Adaptive wavelet threshold denoising method, system and device for downhole das signal and storage medium

Abstract

The application discloses a kind of adaptive wavelet threshold denoising methods, systems, equipment and storage medium for downhole DAS signal, comprising: wavelet decomposition is carried out to downhole DAS signal, and low-frequency approximate coefficient and high-frequency detail coefficient are obtained;High-frequency detail coefficient is processed by threshold through adaptive soft threshold function;Wavelet reconstruction is carried out to the high-frequency detail coefficient and low-frequency approximate coefficient after processing, and the signal-to-noise ratio of reconstructed time domain signal is calculated;Global optimal parameter is obtained according to signal-to-noise ratio through improved GWO-HHO hybrid optimization strategy;Downhole DAS signal is adaptively wavelet threshold denoised based on global optimal parameter.The application constructs continuous improved adaptive soft threshold function, realizes the accurate balance of noise suppression and feature reservation, then accurately obtains global optimal threshold parameter through improved GWO-HHO hybrid optimization strategy, and then improves the denoising effect of downhole DAS signal.

Description

Technical Field

[0001] This invention relates to the field of signal denoising technology, and in particular to an adaptive wavelet threshold denoising method, system, device and storage medium for downhole DAS signals. Background Technology

[0002] The demand for accurate acquisition of downhole two-phase flow parameters is increasingly urgent in modern oil and gas extraction. The non-stationary, high-noise characteristics of distributed acoustic sensing (DAS) signals make feature extraction extremely complex. Traditional wavelet thresholding denoising methods use fixed thresholds, which cannot be dynamically adjusted according to local signal features, easily leading to noise residue or loss of abrupt features. Single intelligent optimization algorithms struggle to balance global search and local exploitation, making it difficult to obtain optimal threshold parameters and resulting in poor denoising performance. Therefore, improving the denoising effect of downhole DAS signals has become a pressing problem. Summary of the Invention

[0003] The main objective of this invention is to provide an adaptive wavelet threshold denoising method, system, device, and storage medium for downhole DAS signals, aiming to solve the technical problem of how to improve the denoising effect of downhole DAS signals.

[0004] To achieve the above objectives, this invention provides an adaptive wavelet thresholding denoising method for downhole DAS signals, the method comprising: S1, perform wavelet decomposition on the downhole DAS signal to obtain low-frequency approximation coefficients and high-frequency detail coefficients; S2, thresholding the high-frequency detail coefficients using an adaptive soft thresholding function; S3, perform wavelet reconstruction on the processed high-frequency detail coefficients and the low-frequency approximation coefficients, and calculate the signal-to-noise ratio of the reconstructed time-domain signal; S4. Based on the signal-to-noise ratio, the globally optimal parameters are obtained through an improved GWO-HHO hybrid optimization strategy. S5, perform adaptive wavelet threshold denoising on the downhole DAS signal based on the global optimal parameters.

[0005] Optionally, S2 includes: S2.1, Determine the search range for the threshold adjustment benchmark and the search range for the exponential decay rate within the adaptive soft threshold function; S2.2, Based on the search interval of the threshold adjustment benchmark and the search interval of the exponential decay rate, generate multiple combinations of threshold parameters to be optimized, and use the combinations of threshold parameters to be optimized as individuals in the population; S2.3, thresholding the high-frequency detail coefficients according to the individual groups using an adaptive soft thresholding function.

[0006] Optionally, the adaptive soft threshold function is:

[0007]

[0008] In the formula, For adaptive dynamic threshold, For high-frequency detail coefficients, Based on a fixed threshold, Adjust the baseline parameters for the threshold. For the exponential decay rate parameter, For symbolic functions, The high-frequency detail coefficients are the processed values.

[0009] Optionally, S4 includes: S4.1, Based on the signal-to-noise ratio, select a preset number of guide individuals from multiple populations and determine the fitness value of each population individual; S4.2 Calculate the standard deviation of population fitness based on the fitness values ​​of individuals in each population group; S4.3, Update the position of individuals in each population using an improved GWO-HHO hybrid optimization strategy based on the population fitness standard deviation; S4.4, Determine whether the current iteration number is greater than or equal to the preset iteration threshold; S4.5 If so, then determine the optimal population individual based on the preset number of guiding individuals, and use the optimal population individual as the global optimal parameter; S4.6 If not, then the population individual with the updated position is taken as the new population individual, and the process returns to S2 iteration.

[0010] Optionally, S4.3 includes: S4.3.1 Determine whether the standard deviation of the population fitness is greater than or equal to a preset switching threshold; S4.3.2, if so, then update the position of individuals in various groups using the GWO global search strategy; S4.3.3, if not, update the position of individuals in each population using the HHO local development strategy.

