Method and device for predicting formation longitudinal wave velocity

Abstract

The present invention discloses a method and device for predicting formation P-wave velocity. The method comprises: performing conventional well logging and full-wavelength acoustic logging on a cased well section to obtain natural gamma, compensated neutron, and array waveform data; processing the array waveform data to extract the P-wave velocity and S-wave velocity of the cased well section; performing data cleaning based on the P-wave velocity, S-wave velocity, natural gamma, and compensated neutron; using the S-wave velocity, natural gamma, and compensated neutron as sample data, and the P-wave velocity as corresponding sample label data, to train a preset cased well acoustic velocity prediction model, and then predicting the P-wave velocity based on the trained preset cased well acoustic velocity prediction model. The method also includes performing data cleaning based on correlation analysis between various formation parameters, and using the cleaned data for training to learn the complex nonlinear mapping relationship between the data, thereby accurately extracting the P-wave velocity of the formation.

Description

Technical Field

[0001] The embodiments of the present invention relate to the field of geophysical acoustic logging, and in particular to a method and device for predicting formation longitudinal wave velocity. Background Art

[0002] In the field of petroleum exploration, multiple cementing cycles are typically performed in oil and gas wells to prevent lost circulation and maintain wellbore integrity. Cementing involves injecting cement slurry downhole through casing. Once at the bottom of the well, it rises along the annulus between the casing and the formation. Over time, the cement solidifies and forms a cement sheath. Obtaining information such as formation sound velocity through the casing and cement sheath is crucial for oil and gas reservoir development. However, when cementing quality is poor, the tube waves propagating through the casing can severely interfere with the formation waves containing valuable information, making accurate formation sound velocity determination impossible using conventional methods.

[0003] Existing technology uses a dual-source flyback instrument. By precisely designing the arrangement position, transmission delay, and amplitude ratio of two sound sources with opposite emission polarities, the interference of casing waves can be offset at the instrument level, thereby obtaining interference-free formation wave signals. However, this type of instrument has not yet been widely used in field data acquisition, and there is insufficient available data. The main existing cased well sound velocity processing method is based on the differences in speed, frequency, energy, and other characteristics between casing waves and formation waves. The two are separated to eliminate casing wave interference and obtain formation sound velocity. However, the above signal processing method has limitations: in fast formations, when the speed of the formation wave is similar to that of the casing wave, the energy of the formation wave is much smaller than that of the casing wave. The effective signal is submerged in the casing wave and other noise, and the formation wave signal cannot be extracted through time domain, frequency domain, and energy differences. Therefore, it is impossible to obtain accurate formation sound velocity, especially the formation longitudinal wave speed. Summary of the Invention

[0004] In view of the above problems, embodiments of the present invention are proposed to provide a method and apparatus for predicting formation longitudinal wave velocity that overcomes the above problems or at least partially solves the above problems.

[0005] According to one aspect of an embodiment of the present invention, a method for predicting formation compressional wave velocity is provided, the method comprising:

[0006] Conduct conventional logging and full-wavelength acoustic logging on the cased hole section to obtain natural gamma ray, compensated neutron and array waveform data;

[0007] Process the array waveform data to extract the longitudinal and shear wave velocities in the cased well section;

[0008] Data cleaning is performed based on longitudinal wave velocity, shear wave velocity, natural gamma and compensated neutrons;

[0009] The shear wave velocity, natural gamma and compensated neutron processed by data cleaning are used as sample data, and the longitudinal wave velocity processed by data cleaning is used as the corresponding sample label data to train a preset cased well acoustic wave velocity prediction model, so as to predict the longitudinal wave velocity based on the trained preset cased well acoustic wave velocity prediction model.

[0010] According to another aspect of an embodiment of the present invention, a device for predicting formation compressional wave velocity is provided, comprising:

[0011] Logging module, suitable for conventional logging and full-wavelength acoustic logging of cased well sections, obtaining natural gamma, compensated neutron and array waveform data;

[0012] The waveform processing module is suitable for processing array waveform data and extracting the longitudinal wave speed and shear wave speed of the cased well section;

[0013] Data cleaning module, suitable for data cleaning processing based on longitudinal wave sound velocity, shear wave sound velocity, natural gamma and compensated neutron;

[0014] The model prediction module is suitable for using the shear wave speed, natural gamma and compensated neutron processed by data cleaning as sample data, and the longitudinal wave speed processed by data cleaning as corresponding sample label data to train a preset cased well acoustic wave speed prediction model, so as to predict the longitudinal wave speed based on the preset cased well acoustic wave speed prediction model obtained by training.

