CN115993345A - SF based on ISFO-VMD-KELM 6 Decomposition component CO 2 Concentration inversion method - Google Patents
SF based on ISFO-VMD-KELM 6 Decomposition component CO 2 Concentration inversion method Download PDFInfo
- Publication number
- CN115993345A CN115993345A CN202210967017.9A CN202210967017A CN115993345A CN 115993345 A CN115993345 A CN 115993345A CN 202210967017 A CN202210967017 A CN 202210967017A CN 115993345 A CN115993345 A CN 115993345A
- Authority
- CN
- China
- Prior art keywords
- flagfish
- optimizer
- concentration
- kelm
- improved
- 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
- 238000000354 decomposition reaction Methods 0.000 title claims abstract description 78
- 238000000034 method Methods 0.000 title claims abstract description 62
- 241001137309 Jordanella floridae Species 0.000 claims abstract description 70
- 238000000862 absorption spectrum Methods 0.000 claims abstract description 30
- 238000001514 detection method Methods 0.000 claims abstract description 10
- 238000002474 experimental method Methods 0.000 claims abstract description 8
- 238000007781 pre-processing Methods 0.000 claims abstract description 8
- 235000019512 sardine Nutrition 0.000 claims description 60
- 241001125048 Sardina Species 0.000 claims description 57
- 238000004422 calculation algorithm Methods 0.000 claims description 27
- 230000006870 function Effects 0.000 claims description 21
- 230000003044 adaptive effect Effects 0.000 claims description 19
- 238000009826 distribution Methods 0.000 claims description 19
- 238000005457 optimization Methods 0.000 claims description 14
- 230000000739 chaotic effect Effects 0.000 claims description 4
- 241000555825 Clupeidae Species 0.000 claims description 3
- 238000001228 spectrum Methods 0.000 abstract description 13
- 238000010586 diagram Methods 0.000 description 13
- 238000010521 absorption reaction Methods 0.000 description 12
- 238000012549 training Methods 0.000 description 10
- 238000012360 testing method Methods 0.000 description 8
- 238000004590 computer program Methods 0.000 description 7
- 238000005516 engineering process Methods 0.000 description 7
- 230000008569 process Effects 0.000 description 6
- 238000012545 processing Methods 0.000 description 6
- 238000004847 absorption spectroscopy Methods 0.000 description 5
- 230000000694 effects Effects 0.000 description 5
- 238000003384 imaging method Methods 0.000 description 5
- 238000012986 modification Methods 0.000 description 5
- 230000004048 modification Effects 0.000 description 5
- 238000001914 filtration Methods 0.000 description 4
- 239000004973 liquid crystal related substance Substances 0.000 description 4
- 239000011810 insulating material Substances 0.000 description 3
- 239000002245 particle Substances 0.000 description 3
- 238000003860 storage Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 210000003495 flagella Anatomy 0.000 description 2
- 210000000540 fraction c Anatomy 0.000 description 2
- 230000002068 genetic effect Effects 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 230000035772 mutation Effects 0.000 description 2
- 239000004065 semiconductor Substances 0.000 description 2
- 238000000926 separation method Methods 0.000 description 2
- 238000004611 spectroscopical analysis Methods 0.000 description 2
- 241000252229 Carassius auratus Species 0.000 description 1
- 108010040721 Flagellin Proteins 0.000 description 1
- 238000002835 absorbance Methods 0.000 description 1
- 230000001133 acceleration Effects 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 230000004913 activation Effects 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000002238 attenuated effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000003763 carbonization Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000005336 cracking Methods 0.000 description 1
- 239000003822 epoxy resin Substances 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000011049 filling Methods 0.000 description 1
- 238000002329 infrared spectrum Methods 0.000 description 1
- 238000009413 insulation Methods 0.000 description 1
- 239000012212 insulator Substances 0.000 description 1
- 238000001285 laser absorption spectroscopy Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000002156 mixing Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 229920000647 polyepoxide Polymers 0.000 description 1
- 244000062645 predators Species 0.000 description 1
- 238000004445 quantitative analysis Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 238000010183 spectrum analysis Methods 0.000 description 1
- 238000004800 variational method Methods 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N21/25—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
- G01N21/31—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry
- G01N21/39—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using tunable lasers
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Software Systems (AREA)
- General Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Molecular Biology (AREA)
- Data Mining & Analysis (AREA)
- Mathematical Physics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Computing Systems (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Chemical & Material Sciences (AREA)
- Optics & Photonics (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The invention discloses an SF based on an ISFO-VMD-KELM 6 Decomposition component CO 2 The concentration inversion method comprises the steps of establishing an improved flagfish optimizer based on a traditional flagfish optimizer; collecting an original absorption spectrum, and preprocessing the original absorption spectrum; according to the pretreatment result, a nuclear extreme learning machine optimized by combining an improved flagfish optimizer is used for establishing the second harmonic amplitude and CO of the absorption spectrum 2 An inversion model of concentration; CO from inversion model 2 Detection experiment, judge SF 6 An operating state of the electrical device. The invention filters the original spectrum signal by adopting the self-adaptive variational modal decomposition combined wavelet threshold method optimized by the improved flagfish optimizer, removes high-frequency noise in the original signal, and then passes through SF 6 Background subtractionThe amplitude of the spectrum can then be read more accurately. Establishing CO using improved flagfish optimizer optimized nuclear extreme learning machine 2 Compared with the traditional concentration inversion method, the concentration inversion model has higher precision and stability.
Description
Technical Field
The invention relates to SF 6 The technical field of detection of internal decomposition components of electrical equipment, in particular to an SF based on ISFO-VMD-KELM 6 Decomposition component CO 2 A method for concentration inversion.
Background
SF 6 Solid insulating parts such as GIS basin-type insulators in electrical equipment are formed by pouring organic insulating materials such as epoxy resin, and the organic insulating materials can be gradually cracked and carbonized under the action of local high temperature of partial discharge or overheat faults, so that the safety of the whole equipment is threatened. CO 2 The gas is the decomposition product of the organic insulating material in the cracking and carbonization process, and the CO is detected 2 The volume fraction of the gas can find SF in time 6 Insulation failure of electrical equipment. In recent years, tunable laser absorption spectroscopy (tunable absorption spectroscopy) has been widely used in the power industry. The tunable absorption spectrum technology utilizes the characteristic of narrow linewidth of a tunable semiconductor laser, and realizes quantitative analysis of target gas by observing the absorption spectrum of the target gas to specific laser, so that the traditional SF 6 The method for inverting the concentration of the decomposition product is to establish a least square method linear or nonlinear equation by utilizing the linear relation between the volume fraction of the target gas and the absorption light intensity.
Along with machine learningMore and more research is applied to the technology of tunable absorption spectroscopy technology by machine learning. In the prior art, an extreme learning machine (Extreme Learning Machine, ELM) is adopted to establish a model of harmonic signals and integral absorbance, but the ELM has strong dependence on parameters and is easy to sink into a local minimum. Still other prior art uses BP neural network to build O 2 The gas concentration inversion model applies a classical genetic algorithm to the parameter adjustment problem of the BP network, but the BP network has the problems of slow learning process and poor stability, and the genetic algorithm has weaker local searching capability and is easy to converge prematurely. Some adopt particle swarm optimization algorithm to optimize Kernel Extreme Learning Machine (KELM), and build CO with higher precision 2 Concentration inversion models, but particle swarm optimization algorithms tend to fall into locally optimal solutions.
Disclosure of Invention
This section is intended to outline some aspects of embodiments of the invention and to briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section as well as in the description summary and in the title of the application, to avoid obscuring the purpose of this section, the description summary and the title of the invention, which should not be used to limit the scope of the invention.
The present invention has been made in view of the above-described problems occurring in the prior art.
Therefore, the invention provides an SF based on ISFO-VMD-KELM 6 Decomposition component CO 2 The concentration inversion method can solve the problems that a learning machine has strong dependence on parameters, is easy to sink into a local minimum value, has weak local searching capability, is easy to converge prematurely, and a particle swarm optimization algorithm is easy to sink into a local optimal solution.
In order to solve the technical problems, the invention provides the following technical scheme that the SF based on the ISFO-VMD-KELM 6 Decomposition component CO 2 A method of concentration inversion comprising:
based on the traditional flagfish optimizer, improving optimizing precision and local searching capability, and establishing an improved flagfish optimizer;
collecting an original absorption spectrum, and preprocessing the original absorption spectrum by combining a variation modal decomposition algorithm optimized by the improved flagfish optimizer with a wavelet threshold method;
according to the pretreatment result, a second harmonic amplitude and CO of an absorption spectrum are established by combining the nuclear extreme learning machine optimized by the improved flagfish optimizer 2 An inversion model of concentration;
CO from the inversion model 2 Detection experiment, judge SF 6 An operating state of the electrical device.
SF based on ISFO-VMD-KELM as described in the present invention 6 Decomposition component CO 2 A preferred embodiment of the method for concentration inversion, wherein: the improved flagfish optimizer includes,
setting algorithm control parameters, and generating initial flagelliforme and sardine populations according to a multi-strategy initialization mode;
calculating a fitness value, and recording global optimal positions of the flagfish and the sardine during the iteration;
updating the position of the sardine according to the attack strength;
respectively carrying out cauchy variation and adaptive t distribution variation on elite individuals of the flagelliforme and the sardine;
comparing the optimal solutions of the flagfish and the sardine, and replacing the positions of the flagfish and the sardine according to the result;
and if the iteration is not finished, calculating the fitness value to continue optimizing, and if the iteration is finished, returning a final result.
SF based on ISFO-VMD-KELM as described in the present invention 6 Decomposition component CO 2 A preferred embodiment of the method for concentration inversion, wherein: the improved flagfish optimizer further includes,
the Tent chaotic sequence is adopted to improve the initial population distribution of the flagfish and the sardine, and the mathematical expression is as follows:
where β is a random number of [0,1 ].
SF based on ISFO-VMD-KELM as described in the present invention 6 Decomposition component CO 2 A preferred embodiment of the method for concentration inversion, wherein: the improved flagfish optimizer further includes,
the mathematical expression of the improved sardine position update is as follows:
Levy＝u/|v| 1/β
wherein the mathematical expression of σ is:
wherein u to N (0, sigma) 2 ) v-N (0, 1), beta is [0,2]Is a random number of (a) in the memory.
SF based on ISFO-VMD-KELM as described in the present invention 6 Decomposition component CO 2 A preferred embodiment of the method for concentration inversion, wherein: the improved flagfish optimizer further comprises optimizing parameters of the adaptive variation modal decomposition and the nuclear extreme learning machine, and performing cauchy variation and adaptive t distribution variation on elite individuals of the flagfish and sardine.
SF based on ISFO-VMD-KELM as described in the present invention 6 Decomposition component CO 2 A preferred embodiment of the method for concentration inversion, wherein: the inversion model comprises the steps of,
initializing basic parameters of an improved flagfish optimizer, and taking the sum of multi-scale permutation entropies of adaptive variation modal decomposition as an fitness function;
generating an initial population, and setting the population proportion of the flagfish and the sardine;
the position information of each group is used as a parameter to be imported into a fitness function to calculate multi-scale arrangement entropy, and the minimum multi-scale arrangement entropy value and the corresponding position vector are used as global solutions and global optimal positions;
updating the positions of the flagstones and the sardines, recalculating the fitness value, and updating the global solution and the global optimal position;
and performing optimization iteration until the maximum iteration times are reached, taking the global optimal position at the last iteration as the optimal parameter of the adaptive variation modal decomposition, and performing the adaptive variation modal decomposition under the parameter condition.
SF based on ISFO-VMD-KELM as described in the present invention 6 Decomposition component CO 2 A preferred embodiment of the method for concentration inversion, wherein: the inversion model further comprises, using a nuclear extreme learning machine to build CO 2 And optimizing regularization coefficients and nuclear parameters of the nuclear extreme learning machine by utilizing an improved flagfish optimizer algorithm, wherein the fitness function is root mean square error of the predicted concentration and the true concentration.
SF based on ISFO-VMD-KELM as described in the present invention 6 Decomposition component CO 2 A preferred embodiment of the method for concentration inversion, wherein: the preprocessing comprises the step of preprocessing an original absorption spectrum by utilizing the adaptive variation modal decomposition after parameter optimization by the improved flagfish optimizer and combining a wavelet threshold method.
SF based on ISFO-VMD-KELM as described in the present invention 6 Decomposition component CO 2 A preferred embodiment of the method for concentration inversion, wherein: the adaptive variational modal decomposition includes,
wherein f (t) is a waveform function, i.e. time sequence, t is time, 1s of waveform is acquired, f 1 (t) is a 40Hz cosine signal of amplitude 1, f 2 (t) is an intermittent sinusoidal signal of amplitude 3 at 150Hz, f 3 And (t) is 20dB of Gaussian white noise.
SF based on ISFO-VMD-KELM as described in the present invention 6 Decomposition component CO 2 A preferred embodiment of the method for concentration inversion, wherein: the inversion model includes taking the second harmonic peak and the gas concentration as a single input and a single output, respectively, of the built model.
The invention has the beneficial effects that: the invention provides an SF based on an ISFO-VMD-KELM 6 Grouping of separationsCO separation 2 The invention adopts the adaptive variational modal decomposition combined wavelet threshold method optimized by the improved flagfish optimizer to filter the original spectrum signal, removes high-frequency noise in the original signal, and then passes through SF 6 The amplitude of the spectrum can be read more accurately after background subtraction. Establishing CO using improved flagfish optimizer optimized nuclear extreme learning machine 2 Compared with the traditional concentration inversion method, the concentration inversion model has higher precision and stability.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings that are needed in the description of the embodiments will be briefly described below, it being obvious that the drawings in the following description are only some embodiments of the present invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art. Wherein:
FIG. 1 shows an SF based on an ISFO-VMD-KELM according to one embodiment of the present invention 6 Decomposition component CO 2 A method flow chart of concentration inversion;
FIG. 2 shows an SF based on ISFO-VMD-KELM according to one embodiment of the present invention 6 Decomposition component CO 2 A self-adaptive variation modal decomposition modal diagram of a concentration inversion method;
FIG. 3 shows an SF based on ISFO-VMD-KELM according to one embodiment of the present invention 6 Decomposition component CO 2 CO of method of concentration inversion 2 An absorption spectrum filtering front-back comparison chart;
FIG. 4 shows an SF based on ISFO-VMD-KELM according to one embodiment of the present invention 6 Decomposition component CO 2 SF of concentration inversion method 6 An absorption spectrum filtering front-back comparison chart;
FIG. 5 is a schematic diagram of an SF based on ISFO-VMD-KELM according to one embodiment of the present invention 6 Decomposition component CO 2 An absorption spectrum diagram after background subtraction of a concentration inversion method;
FIG. 6 is a diagram of an SF based on ISFO-VMD-KELM according to one embodiment of the present invention 6 Grouping of separationsCO separation 2 An absorption spectrum diagram after background subtraction of a concentration inversion method;
FIG. 7 is a diagram of an SF based on ISFO-VMD-KELM according to one embodiment of the present invention 6 Decomposition component CO 2 Linear fitting by a least square method of a concentration inversion method;
Detailed Description
So that the manner in which the above recited objects, features and advantages of the present invention can be understood in detail, a more particular description of the invention, briefly summarized above, may be had by reference to the embodiments, some of which are illustrated in the appended drawings. All other embodiments, which can be made by one of ordinary skill in the art based on the embodiments of the present invention without making any inventive effort, shall fall within the scope of the present invention.
In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention, but the present invention may be practiced in other ways other than those described herein, and persons skilled in the art will readily appreciate that the present invention is not limited to the specific embodiments disclosed below.
Further, reference herein to "one embodiment" or "an embodiment" means that a particular feature, structure, or characteristic can be included in at least one implementation of the invention. The appearances of the phrase "in one embodiment" in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments.
While the embodiments of the present invention have been illustrated and described in detail in the drawings, the cross-sectional view of the device structure is not to scale in the general sense for ease of illustration, and the drawings are merely exemplary and should not be construed as limiting the scope of the invention. In addition, the three-dimensional dimensions of length, width and depth should be included in actual fabrication.
Also in the description of the present invention, it should be noted that the orientation or positional relationship indicated by the terms "upper, lower, inner and outer", etc. are based on the orientation or positional relationship shown in the drawings, are merely for convenience of describing the present invention and simplifying the description, and do not indicate or imply that the apparatus or elements referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus should not be construed as limiting the present invention. Furthermore, the terms "first, second, or third" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance.
The terms "mounted, connected, and coupled" should be construed broadly in this disclosure unless otherwise specifically indicated and defined, such as: can be fixed connection, detachable connection or integral connection; it may also be a mechanical connection, an electrical connection, or a direct connection, or may be indirectly connected through an intermediate medium, or may be a communication between two elements. The specific meaning of the above terms in the present invention will be understood in specific cases by those of ordinary skill in the art.
Example 1
Referring to FIGS. 1-2, a first embodiment of the present invention provides an SF based on ISFO-VMD-KELM 6 Decomposition component CO 2 A method of concentration inversion comprising:
s1: based on the traditional flagfish optimizer, improving optimizing precision and local searching capability, and establishing an improved flagfish optimizer;
still further, the flagelliformer (SFO) is inspired by the natural phenomenon that flagelliformes predate sardines in marine ecology: the flagellates adopt a group-cooperation hunting mode, and the sardine group is forced to float up to the water surface by driving, so that the hunting is unfolded, and the damaged sardine is continuously separated from the team under the attack of the flagellates, so that the flagellates can catch the flagellates.
It should be noted that the position update mathematical expression of the flagelliforme as predator is as follows:
where i represents the current number of iterations,
bit representing updated flag fishPut (I) at>
Representing the position of the head flagellum, which is the flagellum currently possessing the optimal fitness value, < + >>
Indicating the location of an injured sardine, which is currently sardine with the best fitness value, ++>
Represents the current position of the flagpole, r represents a random number between (0, 1), lambda i Representing the iteration coefficients.
It should be noted that the iteration coefficient is related to sardine population density, λ i The mathematical expression is as follows:
λ i ＝2×r×PD-PD
wherein PD represents the density of sardine population, and the PD mathematical expression is as follows:
wherein N is SF 、N S The population numbers of flagelliforme and sardine in each iteration are shown, respectively.
Furthermore, the movement mode of sardine is related to the attack force of flagfish, and the position updating mathematical expression is as follows:
wherein, the liquid crystal display device comprises a liquid crystal display device,
indicating the updated position of sardine, < ->
Representing the current sardine position, AP represents the attack strength of the flagfish at each iteration, and the mathematical expression of AP is as follows:
AP＝A×(1-(2×i×ε))
wherein A represents the maximum attack intensity of the flagfish, epsilon represents the attenuation coefficient, and the attack intensity of the flagfish linearly attenuates from the maximum value A to 0 in the iteration process.
It should be noted that when AP is more than or equal to 0.5, all sardine should update the current position to go to the safe area, and when AP is less than 0.5, only part of sardine is selected to update the current position. The number alpha and the dimension beta need to be considered in partial update, and the update scope is defined as follows:
wherein d i The dimension at the ith iteration is represented, the number of parameters to be optimized is the dimension number, d represents the dimension, i.e. the d parameter, i represents the iteration number, and di represents the d dimension parameter (variable) at the ith iteration. .
Further, the fitness values of the two populations are compared at each iteration and the positions of the flag fish and sardine are replaced according to the formula.
Wherein f (SF) i ) Indicating the fitness value of the current flagfish, f (S i ) And representing the fitness value of the current sardine.
Furthermore, the initial population is generated by the flagelliforme optimizer in a pseudo-random mode, the method cannot ensure the uniformity of population distribution, and the search state is easy to be in stagnation when the optimal solution is not found. The chaos has the characteristics of certainty and randomness, and fig. 2 shows probability distribution histograms of four common chaos sequences, wherein the distribution of the Tent sequence on [0,1] is most uniform, so that the invention adopts the Tent chaos sequence to improve the initial population distribution of the flagfish and sardine, and the mathematical expression is as follows:
wherein, beta is a random number of [0,1], and the value of the invention is 0.7.
Furthermore, the method is further combined with a lens imaging learning mechanism to screen a high-quality population on the basis of obtaining an initial population. The initial population of the flagelliforme and the sardine is generated according to the following steps:
further, generating N population individuals by utilizing the Tent chaotic map, marking the population individuals as a population A, and calculating the fitness value of the population A;
further, using the above formula, the lens imaging solutions of all individuals in the population a are obtained, recorded as the population B, and the fitness value thereof is calculated:
wherein x is i Representing the current solution of the current value,
the mirror solution representing the current solution, N populations (numerical values) are obtained according to the above, a and b represent the minimum and maximum values of the populations, 1 is taken according to the empirical value k, a i And b i Representing the minimum and maximum boundaries of the population, respectively.
It should be noted that, comparing fitness values, if the lens imaging solution is better than its original solution, the original solution in population a is replaced with the corresponding lens imaging solution in population B, and finally N new initial populations are formed.
Furthermore, the sardine updating mode is insufficient in searching strength in the later iteration period, so that the sardine updating mode is easy to sink into a local optimal solution.
It should be noted that, the Levy flight strategy is to make position mutation movement of corresponding distance according to the change of random step length, so in order to increase the searching range of sardine and improve the optimizing performance, the invention introduces Levy random step length into the position updating mode of sardine, increases the searching range of sardine, and the sardine position updating mathematical expression after improvement is insufficient in searching strength in the later stage of iteration in the sardine updating mode of standard flagellin optimizing algorithm, thus being easy to sink into local optimal solution, therefore, the invention improves the sardine position updating mode, and introduces a self-adaptive feedback factor to accelerate the iterative convergence process.
Further, when AP is more than or equal to 0.5, the improved sardine position updating mathematical expression is as follows:
Levy＝u/|v| 1/β
wherein the mathematical expression of σ is:
wherein u to N (0, sigma) 2 ) v-N (0, 1), beta is [0,2]Is 1.5 in this patent.
Furthermore, the optimal solution is used as elite, and the mutation operation is carried out according to the following steps:
cauchy variation was performed on elite individuals of flagelliforme and sardine, respectively, and the mathematical expression was as follows:
wherein cauchy (0, 1) is a 0-1 random number obeying the Cauchy distribution,
representing elite solution->
Solution after cauchy variation.
The adaptive t distribution variation is carried out on individual Chinese goldfish and sardine elite, and the mathematical expression is as follows:
wherein, the liquid crystal display device comprises a liquid crystal display device,
representing elite solution->
T (i) represents student t distribution, and the degree of freedom is the iteration number of the algorithm. />
It should be noted that the elite solution and its variant solution are compared
And (3) selecting a better variant solution to replace a corresponding elite solution. Student t-distribution (Student's t-distribution), which may be referred to simply as t-distribution. Elite solution is defined herein as the optimal solution (the position vector with optimal fitness value is called elite by fitness function value calculation).
Further, the steps of improving the flagelliforme optimizer are as follows:
setting algorithm control parameters, and generating initial flagelliforme and sardine populations according to a multi-strategy initialization mode;
calculating a fitness value, and recording global optimal positions of the flagfish and the sardine during the iteration;
updating the position of the sardine according to the attack strength;
respectively carrying out cauchy variation and adaptive t distribution variation on elite individuals of the flagelliforme and the sardine;
comparing the optimal solutions of the flagfish and the sardine, and replacing the positions of the flagfish and the sardine according to the result;
and if the iteration is not finished, calculating the fitness value to continue optimizing, and if the iteration is finished, returning a final result.
S2: collecting an original absorption spectrum, and preprocessing the original absorption spectrum by combining a variation modal decomposition algorithm optimized by the improved flagfish optimizer with a wavelet threshold method;
furthermore, the variational modal decomposition is a completely non-recursive modal variational method, and the original signal is decomposed into a plurality of intrinsic modes meeting constraint conditions through iterative optimization, so that the problems of modal aliasing and end-point effects of similar EMD (empirical mode decomposition) in processing of non-stationary vibration signals are effectively avoided.
It should be noted that the decomposition of the variation mode requires setting four control parameters, and has high dependency on parameters, wherein K determines the number of decomposition layers, and the penalty factor α determines the spectrum bandwidth of the decomposition mode, and different combinations of (K, α) have different degrees of influence on the decomposition effect finally presented. Therefore, the invention provides the two-dimensional optimization of the (K, alpha) of the variation modal decomposition by utilizing the improved flagfish optimizer, and the self-adaptive capacity and the decomposition effect of the variation modal decomposition are improved.
The flagfish optimizer is only for optimizing parameters, and the Kernel Extreme Learning Machine (KELM) is an algorithm for establishing a CO2 concentration model. The collected spectrum samples are divided into a training set and a testing set, and the training set is used as a sample of a training model. The peak value of the spectrum is used as single input of KELM, the concentration is used as single output, a fixed model is obtained after training, and the input harmonic peak value is used for obtaining a corresponding output concentration (predicted value). And taking the root mean square error of the predicted value and the actual value as a fitness function, performing iterative optimization on the regularization coefficient and the kernel function parameter of the KELM through an ISSFO algorithm to obtain a parameter which minimizes the predicted error of the KELM, and then substituting the parameter to establish a CO2 concentration model. The test set is used to verify the effect of the model.
Compared with the traditional flagfish optimizer, the improved flagfish optimizer can search the optimal fitness value (fewer iterations) faster in the optimizing process, and has higher optimizing precision (smaller fitness value).
Furthermore, the invention adopts multi-scale permutation entropy (MPE) as the fitness function when improving the optimization algorithm of the flagfish optimizer to optimize the decomposition of the two-dimensional parameters of the variation mode. Permutation Entropy (PE) is often used as an evaluation index for determining the degree of disorder of time series data, wherein a larger entropy value indicates a higher degree of disorder of data, whereas a lower randomness of data is indicated, and multi-scale permutation entropy (MPE) is to coarsen time series data and then calculate permutation entropy.
Further, the complete optimization steps are as follows:
initializing basic parameters of an improved flagfish optimizer, taking the sum of the arrangement entropy of decomposition modes of the variation mode decomposition as a fitness function, setting the search boundary of K as [3,9], and the search boundary of alpha as [50,5000];
generating an initial population, and setting the ratio of the flagelliforme population to the sardine population to be 3 to 7;
introducing various group position information as parameters (K, alpha) into an fitness function to calculate permutation entropy, and taking a minimum permutation entropy value and a corresponding position vector as a global solution and a global optimal position;
updating the flagelliforme and sardine locations according to the definition above;
recalculating the fitness value, and updating the global solution and the global optimal position;
performing optimization iteration until the maximum iteration times are reached;
taking the global optimal position (K, alpha) at the last iteration as a variation modal decomposition optimal parameter;
and (3) carrying out variant modal decomposition under the condition of (K, alpha) parameters.
It should be noted that, in order to test the decomposition effect of the improved variation modal decomposition, a test signal shown in the following formula was constructed in MATLAB, a sampling rate of 1kHz was set, and 1000 sample points were acquired in 1 second.
Wherein f (t) is a waveform function, i.e. time sequence, t is time, 1s of waveform is acquired, f 1 (t) is a 40Hz cosine signal of amplitude 1, f 2 (t) is an intermittent sinusoidal signal of amplitude 3 at 150Hz, f 3 And (t) is 20dB of Gaussian white noise.
Furthermore, an improved flagfish optimizer algorithm population is set to be 30, wherein the sardine population accounts for 70%, the flagfish population accounts for 30%, the search dimension is 2 dimensions, the search boundary of the variation modal decomposition layer number K is [3,9], the search boundary of the penalty factor alpha is [50,5000], and K and alpha are subjected to rounding treatment.
It should be noted that, after several iterative optimization, the optimal parameter is (3,228), under the condition of the parameter, the signal is decomposed into 3 eigenmodes, fig. 2 is a time domain and a frequency spectrum diagram corresponding to each decomposition mode, wherein the center frequencies of IMF1 and IMF2 are respectively 40Hz and 150Hz, the two modes respectively correspond to a cosine signal and an intermittent sine signal, and IMF3 is superimposed white gaussian noise.
S3: according to the pretreatment result, a second harmonic amplitude and CO of an absorption spectrum are established by combining the nuclear extreme learning machine optimized by the improved flagfish optimizer 2 An inversion model of concentration;
furthermore, the invention adopts a tunable absorption spectroscopy technology and a Wavelength Modulation Spectroscopy (WMS) as an experimental detection method, and gas molecules have selective absorption characteristics for light with specific wavelength according to the infrared spectrum absorption principle of the gas, and the method is interpreted as follows by Beer-Lambert law:
I＝I 0 exp(-α(v)PCL)
wherein I is 0 The initial light intensity of the laser with the wave number v is represented by I, the light intensity after passing through the gas to be measured, alpha (v) represents the absorption coefficient of the gas to be measured, P represents the gas pressure, C represents the gas volume fraction, and L represents the length of the gas absorption chamber.
Further, the amplitude I of the second harmonic signal is found after the above-mentioned components are unfolded according to the Fourier transform principle 2f The volume fraction C of the measured gas is in direct proportion to the volume fraction C of the measured gas, namely:
I 2f ∝I 0 σ 0 CL
it should be noted that δ 0 Is the absorption cross section, absorption coefficients alpha and delta 0 Can be mutually converted. The light source used a semiconductor laser (VERTILAS, VL-2004-1 m) with an emission wavelength of 2004nm, and the maximum output power was 3mW. The laser wavelength is controlled by adjusting the laser temperature and driving current, and the experimental device adopts a TEC driving chip MAX1968 to control the laser temperature, and the driving current is the superposition signal of 40Hz triangular wave and 72kHz sine wave. The modulated laser passes through an absorption air chamber (the length is 25 cm) after being collimated by a lens (THORLABS, C220 TMD), is selectively absorbed by the detected gas in the air chamber, is received by a photoelectric detector (Binsong, 12181-020) at the other side of the air chamber, and then extracts a second harmonic signal of an absorption spectrum by a lock-in amplifier. The original second harmonic signal collected by the tunable absorption spectrum technology experimental device is at the PC endAnd (3) carrying out data processing in tunable absorption spectrum technology detection software.
S4: CO from the inversion model 2 Detection experiment, judge SF 6 An operating state of the electrical device.
Furthermore, in an improvement strategy of the flagfish optimizer algorithm, the multi-strategy method of Tent chaotic mapping combined lens imaging learning provides more uniform and diverse populations for the flagfish optimizer algorithm, the sardine fused with levy random step length has better optimizing performance, and the capability of the flagfish optimizer algorithm to jump out of local optimum is effectively improved due to self-adaptive t distribution variation and Cauchy variation.
It should be noted that, according to the principle of multi-scale permutation entropy minimization, an improved flagfish optimizer algorithm is adopted to optimize the decomposition layer number K and the penalty factor alpha of the decomposition of the variation mode, so that the problem that the decomposition parameters of the variation mode need to be set manually is solved, and the self-adaptive capacity of the decomposition of the variation mode is improved. The filtering method of the variation modal decomposition combined wavelet threshold can effectively filter interference noise in an absorption spectrum, more accurately read the amplitude of the second harmonic and improve the detection precision.
Example 2
Referring to FIGS. 4-7, for one embodiment of the present invention, an ISFO-VMD-KELM based SF is provided 6 Decomposition component CO 2 In order to verify the beneficial effects of the invention, the concentration inversion method is scientifically demonstrated through a comparison experiment.
To simulate SF 6 SF is used for the experiment of the gas mixing state in the electrical equipment 6 As background gas, CO 2 SF is used in laboratory environment as target detection gas component 6 Dynamic gas distribution system configuration SF 6 And CO 2 CO of (a) mixed sample gas, configured 2 Sample gas fractions included 0.87%, 0.85%, 0.83%, 0.8%, 0.77%, 0.75%, 0.72%, 0.7%, 0.68%, 0.65%, 0.62%, 0.6%, 0.57%, 0.55%, 0.53%, 0.5%. Filling the configuration gas into a tunable absorption spectrum technology detection device one by one for experiment, collecting 60 second harmonic spectrum data of each group of concentration, and collecting 960 spectra in total。
Background gas SF 6 And target gas CO 2 Cross interference exists in the laser spectrum range, and meanwhile, the signals collected by the tunable absorption spectroscopy detection device contain a large amount of high-frequency noise, so that filtering and background subtraction on the original data are indispensable steps of spectrum analysis.
The self-adaptive variation mode provided above is adopted to decompose and decompose the original spectrum, a mode with high correlation is selected to reconstruct according to the correlation coefficient between each decomposition mode and the original data, and finally, a wavelet threshold method is adopted to filter the reconstructed signal again, so that an absorption spectrum line after noise is removed is obtained. FIGS. 4 and 5 are pure SF, respectively 6 And a volume fraction of CO of 0.85% 2 The absorption spectrum is filtered to obtain a front-back comparison chart, and the processed spectrum signal not only removes high-frequency noise in the original signal, but also has smoother waveform.
Pure SF 6 The absorption line of the gas is used as a background signal, and the wavelength modulation spectroscopy is different from the direct absorption spectroscopy in that a background baseline needs to be fitted, and the background signal is directly subtracted from the absorption line of the target gas. For example, the CO in FIG. 4 2 Absorption line minus SF in FIG. 5 6 The absorption spectrum line can obtain CO with the volume fraction of 0.85 percent 2 The second harmonic absorption line after background subtraction. As shown in fig. 6, the interference harmonic components around the center frequency of the second harmonic after background subtraction are effectively attenuated, so that the amplitude of the second harmonic can be more easily and accurately read.
FIG. 7 is a pre-treated 16 groups of different concentrations of CO 2 Absorption spectrum, CO 2 The higher the concentration the higher the amplitude of the second harmonic. Obtaining CO according to least square linear fitting 2 A linear correlation coefficient of the second harmonic peak value with the concentration r2=0.988.
Specific values can be seen in table 1, table 1 concentration inversion results RMSE comparison:
further, the second harmonic amplitude is proportional to the measured gas concentration, so the most commonly used concentration inversion method in the power industry is to use the least square method to establish a linear relation between the second harmonic amplitude and the gas concentration. While the invention selects the Kernel Extreme Learning Machine (KELM) to build the CO 2 And optimizing regularization coefficient C and kernel parameter S of the kernel extreme learning machine by using an improved flagfish optimizer algorithm, wherein the fitness function is Root Mean Square Error (RMSE) of the predicted concentration and the true concentration.
It should be noted that 960 absorption spectra are collected in total in this experiment, 960 second harmonic peaks are extracted after pretreatment, and the second harmonic peaks and the gas concentration are respectively used as single input and single output of the built model. Of these 860 pairs of samples were randomly selected as training sets for model training, and the remaining 100 pairs of samples were used as test sets for model accuracy testing. In order to verify the effectiveness of the improved flagfish optimizer algorithm, the improved SFO-KELM, SFO-KELM and ELM (Sigmoidal activation function, hidden layer node is 20), GA-BP (GA cross probability is 0.7, variation probability is 0.3, BP training frequency is 100, training efficiency is 0.01, hidden layer node is 20), PSO-KELM (PSO inertia factor w is 1, acceleration factors C1 and C2 are 2) and CLS are subjected to model accuracy comparison. The experimental results of each algorithm iterated 30 times are shown in table 1.
Further, the Root Mean Square Error (RMSE) of ISFO-KELM in table 1 is the smallest, where the regularization coefficient c=1000 and the kernel parameter s= 3.4177E-5 of the kernel extreme learning machine model. The root mean square error of the training set of ISFO-KELM is 30.8% less than PSO-KELM and 85.4% less than SFO-KELM. The root mean square error of the test set of ISFO-KELM was 36.9% less than PSO-KELM and 86.8% less than SFO-KELM. The root mean square error of the training set and the test set of the ISFO-KELM are smaller than the GA-BP and the ELM by 2 number levels and smaller than the linear least squares (CLS) by 3 number levels. Therefore, the nuclear extreme learning machine optimized by the improved flagfish optimizer algorithm has excellent performance in precision.
It should be noted that the above embodiments are only for illustrating the technical solution of the present invention and not for limiting the same, and although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical solution of the present invention may be modified or substituted without departing from the spirit and scope of the technical solution of the present invention, which is intended to be covered in the scope of the claims of the present invention.
It should be noted that the above embodiments are only for illustrating the technical solution of the present invention and not for limiting the same, and although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical solution of the present invention may be modified or substituted without departing from the spirit and scope of the technical solution of the present invention, which is intended to be covered in the scope of the claims of the present invention.
It will be appreciated by those skilled in the art that embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein. The solutions in the embodiments of the present application may be implemented in various computer languages, for example, object-oriented programming language Java, and an transliterated scripting language JavaScript, etc.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
While preferred embodiments of the present application have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. It is therefore intended that the following claims be interpreted as including the preferred embodiments and all such alterations and modifications as fall within the scope of the application.
It will be apparent to those skilled in the art that various modifications and variations can be made in the present application without departing from the spirit or scope of the application. Thus, if such modifications and variations of the present application fall within the scope of the claims and the equivalents thereof, the present application is intended to cover such modifications and variations.
Claims (10)
1. SF based on ISFO-VMD-KELM 6 Decomposition component CO 2 A method of concentration inversion, characterized by: comprising the steps of (a) a step of,
based on the traditional flagfish optimizer, improving optimizing precision and local searching capability, and establishing an improved flagfish optimizer;
collecting an original absorption spectrum, and preprocessing the original absorption spectrum by combining a variation modal decomposition algorithm optimized by the improved flagfish optimizer with a wavelet threshold method;
according to the pretreatment result, a second harmonic amplitude and CO of an absorption spectrum are established by combining the nuclear extreme learning machine optimized by the improved flagfish optimizer 2 An inversion model of concentration;
CO from the inversion model 2 Detection experiment, judge SF 6 An operating state of the electrical device.
2. An ISFO-VMD-KELM based SF as claimed in claim 1 6 Decomposition component CO 2 A method of concentration inversion, characterized by: the improved flagfish optimizer includes,
setting algorithm control parameters, and generating initial flagelliforme and sardine populations according to a multi-strategy initialization mode;
calculating a fitness value, and recording global optimal positions of the flagfish and the sardine during the iteration;
updating the position of the sardine according to the attack strength;
respectively carrying out cauchy variation and adaptive t distribution variation on elite individuals of the flagelliforme and the sardine;
comparing the optimal solutions of the flagfish and the sardine, and replacing the positions of the flagfish and the sardine according to the result;
and if the iteration is not finished, calculating the fitness value to continue optimizing, and if the iteration is finished, returning a final result.
3. An ISFO-VMD-KELM based SF as claimed in claim 2 6 Decomposition component CO 2 A method of concentration inversion, characterized by: the improved flagfish optimizer further includes,
the Tent chaotic sequence is adopted to improve the initial population distribution of the flagfish and the sardine, and the mathematical expression is as follows:
where β is a random number of [0,1 ].
4. An ISFO-VMD-KELM based SF as claimed in claim 3 6 Decomposition component CO 2 A method of concentration inversion, characterized by: the improved flagfish optimizer further includes,
the mathematical expression of the improved sardine position update is as follows:
Levy＝u/|v| 1/β
wherein the mathematical expression of σ is:
wherein u to N (0, sigma) 2 ) v-N (0, 1), beta is [0,2]Is a random number of (a) in the memory.
5. An ISFO-VMD-KELM based SF as recited in claim 4 6 Decomposition component CO 2 A method of concentration inversion, characterized by: the improved flagfish optimizer further comprises optimizing parameters of the adaptive variation modal decomposition and the nuclear extreme learning machine, and performing cauchy variation and adaptive t distribution variation on elite individuals of the flagfish and sardine.
6. An ISFO-VMD-KELM based SF as recited in claim 5 6 Decomposition component CO 2 A method of concentration inversion, characterized by: the inversion model comprises the steps of,
initializing basic parameters of an improved flagfish optimizer, and taking the sum of multi-scale permutation entropies of adaptive variation modal decomposition as an fitness function;
generating an initial population, and setting the population proportion of the flagfish and the sardine;
the position information of each group is used as a parameter to be imported into a fitness function to calculate multi-scale arrangement entropy, and the minimum multi-scale arrangement entropy value and the corresponding position vector are used as global solutions and global optimal positions;
updating the positions of the flagstones and the sardines, recalculating the fitness value, and updating the global solution and the global optimal position; and performing optimization iteration until the maximum iteration times are reached, taking the global optimal position at the last iteration as the optimal parameter of the adaptive variation modal decomposition, and performing the adaptive variation modal decomposition under the parameter condition.
7. An ISFO-VMD-KELM based SF as recited in claim 6 6 Decomposition component CO 2 A method of concentration inversion, characterized by: the inversion model also includes taking the second harmonic peak and the gas concentration as a single input and a single output, respectively, of the built model.
8. An ISFO-VMD-KELM based SF as recited in claim 7 6 Decomposition component CO 2 A method of concentration inversion, characterized by: the inversion model further comprises, using a nuclear extreme learning machine to build CO 2 And optimizing regularization coefficients and nuclear parameters of the nuclear extreme learning machine by utilizing an improved flagfish optimizer algorithm, wherein the fitness function is root mean square error of the predicted concentration and the true concentration.
9. An ISFO-VMD-KELM based SF as recited in claim 8 6 Decomposition component CO 2 A method of concentration inversion, characterized by: the preprocessing comprises the step of preprocessing an original absorption spectrum by utilizing the adaptive variation modal decomposition after parameter optimization by the improved flagfish optimizer and combining a wavelet threshold method.
10. An ISFO-VMD-KELM based SF as recited in claim 9 6 Decomposition component CO 2 A method of concentration inversion, characterized by: the adaptive variational modal decomposition includes,
wherein f (t) is a waveform function, i.e. time sequence, t is time, 1s of waveform is acquired, f 1 (t) is a 40Hz cosine signal of amplitude 1, f 2 (t) is an intermittent sinusoidal signal of amplitude 3 at 150Hz, f 3 And (t) is 20dB of Gaussian white noise.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210967017.9A CN115993345A (en) | 2022-08-11 | 2022-08-11 | SF based on ISFO-VMD-KELM 6 Decomposition component CO 2 Concentration inversion method |
| PCT/CN2023/073975 WO2024031938A1 (en) | 2022-08-11 | 2023-01-31 | Method for inverting concentration of sf6 decomposition component co2 on basis of isfo-vmd-kelm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210967017.9A CN115993345A (en) | 2022-08-11 | 2022-08-11 | SF based on ISFO-VMD-KELM 6 Decomposition component CO 2 Concentration inversion method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115993345A true CN115993345A (en) | 2023-04-21 |
Family
ID=85993112
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210967017.9A Pending CN115993345A (en) | 2022-08-11 | 2022-08-11 | SF based on ISFO-VMD-KELM 6 Decomposition component CO 2 Concentration inversion method |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN115993345A (en) |
| WO (1) | WO2024031938A1 (en) |
Family Cites Families (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180240018A1 (en) * | 2016-05-19 | 2018-08-23 | Jiangnan University | Improved extreme learning machine method based on artificial bee colony optimization |
| CN110659621A (en) * | 2019-09-27 | 2020-01-07 | 山东科技大学 | Combined noise reduction method based on variational modal decomposition and permutation entropy |
| CN111693481A (en) * | 2020-06-23 | 2020-09-22 | 贵州电网有限责任公司 | Determination of SF6Method for calibrating non-dispersive infrared absorption spectrum of CO content in gas |
| CN112364758A (en) * | 2020-11-10 | 2021-02-12 | 湖北惠立网络科技有限公司 | Garbage classification recovery method and system based on multi-target image recognition |
| CN112653142B (en) * | 2020-12-18 | 2022-09-02 | 武汉大学 | Wind power prediction method and system for optimizing depth transform network |
| CN113241087B (en) * | 2021-05-10 | 2022-05-27 | 青岛科技大学 | Underwater sound signal denoising method based on combination of improved VMD and improved wavelet threshold method |
| CN114330194A (en) * | 2021-12-31 | 2022-04-12 | 桂林电子科技大学 | Analog circuit fault diagnosis method based on SFO (Small form-factor pluggable) optimization depth limit learning machine |
| CN114565237B (en) * | 2022-02-15 | 2023-09-12 | 石河子大学 | Comprehensive energy low-carbon scheduling method and system for multi-energy combined power station |
| CN114611411A (en) * | 2022-03-31 | 2022-06-10 | 河北工业大学 | IGBT junction temperature prediction method based on ISFO-SVM model |
-
2022
- 2022-08-11 CN CN202210967017.9A patent/CN115993345A/en active Pending
-
2023
- 2023-01-31 WO PCT/CN2023/073975 patent/WO2024031938A1/en unknown
Also Published As
| Publication number | Publication date |
|---|---|
| WO2024031938A1 (en) | 2024-02-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Meng et al. | A modified empirical mode decomposition algorithm in TDLAS for gas detection | |
| Zou et al. | Accuracy improvement of quantitative analysis in laser-induced breakdown spectroscopy using modified wavelet transform | |
| Xu et al. | Deep learning application for predicting soil organic matter content by VIS-NIR spectroscopy | |
| Zhu et al. | A demonstration of long-term memory and climate predictability | |
| Fossati et al. | Non-local thermodynamic equilibrium effects determine the upper atmospheric temperature structure of the ultra-hot Jupiter KELT-9b | |
| Katuwal et al. | Combining X‐ray computed tomography and visible near‐infrared spectroscopy for prediction of soil structural properties | |
| Jiang et al. | Qualitative and quantitative analysis in solid-state fermentation of protein feed by FT-NIR spectroscopy integrated with multivariate data analysis | |
| Si et al. | Hierarchical temperature imaging using pseudoinversed convolutional neural network aided TDLAS tomography | |
| Liu et al. | Continuous background correction using effective points selected in third-order minima segments in low-cost laser-induced breakdown spectroscopy without intensified CCD | |
| Gonzalez et al. | Edge detection methods based on generalized type-2 fuzzy logic | |
| CN114923873A (en) | Gas concentration detection method and device, electronic equipment and storage medium | |
| CN106198321A (en) | Model building method and light scattering component detection method | |
| Kopparla et al. | PCA-based radiative transfer: Improvements to aerosol scheme, vertical layering and spectral binning | |
| Chami et al. | Inversion of oceanic constituents in case I and II waters with genetic programming algorithms | |
| CN115993345A (en) | SF based on ISFO-VMD-KELM 6 Decomposition component CO 2 Concentration inversion method | |
| CN111766210A (en) | Near-shore complex seawater nitrate nitrogen multispectral measurement method | |
| CN113686810B (en) | Near infrared spectrum wavelength selection method based on convolutional neural network | |
| Zhao et al. | Method for extracting Raman spectra characteristic variables of biological sample based on Hilbert–Huang transform | |
| Aukes et al. | Composition Wheels: Visualizing dissolved organic matter using common composition metrics across a variety of Canadian ecozones | |
| CN114062306B (en) | Near infrared spectrum data segmentation preprocessing method | |
| CN115910223A (en) | Irradiation white spirit production process optimization method based on PLS-LSboost gradient lifting tree | |
| Skogholt et al. | Baseline and interferent correction by the Tikhonov regularization framework for linear least squares modeling | |
| Vander Wal et al. | Transfer learning as a method to reproduce high-fidelity non-local thermodynamic equilibrium opacities in simulations | |
| Chen et al. | Artificial Intelligence Assisted Inversion (AIAI): Quantifying the Spectral Features of 56Ni of Type Ia Supernovae | |
| Li et al. | An EEMD-SVD method based on gray wolf optimization algorithm for lidar signal noise reduction |
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 |