[0011] Optionally, updating the positions of individuals in various groups using the GWO global search strategy includes: Calculate the adaptive decay value and dynamic weight coefficient based on the current iteration number and the preset iteration threshold; The coefficient vector is calculated based on the adaptive attenuation value using the enclosing behavior formula; Select a leader individual from the preset number of guiding individuals, and calculate a distance vector based on the leader individual, the coefficient vector, and the individuals of each group; The positions of individuals in each group are updated using a position update formula based on the dynamic weight coefficient, the preset number of guiding individuals, the distance vector, and the coefficient vector.

[0012] Optionally, updating the location of individuals in various populations through the HHO local development strategy includes: The escape energy factor is determined based on the adaptive decay value; The corresponding encirclement strategy is adaptively selected based on the escape energy factor and random number. The encirclement strategy is a soft encirclement strategy, a hard encirclement strategy, a soft encirclement-Levy flight progressive attack strategy, or a hard encirclement-Levy flight progressive attack strategy. The position of individuals in each group is updated according to dynamic weights based on the described encirclement strategy.

[0013] Furthermore, to achieve the above objectives, this invention also proposes an adaptive wavelet threshold denoising system for downhole DAS signals, wherein the adaptive wavelet threshold denoising system for downhole DAS signals includes: The wavelet decomposition module is used to perform wavelet decomposition on downhole DAS signals to obtain low-frequency approximation coefficients and high-frequency detail coefficients. An adaptive soft thresholding module is used to perform thresholding on the high-frequency detail coefficients using an adaptive soft thresholding function. The signal reconstruction module is used to perform wavelet reconstruction on the processed high-frequency detail coefficients and the low-frequency approximation coefficients, and to calculate the signal-to-noise ratio of the reconstructed time-domain signal. The parameter optimization module is used to obtain the globally optimal parameters based on the signal-to-noise ratio by improving the GWO-HHO hybrid optimization strategy; The wavelet threshold denoising module is used to perform adaptive wavelet threshold denoising on the downhole DAS signal based on the globally optimal parameters.

[0014] Furthermore, to achieve the above objectives, the present invention also proposes an adaptive wavelet threshold denoising device for downhole DAS signals. The device includes: a memory, a processor, and an adaptive wavelet threshold denoising program for downhole DAS signals stored in the memory and executable on the processor. The adaptive wavelet threshold denoising program for downhole DAS signals is configured to implement the steps of the adaptive wavelet threshold denoising method for downhole DAS signals as described above.

[0015] Furthermore, to achieve the above objectives, the present invention also proposes a storage medium storing an adaptive wavelet threshold denoising program for downhole DAS signals. When the adaptive wavelet threshold denoising program for downhole DAS signals is executed by a processor, it implements the steps of the adaptive wavelet threshold denoising method for downhole DAS signals as described above.

[0016] This invention first performs wavelet decomposition on the downhole DAS signal to obtain low-frequency approximation coefficients and high-frequency detail coefficients. Then, it performs thresholding on the high-frequency detail coefficients using an adaptive soft thresholding function. The processed high-frequency detail coefficients and low-frequency approximation coefficients are then reconstructed using wavelet decomposition, and the signal-to-noise ratio (SNR) of the reconstructed time-domain signal is calculated. Subsequently, based on the SNR, a globally optimal parameter is obtained through an improved GWO-HHO hybrid optimization strategy. Finally, the downhole DAS signal is subjected to adaptive wavelet threshold denoising based on the globally optimal parameter. This invention constructs a continuous improved adaptive soft threshold function, combined with a dynamic threshold calculation mechanism, which effectively overcomes the amplitude deviation and pseudo-oscillation defects of traditional threshold functions, achieving a precise balance between noise suppression and feature preservation. Then, an improved GWO-HHO collaborative optimization strategy is proposed: by optimizing the convergence factor of GWO to an adaptive decay mode to enhance the global search capability in the early stages, the HHO algorithm is introduced, and iterative adaptive dynamic weights are used to ensure a smooth transition between the two optimization stages. This fully leverages the advantages of HHO in the fine-grained development of the local neighborhood of the optimal solution, thereby accurately obtaining the globally optimal threshold parameters, thus improving the denoising effect on downhole DAS signals and successfully extracting relatively pure downhole flow feature signals. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the structure of an adaptive wavelet threshold denoising device for downhole DAS signals in the hardware operating environment of the embodiment of the present invention; Figure 2 This is a flowchart illustrating the first embodiment of the adaptive wavelet threshold denoising method for downhole DAS signals according to the present invention. Figure 3 This is a schematic diagram of the traditional threshold denoising effect of the first embodiment of the adaptive wavelet threshold denoising method for downhole DAS signals of the present invention. Figure 4 This is a schematic diagram illustrating the adaptive wavelet threshold denoising effect of the first embodiment of the adaptive wavelet threshold denoising method for downhole DAS signals according to the present invention. Figure 5 This is a structural block diagram of the first embodiment of the adaptive wavelet threshold denoising system for downhole DAS signals according to the present invention.