[0015] According to another aspect of an embodiment of the present invention, there is provided a computing device, comprising: a processor, a memory, a communication interface, and a communication bus, wherein the processor, the memory, and the communication interface communicate with each other via the communication bus;

[0016] The memory is used to store at least one executable instruction, and the executable instruction enables the processor to execute operations corresponding to the above-mentioned formation compressional wave velocity prediction method.

[0017] According to another aspect of an embodiment of the present invention, a computer storage medium is provided, wherein the storage medium stores at least one executable instruction, and the executable instruction enables a processor to execute operations corresponding to the above-mentioned formation compressional wave velocity prediction method.

[0018] According to another aspect of an embodiment of the present invention, a computer program product is provided, comprising at least one executable instruction, wherein the executable instruction enables a processor to execute operations corresponding to the above-mentioned method for predicting formation compressional wave velocity.

[0019] According to the method and device for predicting the longitudinal wave velocity of a formation provided by an embodiment of the present invention, the correlation between the longitudinal and shear wave velocities of the formation and various formation parameters such as natural gamma and compensated neutrons is analyzed, and data cleaning processing is performed to remove abnormal data. The data after data cleaning processing are used to train a preset cased well acoustic wave velocity prediction model, and the complex nonlinear mapping relationship between the shear wave velocity, natural gamma, compensated neutrons and longitudinal wave velocity is learned. In this way, when faced with poor cementation quality or the presence of free casing, the interference of casing waves and the lack of effective signals are overcome, and the longitudinal wave velocity of the formation is extracted that is consistent with geological understanding, which also provides solid data support for achieving accurate formation evaluation.

[0020] The above description is only an overview of the technical solutions of the embodiments of the present invention. In order to more clearly understand the technical means of the embodiments of the present invention, they can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the embodiments of the present invention more obvious and easy to understand, the specific implementation methods of the embodiments of the present invention are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Various other advantages and benefits will become apparent to those skilled in the art upon reading the detailed description of the preferred embodiment below. The accompanying drawings are for illustration purposes only and are not to be considered as limiting the embodiments of the present invention. The same reference numerals are used throughout the accompanying drawings to denote the same components. In the accompanying drawings:

[0022] Figure 1 A flow chart of a method for predicting formation longitudinal wave velocity according to an embodiment of the present invention is shown;

[0023] Figure 2 The cross-plot without data cleaning is shown;

[0024] Figure 3 shows the crossplot after data cleaning;

[0025] Figure 4 A schematic diagram of the structure of a preset cased well acoustic wave velocity prediction model is shown;

[0026] Figure 5 A schematic diagram of cased hole formation acoustic logging data processing results is shown;

[0027] Figure 6 A schematic structural diagram of a device for predicting formation longitudinal wave velocity according to an embodiment of the present invention is shown;

[0028] Figure 7 A schematic structural diagram of a computing device according to an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0029] Exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present invention and to fully convey the scope of the present invention to those skilled in the art.

[0030] Figure 1 FIG. 4 shows a flow chart of a method for predicting formation longitudinal wave velocity according to an embodiment of the present invention. Figure 1 As shown, the method includes the following steps:

[0031] Step S101 : performing conventional logging and full-wavelength acoustic logging on the cased well section to obtain natural gamma ray, compensated neutron and array waveform data.

[0032] This embodiment can be used to predict the longitudinal wave velocity of the formation in an environment with poor cement bonding quality or in a free casing environment. It is not affected by casing waves and other noises and can accurately predict the longitudinal wave velocity of the formation.

[0033] Specifically, conventional logging and full-wavelength acoustic logging are performed on the cased hole section using instruments. Conventional logging can generate natural gamma ray (GR) and compensated neutron (CNL), while full-wavelength acoustic logging can generate array waveform data. The array waveform data includes monopole array acoustic wave signals and dipole array acoustic wave signals.

[0034] Step S102 : Process the array waveform data to extract the longitudinal wave velocity and the shear wave velocity of the cased well section.

