailed Description
The invention is further described below with reference to the accompanying drawings. The following examples are only for illustrating the technical solutions of the present invention more clearly, and the protection scope of the present invention is not limited thereby.
The first embodiment is as follows:
as shown in fig. 1, an embodiment of the present invention provides a method for predicting a top-layer oil temperature of a transformer, where the method includes the following steps:
step 1: and calculating direct correlation coefficients of the influencing factors and the top oil temperature of the transformer.
Step 1.1: selecting the influence factors of the top oil temperature of the transformer:
the influencing factors of the oil temperature at the top layer of the transformer comprise gasThe weather factors comprise temperature, humidity, wind speed, air pressure, rainfall and illumination intensity, and the time factors comprise horizontal date and vertical time. Specifically, X1 is definedijTemperature at time j on day i before time to be predicted, X2ijHumidity at the ith and jth day before the moment to be predicted, X3ijThe ith and jth wind speed before the moment to be predicted X4ijPressure at time j on day i before the time to be predicted, X5ijThe rainfall at the ith and jth day before the moment to be predicted is X6ijIllumination intensity at the ith and jth time of day before the time to be predicted, X7ijFor the ith and jth transverse dates, X8, before the time to be predictedijThe ith and jth longitudinal time before the time to be predicted, X9ijIs the ith day, the jth day, Y before the moment to be predictedijThe top-layer oil temperature of the transformer at the ith time and the jth time before the moment to be predicted.
Step 1.2: normalizing the numerical values of the influence factors to obtain numerical values convenient for subsequent calculation:
in the formula (1), XgijRepresenting the g-th influencing factor, C, at the j-th time on the ith day before the time to be predictedxThe normalized coefficient is expressed, and the value range is 0.01-0.45.
Step 1.3: calculating direct correlation coefficients of various influencing factors of the top oil temperature of the transformer:
the direct correlation coefficient is:
in the formula (2), RgRepresenting the direct correlation coefficient, RH, of the g-th influencing factor with the oil temperature at the top layer of the transformergThe transverse direct correlation coefficient of the g influencing factor and the top layer oil temperature of the transformer is shown as follows:
in the formula (3), bhgRepresents the lateral partial regression coefficient, M represents the number of samples, XgijDenotes the g-th influencing factor, Y, at the j-th time on the ith day before the time to be predictedijRepresenting the top-layer oil temperature of the transformer at the jth moment of the ith day before the moment to be predicted;
in the formula (2), RZgThe longitudinal direct correlation coefficient of the g influencing factor and the top layer oil temperature of the transformer is shown as follows:
in the formula (4), bzgRepresenting longitudinal partial regression coefficients;
in the formula (2), R7Direct correlation coefficient, RH, representing the lateral date and the top oil temperature of the transformer7A transverse direct correlation coefficient representing a transverse date and the top layer oil temperature of the transformer; r8Representing the direct correlation coefficient, RH, of the longitudinal time with the oil temperature at the top level of the transformer8And the longitudinal direct correlation coefficient of the longitudinal time and the oil temperature of the top layer of the transformer is represented.
Step 2: and calculating the comprehensive correlation degree of each moment in the research sample and the moment to be predicted.
Step 2.1: calculating the weather correlation A between each moment and the moment to be predictedlij。
Because 6 meteorological factors are normalized and direct correlation coefficients of different meteorological factors are different, an Euclidean distance method based on direct correlation coefficient weighting optimization is adopted when the meteorological correlation degree is researched.
Calculating the Euclidean distance MD of the meteorological factors at the ith and jth moments before the moment to be predictedlij：
In the formula (5), Xg0lTo representThe g-th influencing factor Xg corresponding to the moment l to be predictedijRepresenting the g-th influencing factor, R, at the j-th time on the ith day before the time to be predictedgRepresenting the direct correlation coefficient of the g influencing factor and the top layer oil temperature of the transformer;
maximum Euclidean distance MD for defining meteorological factorsmax：
In formula (6), MDTgThe intermediate Euclidean distance corresponding to the g-th influence factor is represented as follows:
in the formula (7), CxThe normalized coefficient is expressed, and the value range is 0.01-0.45.