[0018] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0019] It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the invention.

[0020] Reference Figure 1 , Figure 1 This is a schematic diagram of the structure of an adaptive wavelet threshold denoising device for downhole DAS signals in the hardware operating environment of the embodiment of the present invention.

[0021] like Figure 1 As shown, the adaptive wavelet threshold denoising device for downhole DAS signals may include: a processor 1001, such as a central processing unit (CPU), a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. The communication bus 1002 is used to establish communication between these components. The user interface 1003 may include a display screen or an input unit such as a keyboard; optionally, the user interface 1003 may also include a standard wired interface or a wireless interface. The network interface 1004 may optionally include a standard wired interface or a wireless interface (such as a Wireless-Fidelity (Wi-Fi) interface). The memory 1005 may be high-speed random access memory (RAM) or stable non-volatile memory (NVM), such as a disk drive. The memory 1005 may also optionally be a storage system independent of the aforementioned processor 1001.

[0022] Those skilled in the art will understand that Figure 1 The structure shown does not constitute a limitation on the adaptive wavelet thresholding denoising device for downhole DAS signals, and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0023] like Figure 1 As shown, the memory 1005, which serves as a storage medium, may include an operating system, a network communication module, a user interface module, and an adaptive wavelet threshold denoising program for downhole DAS signals.

[0024] exist Figure 1In the adaptive wavelet threshold denoising device for downhole DAS signals shown, the network interface 1004 is mainly used for data communication with the network server; the user interface 1003 is mainly used for data interaction with the user; the processor 1001 and memory 1005 in the adaptive wavelet threshold denoising device for downhole DAS signals of the present invention can be set in the adaptive wavelet threshold denoising device for downhole DAS signals. The adaptive wavelet threshold denoising device for downhole DAS signals calls the adaptive wavelet threshold denoising program for downhole DAS signals stored in the memory 1005 through the processor 1001, and executes the adaptive wavelet threshold denoising method for downhole DAS signals provided in the embodiment of the present invention.

[0025] This invention provides an adaptive wavelet thresholding denoising method for downhole DAS signals, referring to... Figure 2 , Figure 2 This is a flowchart illustrating the first embodiment of the adaptive wavelet threshold denoising method for downhole DAS signals according to the present invention.

[0026] In this embodiment, the adaptive wavelet threshold denoising method for downhole DAS signals includes the following steps: S1 performs wavelet decomposition on the downhole DAS signal to obtain low-frequency approximation coefficients and high-frequency detail coefficients.

[0027] It is easy to understand that the execution subject of this embodiment can be an adaptive wavelet threshold denoising system for downhole DAS signals with functions such as data processing, network communication and program execution, or other computer devices with similar functions. This embodiment does not limit it.

[0028] The flow-induced vibration response under different operating conditions was collected using a distributed acoustic sensing (DAS) monitoring system in an actual shale gas production well to obtain DAS signals.

[0029] It should also be noted that the downhole DAS signal is the original noisy DAS vibration signal y(t). A 4-level Symlet4 (sym4) discrete wavelet transform is performed on y(t) to obtain low-frequency approximation coefficients cA4 and 4 levels of high-frequency detail coefficients cD1, cD2, cD3, and cD4. These coefficients are cached to obtain cached wavelet coefficients (cA4, cD1, cD2, cD3, cD4), which are directly used in all subsequent iterations without repeated decomposition.

[0030] It should be noted that, considering the characteristics of downhole DAS signals, which combine transient abrupt changes and strong background noise, this application focuses on the sym4 wavelet. Its fourth-order vanishing moment and the parameter configuration of 7-point tight support length form a good physical fit: the shorter time-domain window ensures accurate positioning of the transient vibration boundary of the flow rate, preventing feature time shift; while the fourth-order vanishing moment exhibits sufficient smoothness in suppressing high-frequency random noise, avoiding the problem of excessive smoothing of weak effective features that is easily caused by higher-order wavelet basis functions. Combined with its approximately symmetrical waveform structure, sym4 can effectively avoid phase distortion during the signal reconstruction stage.

[0031] S2, thresholding the high-frequency detail coefficients using an adaptive soft thresholding function.