[0035] The array waveform data is processed using the STC time-slowness coherence method to obtain the time-slowness coherence diagram of the array waveform data in the time window, as shown in the following formula:

[0036]

[0037] Where ρ(s,T) is the coherence of the array waveform data at time T and the wave propagation slowness is s, T w is the time window length, m is the acoustic receiver serial number, N is the number of receivers, X m (t) is the mth receiver in the time window [T, T+T ω ] is the waveform amplitude at a certain time t, and d is the receiver spacing.

[0038] According to the above formula, the coherence of ρ(s,T) is obtained. The time-slowness coherence diagram of the array waveform data based on the time window T is constructed by using a drawing tool. The formation slowness s at the maximum coherence point is obtained according to the time-slowness coherence diagram. According to the formation slowness s at the maximum coherence point, the array sound velocity can be extracted. The array sound velocity and the formation slowness are inversely proportional to each other. . If X in the formula m When (t) is the waveform amplitude of the monopole array acoustic wave signal of the array waveform data, the formation slowness s at the point of maximum coherence is obtained, and the longitudinal wave speed of the cased well section is determined based on the formation slowness s. The longitudinal wave speed and the formation slowness obtained based on the monopole array acoustic wave signal of the array waveform data are inversely proportional to each other. When X in the formula m When (t) is the waveform amplitude of the dipole array acoustic wave signal of the array waveform data, the formation slowness s at the point of maximum coherence is obtained, and the shear wave speed of the cased well section is determined based on the formation slowness s. The shear wave speed and the formation slowness obtained based on the dipole array acoustic wave signal of the array waveform data are inversely proportional to each other.

[0039] Because the propagation speed of shear waves differs significantly from that of casing waves, their waveforms can be well separated in the time domain. Therefore, the shear wave velocity extracted using the STC time slowness coherence method can eliminate the interference of casing waves, allowing accurate shear wave velocity to be obtained even in poorly cemented sections. Regarding the P-wave velocity, in sections with good cementing quality, the P-wave velocity extracted using the STC time slowness coherence method is the true and accurate P-wave velocity. However, in sections with poor cementing quality, the P-wave velocity is affected by casing waves, resulting in the P-wave velocity being equal to the casing wave propagation velocity, rather than the true and accurate P-wave velocity of the formation. Therefore, the extracted P-wave velocity needs to be processed to remove the P-wave velocity that reflects the casing wave propagation velocity in order to obtain an accurate P-wave velocity for subsequent predictions.

[0040] Step S103 , performing data cleaning processing based on the longitudinal wave velocity, the shear wave velocity, the natural gamma ray and the compensated neutron.

[0041] Since there are inaccurate data in the extracted longitudinal wave sound velocity, the longitudinal wave sound velocity also needs to be cleaned. When cleaning the data, it can be processed through the intersection diagram. For example, the longitudinal wave time difference can be determined based on the longitudinal wave sound velocity, and the shear wave time difference can be determined based on the shear wave sound velocity. An intersection diagram can be constructed based on the longitudinal wave time difference, shear wave time difference, natural gamma and compensated neutron as a group of variables. Furthermore, before constructing the intersection diagram, considering that the data obtained from the well logging may have instrument measurement errors, it is necessary to first perform data invalidation removal on the natural gamma and compensated neutron according to the preset invalid conditions. The preset invalid conditions can be set according to the implementation situation. For example, the natural gamma and compensated neutron obtained at data depths such as -9999, 0, and -9999.25 are invalid data, and they can be subjected to data invalidation removal by setting, for example, an invalid flag. After the data invalidation removal, any two data of the longitudinal wave time difference, shear wave time difference, natural gamma after data invalidation removal, and compensated neutron are used as horizontal and vertical coordinates, respectively, and combined to construct an intersection diagram. For example Figure 2As shown, the first column uses natural gamma as the horizontal axis, and the vertical axes from top to bottom are natural gamma, compensated neutrons, longitudinal wave time difference, and transverse wave time difference; the second column uses compensated neutrons as the horizontal axis, and the vertical axes from top to bottom are natural gamma, compensated neutrons, longitudinal wave time difference, and transverse wave time difference; the third column uses longitudinal wave time difference as the horizontal axis, and the vertical axes from top to bottom are natural gamma, compensated neutrons, longitudinal wave time difference, and transverse wave time difference; the fourth column uses transverse wave time difference as the horizontal axis, and the vertical axes from top to bottom are natural gamma, compensated neutrons, longitudinal wave time difference, and transverse wave time difference. According to Figure 2 The intersection diagram shown in the figure can determine the abnormal longitudinal wave sound velocity based on the correlation of the various data displayed therein, such as the data part with natural gamma greater than 200API, and the data part with longitudinal wave time difference less than 58μs / ft, that is, the data part with longitudinal wave sound velocity greater than 5200m / s has obvious difference in correlation with other data in the intersection diagram. This part of data is not aggregated with other data. It can be determined that this part of data is abnormal, and the data is removed accordingly, that is, the abnormal natural gamma data and abnormal longitudinal wave sound velocity are removed. Here, according to geological knowledge, the data with natural gamma greater than 200API is abnormal due to instrument measurement, and the data with longitudinal wave time difference less than 58μs / ft is the abnormal casing wave sound velocity extracted due to poor cementation quality. Therefore, it is necessary to perform data cleaning on the above abnormal data. According to the longitudinal wave sound velocity, shear wave sound velocity, natural gamma and compensated neutron after data cleaning, the following is obtained: Figure 3 The intersection diagram shown is based on Figure 3 As can be seen from the cross-plot shown, the data signal-to-noise ratio and correlation are significantly enhanced after data cleaning, providing a good data basis for subsequent modeling and prediction.

