CN116068481A - Method, system, equipment and medium for quantitatively evaluating error of current transformer - Google Patents

Method, system, equipment and medium for quantitatively evaluating error of current transformer Download PDF

Info

Publication number
CN116068481A
CN116068481A CN202310239172.3A CN202310239172A CN116068481A CN 116068481 A CN116068481 A CN 116068481A CN 202310239172 A CN202310239172 A CN 202310239172A CN 116068481 A CN116068481 A CN 116068481A
Authority
CN
China
Prior art keywords
error
sample
current transformer
colony
influence factor
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202310239172.3A
Other languages
Chinese (zh)
Inventor
赖国书
黄天富
吴志武
张颖
王春光
黄汉斌
林彤尧
伍翔
曹舒
郭银婷
陈适
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
State Grid Fujian Electric Power Co Ltd
Marketing Service Center of State Grid Fujian Electric Power Co Ltd
Original Assignee
State Grid Fujian Electric Power Co Ltd
Marketing Service Center of State Grid Fujian Electric Power Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by State Grid Fujian Electric Power Co Ltd, Marketing Service Center of State Grid Fujian Electric Power Co Ltd filed Critical State Grid Fujian Electric Power Co Ltd
Priority to CN202310239172.3A priority Critical patent/CN116068481A/en
Publication of CN116068481A publication Critical patent/CN116068481A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R35/00—Testing or calibrating of apparatus covered by the other groups of this subclass
    • G01R35/02—Testing or calibrating of apparatus covered by the other groups of this subclass of auxiliary devices, e.g. of instrument transformers according to prescribed transformation ratio, phase angle, or wattage rating
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G06N3/084—Backpropagation, e.g. using gradient descent
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
    • Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
    • Y04S10/00—Systems supporting electrical power generation, transmission or distribution
    • Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • Artificial Intelligence (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Power Engineering (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention relates to a current transformer error quantitative evaluation method, which comprises the following steps: acquiring a historical error influence factor set and a historical error data set of the electronic current transformer; performing feature selection processing on each sample in the history error influence factor set to obtain a training sample set after feature selection; constructing a WNN wavelet neural network, and optimizing super parameters of the WNN wavelet neural network by using an AAA artificial seaweed algorithm to obtain an AAA-WNN network model; taking the influence factor data of the corresponding historical moment of the training sample set as input, taking the error data of the corresponding historical moment of the historical error data set as output, and carrying out iterative training on the AAA-WNN network model; and acquiring the data of each kind of influence factors at the current moment of the electronic current transformer to be tested, and inputting the data into a trained error evaluation model of the electronic current transformer to be tested to obtain the current error evaluation result of the electronic current transformer to be tested.

