Random noise suppression method, device, equipment and storage medium

Abstract

The application provides a random noise suppression method, device, equipment and storage medium, wherein the method comprises the following steps: configuring the number of components and the penalty factor parameters of variational mode decomposition by using a genetic algorithm; and suppressing random noise by using the variational mode decomposition with the configured parameters, so as to improve the quality of seismic data. The method provided by the application suppresses the random noise in the seismic data by GA-VMD, and improves the quality of the seismic data. The GA-VMD realizes intelligent optimal configuration of the VMD parameters by using GA, reduces the workload of people, avoids the decomposition error caused by the subjective factors of people, and thus improves the noise suppression effect, and creates a good data basis for subsequent inversion and interpretation.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of geophysical exploration, in particular to a random noise suppression method, device, equipment and storage medium. BACKGROUND

[0002] There is inevitably random noise in seismic data, and effectively suppressing noise and improving signal-to-noise ratio and resolution of seismic data are prerequisites for subsequent inversion and interpretation. Seismic data is a kind of nonlinear and non-stationary signal. A large number of signal analysis methods such as Wigner-Ville, wavelet transform and curvelet transform are used to suppress random noise according to this characteristic, although certain effects have been achieved, but there are certain limitations.

[0003] Empirical Mode Decomposition (EMD), the core idea of which is to decompose non-stationary signals into modal components of different frequency bands and then convert them into stationary signals for analysis, can be theoretically applied to the decomposition of any type of signal, so it has obvious advantages in processing nonlinear and non-stationary data. However, the signals decomposed by EMD are prone to modal aliasing and end effect, Ensemble EMD (EEMD) avoids the shortcomings of EMD, and Complete EEMD (CEEMD) effectively overcomes the noise pollution caused by auxiliary white noise introduced by EEMD, but its mathematical theoretical basis is not rigorous, the operation efficiency is relatively low, and part of the effective signal is lost. In the process of obtaining modal components, Variational Mode Decomposition (VMD) determines the frequency center and bandwidth of each component by iteratively searching for the optimal solution of the variational model, which can adaptively realize the frequency domain partitioning of signals and the effective separation of each component. VMD has a solid theoretical basis, can accurately decompose signals, is not prone to modal aliasing, has good algorithm robustness, and has high operation efficiency at the same time. The number of components and the penalty factor are two important parameters of VMD, which are generally given according to expert experience, which may cause errors in signal decomposition and affect the effect of noise suppression. SUMMARY

[0004] In view of the above problems, the present application provides a random noise suppression method, device, equipment and storage medium.

[0005] The present application provides a random noise suppression method, comprising:

[0006] S1: configuring the number of components and the penalty factor parameters of Variational Mode Decomposition by using genetic algorithm;

[0007] S2: suppressing random noise by using Variational Mode Decomposition with completed configuration parameters, and improving the quality of seismic data.

[0008] In some embodiments, the method of configuring the number of components and penalty factor parameters of variational mode decomposition using genetic algorithm comprises:

[0009] (1) encoding the parameters K, C of variational mode decomposition using genetic algorithm;

[0010] (2) performing variational mode decomposition operation on seismic data;

[0011] (3) executing selection operator, crossover operator and mutation operator to generate next generation population, i.e. obtaining new population;

[0012] (4) calculating the fitness value of the individuals in the new population;

[0013] (5) if the stopping criterion is met, the optimal coding string is found, otherwise returning to step (2);

[0014] (6) decoding the optimal coding string into optimized VMD parameters;

[0015] wherein,

[0016] K: the number of components of VMD

[0017] C: penalty factor.

[0018] In some embodiments, the selection operator is:

[0019] wherein, N is the population number, f i is the fitness value;

[0020] The crossover operator is:

[0021] wherein, f′ is the adjusted fitness value, f c ′ is the larger fitness value in the parent to be exchanged;

[0022] The mutation operator is:

[0023] wherein, f m is the fitness value of the mutated individual, k1, k2 are mutation probability adjustment factors;

[0024] f max and f avg correspond to the maximum fitness value and the average fitness value, respectively.

[0025] In some embodiments, the method of calculating the fitness value of the individuals in the new population comprises:

[0026] f' = af + b, where f, f' respectively correspond to fitness value before and after adjustment, a, b are adjustment coefficients, E is error function, and its value appears negative, then adjustment is made according to the following formula:

[0027]

[0028] Δ1 = f max -f avg , Δ2 = f avg -f min , c = 2.0, f max , f min , f avg respectively correspond to maximum fitness value, minimum fitness value, average fitness value; δ k = (z k -y k )y k (1-y k ), z k is the kth expected output, y k is the kth actual output, and S is the number of output units.

[0029] In some embodiments, the genetic algorithm is initialized, initial population size N = 60, crossover operator P c = 0.36, mutation operator P m = 0.15, and maximum iteration number are determined.

[0030] The stop criterion is that the error is less than or equal to 0.001 or the maximum learning number is reached, the maximum learning number is greater than or equal to 500 times, and can be specified by a user.

[0031] In some embodiments, the method for improving seismic data quality by using a variational modal component of a configuration parameter to decompose random noise includes:

[0032] (1) decomposing a seismic signal x(t) into K modal components d k (t);

[0033] (2) performing Hilbert transform on the modal component d k (t) to construct an analytic signal is a convolution operator, δ(k) is an impulse function, and j 2 = -1;

[0034] (3) modulating the spectrum of the modal component to a frequency band centered on ω k ; ω k is an instantaneous frequency;

[0035] (4) estimating the frequency band width of the modal component, that is, and satisfy

[0036] (5) Introducing penalty factor C and Lagrange operator λ(t) to transform the constrained variational problem into an unconstrained variational problem;

[0037] (6) Using Fourier transform to convert the unconstrained variational problem into frequency domain and obtain the frequency domain result of the modal component;

[0038] (7) Using inverse Fourier transform to convert the frequency domain result of the modal component into time domain to obtain the time domain signal of the modal component, and the signal after noise suppression The components of l+1, …, K are caused by noise.

[0039] In some embodiments, the decomposition of the seismic signal x(t) into K modal components d k (t) includes the following specific process:

[0040] d k (t) = A k (t) cos(ψ k (t)), k = 1, 2, 3, …, K, A k (t) is an amplitude, A k (t)' ≥ 0, ψ k (t) is a phase, ψ k (t)' ≥ 0, the instantaneous frequency ω k (t) of d k (t) is dψ k (t) / dt.

[0041] The specific formula for transforming the constrained variational problem into an unconstrained variational problem is:

[0042] δ(t) is an impulse function.

[0043] Embodiments of the present application provide a random noise suppression device, which comprises:

[0044] a parameter module configured for variational modal decomposition and a variational modal decomposition random noise suppression module;

[0045] The parameter module configured for variational modal decomposition uses a genetic algorithm to configure the number of components and the penalty factor parameters of the variational modal decomposition;

[0046] The variational modal decomposition random noise suppression module uses the variational modal decomposition with the configured parameters to suppress random noise and improve the quality of seismic data.

[0047] The embodiment of the present application provides a random noise suppression device, comprising a memory and a processor, and the memory stores a computer program which is executed by the processor to execute the random noise suppression method.

[0048] The embodiment of the present application provides a storage medium which stores a computer program capable of being executed by one or more processors and capable of being used to implement the random noise suppression method.

[0049] The random noise suppression method, device, equipment and storage medium provided by the present application have the following beneficial effects:

[0050] The GA-VMD is used to suppress random noise in seismic data, and the quality of the seismic data is improved. The GA-VMD realizes intelligent optimal configuration of VMD parameters by using GA, reduces the workload of people, avoids decomposition errors caused by subjective factors of people, and thus improves the noise suppression effect, and creates a good data basis for subsequent inversion and interpretation. BRIEF DESCRIPTION OF DRAWINGS

[0051] The present application will be described in more detail below based on the embodiments and with reference to the accompanying drawings.

[0052] Figure 1 The algorithm flowchart of GA-VMD provided by the embodiment of the present application is shown in the figure.

[0053] Figure 2 The random noise suppression method flowchart based on GA-VMD provided by the embodiment of the present application is shown in the figure.

[0054] Fig. 3(a)-(b) is the seismic profile of simulation data and the noisy profile with SNR=2 provided by the embodiment of the present application.

[0055] Fig. 4(a)-(c) is the 10th seismic data, noisy data and Gaussian white noise provided by the embodiment of the present application.

[0056] Fig. 5(a)-(c) is the random noise suppression effect diagram of the 10th seismic data provided by the embodiment of the present application.

[0057] Fig. 6(a)-(b) is the noise suppression seismic profile and noise profile of the CEEMD method provided by the embodiment of the present application.

[0058] Fig. 7(a)-(b) is the noise suppression seismic profile and noise profile of the GA-VMD method provided by the embodiment of the present application.

[0059] Fig. 8(a)-(b) is the comparison diagram before and after noise suppression of the actual single shot record provided by the embodiment of the present application.

[0060] Fig. 9(a)-(b) are contrast diagrams before and after actual superimposed profile noise suppression provided by the embodiment of the present application.

[0061] In the drawings, the same components are designated by the same reference numerals, and the drawings are not drawn according to the actual scale. DETAILED DESCRIPTION

[0062] In order to make the purposes, technical solutions and advantages of the present application more clear, the present application will be further described in detail below with reference to the drawings, and the described embodiments should not be regarded as limiting the present application, and all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0063] In the following description, "some embodiments" are described, which describe a subset of all possible embodiments, but it can be understood that "some embodiments" can be the same subset or different subset of all possible embodiments, and can be combined with each other without conflict.

[0064] If the similar description of "first\second\third" appears in the application file, the following description is added, in the following description, the term "first\second\third" referred to only distinguishes similar objects, and does not represent a specific order of the objects, and it can be understood that "first\second\third" can be interchanged with a specific order or sequence as allowed, so that the embodiments of the present application described herein can be implemented in an order other than that illustrated or described herein.

[0065] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present application belongs. The terms used herein are only for the purpose of describing the embodiments of the present application, and are not intended to limit the present application.

[0066] Before introducing the random noise suppression method provided by the embodiments of the present application, the problems existing in the related art are briefly introduced:

[0067] Random noise inevitably exists in seismic data, and effectively suppressing noise and improving signal-to-noise ratio and resolution of seismic data is a prerequisite for subsequent inversion and interpretation. Seismic data is a kind of nonlinear and non-stationary signal, and a large number of signal analysis methods such as Wigner-Ville, wavelet transform and curvelet transform are used to suppress random noise according to this characteristic, although certain effects have been achieved, but there are also certain limitations.

[0068] Empirical Mode Decomposition (EMD), whose core idea is to decompose non-stationary signals into modal components of different frequency bands and then convert them into stationary signals for analysis, can theoretically be applied to the decomposition of any type of signal, so it has obvious advantages in processing nonlinear and non-stationary data. However, the signals decomposed by EMD are prone to modal aliasing and end effect, Ensemble EMD (EEMD) avoids the shortcomings of EMD, and Complete EEMD (CEEMD) effectively overcomes the noise pollution caused by auxiliary white noise in EEMD, but its mathematical theoretical basis is not rigorous, the operation efficiency is relatively low, and part of the effective signal is lost. Variational Mode Decomposition (VMD) determines the frequency center and bandwidth of each component by iteratively searching for the optimal solution of the variational model in the process of obtaining modal components, which can adaptively realize the frequency domain partitioning of signals and the effective separation of each component. VMD has a solid theoretical basis, can accurately decompose signals, is not prone to modal aliasing, has good algorithm robustness, and has high operation efficiency. The number of components and the penalty factor are two important parameters of VMD, which are generally given according to expert experience, which may cause errors in signal decomposition and affect the effect of noise suppression.

[0069] Based on the problems in the related art, an embodiment of the present application provides a random noise suppression method, which is applied to a random noise suppression device. The random noise suppression device can be an electronic device, such as a computer, a mobile terminal, etc. The function realized by the random noise suppression method provided by the embodiment of the present application can be realized by calling program code by the processor of the electronic device, wherein the program code can be saved in a computer storage medium.

[0070] Example One

[0071] An embodiment of the present application provides a random noise suppression method, Figure 2 The implementation process schematic diagram of the random noise suppression method provided by the embodiment of the present application is shown in Figure 2 as follows, comprising:

[0072] S1: configuring the number of components and the penalty factor parameters of variational mode decomposition by using a genetic algorithm;

[0073] S2: suppressing random noise by using the variational mode decomposition with the completed configuration parameters, and improving the quality of seismic data.

[0074] The random noise suppression method provided in the application suppresses random noise in seismic data through GA-VMD, thereby improving the quality of seismic data. GA-VMD realizes intelligent optimal configuration of VMD parameters by using GA, reduces the workload of people, avoids decomposition errors possibly caused by subjective factors of people, and thereby improves the noise suppression effect, and creates a good data basis for subsequent inversion and interpretation.

[0075] Example Two

[0076] Based on the foregoing embodiments, the embodiments of the application further provide a random noise suppression method, comprising:

[0077] S21: configuring the number of components and the penalty factor parameters of the variational mode decomposition by using a genetic algorithm;

[0078] In some embodiments, the method of configuring the number of components and the penalty factor parameters of the variational mode decomposition by using a genetic algorithm specifically comprises:

[0079] (1) encoding the parameters K and C of the variational mode decomposition by using a genetic algorithm;

[0080] (2) performing variational mode decomposition operation on the seismic data;

[0081] (3) executing an application selection operator, a crossover operator and a mutation operator to generate the next generation population, i.e., a new population;

[0082] (4) calculating the fitness value of the individuals in the new population;

[0083] (5) if the stopping criterion is met, the optimal code string is found, otherwise returning to step (2);

[0084] (6) decoding the optimal code string into the optimized VMD parameters;

[0085] S22: suppressing random noise by using the variational mode decomposition with the configured parameters, and improving the quality of seismic data.

[0086] The random noise suppression method provided in the application suppresses random noise in seismic data through GA-VMD, thereby improving the quality of seismic data. GA-VMD realizes intelligent optimal configuration of VMD parameters by using GA, reduces the workload of people, avoids decomposition errors possibly caused by subjective factors of people, and thereby improves the noise suppression effect, and creates a good data basis for subsequent inversion and interpretation.

[0087] Example Three

[0088] Based on the foregoing embodiments, the embodiments of the application further provide a random noise suppression method, comprising:

[0089] S31: configuring the number of components and the penalty factor parameter of variational mode decomposition by using a genetic algorithm;

[0090] In some embodiments, the method of configuring the number of components and the penalty factor parameter of variational mode decomposition by using a genetic algorithm specifically includes:

[0091] (1) encoding the parameters K and C of variational mode decomposition by using a genetic algorithm;

[0092] wherein,

[0093] K: the number of components of VMD

[0094] C: the penalty factor

[0095] (2) performing variational mode decomposition operation on seismic data;

[0096] (3) executing an application selection operator, a crossover operator and a mutation operator to generate the next generation population, i.e., a new population;

[0097] (4) calculating the fitness value of the individuals in the new population;

[0098] (5) if the stopping criterion is met, the optimal code string is found, otherwise returning to step (2);

[0099] (6) decoding the optimal code string into the optimized VMD parameters;

[0100] In some embodiments, the selection operator is:

[0101] wherein, N is the population number, f i is the fitness value;

[0102] The crossover operator is:

[0103] wherein, f c ' is the larger fitness value in the parent to be exchanged;

[0104] The mutation operator is:

[0105] wherein, f m is the fitness value of the mutated individual, k1 and k2 are mutation probability adjustment factors;

[0106] f max and f avg correspond to the maximum fitness value and the average fitness value, respectively;

[0107] S32: suppressing random noise by using the variational mode decomposition with the configured parameters to improve the quality of seismic data.

[0108] This application provides a random noise suppression method that improves the quality of seismic data by using GA-VMD to suppress random noise in the data. GA-VMD utilizes GA to achieve intelligent optimal configuration of VMD parameters, reducing human workload and avoiding decomposition errors that may be caused by subjective human factors, thereby improving the noise suppression effect and creating a good data foundation for subsequent inversion and interpretation.

[0109] Example Four

[0110] Based on the foregoing embodiments, this application further provides a method for suppressing random noise, including:

[0111] S41: Use a genetic algorithm to configure the number of components and penalty factor parameters for variational mode decomposition;

[0112] In some embodiments, the specific method for configuring the number of components and penalty factor parameters of variational mode decomposition using a genetic algorithm includes:

[0113] (1) Genetic algorithm is used to encode the parameters K and C of variational mode decomposition;

[0114] in,

[0115] K: Number of VMD components

[0116] C: Penalty factor;

[0117] (2) Perform variational mode decomposition on the seismic data;

[0118] (3) Apply the selection operator, crossover operator and mutation operator to generate the next generation population, that is, obtain the new population;

[0119] (4) Calculate the fitness values ​​of individuals in the new population;

[0120] (5) If the stopping criterion is met, the optimal encoded string is found; otherwise, return to step (2).

[0121] (6) Decode the optimal encoded string into optimized VMD parameters;

[0122] In some embodiments, the selection operator is:

[0123] In the formula, N is the population size, f i This is the fitness value;

[0124] The crossover operator is:

[0125] In the formula, f′ is the adjusted fitness value, f cf is the fitness value of the parent to be exchanged;

[0126] The mutation operator is:

[0127] In the formula, f m is the fitness value of the mutated individual, and k1 and k2 are mutation probability adjustment factors;

[0128] f max and f avg respectively correspond to the maximum fitness value and the average fitness value;

[0129] In some embodiments, the specific method of calculating the fitness value of the new population individual includes:

[0130] f' = af + b In the formula, f and f' respectively correspond to the fitness value before and after adjustment, a and b are adjustment coefficients, and E is an error function. When the value of E appears negative, adjustment is performed according to the following formula:

[0131]

[0132] Δ1 = f max -f avg , Δ2 = f avg -f min , c = 2.0, f max , f min , f avg respectively correspond to the maximum fitness value, the minimum fitness value, and the average fitness value; δ k = (z k -y k )y k (1-y k ), z k is the kth expected output, y k is the kth actual output, and S is the number of output units;

[0133] S42: Utilize the variational mode decomposition to suppress random noise with completed configuration parameters, and improve the quality of seismic data.

[0134] The random noise suppression method provided in the application suppresses random noise in seismic data through GA-VMD, thereby improving the quality of seismic data. GA-VMD realizes intelligent optimal configuration of VMD parameters through GA, reduces the workload of people, avoids decomposition errors that may be caused by subjective factors of people, and thereby improves the noise suppression effect, and creates a good data foundation for subsequent inversion and interpretation.

[0135] Example Five

[0136] Based on the foregoing embodiments, the embodiments of the application further provide a random noise suppression method, comprising:

[0137] S51: configuring the number of components and the penalty factor parameter of the variational mode decomposition by using a genetic algorithm;

[0138] In some embodiments, the method of configuring the number of components and the penalty factor parameter of the variational mode decomposition by using a genetic algorithm specifically comprises:

[0139] (1) encoding the parameters K and C of the variational mode decomposition by using a genetic algorithm;

[0140] wherein,

[0141] K: the number of components of the VMD

[0142] C: the penalty factor

[0143] (2) performing variational mode decomposition operation on the seismic data;

[0144] (3) executing an application selection operator, a crossover operator and a mutation operator to generate the next generation population, i.e., a new population;

[0145] (4) calculating the fitness value of the individuals in the new population;

[0146] (5) if the stopping criterion is met, the optimal code string is found, otherwise returning to step (2);

[0147] (6) decoding the optimal code string into the optimized VMD parameters;

[0148] In some embodiments, the selection operator is:

[0149] wherein, N is the population number, f i is the fitness value;

[0150] The crossover operator is:

[0151] wherein, f′ is the adjusted fitness value, f c ′ is the larger fitness value in the parent to be exchanged;

[0152] The mutation operator is:

[0153] wherein, f m is the fitness value of the mutated individual, k1 and k2 are mutation probability adjustment factors;

[0154] f max and f avg correspond to the maximum fitness value and the average fitness value, respectively;

[0155] In some embodiments, the specific method of calculating the fitness value of the new population individuals comprises:

[0156] f' = af + b, wherein f and f' respectively correspond to the fitness values before and after adjustment, a and b are adjustment coefficients, E is an error function, and when the value of E appears negative, adjustment is performed according to the following formula:

[0157]

[0158] Δ1 = f max -f avg , Δ2 = f avg -f min , c = 2.0, f max , f min , f avg respectively correspond to the maximum fitness value, the minimum fitness value, and the average fitness value; δ k = (z k -y k )y k (1-y k ), z k is the kth expected output, y k is the kth actual output, and S is the number of output units;

[0159] In some embodiments, the genetic algorithm is initialized, the initial population size N = 60, the crossover operator P c = 0.36, the mutation operator P m = 0.15, and the maximum number of iterations are determined;

[0160] S52: Random noise is suppressed by using the variational mode decomposition of the completed configuration parameters to improve the quality of seismic data.

[0161] The random noise suppression method provided in the application suppresses random noise in seismic data by GA-VMD, thereby improving the quality of seismic data. GA-VMD uses GA to realize intelligent optimal configuration of VMD parameters, reduces the workload of people, avoids decomposition errors that may be caused by subjective factors of people, and thereby improves the noise suppression effect, and creates a good data foundation for subsequent inversion and interpretation.

[0162] Example Six

[0163] Based on the foregoing embodiments, the embodiments of the application further provide a random noise suppression method, comprising:

[0164] S61: The number of components and the penalty factor parameters of the variational mode decomposition are configured by using a genetic algorithm;

[0165] In some embodiments, the method of configuring the number of components and penalty factor parameters of variational mode decomposition using genetic algorithm comprises:

[0166] (1) encoding the parameters K, C of variational mode decomposition using genetic algorithm;

[0167] wherein,

[0168] K: the number of components of VMD

[0169] C: penalty factor

[0170] (2) performing variational mode decomposition operation on seismic data;

[0171] (3) executing selection operator, crossover operator and mutation operator to generate next generation population, i.e. obtaining new population;

[0172] (4) calculating the fitness value of individuals in new population;

[0173] (5) if the stopping criterion is met, the optimal coding string is found, otherwise returning to step (2);

[0174] (6) decoding the optimal coding string into optimized VMD parameters;

[0175] In some embodiments, the selection operator is:

[0176] wherein, N is the population number, f i is the fitness value;

[0177] The crossover operator is:

[0178] wherein, f′ is the adjusted fitness value, f c ′ is the larger fitness value in the parent to be exchanged;

[0179] The mutation operator is:

[0180] wherein, f m is the fitness value of the mutated individual, k1, k2 are mutation probability adjustment factors;

[0181] f max and f avg correspond to the maximum fitness value and the average fitness value, respectively;

[0182] In some embodiments, the method of calculating the fitness value of individuals in new population comprises:

[0183] f' = af + b, wherein f and f' respectively correspond to the fitness values before and after adjustment, a and b are adjustment coefficients, E is an error function, and when the value of E appears negative, adjustment is performed according to the following formula:

[0184]

[0185] Δ1 = f max -f avg , Δ2 = f avg -f min , c = 2.0, f max , f min , f avg respectively correspond to the maximum fitness value, the minimum fitness value, and the average fitness value; δ k = (z k -y k )y k (1-y k ), z k is the kth expected output, y k is the kth actual output, and S is the number of output units;

[0186] In some embodiments, the stop criterion is that the error is less than or equal to 0.001 or the maximum number of learning times is reached, and the maximum number of learning times is greater than or equal to 500, which can also be specified by a user, such as 1000 times, etc.

[0187] S62: Random noise is suppressed by using the variational mode decomposition of the completed configuration parameters to improve the quality of the seismic data.

[0188] The random noise suppression method provided in the application suppresses random noise in seismic data by GA-VMD, thereby improving the quality of the seismic data. GA-VMD uses GA to realize intelligent optimal configuration of VMD parameters, reduces the workload of people, avoids decomposition errors that can be caused by subjective factors of people, and thereby improves the noise suppression effect, thereby creating a good data foundation for subsequent inversion and interpretation.

[0189] Example Seven

[0190] Based on the foregoing embodiments, the embodiments of the application further provide a random noise suppression method, which comprises:

[0191] S71: The number of components and the penalty factor parameters of the variational mode decomposition are configured by using a genetic algorithm.

[0192] In some embodiments, the method of configuring the number of components and the penalty factor parameters of the variational mode decomposition by using the genetic algorithm comprises:

[0193] (1) encode the parameters K, C of the variational mode decomposition by using the genetic algorithm;

[0194] wherein,

[0195] K: the number of components of VMD

[0196] C: the penalty factor;

[0197] (2) perform the variational mode decomposition operation on the seismic data;

[0198] (3) execute the selection operator, the crossover operator and the mutation operator to generate the next generation population, i.e. obtain a new population;

[0199] (4) calculate the fitness value of the individuals in the new population;

[0200] (5) if the stopping criterion is met, find the optimal code string, otherwise return to step (2);

[0201] (6) decode the optimal code string into the optimized VMD parameters;

[0202] In some embodiments, the selection operator is:

[0203] wherein, N is the population number, f i is the fitness value;

[0204] The crossover operator is:

[0205] wherein, f′ is the adjusted fitness value, f c ′ is the larger fitness value in the parent to be exchanged;

[0206] The mutation operator is:

[0207] wherein, f m is the fitness value of the mutated individual, k1, k2 are mutation probability adjustment factors;

[0208] f max and f avg correspond to the maximum fitness value and the average fitness value respectively;

[0209] In some embodiments, the specific method of calculating the fitness value of the individuals in the new population comprises:

[0210] f′ = af + b wherein, f, f′ correspond to the fitness values before and after adjustment respectively, a, b are adjustment coefficients, E is an error function, and if the value of E appears negative, it is adjusted according to the following formula:

[0211]

[0212] Δ1=f max -f avg Δ2=f avg -f min c = 2.0, f max f min f avg These correspond to the maximum fitness value, minimum fitness value, and average fitness value, respectively. δ k =(z k -y k )y k (1-y k ), z k For the k-th expected output, y k This represents the k-th actual output, and S is the number of output units.

[0213] In some embodiments, the stopping criterion is that the error is less than or equal to 0.001 or the maximum number of learning iterations is reached, wherein the maximum number of learning iterations is greater than or equal to 500 and can be specified by the user;

[0214] S72: Use variational mode decomposition with configured parameters to suppress random noise and improve the quality of seismic data;

[0215] In some embodiments, the specific method for suppressing random noise and improving seismic data quality by utilizing variational mode decomposition with configured parameters includes:

[0216] (1) Decompose the seismic signal x(t) into K modal components d k (t);

[0217] (2) For the modal component d k (t) is subjected to Hilbert transform to construct an analytic signal. * represents the convolution operator, δ(k) is the impulse function, and j 2 =-1;

[0218] (3) Modulate the spectrum of the modal components to a frequency of ω. k On the centered frequency band ω k Instantaneous frequency;

[0219] (4) Estimate the bandwidth of the modal components, i.e. And satisfy

[0220] (5) By introducing the penalty factor C and the Lagrange operator λ(t), the constrained variational problem is transformed into an unconstrained variational problem;

[0221] (6) Fourier transform is used to convert the unconstrained variational problem into the frequency domain, and the frequency domain results of the modal components are obtained;

[0222] (7) Inverse Fourier transform is used to transform the frequency domain results of the modal components into the time domain to obtain the time domain signals of the modal components, and the signal after noise suppression The components of l+1,…,K are caused by noise.

[0223] The random noise suppression method provided in the application suppresses random noise in seismic data through GA-VMD, thereby improving the quality of seismic data. GA-VMD uses GA to realize intelligent optimal configuration of VMD parameters, reduces the workload of people, avoids decomposition errors that may be caused by human subjective factors, and thereby improves the noise suppression effect, thereby creating a good data basis for subsequent inversion and interpretation.

[0224] Example Eight

[0225] Based on the foregoing embodiments, the embodiments of the application further provide a random noise suppression method, comprising:

[0226] S81: configuring the number of components and the penalty factor parameters of variational modal decomposition by using a genetic algorithm;

[0227] In some embodiments, the method of configuring the number of components and the penalty factor parameters of variational modal decomposition by using a genetic algorithm comprises:

[0228] (1) encoding the parameters K and C of variational modal decomposition by using a genetic algorithm;

[0229] wherein,

[0230] K: the number of components of VMD

[0231] C: penalty factor

[0232] (2) performing variational modal decomposition operation on seismic data;

[0233] (3) executing an application selection operator, a crossover operator and a mutation operator to generate the next generation population, i.e., a new population;

[0234] (4) calculating the fitness value of the individuals in the new population;

[0235] (5) if the stopping criterion is met, the optimal code string is found, otherwise returning to step (2);

[0236] (6) decoding the optimal code string into the optimized VMD parameters;

[0237] In some embodiments, the selection operator is:

[0238] where N is the population number, f i is the fitness value;

[0239] The crossover operator is:

[0240] where f' is the adjusted fitness value, f c is the larger fitness value in the parent to be exchanged;

[0241] The mutation operator is:

[0242] where f m is the fitness value of the mutated individual, k1, k2 are mutation probability adjustment factors;

[0243] f max and f avg correspond to the maximum fitness value and the average fitness value, respectively;

[0244] In some embodiments, the specific method of calculating the fitness value of the new population individual comprises:

[0245] f' = af + b where f, f' correspond to the fitness value before and after adjustment, respectively, a, b are adjustment coefficients, and E is an error function, whose value appears negative, and is adjusted according to the following formula:

[0246]

[0247] Δ1 = f max - f avg , Δ2 = f avg - f min , c = 2.0, f max , f min , f avg correspond to the maximum fitness value, the minimum fitness value, and the average fitness value, respectively; δ k = (z k - y k ) y k (1 - y k ), z k is the kth expected output, y k is the kth actual output, and S is the number of output units;

[0248] In some embodiments, the stopping criterion is that the error is less than or equal to 0.001 or the maximum number of learning times is reached, and the maximum number of learning times is greater than or equal to 500, which can be specified by the user;

[0249] S82: Suppress random noise by using variational mode decomposition with completed configuration parameters to improve seismic data quality;

[0250] In some embodiments, the method of suppressing random noise by using variational mode decomposition with completed configuration parameters to improve seismic data quality comprises:

[0251] (1) decomposing the seismic signal x(t) into K modal components d k (t);

[0252] (2) performing Hilbert transform on the modal component d k (t) to construct an analytic signal is a convolution operator, and δ(k) is an impulse function, j 2 = -1;

[0253] (3) modulating the spectrum of the modal component to a frequency band centered at ω k is the instantaneous frequency; ω k is the instantaneous frequency;

[0254] (4) estimating the frequency band width of the modal component, i.e. and satisfies

[0255] (5) introducing a penalty factor C and a Lagrange operator λ(t) to convert the constrained variational problem into an unconstrained variational problem;

[0256] (6) converting the unconstrained variational problem to the frequency domain by using Fourier transform and obtaining the frequency domain result of the modal component;

[0257] (7) converting the frequency domain result of the modal component to the time domain by using inverse Fourier transform to obtain the time domain signal of the modal component, then the signal after noise suppression The components of d k (t) from l+1 to K are caused by noise.

[0258] In some embodiments, the process of decomposing the seismic signal x(t) into K modal components d k (t) comprises:

[0259] k = 1, 2, 3, …, K, A k (t) is the amplitude, A k (t)' ≥ 0, ψ k (t) is the phase, ψ k (t)' ≥ 0, the instantaneous frequency of d k (t) is ω k (t) = dψ k (t) / dt;

[0260] The constraint variational problem is converted into an unconstrained variational problem, and the specific formula is:

[0261] delta(t) is an impulse function.

[0262] The random noise suppression method provided by the application suppresses random noise in seismic data through GA-VMD, thereby improving the quality of seismic data. GA-VMD realizes intelligent optimal configuration of VMD parameters by using GA, reduces the workload of people, avoids decomposition errors that may be caused by human subjective factors, and thereby improves the noise suppression effect, thereby creating a good data foundation for subsequent inversion and interpretation.

[0263] Example Nine

[0264] Based on the method of example eight, as shown in Figure 1 and Figure 2 The application provides an example according to real data:

[0265] S91: configuring the number of components and the penalty factor parameters of variational mode decomposition by using a genetic algorithm;

[0266] In some embodiments, the method of configuring the number of components and the penalty factor parameters of variational mode decomposition by using a genetic algorithm comprises:

[0267] (1) encoding the parameters K and C of variational mode decomposition by using a genetic algorithm;

[0268] wherein,

[0269] K: the number of components of VMD

[0270] C: penalty factor;

[0271] (2) performing variational mode decomposition operation on seismic data;

[0272] (3) executing an application selection operator, a crossover operator and a mutation operator to generate the next generation population, i.e., a new population;

[0273] (4) calculating the fitness value of the individuals in the new population;

[0274] (5) if the stopping criterion is met, the optimal code string is found, otherwise returning to step (2);

[0275] (6) decoding the optimal code string into optimized VMD parameters;

[0276] In some embodiments, the selection operator is:

[0277] wherein, N is the population number, f i is the fitness value;

[0278] The crossover operator is:

[0279] wherein f' is the adjusted fitness value, f is the fitness value before adjustment, f is the fitness value of the parent to be exchanged, f is the fitness value of the other parent to be exchanged, and f is the fitness value of the child to be exchanged. c f is the larger fitness value of the parent to be exchanged;

[0280] The mutation operator is:

[0281] wherein f is the fitness value of the mutated individual, k1 and k2 are mutation probability adjustment factors. m

[0282] f and f correspond to the maximum fitness value and the average fitness value, respectively. max avg

[0283] In some embodiments, the specific method of calculating the fitness value of the new population individual comprises:

[0284] f' = af + b wherein f and f' correspond to the fitness value before and after adjustment, respectively, a and b are adjustment coefficients, and E is an error function.

[0285]

[0286] Δ1 = f - f, Δ2 = f - f, c = 2.0, f, f, f correspond to the maximum fitness value, the minimum fitness value, and the average fitness value, respectively. max avg avg min max min avg δ = (z - y)y (1 - y), z is the kth expected output, y is the kth actual output, and S is the number of output units. k k k k k k k

[0287] In some embodiments, the stopping criterion is that the error is less than or equal to 0.001 or the maximum number of learning times is reached, and the maximum number of learning times is greater than or equal to 500, which can be specified by a user.

[0288] S92: Suppress random noise using variational mode decomposition with completed configuration parameters to improve the quality of seismic data.

[0289] ​​​​​​​​​​​​​​​​​In some embodiments, the specific method for suppressing random noise and improving seismic data quality by utilizing variational mode decomposition with configured parameters includes:

[0290] (1) Decompose the seismic signal x(t) into K modal components d k (t);

[0291] (2) For the modal component d k Perform a Hilbert transform on (t) to construct an analytic signal. * represents the convolution operator, δ(k) is the impulse function, and j 2 =-1;

[0292] (3) Modulate the spectrum of the modal components to a frequency of ω. k On the centered frequency band ω k Instantaneous frequency;

[0293] (4) Estimate the bandwidth of the modal components, i.e. And satisfy

[0294] (5) By introducing the penalty factor C and the Lagrange operator λ(t), the constrained variational problem is transformed into an unconstrained variational problem;

[0295] δ(t) is the impulse function;

[0296] (6) The unconstrained variational problem is transformed into the frequency domain using Fourier transform, and the frequency domain results of the modal components are obtained.

[0297] Find the frequency domain update formula for the quadratic optimization problem:

[0298] τ is the noise tolerance parameter. Represents the Fourier transform;

[0299] (7) The frequency domain result of the modal component IMF is transformed to the time domain using inverse Fourier transform to obtain the time domain signal of the modal component IMF. The signal after noise suppression is then obtained. The components l+1,…,K are caused by noise;

[0300] In some embodiments, the seismic signal x(t) is decomposed into K modal components d k (t) The specific process includes:

[0301] d k (t)=A k (t)cos(ψ k(t)), k = 1, 2, 3, K, A k (t) is amplitude, A k (t)'≥ 0, ψ k (t) is phase, ψ k (t)'≥ 0, d k (t) is instantaneous frequency ω k (t) = dψ k (t) / dt;

[0302] The constraint variational problem is converted into an unconstrained variational problem, and the specific formula is:

[0303] δ(t) is an impulse function.

[0304] When the data volume is large, it is recommended to select part of the data for parameter testing, and then apply the preferred parameter combination to the entire work area to improve the calculation efficiency.

[0305] Figure 3(a) is the simulation data containing 4 seismic events, a total of 501 channels, 501 sampling points, and a time sampling interval of 1 millisecond. Figure 3(b) is the data with added Gaussian white noise, and the signal-to-noise ratio SNR = 2. Figures 4(a)-(c) are the extracted 10th channel data of Figure 3, and the added Gaussian white noise.

[0306] Figures 5(a)-(c) are the effect diagrams of random noise suppression of the 10th seismic data. It can be seen that CEEMD restores the basic form of the original signal, but there is still a small amount of noise, and part of the original effective signal is lost. GA-VMD can effectively restore the original signal, and there is almost no loss of effective signal.

[0307] Figure 6(a) is the seismic profile after noise suppression using the CEEMD method, and Figure 6(b) is its noise profile. Although most of the random noise can be removed, part of the effective signal still remains on the noise profile. Figure 7(a) is the seismic profile after noise suppression using the GA-VMD method, and Figure 7(b) is its noise profile. No effective signal appears on the noise profile, and the random noise is effectively suppressed. Compared with CEEMD, GA-VMD shows better noise suppression effect and high intelligence.

[0308] GA-VMD is applied to random noise suppression of actual seismic data. Figure 8(a) is the original single-shot record, and Figure 8(b) is the single-shot record after noise suppression. Figure 9(a) is the original stack profile, and Figure 9(b) is the stack profile after noise suppression. Severe random noise exists on the single-shot record and the stack profile, which destroys the continuity of the event. After random noise suppression using GA-VMD, the effective signal on the single-shot record is more continuous, and the event on the stack profile is clearer.

[0309] Example Ten

[0310] Based on the foregoing embodiments, the embodiments of the application provide a random noise suppression device, comprising:

[0311] The parameter module configured for the variational mode decomposition and the variational mode decomposition random noise suppression module;

[0312] The parameter module configured for the variational mode decomposition uses a genetic algorithm to configure the number of components and the penalty factor parameters of the variational mode decomposition;

[0313] The variational mode decomposition random noise suppression module uses the variational mode decomposition with the configured parameters to suppress random noise and improve the quality of seismic data.

[0314] It should be noted that, in the embodiments of the application, if the random noise suppression method described above is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the embodiments of the application can be embodied in the form of a software product, and the computer software product is stored in a storage medium, and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the methods described in the embodiments of the application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM, Read Only Memory), a magnetic disk or an optical disk, and various media that can store program codes. Thus, the embodiments of the application are not limited to any specific combination of hardware and software.

[0315] Correspondingly, the embodiments of the application provide a storage medium having a computer program stored thereon, wherein the computer program is executed by a processor to implement the steps of the random noise suppression method provided in the above embodiments.

[0316] Embodiment eleven

[0317] The embodiments of the application provide a random noise suppression device comprising a memory and a processor, wherein the memory has a computer program stored thereon, and the computer program is executed by the processor to configure the processor to execute the program of the random noise suppression method stored in the memory, so as to implement the steps of the random noise suppression method provided in the above embodiments.

[0318] The descriptions of the display device and the storage medium embodiments above are similar to the descriptions of the method embodiments above, and have similar beneficial effects to the method embodiments. For technical details not disclosed in the computer device and storage medium embodiments of the application, please refer to the descriptions of the method embodiments of the application.

[0319] It should be noted that the above description of the storage medium and device embodiments is similar to the description of the method embodiments, and has similar beneficial effects as the method embodiments. For technical details not disclosed in the storage medium and device embodiments of the present application, please refer to the description of the method embodiments for understanding.

[0320] It should be understood that the "one embodiment" or "an embodiment" mentioned throughout the specification means that the specific features, structures or characteristics related to the embodiment are included in at least one embodiment of the present application. Therefore, "in one embodiment" or "in an embodiment" appearing throughout the specification does not necessarily refer to the same embodiment. In addition, these specific features, structures or characteristics can be combined in one or more embodiments in any suitable manner. It should be understood that the size of the sequence number of each process in various embodiments of the present application does not mean the order of execution, and the execution order of each process should be determined according to its function and inherent logic, and should not constitute any limitation on the implementation process of the embodiments of the present application. The sequence number of the above embodiments of the present application is only for description, not representing the advantages and disadvantages of the embodiments.

[0321] It should be noted that in this document, the terms "comprise", "contain" or any other variant thereof are intended to cover non-exclusive inclusion, so that the process, method, article or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed or inherent to such process, method, article or device. Without more limitations, the element defined by the statement "comprises a" does not exclude the presence of additional identical elements in the process, method, article or device including the element.

[0322] In several embodiments provided by the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are only schematic. For example, the division of the units is only a logical function division, and actual implementation can have another division manner, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the displayed or discussed components can be through some interface, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0323] The units described above as separate components can or can not be physically separated, and the components displayed as units can or can not be physical units; they can be located in one place or distributed on multiple network units; some or all of the units can be selected according to actual needs to achieve the purpose of the embodiment scheme.

[0324] In addition, each of the functional units in each embodiment of the present application can be integrated in one processing unit, or each unit can be separately implemented as a unit, or two or more units can be integrated in one unit; the integrated unit can be implemented in the form of hardware or in the form of hardware plus software function unit.

[0325] Those skilled in the art can understand that all or part of the steps of the above-mentioned method embodiments can be completed by program instruction related hardware, and the foregoing program can be stored in a computer readable storage medium, and the program executes the steps including the above-mentioned method embodiments when executed; and the foregoing storage medium includes mobile storage equipment, read only memory (ROM, Read Only Memory), magnetic disc or optical disc and various storage program codes.

[0326] Alternatively, the integrated unit of the present application, if implemented in the form of a software function module and sold or used as an independent product, can also be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the embodiments of the present application can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a plurality of instructions for causing a controller to execute all or part of the methods described in the embodiments of the present application. And the foregoing storage medium includes mobile storage equipment, ROM, magnetic disc or optical disc and various storage program codes.

[0327] The above is only an embodiment of the present application, but the protection scope of the present application is not limited thereto, any person skilled in the art can easily think of changes or replacements within the technical range disclosed in the present application, which should be covered in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for suppressing random noise, characterized in that, include: S1: Use a genetic algorithm to configure the number of components and penalty factor parameters for variational mode decomposition; S2: Use variational mode decomposition with configured parameters to suppress random noise and improve the quality of seismic data; The specific method for configuring the number of components and penalty factor parameters of variational mode decomposition using a genetic algorithm includes: (1) Genetic algorithm is used to encode the parameters K and C of variational mode decomposition; (2) Perform variational mode decomposition on the seismic data; (3) Apply the selection operator, crossover operator and mutation operator to generate the next generation population, that is, obtain the new population; (4) Calculate the fitness values ​​of individuals in the new population; (5) If the stopping criterion is met, the optimal encoded string is found; otherwise, return to step (2). (6) Decode the optimal encoded string into optimized variational mode decomposition (VMD) parameters; Where K: number of VMD components; C: penalty factor; The selection operator is: In the formula, For population size, This is the fitness value; The crossover operator is: In the formula, The adjusted fitness value. The larger fitness value among the parents to be swapped; The mutation operator is: In the formula, The fitness value of the variant individual. , It is a factor for regulating the probability of mutation; and These correspond to the maximum fitness value and the average fitness value, respectively.

2. The method according to claim 1, characterized in that, The specific method for calculating the fitness value of individuals in the new population includes: , In the formula, , These correspond to the fitness values ​​before and after adjustment, respectively. , Here, E is the adjustment coefficient, and E is the error function. If the value of E is negative, the adjustment is made according to the following formula: , c=2.0 , , These correspond to the maximum fitness value, minimum fitness value, and average fitness value, respectively. , For the k-th expected output, This represents the k-th actual output, and S is the number of output units.

3. The method according to claim 1, characterized in that, The initialization of the genetic algorithm specifically involves: determining the initial population size. =60, Cross Operator =0.36, Mutation Operator =0.15 and the maximum number of iterations; The stopping criterion is that the error is less than or equal to 0.001 or the maximum number of learning iterations is reached, where the maximum number of learning iterations is greater than or equal to 500.

4. The method according to claim 1, characterized in that, The specific methods for suppressing random noise and improving seismic data quality by utilizing variational mode decomposition with configured parameters include: (1) The seismic signal Decomposed into K modal components ; (2) For modal components Perform Hilbert transform to construct analytic signal * represents the convolution operator. For impulse functions, =-1; (3) Modulate the spectrum of the modal components to a frequency range of 1000 MHz. On the centered frequency band , Instantaneous frequency; (4) Estimate the bandwidth of the modal components, i.e. and satisfy ; (5) Introduce the penalty factor C and the Lagrange operator This transforms the constrained variational problem into an unconstrained variational problem. (6) The unconstrained variational problem is transformed into the frequency domain using Fourier transform, and the frequency domain results of the modal components are obtained; (7) The frequency domain result of the modal component is transformed to the time domain by using inverse Fourier transform to obtain the time domain signal of the modal component. Then the noise-suppressed signal is... , The components +1, ..., K are caused by noise.

5. The method according to claim 4, characterized in that, The earthquake signal Decomposed into K modal components The specific process includes: , k=1, 2, 3…, K For amplitude, , For phase, , instantaneous frequency ; The specific formula for transforming a constrained variational problem into an unconstrained variational problem is as follows: , It is an impulse function.

6. A random noise suppression device, characterized in that, include: Configure the parameter module for variational mode decomposition and the module for suppressing random noise in variational mode decomposition; The parameter module for configuring variational mode decomposition uses a genetic algorithm to configure the number of components and penalty factor parameters for variational mode decomposition. The variational mode decomposition random noise suppression module: uses variational mode decomposition with configured parameters to suppress random noise and improve the quality of seismic data; The specific method for configuring the number of components and penalty factor parameters of variational mode decomposition using a genetic algorithm includes: (1) Genetic algorithm is used to encode the parameters K and C of variational mode decomposition; (2) Perform variational mode decomposition on the seismic data; (3) Apply the selection operator, crossover operator and mutation operator to generate the next generation population, that is, obtain the new population; (4) Calculate the fitness values ​​of individuals in the new population; (5) If the stopping criterion is met, the optimal encoded string is found; otherwise, return to step (2). (6) Decode the optimal encoded string into optimized variational mode decomposition (VMD) parameters; Where K: number of VMD components; C: penalty factor; The selection operator is: In the formula, For population size, This is the fitness value; The crossover operator is: In the formula, The adjusted fitness value. The larger fitness value among the parents to be swapped; The mutation operator is: In the formula, The fitness value of the variant individual. , It is a factor for regulating the probability of mutation; and These correspond to the maximum fitness value and the average fitness value, respectively.

7. A random noise suppression device, characterized in that, It includes a memory and a processor, wherein the memory stores a computer program that, when executed by the processor, performs the random noise suppression method as described in any one of claims 1 to 5.

8. A storage medium, characterized in that, The computer program stored in the storage medium can be executed by one or more processors and can be used to implement the random noise suppression method as described in any one of claims 1 to 5.

Images