[0032] It should be understood that the core of wavelet thresholding denoising is to process the wavelet coefficients through a threshold function to achieve a balance between noise suppression and effective feature preservation. Although traditional hard and soft thresholding functions are widely used in denoising, they both have obvious drawbacks: traditional hard thresholding functions suffer from signal abrupt changes; traditional soft thresholding functions suffer from loss of detail.

[0033] To address the aforementioned shortcomings, this application proposes an improved adaptive soft thresholding function. Based on the traditional soft thresholding function, the core improvement lies in introducing a dynamic adaptive thresholding calculation mechanism, which enables the threshold to be dynamically adjusted according to the absolute value of the input transform domain coefficients. Its core logic consists of two steps: adaptive thresholding calculation and soft thresholding processing.

[0034] Further, S2 includes: S2.1, determining the search interval of the threshold adjustment benchmark and the search interval of the exponential decay rate within the adaptive soft thresholding function; S2.2, generating multiple combinations of threshold parameters to be optimized based on the search interval of the threshold adjustment benchmark and the search interval of the exponential decay rate, and using the combinations of threshold parameters to be optimized as individuals in the population; S2.3, performing thresholding processing on the high-frequency detail coefficients according to each individual in the population using the adaptive soft thresholding function.

[0035] In this embodiment, the population size N=30 and the maximum number of iterations T=100 are set. The parameters to be optimized are determined to be two key parameters in the improved adaptive soft thresholding function: the threshold adjustment benchmark and the exponential decay rate. Since the wavelet decomposition has four layers, each layer's high-frequency coefficients require an independent set of configurations. Therefore, the encoding length of each individual (gray wolf) is 4 × 2 = 8 dimensions, that is:

[0036] Within the search range, 30 such 8-dimensional vectors are randomly and uniformly generated to form the initial population {X1,X2,…,X30}, where each individual represents a set of threshold parameter combinations to be optimized.

[0037] The adaptive soft threshold function is:

[0038]

[0039] In the formula, For adaptive dynamic threshold, For high-frequency detail coefficients, Based on a fixed threshold, Adjust the baseline parameters for the threshold. For the exponential decay rate parameter, For symbolic functions, The high-frequency detail coefficients are the processed values.

[0040] It should also be noted that, The baseline parameter is adjusted to control the critical point for adaptive threshold changes. As the dividing point between the magnitudes of the coefficients, it determines the starting range for threshold adjustment; the value range (i.e., the search interval) is... ; This is the exponential decay rate parameter, which controls how quickly the threshold changes with the absolute value of the coefficient. The larger the value, the higher the sensitivity of the threshold to changes in the coefficient; the range of values ​​(i.e., the search interval) is... ; The larger the value, the greater the adjustment range of the adaptive threshold for large coefficients; The larger the value, the steeper the threshold change, and the stronger the truncation effect on the noise figure.

[0041] This is a sign function used to preserve the sign of the input coefficients; To truncate the operation, ensure that the absolute value of the coefficient minus the adaptive threshold does not result in a negative number, thus avoiding interference from invalid information.

[0042] When input coefficients The absolute value is much greater than When the exponential term approaches 0, the adaptive threshold... Approaching the base threshold This can retain most of the effective coefficients; when inputting coefficients The absolute value is much smaller than When the exponential term approaches a large positive number, the adaptive threshold increases significantly, which can truncate weak noise coefficients to 0, thus enhancing the noise reduction effect; input coefficients The absolute value equals Time-adaptive threshold is This forms the key dividing point for threshold adjustment.

[0043] exist yes The composite function and the exponential function are continuous, therefore continuous; continuous, continuous, exist The points are continuous, therefore exist Continuous; when hour, For finite values, ,so And both the left and right limits are 0, so in The places are also continuous.

[0044] In the specific implementation, for each individual Xi, its eight parameters are... Substituting the values ​​into the improved adaptive soft thresholding function, thresholding is performed on the high-frequency detail coefficients of the corresponding layers. Specifically: The first layer high-frequency coefficient cD1 uses parameters. ; The second layer high-frequency coefficient cD2 uses parameters. ; The third layer high-frequency coefficient cD3 uses parameters. ; The fourth layer high-frequency coefficient cD4 uses parameters. .

[0045] The low-frequency approximation coefficient cA4 is not processed and is directly retained (because the low-frequency coefficient mainly contains the overall trend information of the signal, while noise energy is mainly distributed in the high-frequency details).

[0046] S3. Perform wavelet reconstruction on the processed high-frequency detail coefficients and the low-frequency approximation coefficients, and calculate the signal-to-noise ratio of the reconstructed time-domain signal.

[0047] In the specific implementation, the processed four-layer high-frequency coefficients and the unprocessed low-frequency approximation coefficients are subjected to inverse wavelet transform (IDWT) to reconstruct the time-domain signal corresponding to the denoised i-th individual. .

[0048] The reference noiseless signal s(t) is a known clean signal in the simulation experiment; in actual signal processing, since it is impossible to obtain a real noiseless signal, the original noisy signal y(t) is used as an approximate reference. At this time, SNR and RMSE are only used as a measure of relative quality and do not represent absolute denoising performance.

[0049] For each individual i, calculate its denoised signal (i.e., the reconstructed time-domain signal). Two evaluation metrics relative to the reference signal s(t): Signal-to-noise ratio (SNR) i :

[0050] Root Mean Square Error (RMSE) i :

[0051] In the formula, N This represents the total number of sampling points for the discrete signal.

[0052] S4. Based on the signal-to-noise ratio, the globally optimal parameters are obtained through an improved GWO-HHO hybrid optimization strategy.

[0053] It should be noted that the key parameters (Threshold adjustment baseline) and The (exponential decay rate) is used as the optimization variable. The optimal parameters are found by using the improved GWO-HHO hybrid intelligent optimization algorithm. The complementary advantages of the early global fast search of GWO and the later local fine optimization of HHO are fully combined. The signal-to-noise ratio (SNR) and root mean square error (RMSE) in the denoising scenario are used as the core evaluation indicators. Through iterative optimization of the hybrid algorithm, the parameter combination that enables the model to achieve the global optimal denoising performance is obtained.

[0054] Further, S4 includes: S4.1, selecting a preset number of guide individuals from multiple populations based on the signal-to-noise ratio, and determining the fitness values ​​of individuals in each population; S4.2, calculating the standard deviation of population fitness based on the fitness values ​​of individuals in each population; S4.3, updating the positions of individuals in each population using a hybrid optimization strategy of the improved Grey Wolf Optimizer (GWO) and Harris Hawks Optimizer (HHO) based on the standard deviation of population fitness; S4.4, determining whether the current iteration number is greater than or equal to a preset iteration threshold; S4.5, if yes, determining the optimal population individual based on the preset number of guide individuals, and using the optimal population individual as the global optimal parameter; S4.6, if no, using the population individual with the updated position as the new population individual, and returning to the S2 iteration process.

[0055] In specific implementations, SNR is used. i As fitness (SNR) i (Higher is better), record the three individuals with the highest fitness in the current population, and name them as follows: Alpha Wolf (Optimal): fitness ; Beta Wolf (Second Best): fitness ; Delta Wolf (Third Best): fitness .

[0056] These three individuals (i.e., a predetermined number of guiding individuals) will serve as "leaders" guiding the group's search direction during the GWO phase.

[0057] Calculate the standard deviation of population fitness P (i.e., the population diversity index):

[0058]

[0059] Further, S4.3 includes: S4.3.1, determining whether the standard deviation of the population fitness is greater than or equal to the preset switching threshold; S4.3.2, if yes, the population distribution is scattered and has not yet converged to the neighborhood of the optimal solution, and the position of each individual in each population is updated through the GWO global search strategy; S4.3.3, if no, the population has converged to the neighborhood of the optimal solution, and the position of each individual in each population is updated through the HHO local development strategy.

[0060] In practice, the preset switching threshold can be set to 0.05.

[0061] The process of updating the position of individuals in various groups using the GWO global search strategy is as follows: Calculate the adaptive decay value and dynamic weight coefficient based on the current iteration number and a preset iteration threshold; calculate the coefficient vector based on the adaptive decay value using the encirclement behavior formula; select a leader individual from a preset number of guide individuals, and calculate the distance vector based on the leader individual, the coefficient vector, and individuals in various groups; update the position of individuals in various groups using the position update formula based on the dynamic weight coefficient, the preset number of guide individuals, the distance vector, and the coefficient vector.

[0062] In the specific implementation, the adaptive convergence factor is calculated. :

[0063] in, This represents the current iteration number. The core advantage of this adaptive decay mode, which maximizes the number of iterations, is to achieve rapid global search in the early stages and precise local optimization in the later stages: in the early stages of iteration ( When it is smaller), The value is close to 2, and the coefficient is... The absolute value is more likely to be greater than 1, allowing the gray wolf pack to perform a global search over a larger range, quickly traversing the parameter space and avoiding missing optimal parameter regions; in the later stages of iteration ( near hour), The value rapidly decays to near 0, the coefficient The absolute value is more likely to be less than 1. The gray wolf group shifts from global search to local optimization, which can accurately approximate the optimal parameter combination, adapt to the multi-scale and nonlinear characteristics of two-phase flow signals, and improve the optimization accuracy and adaptability of threshold parameters.

[0064] Calculate the coefficient vector A and C :

[0065]

[0066] in The convergence factor decreases linearly from 2 to 0 with the number of iterations, and is used to control the contraction and expansion of the search range; rand() is a random number between [0,1], which enhances the randomness of the optimization.

[0067] Calculate dynamic weights :

[0068] in The dynamic weights of wolves, namely Alpha, Beta, and Delta, satisfy the following conditions: In the early stages of iteration Significantly greater than and Alpha wolves (globally optimal candidates) dominate the group search direction, reducing redundant paths; in the later stages of iteration, and , The gap between them narrows, and the group performs a refined search in the neighborhood of the optimal solution, thereby improving the accuracy of the optimization.

[0069] Calculate the distance vector Xleader is a random wolf position (i.e., leader individual) selected from Alpha, Beta, and Delta, used to introduce random perturbation to enhance search diversity.

[0070]

[0071] The updated population positions are {X1(t+1),X2(t+1),…,X30(t+1)}.

[0072] The method for updating the position of individuals in various groups using the HHO local development strategy is as follows: determine the escape energy factor based on the adaptive decay value; adaptively select the corresponding encirclement strategy based on the escape energy factor and random numbers, which can be a soft encirclement strategy, a hard encirclement strategy, a soft encirclement-Levy flight progressive attack strategy, or a hard encirclement-Levy flight progressive attack strategy; update the position of individuals in various groups according to the encirclement strategy and dynamic weights.

[0073] In the specific implementation, in order to smoothly connect with the search behavior in the GWO phase, the escape energy factor E adopts the same as the convergence factor. Forms of collaboration:

[0074] Here, rand() is a random number in the interval [0,1], which allows for the escape energy factor to be optimized. With convergence factor Maintaining a synchronous nonlinear decay law: Early stage of iteration The probability is higher, matching the global search phase of GWO; in the later stages of iteration The probability is higher, which matches the local development stage of HHO, and combines the two algorithms at the core parameter level, rather than simply splicing the stages together.

[0075] Based on |E| and the value of the random number r∈[0,1], one of four capture strategies is adaptively selected: If |E| ≥ 0.5 and r ≥ 0.5: Soft encirclement strategy

[0076] If |E| < 0.5 and r ≥ 0.5: Hard encirclement strategy

[0077] If |E| ≥ 0.5 and r < 0.5: Soft encirclement - Levy flight progressive assault strategy

[0078] If |E| < 0.5 and r < 0.5: Hard encirclement - Levy flight progressive assault strategy

[0079] in As an iterative adaptive dynamic weight, it maintains an inherent consistency with the weight evolution logic of the previous GWO stage, thereby ensuring a smooth connection between the two stages of optimization and reducing oscillations during the convergence process.

[0080] The updated population positions are {X1(t+1),X2(t+1),…,X30(t+1)}.

[0081] In this embodiment, after outputting the updated population position, the current iteration number t is increased by 1: t = t + 1. It is judged whether t reaches the maximum iteration number T: if t < T: return to S2, and perform adaptive soft threshold denoising, fitness calculation and position update again with the updated population; if t ≥ T: stop the iteration, and output the global optimal individual (i.e., the global optimal parameter) recorded during the entire iteration process:

[0082] S5: performing adaptive wavelet threshold denoising on the downhole DAS signal based on the global optimal parameter.

[0083] In specific implementation, the optimal parameter ( ) is substituted into the adaptive soft threshold function to process the first layer high-frequency coefficient cD1; similarly, use ( ) to process cD2, use ( ) to process cD2, use ( ) to process cD3, use ( ) to process cD4.

[0084] The processed 4-layer high-frequency coefficients and the unprocessed low-frequency approximate coefficient cA4 are subjected to inverse wavelet transform to obtain the final high-fidelity denoised signal .

[0085] It should be further noted that, with reference to Figure 3 , Figure 3 is a schematic diagram of the traditional threshold denoising effect of the first embodiment of the adaptive wavelet threshold denoising method for downhole DAS signals of the present invention. The traditional threshold denoising method has the worst denoising effect, the whole waveform has obvious burrs and fluctuations, a large amount of residual noise, and a significant loss of effective signal components—the signal peak is excessively truncated, and the waveform details are smoothed out, so that the pulsation characteristics and key features of the original signal cannot be completely retained. This is because the traditional threshold adopts a fixed threshold mechanism, does not consider the non-stationarity and multi-scale characteristics of the noisy signal, the threshold adjustment lacks adaptability, and it is difficult to balance the denoising effect and the retention of effective signals.