Description

Method, system, equipment and medium for quantitatively evaluating error of current transformer
Technical Field
The invention relates to a method, a system, equipment and a medium for quantitatively evaluating errors of a current transformer, and belongs to the technical field of transformer evaluation.
Background
An Electronic Current Transformer (ECT) is an important device of an intelligent substation, and its operation performance is directly related to the safety of the primary side and the secondary side of the power system and the accuracy of the metering device.
The error evaluation of the electronic current transformer generally adopts an off-line checking or on-line checking method, and the ratio difference and the angle difference of the electronic current transformer are obtained through direct comparison. However, the inspection period of the methods is longer, the field wiring is complex, the working efficiency is low, in order to perfect the error state evaluation system of the electronic current transformer, the design of the error quantitative evaluation method of the electronic current transformer is needed, and the problem of out-of-tolerance error is found in time, so that the detection work of the electronic current transformer is carried out by workers, and the fairness of electric energy metering is ensured.
Disclosure of Invention
In order to solve the problems in the prior art, the invention provides a method, a system, equipment and a medium for quantitatively evaluating the error of a current transformer.
The technical scheme of the invention is as follows:
in one aspect, the invention provides a method for quantitatively evaluating the error of a current transformer, which comprises the following steps:
acquiring a historical error influence factor set and a historical error data set of the electronic current transformer, wherein the historical error influence factor set comprises samples at different historical moments, and each sample comprises a plurality of different types of influence factor data;
performing feature selection processing on each sample in the history error influence factor set to obtain a training sample set after feature selection;
constructing a WNN wavelet neural network, and optimizing super parameters of the WNN wavelet neural network by using an AAA artificial seaweed algorithm to obtain an AAA-WNN network model;
taking influence factor data of the corresponding historical moment of the training sample set as input, taking error data of the corresponding historical moment of the historical error data set as output, and performing iterative training on the AAA-WNN network model to obtain a trained electronic current transformer error evaluation model;
and acquiring the data of each kind of influence factors at the current moment of the electronic current transformer to be tested, and inputting the data into a trained error evaluation model of the electronic current transformer to be tested to obtain the current error evaluation result of the electronic current transformer to be tested.
In a preferred embodiment, the method for performing feature selection processing on each influence factor in the set of influence factors of the historical error specifically adopts a ReliefF algorithm, and specifically includes the following steps:
setting the iteration number as f and setting the weight initial value W of the feature 0 (x l ) Is 0;
in the iterative process, randomly selecting one sample R from the historical error influence factor set i Sample R i Including a set of historical error influencing factorsInfluence factor data of each class;
separately calculate and sample R i Intra-class distance of k nearest neighbor samples of the same class:
Figure SMS_1
and sum sample R i Inter-class distances of k nearest neighbor samples of different classes: />
Figure SMS_2
；
Updating the sample R based on the calculated intra-class distance and inter-class distance i The weight of each feature is obtained;
specifically, if the inter-class distance is larger than the intra-class distance, the weight of the corresponding feature is increased; if the inter-class distance is smaller than or equal to the intra-class distance, the weight of the corresponding feature is reduced, and the calculation formula of the weight is as follows:
Figure SMS_3
；
wherein W is i (x l ) The weight of the ith feature x in the ith sample; h j (j=1, 2, …, k) is the sum of the samples R i The j sample in k nearest neighbor samples of the same class; m is M j (C) (j=1, 2, …, k) is the sum of the samples R i The j-th sample of k nearest neighbor samples of different classes; p (C) represents the same as sample R i The number of similar samples is a proportion of the total number of samples; p (label (R) i ) Is sample R) i The number of samples corresponding to the class number of (1) is the ratio of the total number of samples, wherein label (R i ) Is R i Category number of (c);
and setting a weight threshold, and outputting the characteristics with the weight larger than the weight threshold to obtain a training sample set after characteristic selection.
As a preferred implementation mode, the method for optimizing the super parameters of the WNN wavelet neural network by using the AAA artificial seaweed algorithm specifically comprises the following steps:
initializing parameters of an AAA artificial seaweed algorithm, wherein the parameters comprise a population N, a maximum fitness index FEVs, a dimension D, an upper limit UB/LB, an energy loss le, a fitness evaluation index Ap and a shearing force delta;
randomly initializing the super parameters of the WNN wavelet neural network to form an initial population of microalgae colonies, and calculating the size G and fitness of each microalgae colony
Figure SMS_4
：
Figure SMS_5
；
Figure SMS_6
；
Wherein G is i t Is the size of the ith colony at time t, where v max The maximum growth rate is 1/time, S is nutrient concentration,
Figure SMS_7
is the base saturation constant in mass/volume;
repeating the population updating step until reaching the maximum fitness index FEVs, wherein the population updating step specifically comprises the following steps:
stage of spiral movement: at this stage, 3 colonies were arbitrarily modified:
Figure SMS_8
；
Figure SMS_9
；/>
wherein m, k, l are different algal cells;
Figure SMS_10
surface friction force of colony i at the t-th iteration; alpha and beta are random numbers, and the value range is [0,2 pi ]]；/>
Figure SMS_11
Is a random number, and has a value range of [ -1,1]The method comprises the steps of carrying out a first treatment on the surface of the i, j are different colonies in the population; x is x im t The mth algae cell in the colony i at the t iteration;
propagation phase: the largest and smallest colonies were selected at this stage and the smallest colony was replaced with the largest colony:
Figure SMS_12
；
wherein, smallest t Is the colony with the smallest scale at the t-th iteration; biggest t The colony with the largest model is regulated in the t-th iteration, and m is a randomly selected algae cell;
and (3) an adaptation stage: the most starved colonies were selected:
Figure SMS_13
；
wherein, starving t Is the colony with the maximum starvation degree at the t-th iteration; a is that i t The hunger for colony i is the t-th iteration;
colonies were updated according to the following formula:
Figure SMS_14
wherein Rand 1 Is a first random number with a value range of 0,1]；Rand 2 Is a second random number with a value in the range of 0,1]；A p To adapt to probability;
judging whether a preset iteration stopping condition is met, exiting if the preset iteration stopping condition is met, outputting the super-parameters in the current optimal microalgae colony as the optimal super-parameters, and otherwise, repeatedly executing the steps.
As a preferred embodiment, the different kinds of influencing factors include:
ambient temperature, ambient humidity, vibration, electric field, magnetic field, load, frequency, harmonics, and secondary current.
On the other hand, the invention also provides a current transformer error quantitative evaluation system, which comprises:
the historical data acquisition module is used for acquiring a historical error influence factor set and a historical error data set of the electronic current transformer, wherein the historical error influence factor set comprises samples at different historical moments, and each sample comprises a plurality of different types of influence factor data;
the feature selection module is used for carrying out feature selection processing on each sample in the history error influence factor set to obtain a training sample set after feature selection;
the model building module is used for building a WNN wavelet neural network, optimizing super parameters of the WNN wavelet neural network by utilizing an AAA artificial seaweed algorithm, and obtaining an AAA-WNN network model;
the model training module is used for carrying out iterative training on the AAA-WNN network model by taking the influence factor data of the corresponding historical moment of the training sample set as input and the error data of the corresponding historical moment of the historical error data set as output to obtain a trained electronic current transformer error evaluation model;
the evaluation module is used for collecting various influence factor data of the electronic current transformer to be tested at the current moment, inputting the influence factor data into the trained error evaluation model of the electronic current transformer to be tested, and obtaining the current error evaluation result of the electronic current transformer to be tested.