Minimum Euclidean distance MD considering meteorological factorsminIf the current time is 0, the meteorological correlation A between the time l to be predicted and the ith and jth time before the time to be predicted can be obtained according to the Euclidean distancelij：
Step 2.2: calculating the time correlation B between each moment and the moment to be predictedlij。
According to the calculation principle of the meteorological correlation degree, the time correlation degree B of the time l to be predicted and the ith and jth time before the time l to be predicted can be obtained in the same waylij：
In the formula (9), R7Direct correlation coefficient, R, representing transverse date and top layer oil temperature of transformer8Representing the direct correlation coefficient of the longitudinal time with the oil temperature at the top of the transformer, X70lIndicating to be predictedTransverse date, X7, corresponding to time lilIndicating the transverse date, MDT, of the ith and jth time before the time to be predicted7Intermediate Euclidean distance, X8, representing the correspondence of the horizontal date0lIndicating the longitudinal time corresponding to the time l to be predicted, X8ilIndicating the longitudinal time, MDT, of the ith and jth time of day before the time to be predicted8The intermediate euclidean distance corresponding to the longitudinal time is shown.
Step 2.3: calculating the comprehensive load correlation C between each moment and the moment to be predictedlij。
According to the direct correlation coefficient of the load factor and the top oil temperature of the transformer obtained by the previous calculation, the load correlation C between the moment l to be predicted and the ith and jth moment before the moment to be predicted can be obtainedlij：
In formula (10), X90lIndicating the load factor corresponding to the moment l to be predicted, X9ijRepresenting the load factor, MDT, at the ith and jth time before the time to be predicted9And representing the intermediate Euclidean distance corresponding to the load factors.
Step 2.4: calculating the comprehensive correlation degree D of each time and the time to be predictedlij。
Based on the direct correlation coefficient, weighting the weather correlation, the time correlation and the load correlation to obtain the comprehensive correlation D between the time l to be predicted and the ith and jth time before the time to be predictedlij：
In the formula (11), R9And the direct line pipe coefficient represents the load factor and the top oil temperature of the transformer.
And step 3: and acquiring a similar time set with the maximum comprehensive correlation degree with the time to be predicted from the research sample.
Step 3.1: setting a comprehensive correlation degree threshold value D0Judgment healdSum of correlation limits D0Degree of correlation withlijThe numerical value of (A):
when D is presentlij≥D0Then, the ith time before the time to be predicted is the potential similar time of the time to be predicted; when D is presentlij<D0And the j time on the ith day before the time to be predicted is not the potential similar time of the time l to be predicted.
Step 3.2: and counting the number E of the potential similar moments.
Step 3.3: selecting a similar time set of the time to be predicted:
when E is more than or equal to 20, selecting the comprehensive degree of correlation DlijThe largest 20 potential similar moments are used as a similar moment set of the moments l to be predicted; when E is<20 and E is more than or equal to 10, selecting the comprehensive correlation degree DlijThe maximum 10 potential similar moments are used as a similar moment set of the moment l to be predicted; e<And 10, selecting all the potential similar moments as a similar moment set of the moment l to be predicted.
And 4, step 4: determining main influence factors of the top oil temperature of the transformer:
selecting the minimum value R of direct correlation coefficients of all influencing factors and the top oil temperature of the transformer0；
Judging direct correlation coefficient R of each influencing factorgAnd R0The size of (2):
in the formula (12), RgRepresenting the direct correlation coefficient of the g influencing factor and the oil temperature of the top layer of the transformer, R0And (4) showing.
And 5: the method for determining the optimal parameters of the support vector machine according to the similar time set comprises the following steps:
step 5.1: and determining main influence factors of the top oil temperature of the transformer.
Step 5.2: the research samples are divided into training samples and verification samples.
Step 5.3: calculating the similar time of each time of the verification sample from the training sample to obtain a verification sample similar time set;
step 5.4: the method comprises the steps that main influence factors of all moments in a verification sample similar moment set are used as training input variables of a support vector machine, and transformer top layer oil temperatures corresponding to all moments in the verification sample similar moment set are used as training output variables of the support vector machine;
step 5.5: initializing and setting relevant parameters of the support vector machine, including a kernel function ker, kernel parameters p1 and p2, a parameter C and a Loss function Loss;
step 5.6: leading the constructed input training vector and the constructed output training vector into a support vector machine, and training the support vector machine to obtain a trained support vector machine;
step 5.7: the main influence factors at each moment in the verification sample are used as verification input variables and input into a trained support vector machine to obtain a verification value of the top-layer oil temperature of the transformer at each moment of the verification sample;
step 5.8: establishing a parameter evaluation function based on the sum of squared errors:
in formula (13), Y1jRepresenting the verification value at time j obtained by the support vector machine, Y0jRepresenting the actual value at time j.