[0086] With reference to Figure 4 , Figure 4 is a schematic diagram of the adaptive wavelet threshold denoising effect of the first embodiment of the adaptive wavelet threshold denoising method for downhole DAS signals of the present invention. The denoising effect of the present application is superior: the whole waveform is smooth and regular, without obvious burrs and noise residues, and can weaken the high-frequency noise in the noisy signal; at the same time, the waveform can completely retain all key features of the original signal, including the amplitude and pulsation law of the peak and valley values, has the highest fit with the reference noiseless signal, no obvious signal distortion phenomenon, which verifies the effectiveness of the improvement to a certain extent.

[0087] In this embodiment, the downhole DAS signal is first decomposed by wavelet to obtain low-frequency approximation coefficients and high-frequency detail coefficients. Then, the high-frequency detail coefficients are thresholded by an adaptive soft thresholding function. The processed high-frequency detail coefficients and low-frequency approximation coefficients are then reconstructed by wavelet, and the signal-to-noise ratio of the reconstructed time-domain signal is calculated. Subsequently, the globally optimal parameters are obtained by an improved GWO-HHO hybrid optimization strategy based on the signal-to-noise ratio. Finally, the downhole DAS signal is subjected to adaptive wavelet threshold denoising based on the globally optimal parameters. This embodiment constructs a continuous improved adaptive soft threshold function, combined with a dynamic threshold calculation mechanism, which effectively overcomes the amplitude deviation and pseudo-oscillation defects of traditional threshold functions, achieving a precise balance between noise suppression and feature preservation. Then, an improved GWO-HHO collaborative optimization strategy is proposed: by optimizing the convergence factor of GWO to an adaptive decay mode to enhance the global search capability in the early stages, the HHO algorithm is introduced, and iterative adaptive dynamic weights are used to ensure a smooth transition between the two optimization stages. This fully leverages the advantages of HHO in the fine development of the local neighborhood of the optimal solution, thereby accurately obtaining the globally optimal threshold parameters, thus improving the denoising effect on downhole DAS signals and successfully extracting relatively pure downhole flow feature signals.

[0088] Reference Figure 5 , Figure 5 This is a structural block diagram of the first embodiment of the adaptive wavelet threshold denoising system for downhole DAS signals according to the present invention.

[0089] like Figure 5 As shown, the adaptive wavelet threshold denoising system for downhole DAS signals proposed in this embodiment of the invention includes: The wavelet decomposition module 5001 is used to perform wavelet decomposition on downhole DAS signals to obtain low-frequency approximation coefficients and high-frequency detail coefficients. The adaptive soft thresholding module 5002 is used to perform thresholding on the high-frequency detail coefficients through an adaptive soft thresholding function. The signal reconstruction module 5003 is used to perform wavelet reconstruction on the processed high-frequency detail coefficients and the low-frequency approximation coefficients, and to calculate the signal-to-noise ratio of the reconstructed time-domain signal. The parameter optimization module 5004 is used to obtain the globally optimal parameters based on the signal-to-noise ratio by improving the GWO-HHO hybrid optimization strategy. The wavelet threshold denoising module 5005 is used to perform adaptive wavelet threshold denoising on the downhole DAS signal based on the global optimal parameters.

[0090] Other embodiments or specific implementations of the adaptive wavelet threshold denoising system for downhole DAS signals of the present invention can be found in the above-described method embodiments, and will not be repeated here.

[0091] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element.

[0092] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0093] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as read-only memory / random access memory, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of the present invention.

[0094] The above are merely preferred embodiments of the present invention and do not limit the scope of the patent. Any equivalent structural or procedural transformations made based on the description and drawings of the present invention, or direct or indirect applications in other related technical fields, are similarly included within the scope of patent protection of the present invention.

Claims

1. An adaptive wavelet thresholding denoising method for downhole DAS signals, characterized in that, The method includes the following steps: S1, perform wavelet decomposition on the downhole DAS signal to obtain low-frequency approximation coefficients and high-frequency detail coefficients; S2, thresholding the high-frequency detail coefficients using an adaptive soft thresholding function; S3, perform wavelet reconstruction on the processed high-frequency detail coefficients and the low-frequency approximation coefficients, and calculate the signal-to-noise ratio of the reconstructed time-domain signal; S4, based on the signal-to-noise ratio, the globally optimal parameters are obtained through an improved GWO-HHO hybrid optimization strategy; S5, perform adaptive wavelet threshold denoising on the downhole DAS signal based on the global optimal parameters.