In a preferred embodiment, in the feature selection module, a method for performing feature selection processing on each influence factor in the set of influence factors of the historical error specifically adopts a ReliefF algorithm, and specifically includes the steps of:
setting the iteration number as f and setting the weight initial value W of the feature 0 (x l ) Is 0;
in the iterative process, randomly selecting one sample R from the historical error influence factor set i Sample R i The method comprises the steps of including influence factor data of each category in a history error influence factor set;
separately calculate and sample R i Intra-class distance of k nearest neighbor samples of the same class:
Figure SMS_15
and sum sample R i Inter-class distances of k nearest neighbor samples of different classes: />
Figure SMS_16
；
Updating the sample R based on the calculated intra-class distance and inter-class distance i The weight of each feature is obtained;
specifically, if the inter-class distance is larger than the intra-class distance, the weight of the corresponding feature is increased; if the inter-class distance is smaller than or equal to the intra-class distance, the weight of the corresponding feature is reduced, and the calculation formula of the weight is as follows:
Figure SMS_17
；
wherein W is i (x l ) The weight of the ith feature x in the ith sample; h j (j=1, 2, …, k) is the sum of the samples R i The j sample in k nearest neighbor samples of the same class; m is M j (C) (j=1, 2, …, k) is the sum of the samples R i The j-th sample of k nearest neighbor samples of different classes; p (C) represents the same as sample R i The number of similar samples is a proportion of the total number of samples; p (label (R) i ) Is sample R) i The number of samples corresponding to the class number of (1) is the ratio of the total number of samples, wherein label (R i ) Is R i Category number of (c);
and setting a weight threshold, and outputting the characteristics with the weight larger than the weight threshold to obtain a training sample set after characteristic selection.
In the model building module, as a preferred implementation manner, the method for optimizing the super parameters of the WNN wavelet neural network by using the AAA artificial seaweed algorithm specifically comprises the following steps:
initializing parameters of an AAA artificial seaweed algorithm, wherein the parameters comprise a population N, a maximum fitness index FEVs, a dimension D, an upper limit UB/LB, an energy loss le, a fitness evaluation index Ap and a shearing force delta;
randomly initializing the super parameters of the WNN wavelet neural network to form an initial population of microalgae colonies, and calculating the size G and fitness of each microalgae colony
Figure SMS_18
：
Figure SMS_19
；
Figure SMS_20
；
Wherein G is i t Is the size of the ith colony at time t, where v max The maximum growth rate is 1/time, S is nutrient concentration, K S Is the base saturation constant in mass/volume;
repeating the population updating step until reaching the maximum fitness index FEVs, wherein the population updating step specifically comprises the following steps:
stage of spiral movement: at this stage, 3 colonies were arbitrarily modified:
Figure SMS_21
；
Figure SMS_22
；
wherein m, k, l are different algal cells;
Figure SMS_23
surface friction force of colony i at the t-th iteration; alpha and beta are random numbers, and the value range is [0,2 pi ]]；/>
Figure SMS_24
Is a random number, and has a value range of [ -1,1]The method comprises the steps of carrying out a first treatment on the surface of the i, j are different colonies in the population; x is x im t The mth algae cell in the colony i at the t iteration;
propagation phase: the largest and smallest colonies were selected at this stage and the smallest colony was replaced with the largest colony:
Figure SMS_25
；
wherein, smallest t Is the colony with the smallest scale at the t-th iteration; biggest t T th timeThe colony with the largest model is regulated during iteration, and m is a randomly selected algae cell;
and (3) an adaptation stage: the most starved colonies were selected:
Figure SMS_26
；
wherein, starving t Is the colony with the maximum starvation degree at the t-th iteration; a is that i t The hunger for colony i is the t-th iteration;
colonies were updated according to the following formula:
Figure SMS_27
wherein Rand 1 Is a first random number with a value range of 0,1]；Rand 2 Is a second random number with a value in the range of 0,1]；A p To adapt to probability;
judging whether a preset iteration stopping condition is met, exiting if the preset iteration stopping condition is met, outputting the super-parameters in the current optimal microalgae colony as the optimal super-parameters, and otherwise, repeatedly executing the steps.
As a preferred embodiment, the different kinds of influencing factors include:
ambient temperature, ambient humidity, vibration, electric field, magnetic field, load, frequency, harmonics, and secondary current.
In still another aspect, the present invention further provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor implements the method for quantitatively evaluating a current transformer error according to any embodiment of the present invention when the processor executes the program.
In yet another aspect, the present invention further provides a computer readable storage medium having stored thereon a computer program, wherein the program when executed by a processor implements a method for quantitatively evaluating a current transformer error according to any of the embodiments of the present invention.
The invention has the following beneficial effects:
1. the method for quantitatively evaluating the error of the current transformer can evaluate the error of the electronic current transformer in real time, so that the problem of out-of-tolerance error of the electronic current transformer can be found in time, and workers can conveniently perform detection work of the electronic current transformer, thereby ensuring fairness of electric energy metering.
2. According to the current transformer error quantitative evaluation method, the Relieff algorithm is adopted to conduct feature selection on each influence factor in the historical error influence factor set, and influence factors with evaluation value are screened out.
3. According to the current transformer error quantitative evaluation method, the AAA artificial seaweed algorithm is utilized to optimize the super parameters of the WNN wavelet neural network, so that the performance of the WNN wavelet neural network is improved.
Drawings
FIG. 1 is a flowchart of a method according to an embodiment of the present invention;
fig. 2 is a diagram illustrating an exemplary network structure of a wavelet neural network according to an embodiment of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
It should be understood that the step numbers used herein are for convenience of description only and are not limiting as to the order in which the steps are performed.
It is to be understood that the terminology used in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used in this specification and the appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise.
The terms "comprises" and "comprising" indicate the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.
The term "and/or" refers to any and all possible combinations of one or more of the associated listed items, and includes such combinations.
Embodiment one:
referring to fig. 1, a method for quantitatively evaluating the error of a current transformer includes the following steps:
s100, acquiring a historical error influence factor set and a historical error data set of an electronic current transformer, wherein the historical error influence factor set comprises samples at different historical moments, and each sample comprises a plurality of different types of influence factor data; specifically, a high-precision acquisition device is used for acquiring a historical error influence factor set x of the electronic current transformer:
Figure SMS_28
；
wherein the ambient temperature
Figure SMS_31
Ambient humidity->
Figure SMS_32
Vibration->
Figure SMS_36
Electric field->
Figure SMS_30
Magnetic field->
Figure SMS_33
Load->
Figure SMS_35
Frequency->
Figure SMS_37
Harmonic wave->
Figure SMS_29
Secondary current->
Figure SMS_34
. The historical error data set is obtained through power failure verification and comprises the ratio difference and the angle difference of the electronic current transformer.
S200, performing feature selection processing on each sample in the historical error influence factor set to obtain a training sample set after feature selection;
s300, constructing a WNN wavelet neural network, wherein the wavelet neural network is a neural network formed by taking a common neural network as a basis, taking a wavelet basis function as a nonlinear function of a neuron, and adopting a method of modifying wavelet function parameters by error back propagation. The wavelet neural network combines the good local microscopic features of wavelet transformation and the self-learning capability of the neural network. A specific neural network structure is shown in fig. 2. Wherein:
the output of the ith neuron in the hidden layer is:
Figure SMS_38
；
wherein r is the number of input neurons, x is the input quantity, and x j For j-th input quantity, s l As the number of neurons in the hidden layer, f 1 () An activation function of an implicit layer, the expression of which is
