CN117973206A - Method for identifying degradation state of optical fiber current transformer applicable to energy system - Google Patents
Method for identifying degradation state of optical fiber current transformer applicable to energy system Download PDFInfo
- Publication number
- CN117973206A CN117973206A CN202410155885.6A CN202410155885A CN117973206A CN 117973206 A CN117973206 A CN 117973206A CN 202410155885 A CN202410155885 A CN 202410155885A CN 117973206 A CN117973206 A CN 117973206A
- Authority
- CN
- China
- Prior art keywords
- degradation
- transformation ratio
- optical fiber
- current transformer
- mapping
- 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
Links
- 238000006731 degradation reaction Methods 0.000 title claims abstract description 176
- 230000015556 catabolic process Effects 0.000 title claims abstract description 175
- 239000013307 optical fiber Substances 0.000 title claims abstract description 124
- 238000000034 method Methods 0.000 title claims abstract description 61
- 230000009466 transformation Effects 0.000 claims abstract description 181
- 238000013507 mapping Methods 0.000 claims abstract description 148
- 230000003287 optical effect Effects 0.000 claims abstract description 26
- 238000013528 artificial neural network Methods 0.000 claims abstract description 18
- 239000000835 fiber Substances 0.000 claims description 35
- 238000012360 testing method Methods 0.000 claims description 32
- 230000008569 process Effects 0.000 claims description 26
- 238000004088 simulation Methods 0.000 claims description 26
- 241000282461 Canis lupus Species 0.000 claims description 25
- 230000008859 change Effects 0.000 claims description 23
- 238000012549 training Methods 0.000 claims description 21
- 230000010287 polarization Effects 0.000 claims description 17
- 238000004422 calculation algorithm Methods 0.000 claims description 13
- 241000282421 Canidae Species 0.000 claims description 10
- 238000005457 optimization Methods 0.000 claims description 5
- 238000007781 pre-processing Methods 0.000 claims description 5
- 238000012545 processing Methods 0.000 claims description 4
- 230000000694 effects Effects 0.000 claims description 3
- 238000012804 iterative process Methods 0.000 claims description 3
- 238000010606 normalization Methods 0.000 claims description 3
- 238000001228 spectrum Methods 0.000 claims description 3
- 238000012546 transfer Methods 0.000 claims description 3
- 238000012806 monitoring device Methods 0.000 abstract description 3
- 230000006870 function Effects 0.000 description 13
- 238000010276 construction Methods 0.000 description 10
- 238000010586 diagram Methods 0.000 description 4
- 238000011160 research Methods 0.000 description 3
- 238000006243 chemical reaction Methods 0.000 description 2
- 230000032683 aging Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000009413 insulation Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
Abstract
The invention discloses a degradation state identification method of an optical fiber current transformer suitable for an energy system, which comprises mapping modeling of degradation sources and output transformation ratios of the optical fiber current transformer and rapid identification of degradation states of a deep neural network based on mapping data; the mapping modeling module builds a mapping model of the transformation ratio and each degradation characteristic by building a relation between the transformation ratio and each degradation characteristic; the identification model module obtains each degradation characteristic and a transformation ratio data set through each mapping model on the basis of the mapping modeling module; and classifying the degradation characteristics of the data with the degradation characteristics based on the deep neural network to finish the identification of the degradation state of the optical fiber current transformer. The invention can quickly generate degradation data related to degradation characteristics of the optical fiber current transformer, can quickly identify degradation of a plurality of key optical devices on the premise of not changing original equipment or adding a monitoring device, has the advantages of low cost, easy operation and the like, and improves the operation efficiency of the optical fiber current transformer.
Description
Technical Field
The invention belongs to the technical field of optical fiber current transformers, and particularly relates to a degradation state identification method of an optical fiber current transformer applicable to an energy system.
Background
Compared with the traditional electromagnetic current transformer, the optical fiber current transformer (Fiber Optical Current Transformer, FOCT) has a series of advantages of good insulation property, wide frequency band, large dynamic range, high detection precision, small volume, light weight and the like, is more compatible with modern digital control and protection systems, and becomes an important research direction of the electronic current transformer. As the operating time of the optical fiber current transformer increases, the internal optical devices and electronic devices of the optical fiber current transformer can age, the optical fiber current transformer can gradually deteriorate from a good state until serious faults occur, and the aging and faults of the devices can be accelerated when the optical fiber current transformer operates in a severe environment. The degradation of the device is one of the main reasons of the operation fault of the optical CT field, but due to the lack of fault data and the research of the fault mechanism of the device, researchers cannot judge the fault type of the transformer accurately, so that the identification of the degraded device is difficult.
Disclosure of Invention
The technical problems to be solved are as follows: the invention provides a degradation state identification method of an optical fiber current transformer suitable for an energy system, which is used for generating various degradation data of the optical fiber current transformer and can be used for identifying degradation of key optical devices of 3 types of optical fiber current transformers, namely an SLD light source, a phase modulator or a photoelectric detector; the invention can quickly generate degradation data related to degradation characteristics of the optical fiber current transformer, can realize quick identification of degradation of a plurality of key optical devices on the premise of not changing original equipment or adding a monitoring device, has the advantages of low cost, easy operation and the like, and improves the operation efficiency of the optical fiber current transformer.
The technical scheme is as follows:
An identification method of the degradation state of an optical fiber current transformer suitable for an energy system, the identification method of the degradation state of the optical fiber current transformer comprises the following steps:
S1, analyzing degradation characteristics of key optical devices of an optical fiber current transformer, wherein the degradation characteristics comprise die temperature and driving current of an SLD light source, tail fiber crosstalk and half-wave voltage of a phase modulator, and detector power and detector noise of a photoelectric detector;
S2, constructing a mapping model between the transformation ratio of the optical fiber current transformer and each degradation characteristic of the key optical device by means of experimental simulation and fitting equations, and obtaining a degradation characteristic mapping model of the optical fiber current transformer; the degradation characteristic mapping model of the optical fiber current transformer comprises six mapping models, namely a transformation ratio-tube core temperature mapping model, a transformation ratio-driving current mapping model, a transformation ratio-tail fiber crosstalk mapping model, a transformation ratio-half-wave voltage mapping model, a transformation ratio-detector power mapping model and a transformation ratio-detector noise mapping model;
S3, acquiring each degradation characteristic and transformation ratio data set according to a degradation characteristic mapping model of the optical fiber current transformer, preprocessing each degradation characteristic and transformation ratio data set, and dividing all degradation characteristic and transformation ratio data sets into 7 types of degradation characteristic data sets of characteristic values, namely degradation of die temperature, degradation of driving current, degradation of polarization crosstalk, degradation of half-wave voltage, degradation of detector power, degradation of detector noise and no degradation; randomly scrambling a variable ratio data set of 7 types of degradation characteristics and then dividing the variable ratio data set into a training set and a testing set;
S4, training the training set based on the deep neural network to generate an initial identification model; and combining the test set, optimizing the super-parameter weights and bias items of the input layer, the hidden layer and the output layer of the initial identification model by using a gray wolf algorithm to obtain a rapid identification model of degradation characteristics under the optimal weight parameters, and identifying degradation data of the optical fiber current transformer by using the identification model so as to obtain specific degradation characteristics.