[0042] In step S104, the shear wave velocity, natural gamma, and compensated neutrons processed by data cleaning are used as sample data, and the longitudinal wave velocity processed by data cleaning is used as the corresponding sample label data to train a preset cased well acoustic wave velocity prediction model, so as to predict the longitudinal wave velocity based on the trained preset cased well acoustic wave velocity prediction model.

[0043] The shear wave speed, natural gamma and compensated neutron after data cleaning are used as sample data, and the longitudinal wave speed after data cleaning is used as the corresponding sample label data to construct a sample set.

[0044] A preset cased well acoustic velocity prediction model is constructed based on a bidirectional long short-term memory neural network and a fully connected neural network. The model structure is as follows: Figure 4 As shown, it includes a bidirectional long short-term memory neural network with 256 hidden states and a fully connected neural network with 512 neurons. Figure 4The model also includes the shape change of data during forward propagation in the preset cased-hole acoustic velocity prediction model. The loss function for the preset cased-hole acoustic velocity prediction model is set to the mean square error (MSE) loss function, and the Adam algorithm is used for iterative optimization of model parameters. For example, the initial learning rate is set to 0.001, the batch size is set to 16, and training is performed for 200 epochs. The learning rate decay uses cosine annealing. The model input is a three-dimensional matrix with a shape of 16 × 10 × 3. The first dimension of this matrix, 16, represents the batch size. The second dimension, 10, represents the data of 10 consecutive depth points as a sample. 3 indicates that each sample contains three features: natural gamma, compensated neutrons, and shear wave velocity. The output is a two-dimensional matrix with a shape of (number of samples (e.g., batch size 16 × depth points per sample, 10) × 10) × 1, representing the P-wave velocity corresponding to all input depth points. This sample set enables the preset cased-hole acoustic velocity prediction model to learn the comprehensive impact of other depth points along the well axis on the P-wave velocity of the formation, rather than just considering the influence of a single depth point. The above is an example for explanation. The specific settings will be based on the implementation situation and are not limited here.

[0045] By training a preset cased well acoustic wave velocity prediction model and iteratively optimizing the model parameters, a trained preset cased well acoustic wave velocity prediction model can be obtained. The preset cased well acoustic wave velocity prediction model can be applied to the processing of various formation acoustic waves, and can obtain natural gamma, compensated neutron, and shear wave velocities for the entire well section. The shear wave velocity is extracted and processed based on the array waveform data for the entire well section. Refer to the description of step S102 and will not be repeated here. By inputting the natural gamma, compensated neutron, and shear wave velocities for the entire well section into the trained preset cased well acoustic wave velocity prediction model, the longitudinal wave velocity for the entire well section can be predicted.

[0046] Taking the 200m long well section of cased well formation acoustic logging data as an example, the processing results are as follows: Figure 5 shown. Figure 5The first trace is the depth; the second trace contains the waveform and the arrival time of the longitudinal wave (calculated using the sound velocity obtained by the STC time slowness coherence method). From the waveform and the arrival time of the longitudinal wave, we can clearly see that the depth is 3985~4125 The casing wave in the m well section is strong, and the obtained longitudinal wave is basically a straight line when it arrives, which cannot reflect the formation information; the third track is a curve of the cement bonding index used to evaluate the cementing quality. The higher the value of the cement bonding index curve, the worse the cement bonding quality; the fourth track is the longitudinal wave time difference obtained by the time slowness coherence method and the longitudinal wave time difference obtained by the embodiment of the present application (the longitudinal wave time difference is calculated based on the longitudinal wave velocity). According to the two data, it can be determined that in the well section with poor bonding quality, the longitudinal wave time difference obtained by the STC time slowness coherence method cannot accurately reflect the longitudinal wave sound velocity of the formation, and the longitudinal wave time difference of the formation obtained by the embodiment of the present application is the accurate longitudinal wave sound velocity of the formation; the fifth track is combined with the correlation coefficient of the array sound wave to verify the effectiveness of the embodiment of the present application. In the well section with poor bonding quality, the waveform correlation corresponding to the casing wave velocity is the largest, and the correlation of the formation longitudinal wave velocity is small or basically 0. At places where the P-wave velocity correlation is weak but not zero, it can be seen that the P-wave velocity of the formation obtained using this application (the P-wave time difference is calculated) is consistent with the waveform correlation, and the results at places where the P-wave correlation is basically zero are also consistent with geological understanding.

[0047] According to the method for predicting formation P-wave velocity provided by an embodiment of the present invention, the correlation between the P-wave and S-wave velocity of the formation and various formation parameters such as natural gamma and compensated neutrons is analyzed, and data cleaning is performed to remove abnormal data. The data after data cleaning are used to train a preset cased well acoustic wave velocity prediction model, and the complex nonlinear mapping relationship between S-wave velocity, natural gamma, compensated neutrons and P-wave velocity is learned. In this way, when faced with poor cementation quality or the presence of free casing, the interference of casing waves and the lack of effective signals are overcome, and the P-wave velocity of the formation is extracted that is consistent with geological understanding, which also provides solid data support for achieving accurate formation evaluation.

[0048] Figure 6 FIG. 1 shows a schematic diagram of the structure of a device for predicting the longitudinal wave velocity of a formation provided by an embodiment of the present invention. Figure 6 As shown, the device includes:

[0049] Logging module 610, suitable for performing conventional logging and full-wavelength acoustic logging on cased well sections, obtaining natural gamma, compensated neutron and array waveform data;

[0050] The waveform processing module 620 is adapted to process the array waveform data and extract the longitudinal and shear wave velocities of the cased well section;

[0051] A data cleaning module 630 is adapted to perform data cleaning processing based on longitudinal wave velocity, shear wave velocity, natural gamma and compensated neutrons;

[0052] The model prediction module 640 is suitable for using the shear wave speed, natural gamma and compensated neutron processed by data cleaning as sample data, and the longitudinal wave speed processed by data cleaning as corresponding sample label data to train a preset cased well acoustic wave speed prediction model, so as to predict the longitudinal wave speed based on the preset cased well acoustic wave speed prediction model obtained by training.

[0053] Optionally, the waveform processing module 620 is further adapted to:

[0054] The array waveform data is processed using the time-slowness coherence method to obtain the time-slowness coherence map of the array waveform data in the time window;

[0055] According to the time-slowness coherence diagram, the formation slowness at the point of maximum coherence is obtained;

[0056] The longitudinal and shear wave velocities of the cased well section are determined based on the formation slowness; the longitudinal wave velocity and the formation slowness obtained from the monopole array acoustic signal based on the array waveform data are inversely proportional to each other; and the shear wave velocity and the formation slowness obtained from the dipole array acoustic signal based on the array waveform data are inversely proportional to each other.

[0057] Optionally, the data cleaning module 630 is further adapted to:

[0058] Determining the longitudinal wave time difference based on the longitudinal wave sound velocity, and determining the shear wave time difference based on the shear wave sound velocity;

[0059] According to the longitudinal wave time difference, shear wave time difference, natural gamma and compensated neutron, intersection diagrams are constructed in pairs to determine abnormal data and perform data cleaning on the abnormal data.

[0060] Optionally, the data cleaning module 630 is further adapted to:

[0061] According to the preset invalid conditions, the natural gamma and compensated neutron data are invalidated;

[0062] The intersection diagram is constructed by combining any two data of longitudinal wave time difference, shear wave time difference, natural gamma after data invalidation removal and compensated neutron.

[0063] Optionally, the model prediction module 640 is further adapted to:

[0064] A preset cased well acoustic velocity prediction model is constructed based on a bidirectional long short-term memory neural network and a fully connected neural network;

[0065] The shear wave velocity, natural gamma and compensated neutron after data cleaning are used as sample data to construct a three-dimensional matrix for the input of the preset cased well acoustic wave velocity prediction model; and the longitudinal wave velocity after data cleaning is used as the corresponding sample label data to train the preset cased well acoustic wave velocity prediction model, and the model parameters are iteratively optimized to obtain the trained preset cased well acoustic wave velocity prediction model.

[0066] Optionally, the model prediction module 640 is further adapted to:

[0067] Obtain natural gamma, compensated neutron, and shear wave velocity for the entire well section; the shear wave velocity is extracted by processing the array waveform data for the entire well section;

[0068] The natural gamma, compensated neutron and shear wave velocity of the entire well section are input into the trained preset cased well acoustic wave velocity prediction model to predict the longitudinal wave velocity of the entire well section.

[0069] The description of each module above refers to the corresponding description in the method embodiment and will not be repeated here.

[0070] An embodiment of the present invention further provides a non-volatile computer storage medium storing at least one executable instruction, which can execute operations corresponding to the method for predicting formation compressional wave velocity in any of the above method embodiments.

[0071] An embodiment of the present application provides a computer program product, which includes at least one executable instruction or computer program, and the executable instruction or computer program can enable a processor to perform operations corresponding to the method for predicting formation compressional wave velocity in any of the above method embodiments.

[0072] Figure 7 A schematic structural diagram of a computing device according to an embodiment of the present invention is shown. The specific implementation of the computing device is not limited to the specific implementation of the computing device in the specific embodiment of the present invention.

[0073] like Figure 7 As shown, the computing device may include a processor 702 , a communication interface 704 , a memory 706 , and a communication bus 708 .

[0074] in:

[0075] The processor 702 , the communication interface 704 , and the memory 706 communicate with each other via a communication bus 708 .

[0076] The communication interface 704 is used to communicate with other devices such as clients or other servers.

[0077] The processor 702 is configured to execute the program 710 , and specifically to execute the relevant steps in the above-mentioned embodiment of the method for predicting the longitudinal wave velocity of the formation.

[0078] Specifically, the program 710 may include program codes, which include computer operation instructions.

[0079] Processor 702 may be a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement embodiments of the present invention. The one or more processors included in a computing device may be of the same type, such as one or more CPUs, or may be of different types, such as one or more CPUs and one or more ASICs.

[0080] The memory 706 is used to store the program 710. The memory 706 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), such as at least one disk storage.

[0081] Program 710 can specifically be used to cause processor 702 to execute the formation compressional wave velocity prediction method described in any of the above-described method embodiments. The specific implementation of each step in program 710 can be found in the corresponding descriptions of the corresponding steps and units in the above-described formation compressional wave velocity prediction embodiments, and will not be repeated here. Those skilled in the art will clearly understand that, for ease and brevity of description, the specific operating processes of the devices and modules described above can refer to the corresponding process descriptions in the above-described method embodiments, and will not be repeated here.

[0082] The algorithm or display provided herein is not inherently related to any particular computer, virtual system or other device. Various general-purpose systems can also be used together with the teachings based on this. According to the above description, it is obvious that the structure required for constructing such systems. In addition, the embodiment of the present invention is not directed to any specific programming language. It should be understood that various programming languages ​​can be utilized to implement the content of the embodiment of the present invention described herein, and the above description of specific languages ​​is for the purpose of disclosing the preferred implementation of the embodiment of the present invention.

[0083] In the description provided herein, numerous specific details are described. However, it is understood that embodiments of the present invention may be practiced without these specific details. In some instances, well-known methods, structures, and techniques are not shown in detail so as not to obscure the understanding of this description.

[0084] Similarly, it should be understood that in order to streamline the embodiments of the invention and aid in understanding one or more of the various inventive aspects, in the above description of exemplary embodiments of the invention, various features of the embodiments of the invention are sometimes grouped together into a single embodiment, figure, or description thereof. However, this disclosed approach should not be interpreted as reflecting an intention that the claimed embodiments of the invention require more features than are expressly recited in each claim. Rather, as reflected in the claims below, inventive aspects lie in less than all of the features of the individual embodiments disclosed above. Accordingly, the claims that follow the detailed description are hereby expressly incorporated into this detailed description, with each claim standing on its own as a separate embodiment of the invention.

[0085] Those skilled in the art will appreciate that the modules in the devices of the embodiments can be adaptively modified and installed in one or more devices different from the embodiments. The modules, units, or components in the embodiments can be combined into a single module, unit, or component, and furthermore, they can be divided into multiple sub-modules, sub-units, or sub-components. All features disclosed in this specification (including the accompanying claims, abstract, and drawings), and all processes or units of any method or device disclosed therein, can be combined in any combination, unless at least some of such features and / or processes or units are mutually exclusive. Unless expressly stated otherwise, each feature disclosed in this specification (including the accompanying claims, abstract, and drawings) may be replaced by an alternative feature that serves the same, equivalent, or similar purpose.

[0086] Furthermore, those skilled in the art will appreciate that although some embodiments herein include certain features included in other embodiments but not other features, combinations of features from different embodiments are intended to be within the scope of the present invention and to form different embodiments. For example, in the claims below, any of the claimed embodiments may be used in any combination.

[0087] The various component embodiments of the present invention may be implemented in hardware, as software modules running on one or more processors, or as a combination thereof. Those skilled in the art will appreciate that, in practice, a microprocessor or digital signal processor (DSP) may be used to implement some or all of the functionality of some or all of the components according to the embodiments of the present invention. The embodiments of the present invention may also be implemented as an apparatus or device program (e.g., a computer program or computer program product) for performing part or all of the methods described herein. Such programs implementing the embodiments of the present invention may be stored on a computer-readable medium or in the form of one or more signals. Such signals may be downloaded from an Internet website, provided on a carrier signal, or provided in any other form.

[0088] It should be noted that the above embodiments illustrate rather than limit the embodiments of the invention, and that a person skilled in the art may devise alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between brackets should not be construed as limiting the claims. The word "comprising" does not exclude the presence of elements or steps not listed in the claims. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The embodiments of the invention may be implemented by means of hardware comprising several different elements and by means of a suitably programmed computer. In a unit claim enumerating several means, several of these means may be embodied by the same item of hardware. The use of the words first, second, and third etc. does not indicate any order. These words may be interpreted as names. The steps in the above embodiments should not be understood as limiting the order of execution unless otherwise specified.

Claims

1. A method for predicting the longitudinal wave velocity of a formation, characterized in that the method include: Conduct conventional logging and full-wavelength acoustic logging on the cased hole section to obtain natural gamma ray, compensated neutron and array waveform data; Processing the array waveform data to extract the longitudinal wave velocity and the shear wave velocity of the cased well section; performing data cleaning processing according to the longitudinal wave sound velocity, the shear wave sound velocity, the natural gamma, and the compensated neutron; The shear wave velocity, the natural gamma and the compensated neutron processed by data cleaning are used as sample data, and the longitudinal wave velocity processed by data cleaning is used as the corresponding sample label data to train a preset cased well acoustic wave velocity prediction model, so as to predict the longitudinal wave velocity according to the trained preset cased well acoustic wave velocity prediction model.

2. The method according to claim 1, characterized in that The processing of the array waveform data to extract the longitudinal wave velocity and the shear wave velocity of the cased well section further includes: Processing the array waveform data using a time-slowness coherence method to obtain a time-slowness coherence map of the array waveform data in a time window; Obtaining the formation slowness at the point of maximum coherence according to the time-slowness coherence diagram; The longitudinal wave velocity and the shear wave velocity of the cased well section are determined according to the formation slowness; wherein the longitudinal wave velocity and the formation slowness obtained according to the monopole array acoustic wave signal of the array waveform data are inversely proportional to each other; and the shear wave velocity and the formation slowness obtained according to the dipole array acoustic wave signal of the array waveform data are inversely proportional to each other.

3. The method according to claim 1, characterized in that The data cleaning process according to the longitudinal wave velocity, the shear wave velocity, the natural gamma and the compensation neutron further includes: Determining the longitudinal wave time difference based on the longitudinal wave sound velocity, and determining the shear wave time difference based on the shear wave sound velocity; An intersection diagram is constructed in pairs according to the longitudinal wave time difference, the shear wave time difference, the natural gamma and the compensated neutron, so as to determine abnormal data according to the intersection diagram, and perform data cleaning processing on the abnormal data.

4. The method according to claim 3, characterized in that The constructing of the intersection diagrams according to the longitudinal wave time difference, the shear wave time difference, the natural gamma ray and the compensating neutron further comprises: performing data invalidation removal processing on the natural gamma and the compensating neutron according to a preset invalidation condition; The intersection diagram is constructed by combining any two data of the longitudinal wave time difference, the shear wave time difference, the natural gamma ray after invalid data removal, and the compensated neutron.

5. The method according to claim 1, wherein The method of using the shear wave velocity, the natural gamma, and the compensated neutron processed by data cleaning as sample data, and using the longitudinal wave velocity processed by data cleaning as corresponding sample label data to train a preset cased hole acoustic wave velocity prediction model further includes: A preset cased well acoustic velocity prediction model is constructed based on a bidirectional long short-term memory neural network and a fully connected neural network; The shear wave velocity, the natural gamma, and the compensated neutrons processed by data cleaning are used as sample data to construct a three-dimensional matrix of the preset cased well acoustic wave velocity prediction model input; and the longitudinal wave velocity processed by data cleaning is used as the corresponding sample label data to train the preset cased well acoustic wave velocity prediction model, and the model parameters are iteratively optimized to obtain the trained preset cased well acoustic wave velocity prediction model.

6. The method according to claim 1, characterized in that The method of predicting the longitudinal wave velocity based on the preset cased well acoustic wave velocity prediction model obtained through training further includes: Obtaining natural gamma, compensated neutron, and shear wave velocity for the entire well section; wherein the shear wave velocity is extracted by processing the array waveform data for the entire well section; The natural gamma, compensated neutron and shear wave velocity of the entire well section are input into the trained preset cased well acoustic wave velocity prediction model to predict the longitudinal wave velocity of the entire well section.

7. A device for predicting formation longitudinal wave velocity, characterized in that: The device includes: Logging module, suitable for conventional logging and full-wavelength acoustic logging of cased well sections, obtaining natural gamma, compensated neutron and array waveform data; a waveform processing module adapted to process the array waveform data and extract the longitudinal wave velocity and the shear wave velocity of the cased well section; a data cleaning module, adapted to perform data cleaning processing according to the longitudinal wave sound velocity, the shear wave sound velocity, the natural gamma, and the compensated neutron; The model prediction module is suitable for using the shear wave velocity, the natural gamma and the compensated neutron as sample data, and the longitudinal wave velocity after data cleaning as the corresponding sample label data to train a preset cased well acoustic wave velocity prediction model, so as to predict the longitudinal wave velocity based on the trained preset cased well acoustic wave velocity prediction model.

8. A computing device, characterized in that include: A processor, a memory, a communication interface, and a communication bus, wherein the processor, the memory, and the communication interface communicate with each other via the communication bus; The memory is used to store at least one executable instruction, and the executable instruction enables the processor to perform operations corresponding to the method for predicting formation compressional wave velocity according to any one of claims 1 to 6.

9. A computer storage medium, characterized in that The storage medium stores at least one executable instruction, and the executable instruction enables the processor to execute operations corresponding to the method for predicting formation longitudinal wave velocity according to any one of claims 1 to 6.

10. A computer program product, characterized in that The method comprises at least one executable instruction, wherein the executable instruction enables a processor to execute operations corresponding to the method for predicting formation longitudinal wave velocity according to any one of claims 1 to 6.

Images