Figure SMS_39
, g i To activate the function f 1 () Is the scaling factor, w ij Is an input layer and hidden layer node, b 1i Is the threshold for the ith input neuron.
The output of the p-th neuron of the output layer is:
Figure SMS_40
；
wherein s is 2 To output the number of layer neurons, f 2 () As an activation function of the output layer, w jp To hideLayer and output layer nodes, b 2p Is the threshold of the p-th output neuron.
The error function used in the error back propagation is set as:
Figure SMS_41
；
wherein: q is the number of learning samples, y t An actual error value of the electronic current transformer at the time t, a 2t The prediction error value of the electronic current transformer at the time t.
The weight correction function of the hidden layer is as follows:
Figure SMS_42
wherein, H is a sample set, H-1 is a previous sample set, k is the training times of the network;
Figure SMS_43
micro entropy (gradient) of error versus weight.
The weight correction function only uses the output errors of two sample sets of the network, and when the network performs the k+1st training, the network weight is corrected by differentiating the weight by the output errors of two adjacent sample sets. The error differentiation of the previous sample set H-1 is adopted to correct the weight, the relation among the training samples is fully utilized, and the output errors of two adjacent training sample sets are used each time to continuously correct the weight of the network until the set training precision requirement is met. Wherein,,
Figure SMS_44
and->
Figure SMS_45
The values set by these two parameters can affect the convergence speed of the network weights.
The wavelet neural network is used as a deep learning model, and the performance of the wavelet neural network depends on super parameters (such as in a formula
Figure SMS_46
And->
Figure SMS_47
Equal parameters), in this embodiment, an AAA artificial seaweed algorithm is used to optimize the super parameters of the wavelet neural network, so as to obtain an AAA-WNN network model.
S400, taking influence factor data of the corresponding historical moment of the training sample set as input, taking error data of the corresponding historical moment of the historical error data set as output, and performing iterative training on the AAA-WNN network model to obtain a trained electronic current transformer error evaluation model.
S500, collecting various influence factor data of the electronic current transformer to be tested at the current moment, including real-time environmental temperature, environmental humidity, vibration, electric field, magnetic field, load, frequency, harmonic wave and secondary current data, inputting the data into a trained electronic current transformer error evaluation model, and predicting by the electronic current transformer error evaluation model to obtain the current error evaluation result of the electronic current transformer to be tested.
As a preferred implementation manner of this embodiment, the method for performing feature selection processing on each influence factor in the set of influence factors of the historical error specifically adopts a ReliefF algorithm, which specifically includes the following steps:
setting the iteration number as f and setting the weight initial value W of the feature 0 (x l ) Is 0;
in the iterative process, randomly selecting one sample R from the historical error influence factor set i Sample R i The method comprises the steps of including influence factor data of each category in a history error influence factor set;
separately calculate and sample R i Intra-class distance of k nearest neighbor samples of the same class:
Figure SMS_48
and sum sample R i Inter-class distances of k nearest neighbor samples of different classes: />
Figure SMS_49
；R i Is one sampleThus we choose k and R i Similar neighbor samples are used for calculating the intra-class distance, and k and R are selected at the same time i And calculating the distances between the classes by using the neighbor samples of different classes. Meanwhile, the influence factor data of the same category in each sample is the same category, and the influence factor data of different categories is different categories. And comparing the intra-class distance with the inter-class distance, and further adjusting the characteristic weight to be the characteristic of the ReliefF algorithm.
Updating the sample R based on the calculated intra-class distance and inter-class distance i The weight of each feature is obtained;
specifically, if the inter-class distance is larger than the intra-class distance, the weight of the corresponding feature is increased; if the inter-class distance is smaller than or equal to the intra-class distance, the weight of the corresponding feature is reduced, and the calculation formula of the weight is as follows:
Figure SMS_50
；
wherein W is i (x l ) The weight of the ith feature x in the ith sample; h j (j=1, 2, …, k) is the sum of the samples R i The j sample in k nearest neighbor samples of the same class; m is M j (C) (j=1, 2, …, k) is the sum of the samples R i The j-th sample of k nearest neighbor samples of different classes; p (C) represents the ratio of the number of samples of the same class as the sample Ri to the total number of samples; p (label (R) i ) Is sample R) i The number of samples corresponding to the class number of (1) is the ratio of the total number of samples, wherein label (R i ) Is R i Category number of (c);
and setting a weight threshold, and outputting the characteristics with the weight larger than the weight threshold to obtain a training sample set after characteristic selection.
As a preferred implementation manner of the embodiment, the method for optimizing the super parameters of the WNN wavelet neural network by using the AAA artificial seaweed algorithm specifically comprises the following steps:
initializing parameters of an AAA artificial seaweed algorithm, wherein the parameters comprise a population N, a maximum fitness index FEVs, a dimension D, an upper limit UB/LB, an energy loss le, a fitness evaluation index Ap and a shearing force delta;
randomly initializing the super parameters of the WNN wavelet neural network to form an initial population of microalgae colonies, and calculating the size G and fitness of each microalgae colony
Figure SMS_51
：
Figure SMS_52
；
Figure SMS_53
；
Wherein G is i t Is the size of the ith colony at time t, where v max The maximum growth rate is 1/time, S is nutrient concentration, K S Is the base saturation constant in mass/volume;
repeating the population updating step until reaching the maximum fitness index FEVs, wherein the population updating step specifically comprises the following steps:
stage of spiral movement: at this stage, 3 colonies were arbitrarily modified:
Figure SMS_54
；
Figure SMS_55
；
wherein m, k, l are different algal cells;
Figure SMS_56
surface friction force of colony i at the t-th iteration; alpha and beta are random numbers, and the value range is [0,2 pi ]]；/>
Figure SMS_57
Is a random number, and has a value range of [ -1,1]The method comprises the steps of carrying out a first treatment on the surface of the i, j are different colonies in the population; x is x im t The mth algae cell in the colony i at the t iteration;
propagation phase: the largest and smallest colonies were selected at this stage and the smallest colony was replaced with the largest colony:
Figure SMS_58
；
wherein, smallest t Is the colony with the smallest scale at the t-th iteration; biggest t The colony with the largest model is regulated in the t-th iteration, and m is a randomly selected algae cell;
and (3) an adaptation stage: the most starved colonies were selected:
Figure SMS_59
；/>
wherein, starving t Is the colony with the maximum starvation degree at the t-th iteration; a is that i t The hunger for colony i is the t-th iteration;
colonies were updated according to the following formula:
Figure SMS_60
wherein Rand 1 Is a first random number with a value range of 0,1]；Rand 2 Is a second random number with a value in the range of 0,1]；A p To adapt to probability;
judging whether a preset iteration stopping condition is met, exiting, outputting the super parameter in the current optimal microalgae colony as the optimal super parameter, setting a maximum iteration number by the iteration stopping condition, outputting the last optimal microalgae colony as the optimal super parameter if the iteration number meets the requirement, and repeating the steps if the iteration number does not meet the requirement.
Through the optimization process, the optimal super parameters of the wavelet neural network are obtained, and an AAA-WNN network model is constructed.
Embodiment two:
the embodiment provides a current transformer error quantitative evaluation system, including:
the historical data acquisition module is used for acquiring a historical error influence factor set and a historical error data set of the electronic current transformer, wherein the historical error influence factor set comprises samples at different historical moments, and each sample comprises a plurality of different types of influence factor data; the module is used for implementing the function of step S100 in the first embodiment, and will not be described here again;
the feature selection module is used for carrying out feature selection processing on each sample in the history error influence factor set to obtain a training sample set after feature selection; the module is used for implementing the function of step S200 in the first embodiment, and will not be described in detail herein;