Further, in step S2, the process of constructing the transformation ratio-die temperature mapping model includes the following sub-steps:
S211, constructing and obtaining a functional relation between the transformation ratio K of the optical fiber current transformer and the central wavelength of the SLD light source, wherein the functional relation is as follows:
Wherein m is the number of D/A converters, N is the number of optical fiber turns of the optical fiber sensing ring, N is the refractive index of the sensing optical fiber, lambda is the center wavelength of the light source, and C is a constant;
s212, introducing a light source driver and a spectrum analyzer to acquire a change rule of the central wavelength of the SLD light source along with the die temperature, and establishing a fitting equation of the central wavelength of the light source and the die temperature by utilizing a Matlab optimizing tool box to acquire a relational expression between the central wavelength of the SLD light source and the die temperature of the SLD light source;
S213, constructing a mapping model of the transformation ratio and the die temperature based on a functional relation of the transformation ratio and the SLD light source central wavelength and a fitting equation of the SLD light source central wavelength and the die temperature, and obtaining a simulation curve of the relation between the transformation ratio and the die temperature;
S214, adjusting TEC temperature control current of the light source driver to change the temperature of the tube core and synchronously testing the transformation ratio data, and constructing a test curve of the relation between the transformation ratio and the temperature of the tube core; specifically, assuming that the relation function between the die temperature and the transformation ratio is k=at+b, parameters a and b are adjusted by comparing a simulation curve and a test curve of the relation between the transformation ratio and the die temperature, so that the mapping model of the transformation ratio and the die temperature is further optimized.
Further, in step S2, the process of constructing the transformation ratio-driving current mapping model includes the following sub-steps:
S221, selecting the driving current at the die temperature of 25 DEG as a threshold current I 0, and testing a plurality of correlation relations between the output power and the driving current under different temperature conditions by controlling the die temperature change; specifically, controlling the die temperature T to be constant, and obtaining the variation relation between the driving current and the output power at the die temperature by changing the driving current of the SLD light source and testing the output light power; the tube core temperature T is changed, the tube core temperature is changed within a normal temperature range, and the test is repeated, so that the change relation between the output power and the driving current corresponding to different tube core temperatures is obtained;
Fitting the influence of the die temperature to a temperature influence coefficient K F, and constructing an empirical formula between the SLD light source die output power P and the driving current I in a normal temperature range:
P＝M·[I-I0·KF)]+O
Wherein M, O is a fitting equation parameter, which can be solved by fitting software, K F is a temperature influence coefficient, and I 0 is a threshold current;
s222, constructing a transformation ratio-driving current fitting equation through a Matlab tool box according to transformation ratio and driving current data acquired under the actual application condition of the optical fiber current transformer in the earlier stage, further constructing a transformation ratio-driving current mapping model, and carrying out simulation according to the mapping relation to acquire simulation data between the transformation ratio and the driving current of the optical fiber current transformer;
s223, obtaining a change curve between the transformation ratio of the optical fiber current transformer and the driving current by adjusting the driving current of the SLD light source, comparing the change curve with a simulation curve between the transformation ratio of the optical fiber current transformer and the driving current, and optimizing a transformation ratio-driving current mapping model by adjusting an average wavelength normalization ratio coefficient and a driving current size adjustment coefficient in a transformation ratio-driving current fitting equation according to a comparison result;
S224, based on a correlation model of the driving current and the SLD light source output light power and a mapping model of the transformation ratio and the driving current, a functional relation between the transformation ratio of the optical fiber current transformer and the SLD light source output light power is established.
Further, in step S2, the process of constructing the transformation ratio-pigtail crosstalk mapping model includes the following sub-steps:
S231, constructing an analytical expression of the received interference light intensity of the photoelectric detector based on a Jones vector method, and mapping polarization crosstalk of the input and output tail fibers of the phase modulator with interference sub-items in the analytical expression;
S232, taking interference main peak and secondary peak not being aliased as targets, and acquiring the shortest length of an input tail fiber and an output tail fiber of the modulator based on the decoherence length of the SLD light source;
s233, adopting a scanning reflector to continuously scan and draw an interference pattern, fusing analysis mapping of polarization crosstalk to determine interference secondary peaks corresponding to the crosstalk of the input and output tail fibers in the interference pattern, and equating peak values of the secondary peaks to polarization crosstalk values;
s234, introducing a high-low temperature experimental box to simulate the temperature changing condition of the modulator, testing polarization crosstalk of the input and output tail fibers of the modulator at different temperatures, reconstructing a transformation ratio equation of the optical fiber current transformer, acquiring a temperature curve of the transformation ratio of the optical fiber current transformer when the crosstalk is changed, and constructing a transformation ratio-polarization crosstalk mapping model.
Further, in step S2, the process of constructing the transformation ratio-half-wave voltage mapping model includes the following sub-steps:
S241, establishing a correlation model of the transformation ratio and the half-wave voltage of the modulator according to the basic theory of the electro-optic effect and the transformation ratio process of the optical fiber current transformer;
s242, adjusting the height of the output ladder wave of the D/A converter in the signal processing unit circuit, changing the half-wave voltage value analog phase/voltage mismatch degradation of the phase modulator electrode, and obtaining the change rule of the transformation ratio along with different half-wave voltages;
S243, constructing a transformation ratio-half-wave voltage mapping model based on a transformation ratio variation rule with half-wave voltage obtained by a simulation experiment.
Further, in step S2, the process of constructing the transformation ratio-detector power mapping model includes the following sub-steps:
Introducing an adjustable optical fiber attenuator into an input tail fiber of the photoelectric detector, adjusting the adjustable optical fiber attenuator to change the receiving power of the photoelectric detector, and obtaining mapping data of a transformation ratio and the receiving power of the detector; a transformation ratio-detector power mapping model is constructed based on the mapping data.
