Deep learning interval transit time curve prediction method and system based on Frust equation constraint

By constructing the Faust model and deep learning model, reconstructing and predicting the acoustic wave time difference curve, the problem of missing acoustic wave time difference curve is solved, the accuracy of prediction and the sensitivity of the model are improved, and technical support is provided for complex exploration and development.

CN119064992APending Publication Date: 2024-12-03PETROCHINA CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202310644145.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2023-06-01
Publication Date
2024-12-03

AI Technical Summary

Technical Problem

In different eras and different admission environments, the sound wave time difference curve is easily missing, making it difficult to directly use deep learning for prediction, which in turn affects research such as logging formation evaluation and rock physics analysis.

Method used

By obtaining the sound wave time difference curve samples, building a Faust model with deep variable parameters, reconstructing the sound wave time difference curve, and combining with the deep learning model for training, the prediction of the well of the measured sound wave time difference curve is achieved.

Benefits of technology

The problem of missing sound wave time difference curve is solved, the input curve sensitivity of the deep learning model is improved, and the prediction of sound wave time difference curve in the admission environment of complex logging data is realized, providing important technical support for exploration and development.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119064992A_ABST
    Figure CN119064992A_ABST
Patent Text Reader

Abstract

The invention relates to a deep learning interval transit time curve prediction method and system based on Frust equation constraint. The method comprises the following steps: obtaining an interval transit time curve sample; on the basis of the interval transit time curve sample, aiming at different depth sections, constructing a Frust model of depth variation parameters; constructing a deep learning model and training to obtain a final model; acquiring a Frust reconstructed interval transit time curve of a to-be-predicted interval transit time curve missing well; and inputting the Frust reconstructed interval transit time curve of the interval transit time curve missing well to be predicted and the interval transit time curve of the interval transit time curve missing well to be predicted into the final model to obtain a predicted interval transit time curve. The interval transit time curve prediction method solves the problem of interval transit time curve missing caused by special times such as old wells and horizontal wells and measurement modes, realizes interval transit time curve prediction based on physical constraints and a deep learning model, and provides technical support for exploration evaluation of complex lithologic reservoirs and fine description of oil and gas reservoirs.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of oil and gas exploration, and in particular relates to a deep learning acoustic time difference curve prediction method and system based on Faust equation constraints. Background Art

[0002] In different eras and under different recording environments, the acoustic time difference curve is missing, and deep learning and other methods are difficult to be directly used for acoustic time difference curve prediction due to the lack of sensitive samples, which leads to the lack of basic data for research work such as logging formation evaluation, rock physics analysis, and well seismic calibration. Combined with the deep rock physics trend of the acoustic time difference curve, the identification of high-quality acoustic time difference curve wells is realized, and the parameter scanning method is further used to establish a long well section variable depth Faust model, and the lithology, electrical properties and depth curves are converted into physically meaningful reconstructed acoustic time difference curves to improve the target sensitivity of the deep learning input curve, thereby realizing the training of the deep learning model and using it for the acoustic time difference curve prediction of wells without measured acoustic time difference curves, which is of great significance for exploration and development in complex logging data recording environments. Summary of the invention

[0003] The purpose of the present invention is to provide a deep learning acoustic time difference curve prediction method and system based on Faust equation constraints in order to solve the above problems.

[0004] The present invention achieves the above-mentioned purpose through the following technical solutions:

[0005] A deep learning acoustic time difference curve prediction method based on Faust equation constraints includes the following steps:

[0006] Obtain a sample of the acoustic wave time difference curve;

[0007] Based on the acoustic time difference curve samples, a Faust model with deep variable parameters is constructed for different depth sections, and the Faust reconstructed acoustic time difference curve of the acoustic time difference well sample is obtained;

[0008] Constructing a deep learning model, inputting the acoustic wave time difference samples and the Faust reconstructed acoustic wave time difference curves of the acoustic wave time difference curve samples into the deep learning model for training, and obtaining a final model;

[0009] Acquiring an acoustic time difference curve of a well with a missing acoustic time difference curve to be predicted and processing it through the Faust model to obtain a Faust reconstructed acoustic time difference curve of the well with a missing acoustic time difference curve to be predicted;

[0010] The Faust reconstructed acoustic time difference curve of the well with missing acoustic time difference curve to be predicted and the acoustic time difference curve of the well with missing acoustic time difference curve to be predicted are input into the final model to obtain the predicted acoustic time difference curve.

[0011] As a further optimization solution of the present invention, the specific process of obtaining the acoustic wave time difference curve sample is as follows:

[0012] Based on wells with measured acoustic time difference curves, the quality of the acoustic time difference curves is evaluated in combination with acoustic velocity and deep rock physical trends, and the acoustic time difference curves evaluated as excellent are selected to obtain acoustic time difference samples.

[0013] As a further optimization scheme of the present invention, the specific process of quality evaluation of the acoustic time difference curve is as follows:

[0014] The acoustic time difference curve is compared with the natural gamma curve and the electrical curve to determine whether the vertical change trend and numerical dynamic range of the acoustic time difference curve exceed the preset threshold. If not, it is evaluated as excellent; if so, it is evaluated as poor.

[0015] As a further optimization scheme of the present invention, based on the acoustic time difference curve sample, for different depth segments, a Faust model with deep variable parameters is constructed, and the specific process of obtaining the Faust reconstructed acoustic time difference curve of the acoustic time difference curve sample is as follows:

[0016] For different depth sections, the acoustic time difference curve samples are subjected to parameter scanning to determine the optimal Faust equation parameters at different depths, and to construct a Faust model with deep variable parameters;

[0017] The acoustic wave time difference curve sample is input into the Faust model of the deep variable parameter to obtain the Faust reconstructed acoustic wave time difference curve of the acoustic wave time difference curve sample.

[0018] As a further optimization scheme of the present invention, the Faust model of the deep variable parameters is expressed by the following formula:

[0019] V p =a*(R t *Z) b ;

[0020] Among them, R t represents the formation resistivity, Z represents the depth, and a and b represent the model parameters.

[0021] As a further optimization scheme of the present invention, the deep learning model includes an input layer, an LSTM layer, a fully connected layer, a pooling layer and an output layer connected in sequence.

[0022] As a further optimization scheme of the present invention, the parameters of the deep learning model are solved using the Adam method.

[0023] A deep learning acoustic time difference curve prediction system based on Faust equation constraints, comprising:

[0024] An acquisition unit, used for acquiring acoustic time difference curve samples and acoustic time difference curves of wells with missing acoustic time difference curves to be predicted;

[0025] A model building unit, used to build a Faust model with deep variable parameters for different depth segments based on the acoustic wave time difference curve samples; and also used to build a deep learning model;

[0026] A processing unit, configured to obtain the Faust reconstructed acoustic time difference curve of the acoustic time difference well sample and the Faust reconstructed acoustic time difference curve of the well with missing acoustic time difference curve to be predicted based on the Faust model of the deep variable parameter;

[0027] A training unit, used for inputting the acoustic wave time difference samples and the Faust reconstructed acoustic wave time difference curves of the acoustic wave time difference curve samples into the deep learning model for training to obtain a final model;

[0028] The output unit is used to input the Faust reconstructed acoustic time difference curve of the well with missing acoustic time difference curve to be predicted and the acoustic time difference curve of the well with missing acoustic time difference curve to be predicted into the final model to obtain the predicted acoustic time difference curve.

[0029] An electronic device comprises a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other via the communication bus;

[0030] Memory, used to store computer programs;

[0031] The processor is used to implement the deep learning acoustic wave time difference curve prediction method based on Faust equation constraints when executing the program stored in the memory.

[0032] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the deep learning acoustic time difference curve prediction method based on Faust equation constraints.

[0033] The beneficial effects of the present invention are:

[0034] The present invention solves the problem of missing acoustic time difference curves caused by special eras and measurement methods such as old wells and horizontal wells, and realizes the prediction of acoustic time difference curves based on physical constraints and deep learning models, providing technical support for the exploration and evaluation of complex lithology reservoirs and the detailed description of oil and gas reservoirs. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 It is a diagram of the method steps of the present invention;

[0036] Figure 2 This is a well quality classification diagram of a certain area in the Tarim Basin of the present invention;

[0037] Figure 3 It is a diagram of a deep scan example of Faust model parameters of the present invention;

[0038] Figure 4 This is a diagram showing an example of reconstructing an acoustic time difference curve of a well without a measured acoustic time difference curve according to the present invention using a variable parameter Faust model;

[0039] Figure 5 It is a schematic diagram of the deep learning model structure of the present invention;

[0040] Figure 6 This is a diagram of an example of the well logging curve deep learning model training verification of the present invention;

[0041] Figure 7 This is a diagram of an example of deep learning prediction of the acoustic time difference curve of the present invention;

[0042] Figure 8 is a structural block diagram of a prediction system of the present invention;

[0043] Fig. 9 It is a structural block diagram of the electronic device of the present invention. DETAILED DESCRIPTION

[0044] The present application is further described in detail below in conjunction with the accompanying drawings. It is necessary to point out here that the following specific implementation methods are only used to further illustrate the present application and cannot be understood as limiting the scope of protection of the present application. Technical personnel in this field can make some non-essential improvements and adjustments to the present application based on the above application content.

[0045] like Figure 1 As shown, a deep learning acoustic time difference curve prediction method based on Faust equation constraints includes the following steps:

[0046] S101: Obtaining a sound wave time difference curve sample;

[0047] S102: Based on the acoustic time difference curve samples, for different depth sections, construct a Faust model with deep variable parameters, and obtain the Faust reconstructed acoustic time difference curve of the acoustic time difference well sample;

[0048] S103: constructing a deep learning model, inputting the acoustic wave time difference samples and the Faust reconstructed acoustic wave time difference curves of the acoustic wave time difference curve samples into the deep learning model for training, and obtaining a final model;

[0049] S104: Acquire the acoustic time difference curve of the well with missing acoustic time difference curve to be predicted and process it through the Faust model to obtain the Faust reconstructed acoustic time difference curve of the well with missing acoustic time difference curve to be predicted;

[0050] S105: Inputting the Faust reconstructed acoustic time difference curve of the well with missing acoustic time difference curve to be predicted and the acoustic time difference curve of the well with missing acoustic time difference curve to be predicted into the final model to obtain a predicted acoustic time difference curve.

[0051] The specific process of obtaining the acoustic time difference curve sample is as follows:

[0052] Based on wells with measured acoustic time difference curves, the quality of the acoustic time difference curves is evaluated in combination with acoustic velocity and deep rock physical trends, and the acoustic time difference curves evaluated as excellent are selected to obtain acoustic time difference samples.

[0053] The specific process of quality assessment of the acoustic time difference curve is as follows:

[0054] The acoustic time difference curve is compared with the natural gamma curve and the electrical curve to determine whether the vertical change trend and numerical dynamic range of the acoustic time difference curve exceed the preset threshold. If not, it is evaluated as excellent; if so, it is evaluated as poor.

[0055] Based on the acoustic time difference curve sample, for different depth segments, a Faust model with deep variable parameters is constructed, and the specific process of obtaining the Faust reconstructed acoustic time difference curve of the acoustic time difference curve sample is as follows:

[0056] For different depth sections, the acoustic time difference curve samples are subjected to parameter scanning to determine the optimal Faust equation parameters at different depths, and to construct a Faust model with deep variable parameters;

[0057] The acoustic wave time difference curve sample is input into the Faust model of the deep variable parameter to obtain the Faust reconstructed acoustic wave time difference curve of the acoustic wave time difference curve sample.

[0058] The Faust model of the deep variable parameters is expressed by the following formula:

[0059] V p =a*(R t *Z) b ;

[0060] Among them, R t represents the formation resistivity, Z represents the depth, and a and b represent the model parameters.

[0061] The deep learning model includes an input layer, an LSTM layer, a fully connected layer, a pooling layer and an output layer connected in sequence.

[0062] The parameters of the deep learning model are solved using the Adam method.

[0063] In this embodiment, the specific process is as follows:

[0064] The method includes five key steps: optimization of acoustic time difference curve prediction well samples, Faust equation parameter scanning and model construction in depth segments, Faust model acoustic time difference curve reconstruction of wells with missing acoustic time difference curves, deep learning model training of acoustic time difference curves, and deep learning prediction of acoustic time difference curves.

[0065] Step 1: For wells with measured acoustic transit time curves, evaluate the quality of the acoustic transit time curves in combination with the acoustic velocity depth and rock physics trends, and select high-quality acoustic transit time wells as samples.

[0066] Step 2: For high-quality acoustic time difference well samples, carry out the reconstruction test of the acoustic time difference curve of the Faust equation with different parameters, calculate the correlation with the measured acoustic time difference, determine the optimal Faust equation parameters at different depths, establish a deep variable parameter Faust model, and realize the Faust reconstruction of the acoustic time difference curve.

[0067] Step 3: For wells with missing acoustic time difference curves, the deep variable parameter Faust model established using high-quality acoustic time difference well samples is used to achieve Faust reconstruction of the acoustic time difference curves.

[0068] Step 4: For high-quality acoustic time difference well samples, the conventional curve of the high-quality well and the Faust equation are used to reconstruct the acoustic time difference curve as input, and the measured acoustic time difference curve is used as the target to realize the deep learning model training of the acoustic time difference.

[0069] Step 5: For wells with missing acoustic time difference curves, use the deep learning model of acoustic time difference to reconstruct the acoustic time difference curve with conventional curves and Faust equation as input to realize the prediction of acoustic time difference curves.

[0070] Application example: Deep learning acoustic time difference curve prediction based on Faust equation constraints in a shallow layer of a certain area in the Tarim Basin

[0071] (1) Selection of high-quality acoustic time difference samples based on acoustic velocity depth rock physics trends

[0072] like Figure 2As shown in the figure, a) and b) are typical examples of high-quality acoustic transit time wells and low-quality acoustic transit time wells in a certain area of ​​the Tarim Basin, respectively. In a) and b), the curves in the first column on the right are acoustic transit time curves. In a), the depth variation of the acoustic transit time curve is consistent with the natural gamma curve (first column on the left) and the electrical curve (second column on the right). In contrast, the acoustic transit time curve in b) conflicts with the natural gamma curve (first column on the left) and the electrical curve (second column on the right) in terms of both vertical variation trend and numerical dynamic range. In order to ensure the rationality of the deep learning model, high-quality acoustic transit time curve wells are selected as samples for subsequent model training.

[0073] (2) Parameter scanning and model construction of Faust equation in different depth segments

[0074] The mathematical form of the Faust model is V p =a*(R t *Z) b , where R t is the formation resistivity, Z is the depth, and a and b are model parameters. Parameter a is negatively correlated with the formation water resistivity. Compared with conventional methods such as multivariate linear regression, the Faust model derived from theoretical and experimental data can better control the mathematical form of the acoustic velocity (acoustic time derivative) and resistivity and depth curves. Figure 3 In the four columns on the right, the dark curves are the acoustic time difference curves reconstructed using the Faust equation with different parameters, and the light curves are the measured acoustic time difference curves. The reconstructed acoustic time difference curves with different parameters have different sensitivities to the measured acoustic time difference curves at different depths. Therefore, the Faust model with variable parameters can be constructed for different depth sections by using parameter scanning. The correlations between the measured resistivity curve and the acoustic time difference curve reconstructed by the variable parameter Faust model and the measured acoustic time difference curve are 47.63% and 87.88% respectively, which fully demonstrates that the method of reconstructing the curve using the variable parameter Faust model has a stronger sensitivity.

[0075] (3) Reconstruction of the Faust model of acoustic transit time curves of wells with missing acoustic transit time

[0076] In order to improve the target sensitivity of the input curve of the deep learning model, the variable parameter Faust model established for high-quality acoustic transit time wells is used to reconstruct the sensitivity curve of wells with missing acoustic transit time. Figure 4 This is an example of the reconstruction of the acoustic transit time curve using the variable parameter Faust model for a typical well without measured acoustic transit time curves. The second column on the left side of the figure is the measured resistivity curve (dark color is deep resistivity, light color is shallow resistivity), and the first column on the right side is the acoustic transit time curve reconstructed using the variable parameter Faust model.

[0077] (4) Deep learning model training of acoustic time difference curve

[0078] The training of the deep learning model of well logging curves mainly includes the construction of the deep learning model structure and the determination of machine learning parameters. Figure 5 As shown in the figure, in this step, the machine learning model structure used includes input layer, LSTM layer (long short-term memory), fully connected layer, pooling layer and output layer. The machine learning model parameters are solved using the Adam (Adaptive moment estimation) method. Figure 6 This is an example of a well verification using a trained deep learning model. The first column of lithology curves (natural gamma curves), the second column of electrical curves (resistivity curves), and the third column of Faust model reconstructed acoustic time difference curves on the left side of the figure are the input curves of the deep learning model. The dark measured acoustic time difference curves in the first column on the right are the learning target curves of the deep learning model, and the black ones are the acoustic time difference curves predicted by the deep learning model. The predicted black curves in the first column on the right are highly consistent with the measured dark curves, which verifies the predictive and generalization capabilities of the model.

[0079] (5) Deep learning prediction of acoustic time difference curve

[0080] Figure 7 The logging curve deep learning model established in the first four steps is applied to the same Figure 2 An example of a well with missing acoustic time difference curves. In the figure, the first column on the left is the lithology curve column (natural gamma curve), the second column is the electrical curve column (resistivity curve), and the third column is the Faust model reconstructed acoustic time difference curve. The fourth column is the acoustic time difference curve predicted by the deep learning model, which realizes the prediction of the acoustic time difference curve of lithology and typical curves under the constraints of the Faust equation.

[0081] like Figure 8 As shown, a deep learning acoustic wave time difference curve prediction system based on Faust equation constraints includes:

[0082] An acquisition unit, used for acquiring acoustic time difference curve samples and acoustic time difference curves of wells with missing acoustic time difference curves to be predicted;

[0083] A model building unit, used to build a Faust model with deep variable parameters for different depth segments based on the acoustic wave time difference curve samples; and also used to build a deep learning model;

[0084] A processing unit, configured to obtain the Faust reconstructed acoustic time difference curve of the acoustic time difference well sample and the Faust reconstructed acoustic time difference curve of the well with missing acoustic time difference curve to be predicted based on the Faust model of the deep variable parameter;

[0085] A training unit, used for inputting the acoustic wave time difference samples and the Faust reconstructed acoustic wave time difference curves of the acoustic wave time difference curve samples into the deep learning model for training to obtain a final model;

[0086] The output unit is used to input the Faust reconstructed acoustic time difference curve of the well with missing acoustic time difference curve to be predicted and the acoustic time difference curve of the well with missing acoustic time difference curve to be predicted into the final model to obtain the predicted acoustic time difference curve.

[0087] The implementation process of the functions and effects of each unit in the above system is specifically described in the implementation process of the corresponding steps in the above method, which will not be repeated here.

[0088] As for the system embodiment, since it basically corresponds to the method embodiment, the relevant parts can refer to the partial description of the method embodiment. The system embodiment described above is only schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of the present invention. Ordinary technicians in this field can understand and implement it without paying creative work.

[0089] In the above-mentioned embodiment, any multiple of all units can be combined in one module for implementation, or any one of the modules can be split into multiple modules. Alternatively, at least some functions of one or more of these modules can be combined with at least some functions of other modules and implemented in one module. At least one of all units can be at least partially implemented as a hardware circuit, such as a field programmable gate array (FPGA), a programmable logic array (PLA), a system on a chip, a system on a substrate, a system on a package, an application specific integrated circuit (ASIC), or can be implemented by hardware or firmware such as any other reasonable way of integrating or packaging the circuit, or implemented in any one of the three implementation modes of software, hardware and firmware or in a suitable combination of any of them. Alternatively, at least one of all units can be at least partially implemented as a computer program module, and when the computer program module is run, the corresponding function can be executed.

[0090] like Fig. 9 As shown, an electronic device includes a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other through the communication bus;

[0091] Memory, used to store computer programs;

[0092] The processor is used to implement the deep learning acoustic wave time difference curve prediction method based on Faust equation constraints when executing the program stored in the memory.

[0093] The above communication bus may be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. The communication bus may be divided into an address bus, a data bus, a control bus, etc.

[0094] The communication interface is used for communication between the above electronic device and other devices.

[0095] The memory may include a random access memory (RAM) or a non-volatile memory, such as at least one disk memory. Optionally, the memory may also be at least one storage device located away from the aforementioned processor.

[0096] The above-mentioned processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0097] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the deep learning acoustic time difference curve prediction method based on Faust equation constraints.

[0098] The computer-readable storage medium may be included in the device / apparatus described in the above embodiment; or it may exist independently without being assembled into the device / apparatus. The above computer-readable storage medium carries one or more programs. When the above one or more programs are executed, the deep learning acoustic wave time difference curve prediction method based on Faust equation constraint according to the embodiment of the present disclosure is implemented.

[0099] According to an embodiment of the present disclosure, a computer-readable storage medium may be a non-volatile computer-readable storage medium, for example, may include but is not limited to: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in conjunction with an instruction execution system, apparatus, or device.

[0100] The above-mentioned embodiments only express several implementation methods of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the present invention. It should be pointed out that for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention.

Claims

1. A deep learning acoustic time difference curve prediction method based on Faust equation constraints, characterized in that: The following steps are involved: Obtain a sample of the acoustic wave time difference curve; Based on the acoustic time difference curve samples, a Faust model with deep variable parameters is constructed for different depth sections, and the Faust reconstructed acoustic time difference curve of the acoustic time difference well sample is obtained; Constructing a deep learning model, inputting the acoustic wave time difference samples and the Faust reconstructed acoustic wave time difference curves of the acoustic wave time difference curve samples into the deep learning model for training, and obtaining a final model; Acquiring an acoustic time difference curve of a well with a missing acoustic time difference curve to be predicted and processing it through the Faust model to obtain a Faust reconstructed acoustic time difference curve of the well with a missing acoustic time difference curve to be predicted; The Faust reconstructed acoustic time difference curve of the well with missing acoustic time difference curve to be predicted and the acoustic time difference curve of the well with missing acoustic time difference curve to be predicted are input into the final model to obtain the predicted acoustic time difference curve.

2. According to claim 1, a deep learning acoustic wave time difference curve prediction method based on Faust equation constraint is characterized in that: The specific process of obtaining the acoustic time difference curve sample is as follows: Based on wells with measured acoustic time difference curves, the quality of the acoustic time difference curves is evaluated in combination with acoustic velocity and deep rock physical trends, and the acoustic time difference curves evaluated as excellent are selected to obtain acoustic time difference samples.

3. The deep learning acoustic time difference curve prediction method based on Faust equation constraint according to claim 2 is characterized in that: The specific process of quality assessment of the acoustic time difference curve is as follows: The acoustic time difference curve is compared with the natural gamma curve and the electrical curve to determine whether the vertical change trend and numerical dynamic range of the acoustic time difference curve exceed the preset threshold. If not, it is evaluated as excellent; if so, it is evaluated as poor.

4. The deep learning acoustic time difference curve prediction method based on Faust equation constraint according to claim 1 is characterized in that: Based on the acoustic time difference curve sample, for different depth segments, a Faust model with deep variable parameters is constructed, and the specific process of obtaining the Faust reconstructed acoustic time difference curve of the acoustic time difference curve sample is as follows: For different depth segments, the acoustic time difference curve samples are subjected to parameter scanning to determine the optimal Faust equation parameters at different depths, and to construct a Faust model with deep variable parameters; The acoustic wave time difference curve sample is input into the Faust model of the deep variable parameter to obtain the Faust reconstructed acoustic wave time difference curve of the acoustic wave time difference curve sample.

5. The deep learning acoustic time difference curve prediction method based on Faust equation constraint according to claim 4 is characterized in that: The Faust model of the deep variable parameters is expressed by the following formula: V p =*(R t *) b ; Among them, R t represents the formation resistivity, Z represents the depth, and a and b represent the model parameters.

6. The deep learning acoustic time difference curve prediction method based on Faust equation constraint according to claim 1 is characterized in that: The deep learning model includes an input layer, an LSTM layer, a fully connected layer, a pooling layer and an output layer connected in sequence.

7. The deep learning acoustic time difference curve prediction method based on Faust equation constraint according to claim 6 is characterized in that: The parameters of the deep learning model are solved using the Adam method.

8. A deep learning acoustic time difference curve prediction system based on Faust equation constraints, characterized in that: include: An acquisition unit, used for acquiring acoustic time difference curve samples and acoustic time difference curves of wells with missing acoustic time difference curves to be predicted; A model building unit, used to build a Faust model with deep variable parameters for different depth segments based on the acoustic wave time difference curve samples; and also used to build a deep learning model; A processing unit, configured to obtain the Faust reconstructed acoustic time difference curve of the acoustic time difference well sample and the Faust reconstructed acoustic time difference curve of the well with missing acoustic time difference curve to be predicted based on the Faust model of the deep variable parameter; A training unit, used for inputting the acoustic wave time difference samples and the Faust reconstructed acoustic wave time difference curves of the acoustic wave time difference curve samples into the deep learning model for training to obtain a final model; The output unit is used to input the Faust reconstructed acoustic time difference curve of the well with missing acoustic time difference curve to be predicted and the acoustic time difference curve of the well with missing acoustic time difference curve to be predicted into the final model to obtain the predicted acoustic time difference curve.

9. An electronic device, characterized in that: It includes a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other via the communication bus; Memory, used to store computer programs; The processor is used to implement the deep learning acoustic wave time difference curve prediction method based on Faust equation constraints as described in any one of claims 1 to 7 when executing the program stored in the memory.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the deep learning acoustic wave time difference curve prediction method based on Faust equation constraints described in any one of claims 1 to 7 is implemented.