the model building module is used for building a WNN wavelet neural network, optimizing super parameters of the WNN wavelet neural network by utilizing an AAA artificial seaweed algorithm, and obtaining an AAA-WNN network model; the module is used for implementing the function of step S300 in the first embodiment, and will not be described in detail herein;
the model training module is used for carrying out iterative training on the AAA-WNN network model by taking the influence factor data of the corresponding historical moment of the training sample set as input and the error data of the corresponding historical moment of the historical error data set as output to obtain a trained electronic current transformer error evaluation model; the module is used for realizing the function of step S400 in the first embodiment, and will not be described in detail herein;
the evaluation module is used for collecting various influence factor data of the electronic current transformer to be tested at the current moment, inputting the influence factor data into the trained error evaluation model of the electronic current transformer to be tested, and obtaining the current error evaluation result of the electronic current transformer to be tested; the module is used to implement the function of step S500 in the first embodiment, and will not be described herein.
As a preferred implementation manner of this embodiment, in the feature selection module, a method for performing feature selection processing on each influence factor in the set of influence factors of the history error specifically adopts a ReliefF algorithm, which specifically includes the steps of:
setting the iteration number as f and setting the weight initial value W of the feature 0 (x l ) Is 0;
in the iterative process, randomly selecting one sample R from the historical error influence factor set i Sample R i The number of influence factors of each category in the history error influence factor set is includedAccording to the above;
separately calculate and sample R i Intra-class distance of k nearest neighbor samples of the same class:
Figure SMS_61
and sum sample R i Inter-class distances of k nearest neighbor samples of different classes: />
Figure SMS_62
；
Updating the sample R based on the calculated intra-class distance and inter-class distance i The weight of each feature is obtained;
specifically, if the inter-class distance is larger than the intra-class distance, the weight of the corresponding feature is increased; if the inter-class distance is smaller than or equal to the intra-class distance, the weight of the corresponding feature is reduced, and the calculation formula of the weight is as follows:
Figure SMS_63
；
wherein W is i (x l ) The weight of the ith feature x in the ith sample; h j (j=1, 2, …, k) is the sum of the samples R i The j sample in k nearest neighbor samples of the same class; m is M j (C) (j=1, 2, …, k) is the sum of the samples R i The j-th sample of k nearest neighbor samples of different classes; p (C) represents the same as sample R i The number of similar samples is a proportion of the total number of samples; p (label (R) i ) Is sample R) i The number of samples corresponding to the class number of (1) is the ratio of the total number of samples, wherein label (R i ) Is R i Category number of (c);
and setting a weight threshold, and outputting the characteristics with the weight larger than the weight threshold to obtain a training sample set after characteristic selection.
As a preferred implementation manner of the embodiment, in the model building module, the method for optimizing the super parameters of the WNN wavelet neural network by using the AAA artificial seaweed algorithm specifically comprises the following steps:
initializing parameters of an AAA artificial seaweed algorithm, wherein the parameters comprise a population N, a maximum fitness index FEVs, a dimension D, an upper limit UB/LB, an energy loss le, a fitness evaluation index Ap and a shearing force delta;
randomly initializing the super parameters of the WNN wavelet neural network to form an initial population of microalgae colonies, and calculating the size G and fitness of each microalgae colony
Figure SMS_64
：
Figure SMS_65
；
Figure SMS_66
；
Wherein G is i t Is the size of the ith colony at time t, where v max The maximum growth rate is 1/time, S is nutrient concentration, K S Is the base saturation constant in mass/volume;
repeating the population updating step until reaching the maximum fitness index FEVs, wherein the population updating step specifically comprises the following steps:
stage of spiral movement: at this stage, 3 colonies were arbitrarily modified:
Figure SMS_67
；
Figure SMS_68
；/>
wherein m, k, l are different algal cells;
Figure SMS_69
surface friction force of colony i at the t-th iteration; alpha and beta are random numbers, and the value range is [0,2 pi ]]；/>
Figure SMS_70
Is a random number, and has a value range of [ -1,1]The method comprises the steps of carrying out a first treatment on the surface of the i, j are different colonies in the population; x is x im t The mth algae cell in the colony i at the t iteration;
propagation phase: the largest and smallest colonies were selected at this stage and the smallest colony was replaced with the largest colony:
Figure SMS_71
；
wherein, smallest t Is the colony with the smallest scale at the t-th iteration; biggest t The colony with the largest model is regulated in the t-th iteration, and m is a randomly selected algae cell;
and (3) an adaptation stage: the most starved colonies were selected:
Figure SMS_72
；
wherein, starving t Is the colony with the maximum starvation degree at the t-th iteration; a is that i t The hunger for colony i is the t-th iteration;
colonies were updated according to the following formula:
Figure SMS_73
wherein Rand 1 Is a first random number with a value range of 0,1]；Rand 2 Is a second random number with a value in the range of 0,1]；A p To adapt to probability;
judging whether a preset iteration stopping condition is met, exiting if the preset iteration stopping condition is met, outputting the super-parameters in the current optimal microalgae colony as the optimal super-parameters, and otherwise, repeatedly executing the steps.
As a preferred implementation of this embodiment, the different types of influencing factors include:
ambient temperature, ambient humidity, vibration, electric field, magnetic field, load, frequency, harmonics, and secondary current.
Embodiment III:
the embodiment provides an electronic device, which comprises a memory, a processor and a computer program stored on the memory and capable of running on the processor, wherein the processor realizes the quantitative evaluation method for the error of the current transformer according to any embodiment of the invention when executing the program.
Embodiment four:
the present embodiment proposes a computer-readable storage medium, on which a computer program is stored, which when executed by a processor implements a method for quantitatively evaluating a current transformer error according to any of the embodiments of the present invention.
In the embodiments of the present application, "at least one" means one or more, and "a plurality" means two or more. "and/or", describes an association relation of association objects, and indicates that there may be three kinds of relations, for example, a and/or B, and may indicate that a alone exists, a and B together, and B alone exists. Wherein A, B may be singular or plural. The character "/" generally indicates that the context-dependent object is an "or" relationship. "at least one of the following" and the like means any combination of these items, including any combination of single or plural items. For example, at least one of a, b and c may represent: a, b, c, a and b, a and c, b and c or a and b and c, wherein a, b and c can be single or multiple.
Those of ordinary skill in the art will appreciate that the various elements and algorithm steps described in the embodiments disclosed herein can be implemented as a combination of electronic hardware, computer software, and electronic hardware. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application.
It will be clear to those skilled in the art that, for convenience and brevity of description, specific working procedures of the above-described systems, apparatuses and units may refer to corresponding procedures in the foregoing method embodiments, and are not repeated herein.
In several embodiments provided herein, any of the functions, if implemented in the form of software functional units and sold or used as a stand-alone product, may be stored in a computer-readable storage medium. Based on such understanding, the technical solution of the present application may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution, in the form of a software product stored in a storage medium, including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in the embodiments of the present application. And the aforementioned storage medium includes: a usb disk, a removable hard disk, a Read-only memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk, etc., which can store program codes.
The foregoing description is only illustrative of the present invention and is not intended to limit the scope of the invention, and all equivalent structures or equivalent processes or direct or indirect application in other related technical fields are included in the scope of the present invention.