Further, in step S2, the process of constructing the transformation ratio-detector noise mapping model includes the following sub-steps:
S251, simulating signals received by the photoelectric detector by adopting Gaussian white noise, 1/f fractal noise and an integral and slope function of the Gaussian white noise;
S252, a small deviation linearization method in a closed-loop control theory is adopted, physical models of all components in the optical fiber current transformer are fused, a system discrete transfer function is calculated according to the connection relation between the components, and a dynamic model of the optical fiber current transformer is constructed;
S253, the received signal of the superimposed noise obtained in the step S251 is used as a dynamic model input, mapping data of the transformation ratio of the optical fiber current transformer under different photoelectric detector noises are obtained through model output, and a transformation ratio-detector noise mapping model is constructed.
Further, in step S4, the process of optimizing the super-parameter weights and bias terms of the input layer, the hidden layer and the output layer of the initial recognition model by using the wolf algorithm includes the following steps:
S41, initializing parameters, wherein the initialized parameters comprise a wolf population, the dimension dim of wolf individual position information (the number of optimized parameters), the upper bound lb and the lower bound ub of the wolf dimension, and the maximum iteration number t max;
s42, taking weights and bias items u(ωih(1,1),···,ωih(i,m),θ1,1 1,···,θi,m 2,ωhh(1,1),···,ωhh(m,m),ωho(1,1),···,ωho(m,n),θ3,1 3,···,θ3,k 3) of an input layer, an hidden layer and an output layer in the deep neural network as initial positions of the wolves, namely positions of the wolves X(u)＝X(ωih(1,1),···,ωih(i,m),θ1,1 1,···,θi,m 2,ωhh(1,1),···,ωhh(m,m),ωho(1,1),···,ωho(m,n),θ3,1 3,···,θ3,k 3);
S43, taking the sum of the classification accuracy of the training set and the test set as a fitness function f (u), determining Xα(ωih(1,1),···,ωih(i,m),θ1,1 1,···,θi,m 2,ωhh(1,1),···,ωhh(m,m),ωho(1,1),···,ωho(m,n),θ3,1 3,···,θ3,k 3)、 when the fitness function is maximum, Xβ(ωih(1,1),···,ωih(i,m),θ1,1 1,···,θi,m 2,ωhh(1,1),···,ωhh(m,m),ωho(1,1),···,ωho(m,n),θ3,1 3,···,θ3,k 3)、 when the fitness function is second and Xδ(ωih(1,1),···,ωih(i,m),θ1,1 1,···,θi,m 2,ωhh(1,1),···,ωhh(m,m),ωho(1,1),···,ωho(m,n),θ3,1 3,···,θ3,k 3), when the fitness function is third according to the fitness, and calculating the forward position of the gray wolf according to the following formula:
X1＝Xα-A·|D·Xα-X|
X2＝Xβ-A·|D·Xβ-X|
X3＝Xδ-A·|D·Xδ-X|
Wherein X 1,X2,X3 is the forward position of the wolf, A is a random number between (-a, a), a is linearly reduced to 0 from 2 along with the iterative process, and D is a random number between (-1, 1);
Judging whether X 1、X2、X3 exceeds an upper limit or a lower limit, if yes, replacing the value of X 1、X2 or X 3 which is out of range with an upper limit lb or a lower limit ub, updating the position X (t+1) = (X 1+X2+X3)/3 of the gray wolves, wherein X (t+1) is the updated position of the gray wolves;
S44, judging the iteration times, namely if the iteration times are smaller than the maximum iteration times t max, returning to the step S42, continuing the next iteration until the algorithm is ended when the iteration times are greater than or equal to the maximum iteration times t max, and outputting Xα(ωih(1,1),···,ωih(i,m),θ1,1 1,···,θi,m 2,ωhh(1,1),···,ωhh(m,m),ωho(1,1),···,ωho(m,n),θ3,1 3,···,θ3,k 3),, wherein the weight and the bias term in the X α are the optimal.
Accordingly, an optical fiber current transformer degradation state identification system suitable for an energy system comprises:
The degradation characteristic module is used for analyzing degradation characteristics of key optical devices of the optical fiber current transformer, wherein the degradation characteristics comprise a die temperature and a driving current for an SLD light source, a tail fiber crosstalk and a half-wave voltage for a phase modulator, and a detector power and a detector noise for a photoelectric detector;
The degradation characteristic mapping module is used for constructing a mapping model between the transformation ratio of the optical fiber current transformer and each degradation characteristic of the key optical device by means of experimental simulation and fitting equations to obtain a degradation characteristic mapping model of the optical fiber current transformer; the degradation characteristic mapping model of the optical fiber current transformer comprises six mapping models, namely a transformation ratio-tube core temperature mapping model, a transformation ratio-driving current mapping model, a transformation ratio-tail fiber crosstalk mapping model, a transformation ratio-half-wave voltage mapping model, a transformation ratio-detector power mapping model and a transformation ratio-detector noise mapping model;
The transformation ratio data set dividing module is used for acquiring each degradation characteristic and transformation ratio data set according to the degradation characteristic mapping model of the optical fiber current transformer, preprocessing each degradation characteristic and transformation ratio data set, and dividing all degradation characteristic and transformation ratio data sets into 7 types of degradation characteristic values, namely degradation of die temperature, degradation of driving current, degradation of polarization crosstalk, degradation of half-wave voltage, degradation of detector power, degradation of detector noise and no degradation; randomly scrambling a variable ratio data set of 7 types of degradation characteristics and then dividing the variable ratio data set into a training set and a testing set;
The deep neural network module is used for training the training set based on the deep neural network to generate an initial identification model; and combining the test set, optimizing the super-parameter weights and bias items of the input layer, the hidden layer and the output layer of the initial identification model by using a gray wolf algorithm to obtain a rapid identification model of degradation characteristics under the optimal weight parameters, and identifying degradation data of the optical fiber current transformer by using the identification model to further obtain specific degradation characteristics.
The beneficial effects are that:
first, the method for identifying the degradation state of the optical fiber current transformer suitable for the energy system can construct a mapping model between the transformation ratio of the optical fiber current transformer and the degradation characteristics of the key optical device by using methods such as an intermediate variable method, a fitting equation method, an analog simulation experiment and the like, and can quickly and accurately generate various degradation data of the optical fiber current transformer based on various mapping models.
Secondly, the degradation state identification method of the optical fiber current transformer suitable for the energy system is used for monitoring the degradation of the optical device in the optical fiber current transformer, and the degradation identification of the optical device can be realized only by processing an output signal without changing original equipment or adding a monitoring device.
Drawings
FIG. 1 is a flowchart of a method for identifying a degradation state of an optical fiber current transformer suitable for an energy system according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of a fiber optic current transformer;
FIG. 3 is a schematic diagram of a method for constructing a degradation characteristic mapping model of an optical fiber current transformer;
Fig. 4 is a schematic diagram of a method for quickly identifying a degradation state of a deep neural network based on mapping data.