2. The method as described in claim 1, characterized in that, S2 includes: S2.1, Determine the search range for the threshold adjustment benchmark and the search range for the exponential decay rate within the adaptive soft threshold function; S2.2, Based on the search interval of the threshold adjustment benchmark and the search interval of the exponential decay rate, generate multiple combinations of threshold parameters to be optimized, and use the combinations of threshold parameters to be optimized as individuals in the population; S2.3, thresholding the high-frequency detail coefficients according to the individual groups using an adaptive soft thresholding function.

3. The method as described in claim 2, characterized in that, The adaptive soft threshold function is: In the formula, For adaptive dynamic threshold, For high-frequency detail coefficients, Based on a fixed threshold, Adjust the baseline parameters for the threshold. For the exponential decay rate parameter, For symbolic functions, The high-frequency detail coefficients are the processed values.

4. The method as described in claim 1, characterized in that, The S4 includes: S4.1, Based on the signal-to-noise ratio, select a preset number of guide individuals from multiple populations and determine the fitness value of each population individual; S4.2 Calculate the standard deviation of population fitness based on the fitness values ​​of individuals in each population group; S4.3, Update the position of individuals in each population using an improved GWO-HHO hybrid optimization strategy based on the population fitness standard deviation; S4.4, Determine whether the current iteration number is greater than or equal to the preset iteration threshold; S4.5 If so, then determine the optimal population individual based on the preset number of guiding individuals, and use the optimal population individual as the global optimal parameter; S4.6 If not, then the population individual with the updated position is taken as the new population individual, and the process returns to S2 iteration.

5. The method as described in claim 4, characterized in that, S4.3 includes: S4.3.1 Determine whether the standard deviation of the population fitness is greater than or equal to a preset switching threshold; S4.3.2, if so, then update the position of individuals in various groups using the GWO global search strategy; S4.3.3, if not, update the position of individuals in each population using the HHO local development strategy.

6. The method as described in claim 5, characterized in that, The method of updating the positions of various group individuals using the GWO global search strategy includes: Calculate the adaptive decay value and dynamic weight coefficient based on the current iteration number and the preset iteration threshold; The coefficient vector is calculated based on the adaptive attenuation value using the enclosing behavior formula; Select a leader individual from the preset number of guiding individuals, and calculate a distance vector based on the leader individual, the coefficient vector, and the individuals of each group; The positions of individuals in each group are updated using a position update formula based on the dynamic weight coefficient, the preset number of guiding individuals, the distance vector, and the coefficient vector.

7. The method as described in claim 5, characterized in that, The method of updating the location of individuals in various populations through the HHO local development strategy includes: The escape energy factor is determined based on the adaptive decay value; The corresponding encirclement strategy is adaptively selected based on the escape energy factor and random number. The encirclement strategy is a soft encirclement strategy, a hard encirclement strategy, a soft encirclement-Levy flight progressive attack strategy, or a hard encirclement-Levy flight progressive attack strategy. The position of individuals in each group is updated according to dynamic weights based on the described encirclement strategy.

8. An adaptive wavelet threshold denoising system for downhole DAS signals, characterized in that, The system includes: The wavelet decomposition module is used to perform wavelet decomposition on downhole DAS signals to obtain low-frequency approximation coefficients and high-frequency detail coefficients. An adaptive soft thresholding module is used to perform thresholding on the high-frequency detail coefficients using an adaptive soft thresholding function. The signal reconstruction module is used to perform wavelet reconstruction on the processed high-frequency detail coefficients and the low-frequency approximation coefficients, and to calculate the signal-to-noise ratio of the reconstructed time-domain signal. The parameter optimization module is used to obtain the globally optimal parameters based on the signal-to-noise ratio by improving the GWO-HHO hybrid optimization strategy; The wavelet threshold denoising module is used to perform adaptive wavelet threshold denoising on the downhole DAS signal based on the globally optimal parameters.

9. An adaptive wavelet threshold denoising device for downhole DAS signals, characterized in that, The device includes: a memory, a processor, and an adaptive wavelet thresholding denoising program for downhole DAS signals stored in the memory and executable on the processor, the adaptive wavelet thresholding denoising program for downhole DAS signals being configured to implement the steps of the adaptive wavelet thresholding denoising method for downhole DAS signals as claimed in any one of claims 1 to 7.

10. A storage medium, characterized in that, The storage medium stores an adaptive wavelet threshold denoising program for downhole DAS signals. When the processor executes the adaptive wavelet threshold denoising program for downhole DAS signals, it implements the steps of the adaptive wavelet threshold denoising method for downhole DAS signals as described in any one of claims 1 to 7.

Images