Step 5.9: when the parameter evaluation function obtains the minimum value, the related parameters of the support vector machine are the optimal parameters of the support vector machine, otherwise, the kernel function ker, the kernel parameters p1 and p2, the parameter C and the Loss function Loss of the support vector machine are updated, and the support vector machine is retrained until the parameter evaluation function obtains the minimum value; and outputting the optimal parameters of the support vector machine to obtain a prediction support vector machine which can be used for predicting the top oil temperature of the transformer.
It should be noted that the research samples are divided into training samples and verification samples according to the time sequence, the training samples are in the first time sequence, and the verification samples are in the last time sequence. The training sample comprises main influencing factors of the similar time sets of the verification samples in the training samples at corresponding moments and corresponding transformer top oil temperature, and the verification samples comprise the main influencing factors of the verification samples at the moments and the corresponding transformer top oil temperature.
Step 6: and predicting the top oil temperature of the transformer.
The method for predicting the top-layer oil temperature of the transformer is selected according to the number of similar moments in a similar moment set:
when the number of the similar moments in the similar moment set is less than or equal to 10, processing the top-layer oil temperature of the transformer corresponding to each moment in the similar moment set by adopting a linear weighting algorithm to obtain a predicted value of the top-layer oil temperature of the transformer;
and when the number of similar moments in the similar moment set is 20, predicting the top oil temperature of the transformer by using a prediction support vector machine to obtain a predicted value of the top oil temperature of the transformer.
Specifically, a linear weighting algorithm is adopted to process the top-layer oil temperature of the transformer corresponding to each time in the similar time set, and a predicted value of the top-layer oil temperature of the transformer is obtained, and the method comprises the following steps:
configuring different weight coefficients p for the similar moments according to the comprehensive correlation degree of the similar moments in the similar moment set and the moments to be predictedxAnd 0 < px＜1；
Calculating a predicted value of the top oil temperature of the transformer:
in the formula (14), Y represents a predicted value of the top-layer oil temperature of the transformer; y0xAnd representing the actual value of the top-layer oil temperature of the transformer at the x-th similar moment.
Example two:
the embodiment of the invention provides a system for predicting top layer oil temperature of a transformer, which comprises:
a first calculation module: the method comprises the steps of calculating the comprehensive correlation degree of each moment and the moment to be predicted in a research sample, and obtaining a similar moment set with the maximum comprehensive correlation degree with the moment to be predicted;
an optimal parameter determination module: the optimal parameters for determining the support vector machine;
the transformer top oil temperature prediction module: and predicting the top oil temperature of the transformer according to the size of the similar time set and the optimal parameters of the support vector machine.
As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
The above description is only a preferred embodiment of the present invention, and it should be noted that, for those skilled in the art, several modifications and variations can be made without departing from the technical principle of the present invention, and these modifications and variations should also be regarded as the protection scope of the present invention.
Claims (10)
1. A method for predicting top layer oil temperature of a transformer is characterized by comprising the following steps:
calculating the comprehensive correlation degree of each moment and the moment to be predicted in the research sample to obtain a similar moment set with the maximum comprehensive correlation degree with the moment to be predicted;
determining optimal parameters of a support vector machine based on a research sample;
and predicting the top oil temperature of the transformer according to the size of the similar time set and the optimal parameters of the support vector machine.
2. The method for predicting the top-layer oil temperature of the transformer according to claim 1, wherein the step of calculating the comprehensive correlation between each time in the research sample and the time to be predicted comprises the following steps:
calculating direct correlation coefficients of various influencing factors of the top oil temperature of the transformer;
calculating the weather correlation, the time correlation and the load correlation of each moment and the moment to be predicted in the research sample according to the calculated direct correlation coefficient;
and carrying out linear weighting based on the meteorological correlation, the time correlation and the load correlation to obtain the comprehensive correlation between each moment and the moment to be predicted in the research sample.
3. The method for predicting the top-layer oil temperature of the transformer according to claim 2, wherein direct correlation coefficients of various influencing factors of the top-layer oil temperature of the transformer are calculated, and the direct correlation coefficients are as follows:
in the formula (1), RgRepresenting the direct correlation coefficient, RH, of the g-th influencing factor with the oil temperature at the top layer of the transformergThe transverse direct correlation coefficient of the g influencing factor and the top layer oil temperature of the transformer is shown as follows:
in the formula (2), bhgRepresents the lateral partial regression coefficient, M represents the number of samples, XgijDenotes the g-th influencing factor, Y, at the j-th time on the ith day before the time to be predictedijRepresenting the top-layer oil temperature of the transformer at the jth moment of the ith day before the moment to be predicted;
in the formula (1), RZgThe longitudinal direct correlation coefficient of the g influencing factor and the top layer oil temperature of the transformer is shown as follows:
in the formula (3), bzgRepresenting longitudinal partial regression coefficients;
in the formula (1), R7Direct correlation coefficient, RH, representing the lateral date and the top oil temperature of the transformer7A transverse direct correlation coefficient representing a transverse date and the top layer oil temperature of the transformer; r8Representing the direct correlation coefficient, RH, of the longitudinal time with the oil temperature at the top level of the transformer8And the longitudinal direct correlation coefficient of the longitudinal time and the oil temperature of the top layer of the transformer is represented.
4. The method for predicting the top-layer oil temperature of the transformer according to claim 2, wherein the comprehensive correlation is as follows:
in the formula (4), DlijRepresenting the comprehensive degree of correlation, R, between the time l to be predicted and the time j on the ith day before the time to be predictedgRepresenting the direct correlation coefficient of the g influencing factor and the oil temperature of the top layer of the transformer, R9Representing the direct correlation coefficient of the load factor with the top oil temperature of the transformer, AlijRepresenting the weather correlation degree of the moment l to be predicted and the ith and jth moment before the moment to be predicted:
in formula (5), MDlijRepresenting Euclidean distance between a time l to be predicted and a j-th day before the time to be predicted:
in formula (6), Xg0lIndicating the g-th influencing factor, Xg, corresponding to the moment l to be predictedijRepresenting the g type influence factor of the j time on the ith day before the time to be predicted;
in formula (5), MDmaxMaximum euclidean distance representing meteorological factors:
in the formula (7), CxExpressing the normalized coefficient, the value range is 0.01-0.45, MDTgThe intermediate Euclidean distance corresponding to the g-th influence factor is represented as follows:
in the formula (4), BlijRepresenting the time correlation between the time l to be predicted and the ith and jth time before the time to be predicted:
in the formula (9), R7Direct correlation coefficient, R, representing transverse date and top layer oil temperature of transformer8Representing the direct correlation coefficient of the longitudinal time with the oil temperature at the top of the transformer, X70lIndicating the horizontal date corresponding to the moment l to be predicted, X7ilIndicating the transverse date, MDT, of the ith and jth time before the time to be predicted7Intermediate Euclidean distance, X8, representing the correspondence of the horizontal date0lIndicating the longitudinal time corresponding to the time l to be predicted, X8ilIndicating the longitudinal time, MDT, of the ith and jth time of day before the time to be predicted8Representing the intermediate Euclidean distance corresponding to the longitudinal time;
in the formula (4), ClijRepresenting the load correlation degree of the moment l to be predicted and the ith and jth moment before the moment to be predicted:
in formula (10), X90lIndicating the load factor corresponding to the moment l to be predicted, X9ilRepresenting the load factor, MDT, at the ith and jth time before the time to be predicted9And representing the intermediate Euclidean distance corresponding to the load factors.
5. The method for predicting the top-layer oil temperature of the transformer according to claim 4, wherein the step of obtaining a similar time set with the maximum comprehensive correlation with the time to be predicted from a research sample comprises the following steps:
setting a comprehensive correlation degree threshold value D0；
Judging comprehensive correlation degree threshold value D0Degree of correlation withlijThe numerical value of (A):
when D is presentlij≥D0Then, the ith time before the time to be predicted is the potential similar time of the time to be predicted; when D is presentlij<D0Then, the ith time before the time to be predicted is not the potential similar time of the time to be predicted;
counting the number E of the potential similar moments;
selecting a similar time set of the time to be predicted:
when E is more than or equal to 20, selecting the comprehensive degree of correlation DlijThe largest 20 potential similar moments are used as a similar moment set of the moments l to be predicted; when E is<20 and E is more than or equal to 10, selecting the comprehensive correlation degree DlijThe maximum 10 potential similar moments are used as a similar moment set of the moment l to be predicted; e<And 10, selecting all the potential similar moments as a similar moment set of the moment l to be predicted.
6. The method of claim 1, further comprising determining a primary factor affecting the top-level oil temperature of the transformer, comprising:
selecting the minimum value R of direct correlation coefficients of all influencing factors and the top oil temperature of the transformer0；
Judging direct correlation coefficient R of each influencing factorgAnd R0The size of (2):
in the formula (11), RgRepresenting the direct correlation coefficient of the g influencing factor and the oil temperature of the top layer of the transformer, R0And (4) showing.
7. The method for predicting the top-layer oil temperature of the transformer according to claim 6, wherein the step of determining the optimal parameters of the support vector machine based on the research samples comprises the following steps:
determining main influence factors of the top oil temperature of the transformer;
separating a training sample and a verification sample from a research sample;
calculating the similar time of each time of the verification sample from the training sample to obtain a verification sample similar time set;
the method comprises the steps that main influence factors of all moments in a verification sample similar moment set are used as training input variables of a support vector machine, and transformer top layer oil temperatures corresponding to all moments in the verification sample similar moment set are used as training output variables of the support vector machine;
initializing and setting relevant parameters of the support vector machine, including a kernel function ker, kernel parameters p1 and p2, a parameter C and a Loss function Loss;
leading the constructed input training vector and the constructed output training vector into a support vector machine, and training the support vector machine to obtain a trained support vector machine;
the main influence factors at each moment in the verification sample are used as verification input variables and input into a trained support vector machine to obtain a verification value of the top-layer oil temperature of the transformer at each moment of the verification sample;
establishing a parameter evaluation function based on the sum of squared errors:
in formula (12), Y1jRepresenting the verification value at time j obtained by the support vector machine, Y0jRepresents the actual value at time j;
when the parameter evaluation function obtains the minimum value, the related parameters of the support vector machine are the optimal parameters of the support vector machine, otherwise, the kernel function ker, the kernel parameters p1 and p2, the parameter C and the Loss function Loss of the support vector machine are updated, and the support vector machine is retrained until the parameter evaluation function obtains the minimum value;
and outputting the optimal parameters of the support vector machine to obtain a prediction support vector machine which can be used for predicting the top oil temperature of the transformer.
8. The method of predicting the transformer top layer oil temperature as recited in claim 1, wherein the predicting the transformer top layer oil temperature comprises:
the method for predicting the top-layer oil temperature of the transformer is selected according to the number of similar moments in a similar moment set:
when the number of the similar moments in the similar moment set is less than or equal to 10, processing the top-layer oil temperature of the transformer corresponding to each moment in the similar moment set by adopting a linear weighting algorithm to obtain a predicted value of the top-layer oil temperature of the transformer;
and when the number of similar moments in the similar moment set is 20, predicting the top oil temperature of the transformer by using a support vector machine to obtain a predicted value of the top oil temperature of the transformer.
9. The method for predicting the top-layer oil temperature of the transformer according to claim 8, wherein the linear weighting algorithm is as follows:
configuring different weight coefficients p for the similar moments according to the comprehensive correlation degree of the similar moments in the similar moment set and the moments to be predictedxAnd 0 < px＜1；
Calculating a predicted value of the top oil temperature of the transformer:
in the formula (13), Y represents a predicted value of the top-layer oil temperature of the transformer; y0xRepresenting the actual value of the transformer top layer oil temperature at the x-th similar moment.
10. A prediction system for top layer oil temperature of a transformer, comprising:
a first calculation module: the method comprises the steps of calculating the comprehensive correlation degree of each moment and the moment to be predicted in a research sample, and obtaining a similar moment set with the maximum comprehensive correlation degree with the moment to be predicted;
an optimal parameter determination module: the optimal parameters for determining the support vector machine;
the transformer top oil temperature prediction module: and predicting the top oil temperature of the transformer according to the size of the similar time set and the optimal parameters of the support vector machine.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011178764.1A CN112347695B (en) | 2020-10-29 | 2020-10-29 | Method and system for predicting top layer oil temperature of transformer |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011178764.1A CN112347695B (en) | 2020-10-29 | 2020-10-29 | Method and system for predicting top layer oil temperature of transformer |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112347695A true CN112347695A (en) | 2021-02-09 |
| CN112347695B CN112347695B (en) | 2023-12-01 |
Family
ID=74356969
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011178764.1A Active CN112347695B (en) | 2020-10-29 | 2020-10-29 | Method and system for predicting top layer oil temperature of transformer |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112347695B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20170206291A1 (en) * | 2016-01-20 | 2017-07-20 | Soliton Holdings Corporation, Delaware Corporation | Method for computational fluid dynamics and apparatuses for jet-effect use |
| CN108664682A (en) * | 2017-11-30 | 2018-10-16 | 全球能源互联网研究院有限公司 | A kind of prediction technique and its system of transformer top-oil temperature |
| CN111666711A (en) * | 2020-05-12 | 2020-09-15 | 杭州安脉盛智能技术有限公司 | Transformer top layer oil temperature abnormity monitoring method based on multi-dimensional information |
| CN111680712A (en) * | 2020-04-16 | 2020-09-18 | 国网江苏省电力有限公司检修分公司 | Transformer oil temperature prediction method, device and system based on similar moments in the day |
-
2020
- 2020-10-29 CN CN202011178764.1A patent/CN112347695B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20170206291A1 (en) * | 2016-01-20 | 2017-07-20 | Soliton Holdings Corporation, Delaware Corporation | Method for computational fluid dynamics and apparatuses for jet-effect use |
| CN108664682A (en) * | 2017-11-30 | 2018-10-16 | 全球能源互联网研究院有限公司 | A kind of prediction technique and its system of transformer top-oil temperature |
| CN111680712A (en) * | 2020-04-16 | 2020-09-18 | 国网江苏省电力有限公司检修分公司 | Transformer oil temperature prediction method, device and system based on similar moments in the day |
| CN111666711A (en) * | 2020-05-12 | 2020-09-15 | 杭州安脉盛智能技术有限公司 | Transformer top layer oil temperature abnormity monitoring method based on multi-dimensional information |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112347695B (en) | 2023-12-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111639815B (en) | Method and system for predicting power grid defect materials through multi-model fusion | |
| CN103853106B (en) | A kind of energy consumption Prediction Parameters optimization method of building energy supplied equipment | |
| CN112700027B (en) | Multi-precision index comprehensive evaluation method and system for electric power prediction model | |
| CN109583585B (en) | Construction method of power station boiler wall temperature prediction neural network model | |
| CN109976147B (en) | Intelligent learning-based large-volume concrete temperature control method | |
| CN102183621A (en) | Aquaculture dissolved oxygen concentration online forecasting method and system | |
| CN111639783A (en) | Line loss prediction method and system based on LSTM neural network | |
| CN104281779A (en) | Abnormal data judging and processing method and device | |
| Feng et al. | End temperature prediction of molten steel in LF based on CBR–BBN | |
| CN111680712B (en) | Method, device and system for predicting oil temperature of transformer based on similar time in day | |
| JP7085067B2 (en) | Corrosion generation prediction model under heat insulating material and plant maintenance support device | |
| CN110969306A (en) | Power distribution low-voltage distribution area load prediction method and device based on deep learning | |
| CN115167562A (en) | Method and device for controlling temperature of machine room | |
| CN101825869A (en) | Method for identifying superheater model parameters based on data drive | |
| CN112347695B (en) | Method and system for predicting top layer oil temperature of transformer | |
| CN116862077A (en) | Electric heating operation cost prediction method and medium based on multi-mode combination model | |
| CN109027700B (en) | Method for evaluating leakage detection effect of leakage point | |
| CN104515591B (en) | A kind of high-precision coal pulverizer vibration detection process | |
| CN110210642A (en) | A kind of city electric energy substitution amount prediction technique and device | |
| CN111859783B (en) | Water pressure prediction method, system, storage medium, equipment and urban water supply system | |
| CN112801388B (en) | Power load prediction method and system based on nonlinear time series algorithm | |
| CN116151638A (en) | Distributed photovoltaic operation and maintenance effect evaluation method and system | |
| CN114881382B (en) | Optimized adjusting method for autonomous prediction of heat supply demand load | |
| CN111682574B (en) | Method for identifying running scene of alternating current-direct current hybrid system, storage medium and equipment | |
| CN112215383A (en) | Distributed photovoltaic power generation power prediction method and system |
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 |