Detailed Description
The following examples will provide those skilled in the art with a more complete understanding of the invention, but are not intended to limit the invention in any way.
As shown in fig. 1, the method for identifying the degradation state of the optical fiber current transformer suitable for the energy system comprises the following steps:
Constructing a mapping model between the transformation ratio of the optical fiber current transformer and key degradation characteristics based on experimental simulation, fitting equations and other methods;
constructing a degradation characteristic mapping data set of the transformation ratio of the optical fiber current transformer based on each mapping model;
Preprocessing a transformation ratio data set, and dividing a degradation data set into seven types in sequence;
randomly scrambling seven types of degradation characteristic transformation ratio data sets, and dividing the seven types of degradation characteristic transformation ratio data sets into training sets and data sets according to a ratio of 9:1;
Training the data set based on a deep neural network algorithm;
And (3) carrying out weight parameter optimization on the initial identification model by using a gray wolf algorithm to obtain an optimal identification model for real-time degradation identification.
Fig. 2 is a schematic diagram of a fiber optic current transformer. The main degradation optical devices in the structure of the optical fiber current transformer are an SLD light source, a phase modulator and a detector.
The method for identifying the degradation state of the optical fiber current transformer suitable for the energy system in the embodiment comprises the steps of constructing an optical fiber current transformer degradation characteristic mapping model and identifying the degradation state based on mapping model data, as shown in fig. 2. By analyzing the degradation characteristics of the key optical device of the optical fiber current transformer, constructing a mapping model between the transformation ratio of the optical fiber current transformer and the degradation characteristics of the key optical device by means of experimental simulation, fitting equations and the like, generating data based on the mapping model, and realizing a rapid identification method of the degradation of the key optical device of the optical fiber current transformer by means of a deep neural network.
The degradation characteristic mapping model of the optical fiber current transformer comprises six mapping models, namely a transformation ratio-tube core temperature mapping model, a transformation ratio-driving current mapping model, a transformation ratio-tail fiber crosstalk mapping model, a transformation ratio-half-wave voltage mapping model, a transformation ratio-detector power mapping model and a transformation ratio-detector noise mapping model. Specifically, referring to fig. 3, the construction process of the degradation characteristic mapping model of the optical fiber current transformer includes:
Conversion ratio-die temperature mapping model
The construction process of the transformation ratio-die temperature mapping model comprises the following substeps:
A: the relation between the transformation ratio K of the optical fiber current transformer and the central wavelength of the light source is obtained through deduction:
where m is the number of D/A converters, N is the number of fiber turns of the fiber sensing ring, N is the refractive index of the sensing fiber, λ is the center wavelength of the light source, and C is a constant. Thus, the relation between the transformation ratio of the optical fiber current transformer and the central wavelength of the SLD light source can be constructed.
B: the change rule of the SLD light source center wavelength along with the die temperature is researched by introducing a light source driver and a spectrum analyzer, and a Matlab optimization tool box is utilized to establish a fitting equation of the light source center wavelength and the die temperature, so that the relation between the SLD light source center wavelength and the SLD light source die temperature is obtained.
C: based on the function relation between the transformation ratio and the SLD light source center wavelength and the fitting equation of the center wavelength and the die temperature, a mapping model of the transformation ratio and the die temperature can be constructed, a relation curve of the transformation ratio and the die temperature is obtained through simulation, the die temperature is changed by adjusting TEC temperature control current of a light source driver, transformation ratio data is synchronously tested, and the simulation and the test curve are compared to improve the construction precision of the mapping model, so that the high-precision construction of the transformation ratio-die temperature mapping model is completed. Specifically, through earlier work research, we can know that the relationship between the die temperature and the transformation ratio is approximately linear, so we can simply assume that the relationship between the die temperature and the transformation ratio is k=at+b, and then adjust parameters in the function, namely, assume parameters a or b in the equation, through comparing the simulation curve and the test curve of the relationship between the transformation ratio and the die temperature, so that the mapping model of the transformation ratio and the die temperature is further optimized.
(II) transformation ratio-output power mapping model
The construction process of the transformation ratio-output power mapping model comprises the following substeps:
A: the normal operating temperature is typically 25 °, so that the driving current at 25 ° can be selected as the threshold current I 0, and then the relationship between the output power and the driving current under different temperature conditions can be tested by controlling the die temperature variation. It can be divided into the following two steps: firstly, controlling a die temperature parameter T to be constant (for example, keeping the temperature T=20°), and then obtaining a change relation between the driving current and the output power at the die temperature by changing the driving current of the SLD light source and testing the output optical power; then we change the tube core temperature T to make the temperature change in the normal temperature range, repeat the above process to obtain the change relation between the output power and the driving current in the normal variation range of the tube core temperature, at this time we can obtain a plurality of correlation relations, we can fit the influence of the tube core temperature into the temperature influence coefficient K F, and then construct the empirical formula between the output power P and the driving current I of the SLD light source tube core in the normal temperature range:
P＝M·[I-I0·KF)]+O
Wherein M, O is a fitting equation parameter, which can be solved by fitting software, K F is a temperature influence coefficient, and I 0 is a threshold current;
B: the method comprises the steps of constructing a transformation ratio-driving current fitting equation through transformation ratio and driving current data acquired under the actual application condition of an optical fiber current transformer in the early stage, constructing a transformation ratio-driving current mapping model, and performing simulation through a mapping relation to acquire simulation data between the transformation ratio and the driving current of the optical fiber current transformer;
C: the variation curve between the transformation ratio of the optical fiber current transformer and the driving current is obtained by adjusting the driving current of the SLD light source, the variation curve is compared with the simulation curve between the transformation ratio of the optical fiber current transformer and the driving current, and the transformation ratio-driving current mapping model is optimized by adjusting the average wavelength normalization proportionality coefficient and the driving current size adjustment coefficient in the transformation ratio-driving current fitting equation according to the comparison result;
D: and establishing a functional relation between the transformation ratio of the optical fiber current transformer and the output optical power of the SLD light source based on a correlation model of the driving current and the output optical power of the SLD light source and a mapping model of the transformation ratio and the driving current.
(III) transformation ratio-tail fiber crosstalk mapping model
The construction process of the transformation ratio-tail fiber crosstalk mapping model comprises the following substeps:
A: and constructing an analytical expression of the received interference light intensity of the photoelectric detector based on a Jones vector method, and mapping the polarization crosstalk of the input and output tail fibers of the phase modulator with interference sub-terms in the analytical expression.