Claims (10)

1. The method for quantitatively evaluating the error of the current transformer is characterized by comprising the following steps of:
acquiring a historical error influence factor set and a historical error data set of the electronic current transformer, wherein the historical error influence factor set comprises samples at different historical moments, and each sample comprises a plurality of different types of influence factor data;
performing feature selection processing on each sample in the history error influence factor set to obtain a training sample set after feature selection;
constructing a WNN wavelet neural network, and optimizing super parameters of the WNN wavelet neural network by using an AAA artificial seaweed algorithm to obtain an AAA-WNN network model;
taking influence factor data of the corresponding historical moment of the training sample set as input, taking error data of the corresponding historical moment of the historical error data set as output, and performing iterative training on the AAA-WNN network model to obtain a trained electronic current transformer error evaluation model;
and acquiring the data of each kind of influence factors at the current moment of the electronic current transformer to be tested, and inputting the data into a trained error evaluation model of the electronic current transformer to be tested to obtain the current error evaluation result of the electronic current transformer to be tested.
2. The method for quantitatively evaluating the error of the current transformer according to claim 1, wherein the method for performing feature selection processing on each influence factor in the set of influence factors of the historical error specifically adopts a ReliefF algorithm, and specifically comprises the following steps:
setting the iteration number as f and setting the weight initial value W of the feature 0 (x l ) Is 0;
in the iterative process, randomly selecting one sample R from the historical error influence factor set i Sample R i The method comprises the steps of including influence factor data of each category in a history error influence factor set;
separately calculate and sample R i Intra-class distance of k nearest neighbor samples of the same class:
Figure QLYQS_1
and sum sample R i Inter-class distances of k nearest neighbor samples of different classes: />
Figure QLYQS_2
；
Updating the sample R based on the calculated intra-class distance and inter-class distance i The weight of each feature is obtained;
specifically, if the inter-class distance is larger than the intra-class distance, the weight of the corresponding feature is increased; if the inter-class distance is smaller than or equal to the intra-class distance, the weight of the corresponding feature is reduced, and the calculation formula of the weight is as follows:
Figure QLYQS_3
；
wherein W is i (x l ) The weight of the ith feature x in the ith sample; h j (j=1,2…, k) is the sum of the samples R i The j sample in k nearest neighbor samples of the same class; m is M j (C) (j=1, 2, …, k) is the sum of the samples R i The j-th sample of k nearest neighbor samples of different classes; p (C) represents the same as sample R i The number of similar samples is a proportion of the total number of samples; p (label (R) i ) Is sample R) i The number of samples corresponding to the class number of (1) is the ratio of the total number of samples, wherein label (R i ) Is R i Category number of (c);
and setting a weight threshold, and outputting the characteristics with the weight larger than the weight threshold to obtain a training sample set after characteristic selection.
3. The method for quantitatively evaluating the error of the current transformer according to claim 1, wherein the method for optimizing the super parameters of the WNN wavelet neural network by using the AAA artificial seaweed algorithm is specifically as follows:
initializing parameters of an AAA artificial seaweed algorithm, wherein the parameters comprise a population N, a maximum fitness index FEVs, a dimension D, an upper limit UB/LB, an energy loss le, a fitness evaluation index Ap and a shearing force delta;
randomly initializing the super parameters of the WNN wavelet neural network to form an initial population of microalgae colonies, and calculating the size G and fitness of each microalgae colony
Figure QLYQS_4
：
Figure QLYQS_5
；
Figure QLYQS_6
；
Wherein G is i t Is the size of the ith colony at time t, where v max The maximum growth rate is 1/time, S is nutrient concentration,
Figure QLYQS_7
is saturated with baseAnd a constant in mass/volume;
repeating the population updating step until reaching the maximum fitness index FEVs, wherein the population updating step specifically comprises the following steps:
stage of spiral movement: at this stage, 3 colonies were arbitrarily modified:
Figure QLYQS_8
；
Figure QLYQS_9
；
wherein m, k, l are different algal cells;
Figure QLYQS_10
surface friction force of colony i at the t-th iteration; alpha and beta are random numbers, and the value range is [0,2 pi ]]；/>
Figure QLYQS_11
Is a random number, and has a value range of [ -1,1]The method comprises the steps of carrying out a first treatment on the surface of the i, j are different colonies in the population; x is x im t The mth algae cell in the colony i at the t iteration;
propagation phase: the largest and smallest colonies were selected at this stage and the smallest colony was replaced with the largest colony:
Figure QLYQS_12
；
wherein, smallest t Is the colony with the smallest scale at the t-th iteration; biggest t The colony with the largest model is regulated in the t-th iteration, and m is a randomly selected algae cell;
and (3) an adaptation stage: the most starved colonies were selected:
Figure QLYQS_13
；
wherein, starving t Hunger at the t th iterationColonies with maximum hunger; a is that i t The hunger for colony i is the t-th iteration;
colonies were updated according to the following formula:
Figure QLYQS_14
wherein Rand 1 Is a first random number with a value range of 0,1]；Rand 2 Is a second random number with a value in the range of 0,1]；A p To adapt to probability;
judging whether a preset iteration stopping condition is met, exiting if the preset iteration stopping condition is met, outputting the super-parameters in the current optimal microalgae colony as the optimal super-parameters, and otherwise, repeatedly executing the steps.
4. The method for quantitatively evaluating the error of the current transformer according to claim 1, wherein the different types of influencing factors comprise:
ambient temperature, ambient humidity, vibration, electric field, magnetic field, load, frequency, harmonics, and secondary current.
5. A current transformer error quantitative assessment system, comprising:
the historical data acquisition module is used for acquiring a historical error influence factor set and a historical error data set of the electronic current transformer, wherein the historical error influence factor set comprises samples at different historical moments, and each sample comprises a plurality of different types of influence factor data;