B: the shortest length calculation method for researching the input and output tail fibers of the modulator based on the decoherence length of the SLD light source aims at the interference main peak and the secondary peak and the non-aliasing of the secondary peak and the secondary peak.
C: and continuously scanning the scanning reflector, drawing an interference pattern, fusing analysis mapping of polarization crosstalk, determining interference secondary peaks corresponding to the crosstalk of the input and output tail fibers in the interference pattern, and equating the peak value of the secondary peaks to the polarization crosstalk value.
D: and finally, introducing a high-low temperature experimental box to simulate the temperature changing condition of the modulator, testing polarization crosstalk of the input and output tail fibers of the modulator at different temperatures, reconstructing a transformation ratio equation of the optical fiber current transformer, acquiring a temperature curve of the transformation ratio of the optical fiber current transformer when the crosstalk is changed, and constructing a transformation ratio-polarization crosstalk mapping model.
(IV) transformation ratio-half-wave voltage mapping model
The construction process of the transformation ratio-half-wave voltage mapping model comprises the following substeps:
A: and establishing a correlation model of the transformation ratio and the half-wave voltage of the modulator according to the basic theory of the electro-optic effect and the transformation ratio equation of the optical fiber current transformer.
B: the D/A converter in the signal processing unit circuit is adjusted to output the step wave height, the half-wave voltage value of the modulator electrode is changed to simulate the phase/voltage mismatch degradation, and the change rule of the transformation ratio along with different half-wave voltages is obtained.
C: and constructing a transformation ratio-half-wave voltage mapping model based on a transformation ratio variation rule of the simulation experiment along with half-wave voltage.
Fifth transformation ratio-detector power mapping model
The construction process of the transformation ratio-detector power mapping model comprises the following substeps:
A: an adjustable optical fiber attenuator is introduced into an input tail fiber of the photoelectric detector, the attenuator is adjusted to change the receiving power of the detector, and mapping data of the transformation ratio and the receiving power of the detector are obtained.
B: a transformation ratio-detector power mapping model is constructed based on the mapping data.
Sixth transformation ratio-detector noise mapping model
The construction process of the transformation ratio-detector noise mapping model comprises the following substeps:
a: the signal received by the detector is simulated by Gaussian white noise, 1/f fractal noise, and an integral and ramp function of Gaussian white noise.
B: the method is characterized in that a small deviation linearization method in a closed-loop control theory is adopted, a specific physical model and a working principle of each component in the optical fiber current transformer are fused, a discrete transfer function of a system is calculated according to a connection relation between the components, and then a dynamic model of the optical fiber current transformer is constructed.
C: and taking the received signal with superimposed noise as a dynamic model input, and obtaining mapping data of the transformation ratio of the optical fiber current transformer under different detector noises through model output so as to construct a transformation ratio-detector noise mapping model.
The fast identification of the degradation state of the deep neural network based on the mapping data refers to obtaining each degradation characteristic and transformation ratio data set according to a degradation characteristic mapping model of the optical fiber current transformer, generating an initial identification model based on the deep neural network, obtaining weights and bias items of layers (super parameters of an input layer, an hidden layer and an output layer) after the initial model is optimized, and utilizing the obtained super parameter optimal value reconstruction identification model to classify the degradation characteristic of the data with the degradation characteristic so as to complete the identification model construction. Specifically, referring to fig. 4, the process for fast identifying the degradation state of the deep neural network based on the mapping data specifically includes the following steps:
a: based on the data produced by the various mapping models mentioned above, a fiber optic current transformer transformation ratio-degradation characteristic mapping dataset is constructed.
B: the conversion ratio data set with different degradation characteristics is preprocessed, all degradation characteristics are divided into 7 types of characteristic values of degradation of die temperature, degradation of driving current, degradation of polarization crosstalk, degradation of half-wave voltage, degradation of detector power, degradation of detector noise and no degradation, and the 7 types of degradation characteristic categories are identified by numerals 1-7 in turn.
C: the variable ratio data set of the 7 types of degradation characteristics is randomly disordered and divided into two parts according to the proportion of 9:1, wherein one part is a training set, the other part is a test set, the total number of samples of the variable ratio data set is N, the number of samples of the training set is N1 (N1=9N/10), and the number of samples of the test set is N2 (N2=N/10).
D: and (3) model learning is carried out on the training set based on a deep neural network algorithm, the training set is input into transformation ratio data, the transformation ratio data are output into degradation characteristic categories 1-7, an initial degradation characteristic identification model is obtained, degradation characteristic prediction is carried out on the testing set by utilizing the initial identification model, and the accuracy of a predicted value and a true value is analyzed.
E: and carrying out weight parameter optimization on the initial identification model by adopting a gray wolf algorithm to obtain a rapid identification model of degradation characteristics under the optimal weight parameters, identifying degradation data of the optical fiber current transformer by using the identification model to obtain specific degradation characteristics, positioning a degradation optical device, and realizing rapid identification of key optical device degradation of the optical fiber current transformer.
The specific parameter optimization process comprises the following steps:
1) Initializing parameters, wherein the initialized parameters comprise a wolf population, the dimension dim (the number of optimized parameters) of individual position information of the wolves, the upper bound lb and the lower bound ub of the dimension of the wolves, and the maximum iteration number t max;
2) The weights and bias items u(ωih(1,1),···,ωih(i,m),θ1,1 1,···,θi,m 2,ωhh(1,1),···,ωhh(m,m),ωho(1,1),···,ωho(m,n),θ3,1 3,···,θ3,k 3) of the input layer, the hidden layer and the output layer in the deep neural network are used as the initial position of the wolf, namely the position of the wolf X(u)＝X(ωih(1,1),···,ωih(i,m),θ1,1 1,···,θi,m 2,ωhh(1,1),···,ωhh(m,m),ωho(1,1),···,ωho(m,n),θ3,1 3,···,θ3,k 3);
3) Taking the sum of classification accuracy of the training set and the test set as a fitness function f (u), determining Xα(ωih(1,1),···,ωih(i,m),θ1,1 1,···,θi,m 2,ωhh(1,1),···,ωhh(m,m),ωho(1,1),···,ωho(m,n),θ3,1 3,···,θ3,k 3)、 when the fitness function is maximum, Xβ(ωih(1,1),···,ωih(i,m),θ1,1 1,···,θi,m 2,ωhh(1,1),···,ωhh(m,m),ωho(1,1),···,ωho(m,n),θ3,1 3,···,θ3,k 3)、 when the fitness is second and Xδ(ωih(1,1),···,ωih(i,m),θ1,1 1,···,θi,m 2,ωhh(1,1),···,ωhh(m,m),ωho(1,1),···,ωho(m,n),θ3,1 3,···,θ3,k 3), when the fitness is third according to the fitness, and calculating the forward position of the gray wolf according to the following formula:
X1＝Xα-A·|D·Xα-X|
X2＝Xβ-A·|D·Xβ-X|
X3＝Xδ-A·|D·Xδ-X|
Wherein X 1,X2,X3 is the forward position of the wolf, A is a random number between (-a, a), a is linearly reduced to 0 from 2 along with the iterative process, and D is a random number between (-1, 1);
Judging whether X 1,X2,X3 exceeds the upper/lower limit, if yes, replacing the value of X 1,X2,X3 with an upper limit lb or a lower limit ub, updating the position X (t+1) = (X 1+X2+X3)/3 of the gray wolf, wherein X (t+1) is the updated position of the gray wolf;
4) And (3) judging the iteration times, namely if the iteration times are smaller than the maximum iteration times t max, repeating the step (2), continuing the next iteration until the algorithm is ended when the iteration times are greater than or equal to the maximum iteration times t max, and outputting Xα(ωih(1,1),···,ωih(i,m),θ1,1 1,···,θi,m 2,ωhh(1,1),···,ωhh(m,m),ωho(1,1),···,ωho(m,n),θ3,1 3,···,θ3,k 3),, wherein the weight and the bias term in the X α are optimal.
The foregoing is merely a preferred embodiment of the present invention, and is not intended to limit the present invention in any way, and all other embodiments obtained by those skilled in the art without making any inventive effort based on the embodiments of the present invention are within the scope of the present invention; any simple modification or equivalent variation of the above embodiments falls within the scope of the present invention.
Claims (9)
1. The method for identifying the degradation state of the optical fiber current transformer suitable for the energy system is characterized by comprising the following steps of:
S1, analyzing degradation characteristics of key optical devices of an optical fiber current transformer, wherein the degradation characteristics comprise die temperature and driving current of an SLD light source, tail fiber crosstalk and half-wave voltage of a phase modulator, and detector power and detector noise of a photoelectric detector;
S2, constructing a mapping model between the transformation ratio of the optical fiber current transformer and each degradation characteristic of the key optical device by means of experimental simulation and fitting equations, and obtaining a degradation characteristic mapping model of the optical fiber current transformer; the degradation characteristic mapping model of the optical fiber current transformer comprises six mapping models, namely a transformation ratio-tube core temperature mapping model, a transformation ratio-driving current mapping model, a transformation ratio-tail fiber crosstalk mapping model, a transformation ratio-half-wave voltage mapping model, a transformation ratio-detector power mapping model and a transformation ratio-detector noise mapping model;
S3, acquiring each degradation characteristic and transformation ratio data set according to a degradation characteristic mapping model of the optical fiber current transformer, preprocessing each degradation characteristic and transformation ratio data set, and dividing all degradation characteristic and transformation ratio data sets into 7 types of degradation characteristic data sets of characteristic values, namely degradation of die temperature, degradation of driving current, degradation of polarization crosstalk, degradation of half-wave voltage, degradation of detector power, degradation of detector noise and no degradation; randomly scrambling a variable ratio data set of 7 types of degradation characteristics and then dividing the variable ratio data set into a training set and a testing set;
S4, training the training set based on the deep neural network to generate an initial identification model; and combining the test set, optimizing the super-parameter weights and bias items of the input layer, the hidden layer and the output layer of the initial identification model by using a gray wolf algorithm to obtain a rapid identification model of degradation characteristics under the optimal weight parameters, and identifying degradation data of the optical fiber current transformer by using the identification model to further obtain specific degradation characteristics.
2. The method for identifying the degradation state of an optical fiber current transformer applicable to an energy system according to claim 1, wherein in step S2, the process of constructing the transformation ratio-die temperature mapping model comprises the following sub-steps:
S211, constructing and obtaining a functional relation between the transformation ratio K of the optical fiber current transformer and the central wavelength of the SLD light source, wherein the functional relation is as follows:
Wherein m is the number of D/A converters, N is the number of optical fiber turns of the optical fiber sensing ring, N is the refractive index of the sensing optical fiber, lambda is the center wavelength of the light source, and C is a constant;
s212, introducing a light source driver and a spectrum analyzer to acquire a change rule of the central wavelength of the SLD light source along with the die temperature, and establishing a fitting equation of the central wavelength of the light source and the die temperature by utilizing a Matlab optimizing tool box to acquire a relational expression between the central wavelength of the SLD light source and the die temperature of the SLD light source;
S213, constructing a mapping model of the transformation ratio and the die temperature based on a functional relation of the transformation ratio and the SLD light source central wavelength and a fitting equation of the SLD light source central wavelength and the die temperature, and obtaining a simulation curve of the relation between the transformation ratio and the die temperature;
S214, adjusting TEC temperature control current of the light source driver to change the temperature of the tube core and synchronously testing the transformation ratio data, and constructing a test curve of the relation between the transformation ratio and the temperature of the tube core; specifically, assuming that the relation function between the die temperature and the transformation ratio is k=at+b, parameters a and b are adjusted by comparing a simulation curve and a test curve of the relation between the transformation ratio and the die temperature, so that the mapping model of the transformation ratio and the die temperature is further optimized.
3. The method for identifying the degradation state of an optical fiber current transformer applicable to an energy system according to claim 1, wherein in step S2, the process of constructing the transformation ratio-driving current mapping model comprises the following sub-steps:
S221, selecting the driving current at the die temperature of 25 DEG as a threshold current I 0, and testing a plurality of correlation relations between the output power and the driving current under different temperature conditions by controlling the die temperature change; specifically, controlling the die temperature T to be constant, and obtaining the variation relation between the driving current and the output power at the die temperature by changing the driving current of the SLD light source and testing the output light power; the tube core temperature T is changed, the tube core temperature is changed within a normal temperature range, and the test is repeated, so that the change relation between the output power and the driving current corresponding to different tube core temperatures is obtained;
Fitting the influence of the die temperature to a temperature influence coefficient K F, and constructing an empirical formula between the SLD light source die output power P and the driving current I in a normal temperature range:
P＝M·[I-I0·KF)]+O
Wherein M, O is a fitting equation parameter, which can be solved by fitting software, K F is a temperature influence coefficient, and I 0 is a threshold current;
s222, constructing a transformation ratio-driving current fitting equation through a Matlab tool box according to transformation ratio and driving current data acquired under the actual application condition of the optical fiber current transformer in the earlier stage, further constructing a transformation ratio-driving current mapping model, and carrying out simulation according to the mapping relation to acquire simulation data between the transformation ratio and the driving current of the optical fiber current transformer;
s223, obtaining a change curve between the transformation ratio of the optical fiber current transformer and the driving current by adjusting the driving current of the SLD light source, comparing the change curve with a simulation curve between the transformation ratio of the optical fiber current transformer and the driving current, and optimizing a transformation ratio-driving current mapping model by adjusting an average wavelength normalization ratio coefficient and a driving current size adjustment coefficient in a transformation ratio-driving current fitting equation according to a comparison result;
S224, based on a correlation model of the driving current and the SLD light source output light power and a mapping model of the transformation ratio and the driving current, a functional relation between the transformation ratio of the optical fiber current transformer and the SLD light source output light power is established.