the feature selection module is used for carrying out feature selection processing on each sample in the history error influence factor set to obtain a training sample set after feature selection;
the model building module is used for building a WNN wavelet neural network, optimizing super parameters of the WNN wavelet neural network by utilizing an AAA artificial seaweed algorithm, and obtaining an AAA-WNN network model;
the model training module is used for carrying out iterative training on the AAA-WNN network model by taking the influence factor data of the corresponding historical moment of the training sample set as input and the error data of the corresponding historical moment of the historical error data set as output to obtain a trained electronic current transformer error evaluation model;
the evaluation module is used for collecting various influence factor data of the electronic current transformer to be tested at the current moment, inputting the influence factor data into the trained error evaluation model of the electronic current transformer to be tested, and obtaining the current error evaluation result of the electronic current transformer to be tested.
6. The system for quantitatively evaluating the error of the current transformer according to claim 5, wherein in the feature selection module, a method for performing feature selection processing on each influence factor in the set of influence factors of the historical error specifically adopts a ReliefF algorithm, and specifically comprises the following steps:
setting the iteration number as f and setting the weight initial value W of the feature 0 (x l ) Is 0;
in the iterative process, randomly selecting one sample R from the historical error influence factor set i Sample R i The method comprises the steps of including influence factor data of each category in a history error influence factor set;
separately calculate and sample R i Intra-class distance of k nearest neighbor samples of the same class:
Figure QLYQS_15
and sum sample R i Inter-class distances of k nearest neighbor samples of different classes: />
Figure QLYQS_16
；
Updating the sample R based on the calculated intra-class distance and inter-class distance i The weight of each feature is obtained;
specifically, if the inter-class distance is larger than the intra-class distance, the weight of the corresponding feature is increased; if the inter-class distance is smaller than or equal to the intra-class distance, the weight of the corresponding feature is reduced, and the calculation formula of the weight is as follows:
Figure QLYQS_17
；
wherein W is i (x l ) The weight of the ith feature x in the ith sample; h j (j=1, 2, …, k) is the sum of the samples R i The j sample in k nearest neighbor samples of the same class; m is M j (C) (j=1, 2, …, k) is the sum of the samples R i The j-th sample of k nearest neighbor samples of different classes; p (C) represents the same as sample R i The number of similar samples is a proportion of the total number of samples; p (label (R) i ) Is sample R) i The number of samples corresponding to the class number of (1) is the ratio of the total number of samples, wherein label (R i ) Is R i Category number of (c);
and setting a weight threshold, and outputting the characteristics with the weight larger than the weight threshold to obtain a training sample set after characteristic selection.
7. The system for quantitatively evaluating the error of the current transformer according to claim 5, wherein in the model building module, the method for optimizing the super parameters of the WNN wavelet neural network by using the AAA artificial seaweed algorithm comprises the following steps:
initializing parameters of an AAA artificial seaweed algorithm, wherein the parameters comprise a population N, a maximum fitness index FEVs, a dimension D, an upper limit UB/LB, an energy loss le, a fitness evaluation index Ap and a shearing force delta;
randomly initializing the super parameters of the WNN wavelet neural network to form an initial population of microalgae colonies, and calculating the size G and fitness of each microalgae colony
Figure QLYQS_18
：
Figure QLYQS_19
；
Figure QLYQS_20
；
Wherein G is i t Is the size of the ith colony at time t, where v max The maximum growth rate is 1/time, S is nutrient concentration, K S Is the base saturation constant in mass/volume;
repeating the population updating step until reaching the maximum fitness index FEVs, wherein the population updating step specifically comprises the following steps:
stage of spiral movement: at this stage, 3 colonies were arbitrarily modified:
Figure QLYQS_21
；
Figure QLYQS_22
；
wherein m, k, l are different algal cells;
Figure QLYQS_23
surface friction force of colony i at the t-th iteration; alpha and beta are random numbers, and the value range is [0,2 pi ]]；/>
Figure QLYQS_24
Is a random number, and has a value range of [ -1,1]The method comprises the steps of carrying out a first treatment on the surface of the i, j are different colonies in the population; x is x im t The mth algae cell in the colony i at the t iteration;
propagation phase: the largest and smallest colonies were selected at this stage and the smallest colony was replaced with the largest colony:
Figure QLYQS_25
；
wherein, smallest t Is the colony with the smallest scale at the t-th iteration; biggest t The colony with the largest model is regulated in the t-th iteration, and m is a randomly selected algae cell;
and (3) an adaptation stage: the most starved colonies were selected:
Figure QLYQS_26
；
wherein, starving t Is the colony with the maximum starvation degree at the t-th iteration; a is that i t The hunger for colony i is the t-th iteration;
colonies were updated according to the following formula:
Figure QLYQS_27
；
wherein Rand 1 Is a first random number with a value range of 0,1]；Rand 2 Is a second random number with a value in the range of 0,1]；A p To adapt to probability;
judging whether a preset iteration stopping condition is met, exiting if the preset iteration stopping condition is met, outputting the super-parameters in the current optimal microalgae colony as the optimal super-parameters, and otherwise, repeatedly executing the steps.
8. The current transformer error quantitative assessment system of claim 5, wherein the different classes of influencing factors comprise:
ambient temperature, ambient humidity, vibration, electric field, magnetic field, load, frequency, harmonics, and secondary current.
9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the current transformer error quantitative assessment method of any one of claims 1 to 4 when the program is executed by the processor.
10. A computer-readable storage medium, on which a computer program is stored, characterized in that the program, when being executed by a processor, implements the current transformer error quantitative assessment method according to any one of claims 1 to 4.
CN202310239172.3A 2023-03-14 2023-03-14 Method, system, equipment and medium for quantitatively evaluating error of current transformer Pending CN116068481A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202310239172.3A CN116068481A (en) 2023-03-14 2023-03-14 Method, system, equipment and medium for quantitatively evaluating error of current transformer

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310239172.3A CN116068481A (en) 2023-03-14 2023-03-14 Method, system, equipment and medium for quantitatively evaluating error of current transformer

Publications (1)

Publication Number Publication Date
CN116068481A true CN116068481A (en) 2023-05-05

Family

ID=86175167

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202310239172.3A Pending CN116068481A (en) 2023-03-14 2023-03-14 Method, system, equipment and medium for quantitatively evaluating error of current transformer

Country Status (1)

Country Link
CN (1) CN116068481A (en)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116485049A (en) * 2023-06-25 2023-07-25 佛山市龙生光启科技有限公司 Electric energy metering error prediction and optimization system based on artificial intelligence
CN117073728A (en) * 2023-10-17 2023-11-17 天津易泰炬业科技有限公司 Flexible capacitive touch sensor
CN117233687A (en) * 2023-11-13 2023-12-15 威胜集团有限公司 CVT initial error assessment method, medium and terminal based on historical data
CN117388623A (en) * 2023-12-12 2024-01-12 国网江西省电力有限公司电力科学研究院 Comprehensive diagnosis analyzer and method for power transformer without disassembling lead

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR970076952A (en) * 1996-05-14 1997-12-12 이종훈 Apparatus and method for restoring error of transformer using neural network
CN109038602A (en) * 2018-07-20 2018-12-18 国家电网有限公司 It is a kind of meter and transmission losses UPFC optimal configuration method
CN111814390A (en) * 2020-06-18 2020-10-23 三峡大学 Voltage transformer error prediction method based on transfer entropy and wavelet neural network
CN113536662A (en) * 2021-06-16 2021-10-22 三峡大学 Electronic transformer error state prediction method based on firefly optimized LightGBM algorithm

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR970076952A (en) * 1996-05-14 1997-12-12 이종훈 Apparatus and method for restoring error of transformer using neural network
CN109038602A (en) * 2018-07-20 2018-12-18 国家电网有限公司 It is a kind of meter and transmission losses UPFC optimal configuration method
CN111814390A (en) * 2020-06-18 2020-10-23 三峡大学 Voltage transformer error prediction method based on transfer entropy and wavelet neural network
CN113536662A (en) * 2021-06-16 2021-10-22 三峡大学 Electronic transformer error state prediction method based on firefly optimized LightGBM algorithm

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
史建青等: "SBP探测技术及其在海底勘查中的应用", 中国矿业大学出版社, pages: 104 - 107 *
李世煜: ""人工藻类算法在电力系统经济调度中的研究应用"", 《万方数据》, pages 11 - 16 *

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116485049A (en) * 2023-06-25 2023-07-25 佛山市龙生光启科技有限公司 Electric energy metering error prediction and optimization system based on artificial intelligence
CN116485049B (en) * 2023-06-25 2024-04-19 陕西银河电力仪表股份有限公司 Electric energy metering error prediction and optimization system based on artificial intelligence
CN117073728A (en) * 2023-10-17 2023-11-17 天津易泰炬业科技有限公司 Flexible capacitive touch sensor
CN117073728B (en) * 2023-10-17 2024-01-23 天津易泰炬业科技有限公司 Flexible capacitive touch sensor
CN117233687A (en) * 2023-11-13 2023-12-15 威胜集团有限公司 CVT initial error assessment method, medium and terminal based on historical data
CN117233687B (en) * 2023-11-13 2024-02-06 威胜集团有限公司 CVT initial error assessment method, medium and terminal based on historical data
CN117388623A (en) * 2023-12-12 2024-01-12 国网江西省电力有限公司电力科学研究院 Comprehensive diagnosis analyzer and method for power transformer without disassembling lead
CN117388623B (en) * 2023-12-12 2024-05-14 国网江西省电力有限公司电力科学研究院 Comprehensive diagnosis analyzer and method for power transformer without disassembling lead

Similar Documents

Publication Publication Date Title
CN116068481A (en) Method, system, equipment and medium for quantitatively evaluating error of current transformer
CN109002686B (en) Multi-grade chemical process soft measurement modeling method capable of automatically generating samples
CN111340273A (en) Short-term load prediction method for power system based on GEP parameter optimization XGboost
CN110348075A (en) A kind of grinding surface roughness prediction technique based on improvement algorithm of support vector machine
CN110286586A (en) A kind of MR damper hybrid modeling method
CN103105246A (en) Greenhouse environment forecasting feedback method of back propagation (BP) neural network based on improvement of genetic algorithm
Aladag et al. A new seasonal fuzzy time series method based on the multiplicative neuron model and SARIMA
CN106779177A (en) Multiresolution wavelet neutral net electricity demand forecasting method based on particle group optimizing
CN116542382A (en) Sewage treatment dissolved oxygen concentration prediction method based on mixed optimization algorithm
CN115841278B (en) Method, system, equipment and medium for evaluating running error state of electric energy metering device
CN110610308A (en) Method for evaluating environmental technology based on benchmarking method
CN106096723A (en) A kind of based on hybrid neural networks algorithm for complex industrial properties of product appraisal procedure
CN114707712A (en) Method for predicting requirement of generator set spare parts
CN114781692A (en) Short-term power load prediction method and device and electronic equipment
Karlov et al. Application of high-speed algorithms for training neural networks for forecasting financial markets
CN116187836A (en) Electric power material quality evaluation and spot check method
Khodadadi et al. Multi-objective moth-flame optimization algorithm for engineering problems
Varkeshi et al. Predicting the performance of Gorgan wastewater treatment plant using ANN-GA, CANFIS, and ANN models
Soltany Mahboob et al. Variable Length IPO and its application in concurrent design and train of ANFIS systems
Peng et al. Research on fault diagnosis method for transformer based on fuzzy genetic algorithm and artificial neural network
CN115185184A (en) Multi-parameter identification method for electro-hydraulic position servo system
CN107766887A (en) A kind of local weighted deficiency of data mixes clustering method
CN114638421A (en) Method for predicting requirement of generator set spare parts
Mottaghi-Kashtiban et al. Optimization of rational-powered membership functions using extended Kalman filter
Zhu Handwritten Chinese Characters Recognition Based on QPSO-BP

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
WD01 Invention patent application deemed withdrawn after publication

Application publication date: 20230505

WD01 Invention patent application deemed withdrawn after publication