4. The method for identifying the degradation state of an optical fiber current transformer applicable to an energy system according to claim 1, wherein in the step S2, the process of constructing the transformation ratio-pigtail crosstalk mapping model comprises the following sub-steps:
S231, constructing an analysis expression of the received interference light intensity of the photoelectric detector based on a Jones vector method, and mapping polarization crosstalk of the input and output tail fibers of the phase modulator with interference sub-items in the analysis expression;
S232, taking interference main peak and secondary peak not being aliased as targets, and acquiring the shortest length of an input tail fiber and an output tail fiber of the modulator based on the decoherence length of the SLD light source;
s233, adopting a scanning reflector to continuously scan and draw an interference pattern, fusing analysis mapping of polarization crosstalk to determine interference secondary peaks corresponding to the crosstalk of the input and output tail fibers in the interference pattern, and equating peak values of the secondary peaks to polarization crosstalk values;
s234, introducing a high-low temperature experimental box to simulate the temperature changing condition of the modulator, testing polarization crosstalk of the input and output tail fibers of the modulator at different temperatures, reconstructing a transformation ratio equation of the optical fiber current transformer, acquiring a temperature curve of the transformation ratio of the optical fiber current transformer when the crosstalk is changed, and constructing a transformation ratio-polarization crosstalk mapping model.
5. The method for identifying the degradation state of an optical fiber current transformer applicable to an energy system according to claim 1, wherein in the step S2, the process of constructing the transformation ratio-half-wave voltage mapping model comprises the following sub-steps:
S241, establishing a correlation model of the transformation ratio and the half-wave voltage of the modulator according to the basic theory of the electro-optic effect and the transformation ratio process of the optical fiber current transformer;
s242, adjusting the height of the output ladder wave of the D/A converter in the signal processing unit circuit, changing the half-wave voltage value analog phase/voltage mismatch degradation of the phase modulator electrode, and obtaining the change rule of the transformation ratio along with different half-wave voltages;
S243, constructing a transformation ratio-half-wave voltage mapping model based on a transformation ratio variation rule with half-wave voltage obtained by a simulation experiment.
6. The method for identifying the degradation state of an optical fiber current transformer applicable to an energy system according to claim 1, wherein in step S2, the process of constructing the transformation ratio-detector power mapping model comprises the following sub-steps:
Introducing an adjustable optical fiber attenuator into an input tail fiber of the photoelectric detector, adjusting the adjustable optical fiber attenuator to change the receiving power of the photoelectric detector, and obtaining mapping data of a transformation ratio and the receiving power of the detector; a transformation ratio-detector power mapping model is constructed based on the mapping data.
7. The method for identifying the degradation state of an optical fiber current transformer applicable to an energy system according to claim 1, wherein in step S2, the process of constructing the transformation ratio-detector noise mapping model comprises the following sub-steps:
S251, simulating signals received by the photoelectric detector by adopting Gaussian white noise, 1/f fractal noise and an integral and slope function of the Gaussian white noise;
S252, a small deviation linearization method in a closed-loop control theory is adopted, physical models of all components in the optical fiber current transformer are fused, a system discrete transfer function is calculated according to the connection relation between the components, and a dynamic model of the optical fiber current transformer is constructed;
S253, the received signal of the superimposed noise obtained in the step S251 is used as a dynamic model input, mapping data of the transformation ratio of the optical fiber current transformer under different photoelectric detector noises are obtained through model output, and a transformation ratio-detector noise mapping model is constructed.
8. The method for identifying the degradation state of the optical fiber current transformer applicable to the energy system according to claim 1, wherein in the step S4, the process of optimizing the super-parameter weights and bias items of the input layer, the hidden layer and the output layer of the initial identification model by using the gray wolf algorithm comprises the following steps:
S41, initializing parameters, wherein the initialized parameters comprise the number of the wolf population, the dimension dim optimization parameters of the wolf individual position information, the upper bound lb and the lower bound ub of the wolf dimension, and the maximum iteration number t max;
s42, taking weights and bias items u(ωih(1,1),···,ωih(i,m),θ1,1 1,···,θi,m 2,ωhh(1,1),···,ωhh(m,m),ωho(1,1),···,ωho(m,n),θ3,1 3,···,θ3,k 3) of an input layer, an hidden layer and an output layer in the deep neural network as initial positions of the wolves, namely positions of the wolves X(u)＝X(ωih(1,1),···,ωih(i,m),θ1,1 1,···,θi,m 2,ωhh(1,1),···,ωhh(m,m),ωho(1,1),···,ωho(m,n),θ3,1 3,···,θ3,k 3);
S43, taking the sum of the classification accuracy of the training set and the test set as a fitness function f (u), determining Xα(ωih(1,1),···,ωih(i,m),θ1,1 1,···,θi,m 2,ωhh(1,1),···,ωhh(m,m),ωho(1,1),···,ωho(m,n),θ3,1 3,···,θ3,k 3)、 when the fitness function is maximum, Xβ(ωih(1,1),···,ωih(i,m),θ1,1 1,···,θi,m 2,ωhh(1,1),···,ωhh(m,m),ωho(1,1),···,ωho(m,n),θ3,1 3,···,θ3,k 3)、 when the fitness function is second and Xδ(ωih(1,1),···,ωih(i,m),θ1,1 1,···,θi,m 2,ωhh(1,1),···,ωhh(m,m),ωho(1,1),···,ωho(m,n),θ3,1 3,···,θ3,k 3), when the fitness function is third according to the fitness, and calculating the forward position of the gray wolf according to the following formula:
X1＝Xα-A·|D·Xα-X|
X2＝Xβ-A·|D·Xβ-X|
X3＝Xδ-A·|D·Xδ-X|
Wherein X 1,X2,X3 is the forward position of the wolf, A is a random number between (-a, a), a is linearly reduced to 0 from 2 along with the iterative process, and D is a random number between (-1, 1);
Judging whether X 1、X2、X3 exceeds an upper limit or a lower limit, if yes, replacing the value of X 1、X2 or X 3 which is out of range with an upper limit lb or a lower limit ub, updating the position X (t+1) = (X 1+X2+X3)/3 of the gray wolves, wherein X (t+1) is the updated position of the gray wolves;
S44, judging the iteration times, namely if the iteration times are smaller than the maximum iteration times t max, returning to the step S42, continuing the next iteration until the algorithm is ended when the iteration times are greater than or equal to the maximum iteration times t max, and outputting Xα(ωih(1,1),···,ωih(i,m),θ1,1 1,···,θi,m 2,ωhh(1,1),···,ωhh(m,m),ωho(1,1),···,ωho(m,n),θ3,1 3,···,θ3,k 3),, wherein the weight and the bias term in the X α are the optimal.
9. An optical fiber current transformer degradation state identification system suitable for an energy system, comprising:
The degradation characteristic module is used for analyzing degradation characteristics of key optical devices of the optical fiber current transformer, wherein the degradation characteristics comprise a die temperature and a driving current for an SLD light source, a tail fiber crosstalk and a half-wave voltage for a phase modulator, and a detector power and a detector noise for a photoelectric detector;
The degradation characteristic mapping module is used for constructing a mapping model between the transformation ratio of the optical fiber current transformer and each degradation characteristic of the key optical device by means of experimental simulation and fitting equations to obtain a degradation characteristic mapping model of the optical fiber current transformer; the degradation characteristic mapping model of the optical fiber current transformer comprises six mapping models, namely a transformation ratio-tube core temperature mapping model, a transformation ratio-driving current mapping model, a transformation ratio-tail fiber crosstalk mapping model, a transformation ratio-half-wave voltage mapping model, a transformation ratio-detector power mapping model and a transformation ratio-detector noise mapping model;
The transformation ratio data set dividing module is used for acquiring each degradation characteristic and transformation ratio data set according to the degradation characteristic mapping model of the optical fiber current transformer, preprocessing each degradation characteristic and transformation ratio data set, and dividing all degradation characteristic and transformation ratio data sets into 7 types of degradation characteristic values, namely degradation of die temperature, degradation of driving current, degradation of polarization crosstalk, degradation of half-wave voltage, degradation of detector power, degradation of detector noise and no degradation; randomly scrambling a variable ratio data set of 7 types of degradation characteristics and then dividing the variable ratio data set into a training set and a testing set;
The deep neural network module is used for training the training set based on the deep neural network to generate an initial identification model; and combining the test set, optimizing the super-parameter weights and bias items of the input layer, the hidden layer and the output layer of the initial identification model by using a gray wolf algorithm to obtain a rapid identification model of degradation characteristics under the optimal weight parameters, and identifying degradation data of the optical fiber current transformer by using the identification model to further obtain specific degradation characteristics.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310549240.6A CN116776549A (en) | 2023-05-16 | 2023-05-16 | Method for identifying degradation state of optical fiber current transformer applicable to energy system |
| CN2023105492406 | 2023-05-16 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117973206A true CN117973206A (en) | 2024-05-03 |
Family
ID=87993945
Family Applications (2)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310549240.6A Pending CN116776549A (en) | 2023-05-16 | 2023-05-16 | Method for identifying degradation state of optical fiber current transformer applicable to energy system |
| CN202410155885.6A Pending CN117973206A (en) | 2023-05-16 | 2024-02-04 | Method for identifying degradation state of optical fiber current transformer applicable to energy system |
Family Applications Before (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310549240.6A Pending CN116776549A (en) | 2023-05-16 | 2023-05-16 | Method for identifying degradation state of optical fiber current transformer applicable to energy system |
Country Status (1)
| Country | Link |
|---|---|
| CN (2) | CN116776549A (en) |
-
2023
- 2023-05-16 CN CN202310549240.6A patent/CN116776549A/en active Pending
-
2024
- 2024-02-04 CN CN202410155885.6A patent/CN117973206A/en active Pending
Also Published As
| Publication number | Publication date |
|---|---|
| CN116776549A (en) | 2023-09-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Lu et al. | Construction of prediction intervals for carbon residual of crude oil based on deep stochastic configuration networks | |
| CN111860982A (en) | Wind power plant short-term wind power prediction method based on VMD-FCM-GRU | |
| CN109902745A (en) | A kind of low precision training based on CNN and 8 integers quantization inference methods | |
| CN110956312B (en) | Photovoltaic power distribution network voltage prediction method based on EMD-CNN deep neural network | |
| Zilong et al. | A novel immune evolutionary algorithm incorporating chaos optimization | |
| GB2610319A (en) | Automatic compression method and platform for multilevel knowledge distillation-based pre-trained language model | |
| CN112966667A (en) | Method for identifying one-dimensional distance image noise reduction convolution neural network of sea surface target | |
| CN114201987A (en) | Active interference identification method based on self-adaptive identification network | |
| CN112560342A (en) | DNN-based atmospheric waveguide parameter estimation method | |
| Cheng et al. | Two novel reconstruction methods of sparsity adaptive adjustment for road roughness compressive signal based on I-SA and GSM | |
| CN110070004A (en) | A kind of field hyperspectrum Data expansion method applied to deep learning | |
| CN113536662A (en) | Electronic transformer error state prediction method based on firefly optimized LightGBM algorithm | |
| CN111539157B (en) | Dynamic load time domain identification method based on multilayer perceptron | |
| CN117973206A (en) | Method for identifying degradation state of optical fiber current transformer applicable to energy system | |
| CN113222250A (en) | High-power laser device output waveform prediction method based on convolutional neural network | |
| Cheng et al. | SensorFormer: Efficient many-to-many sensor calibration with learnable input subsampling | |
| CN106067065A (en) | Spectrum analysis Model transfer technology based on radial base neural net | |
| CN115014404A (en) | High-precision high-speed fiber grating demodulation method based on deep learning | |
| Liao et al. | Intra-pulse modulation recognition of radar signals based on multi-feature random matching fusion network | |
| CN111506868B (en) | Ultra-short-term wind speed prediction method based on HHT weight optimization | |
| CN116187206B (en) | COD spectrum data migration method based on generation countermeasure network | |
| CN113113776B (en) | Terahertz composite metamaterial multi-band absorber and bidirectional design method | |
| AU2021101713A4 (en) | Remote sensing image building recognition method based on activated representational substitution | |
| Chang et al. | Predicting Aerosol Extinction Coefficient With LiDAR Data Based on Deep Belief Network | |
| Ji et al. | Regression Prediction of Effective Temperature of Stellar Spectrum Based on Kernel Method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication |