Vapor channeling phenomenon identification method and device and electronic equipment

Abstract

The invention discloses a steam channeling phenomenon identification method and device and electronic equipment. The method comprises the following steps: receiving a steam channeling phenomenon identification request; in response to the steam channeling phenomenon identification request, obtaining target well data of a target adjacent well; and inputting the target steam injection data and the target well data into the target identification model to obtain a target identification result about whether a steam channeling phenomenon occurs in the target adjacent well. According to the method and the device, the technical problem of low identification accuracy of whether the steam channeling phenomenon occurs in the adjacent well or not when the steam is injected into the steam injection well in the related technology is solved.

Description

Technical Field

[0001] The present invention relates to the field of petroleum exploitation, and in particular to a method, a device and an electronic device for identifying steam channeling. Background Art

[0002] Steam channeling in oil wells refers to the formation of high permeability channels between oil layers of different heavy oil wells, where steam directly enters the production well from the injection well, affecting the normal production of the production well. In oilfield blocks with a certain history of exploitation, due to the generally small well spacing, increasing steam throughput cycles, and reduced formation pressure, the steam channeling phenomenon is becoming more and more serious. The occurrence of steam channeling causes uneven heating of the reservoir, which has become the main challenge restricting the improvement of recovery.

[0003] However, when using the related methods provided in the related art to inject steam into a steam injection well, the problem of low accuracy in identifying whether steam cross-talk occurs in an adjacent well is solved.

[0004] To address the above-mentioned problems, no effective solution has been proposed yet. Summary of the invention

[0005] The embodiments of the present invention provide a method, device and electronic device for identifying steam crosstalk, so as to at least solve the technical problem in the related art of low accuracy in identifying whether steam crosstalk occurs in adjacent wells when injecting steam into a steam injection well.

[0006] According to one aspect of an embodiment of the present invention, a method for identifying a steam channeling phenomenon is provided, comprising: receiving a steam channeling phenomenon identification request, wherein the steam channeling phenomenon identification request is used to identify whether a steam channeling phenomenon occurs in a target adjacent well when steam is injected into a target steam injection well with target steam injection data, the target adjacent well being a well whose distance from the target steam injection well is less than a predetermined threshold, and the identification request carries the target steam injection data; in response to the steam channeling phenomenon identification request, acquiring target well data of the target adjacent well; inputting the target steam injection data and the target well data into a target identification model to obtain a target identification result of whether the steam channeling phenomenon occurs in the target adjacent well, wherein the target identification model is The positive sample data group and the negative sample data group are obtained by training the initial model, wherein the positive sample data group and the negative sample data group are determined based on steam crosstalk parameter indicators and historical well logging data, the positive sample data group includes positive sample steam injection data, positive sample well data and positive sample recognition results, the negative sample data group includes negative sample steam injection data, negative sample well data and negative sample recognition results, the positive sample recognition result is the recognition result of the steam crosstalk phenomenon occurring in the adjacent wells of the positive sample when steam is injected into the positive sample steam injection well with the positive sample steam injection data, and the negative sample recognition result is the recognition result of the steam crosstalk phenomenon not occurring in the adjacent wells of the negative sample when steam is injected into the negative sample steam injection well with the negative sample steam injection data.

[0007] Optionally, before inputting the target steam injection data and the target well data into the target recognition model to obtain the target recognition result of whether the steam channeling phenomenon occurs in the target adjacent well, it also includes: acquiring the historical logging data, wherein the historical logging data includes logging data related to the sample steam injection well and logging data related to the sample adjacent well; determining the steam channeling parameter value corresponding to the steam channeling parameter index from the historical logging data; determining the sample result of whether the steam channeling phenomenon occurs in the sample adjacent well based on the steam channeling parameter value; based on the sample result, determining the positive sample steam injection data and the positive sample well data, and the negative sample steam injection data and the negative sample well data from the historical logging data; determining the positive sample recognition result in the sample result, the positive sample steam injection data and the positive sample well data are the positive sample data group, and determining the negative sample recognition result in the sample result, the negative sample steam injection data and the negative sample well data are the negative sample data group.

[0008] Optionally, determining the sample result of whether steam channeling occurs in the sample adjacent wells based on the steam channeling parameter value includes: determining a steam channeling parameter threshold corresponding to the steam channeling parameter index; and determining the sample result of whether steam channeling occurs in the sample adjacent wells based on the steam channeling parameter value and the steam channeling parameter threshold.

[0009] Optionally, obtaining the historical logging data includes: obtaining historical initial logging data; screening out abnormal logging data that meets predetermined conditions from the historical initial logging data to obtain the historical logging data, wherein the predetermined conditions include at least one of the following: the number of dates between the steam crosstalk date and the steam injection date exceeds a first threshold, and a fluctuation index corresponding to the steam crosstalk parameter value exceeds a second threshold.

[0010] Optionally, the inputting the target steam injection data and the target well data into the target recognition model to obtain the target recognition result of whether the steam channeling phenomenon occurs in the target adjacent wells includes: acquiring a steam channeling database, wherein the steam channeling database is a database that records the recognition result of whether the steam channeling phenomenon occurs in predetermined adjacent wells of predetermined well data when steam is injected into the target steam injection well with predetermined steam injection data, the predetermined adjacent wells being wells whose distance from the target steam injection well is less than the predetermined threshold, and the predetermined adjacent wells include the target adjacent wells; inputting the steam channeling database, the target steam injection data and the target well data into the target recognition model to obtain the target recognition result of whether the steam channeling phenomenon occurs in the target adjacent wells.

[0011] Optionally, after inputting the steam channeling database, the target steam injection data and the target well data into the target recognition model to obtain the target recognition result of whether the steam channeling phenomenon occurs in the target adjacent well, it also includes: when the target recognition result is the recognition result of the steam channeling phenomenon occurring in the target well data, determining whether there is a predetermined recognition result corresponding to the target adjacent well of the target well data when steam is injected into the target steam injection well with the target steam injection data in the steam channeling database, and obtaining an existence result; when the existence result is that the predetermined recognition result does not exist in the steam channeling database, adding the target recognition result of the steam channeling phenomenon sent by the target adjacent well of the target well data when steam is injected into the target steam injection well with the target steam injection data to the steam channeling database, and obtaining an updated steam channeling database.

[0012] Optionally, after inputting the target steam injection data and the target well data into the target recognition model to obtain a target recognition result of whether the steam channeling phenomenon occurs in the target adjacent well, it also includes: when the target recognition result is a recognition result that the steam channeling phenomenon does not occur in the target adjacent well, adding the target steam injection data, the target well data, and the target recognition result to the negative sample data group to obtain an updated negative sample data group.

[0013] According to one aspect of an embodiment of the present invention, a steam channeling phenomenon identification device is provided, comprising: a receiving module for receiving a steam channeling phenomenon identification request, wherein the steam channeling phenomenon identification request is used to identify whether a steam channeling phenomenon occurs in a target adjacent well when steam is injected into a target steam injection well with target steam injection data, the target adjacent well being a well whose distance from the target steam injection well is less than a predetermined threshold, and the identification request carries the target steam injection data; an acquisition module for acquiring target well data of the target adjacent well in response to the steam channeling phenomenon identification request; an input module for inputting the target steam injection data and the target well data into a target identification model to obtain a target identification result of whether the steam channeling phenomenon occurs in the target adjacent well, wherein The target recognition model is obtained by training an initial model with a positive sample data group and a negative sample data group. The positive sample data group and the negative sample data group are determined based on steam channeling parameter indicators and historical well logging data. The positive sample data group includes positive sample steam injection data, positive sample well data and positive sample recognition results. The negative sample data group includes negative sample steam injection data, negative sample well data and negative sample recognition results. The positive sample recognition result is the recognition result of the steam channeling phenomenon occurring in the adjacent wells of the positive sample when the positive sample steam injection data is used to inject steam into the positive sample steam injection well. The negative sample recognition result is the recognition result of the steam channeling phenomenon not occurring in the adjacent wells of the negative sample when the negative sample steam injection data is used to inject steam into the negative sample steam injection well.

[0014] According to one aspect of an embodiment of the present invention, an electronic device is provided, comprising: a processor; and a memory for storing requests executable by the processor; wherein the processor is configured to execute the requests to implement a method for identifying a steam crosstalk phenomenon as described in any one of the above.

[0015] According to one aspect of an embodiment of the present invention, a computer-readable storage medium is provided, comprising: when a request in the computer-readable storage medium is executed by a processor of an electronic device, the electronic device is enabled to execute any of the above-mentioned methods for identifying steam crosstalk.

[0016] In an embodiment of the present invention, a steam channeling phenomenon identification request is received, wherein the steam channeling phenomenon identification request is used to identify whether steam channeling occurs in a target adjacent well when steam is injected into a target steam injection well with target steam injection data, the target adjacent well being a well whose distance from the target steam injection well is less than a predetermined threshold, and the identification request carries the target steam injection data. In response to a steam channeling phenomenon identification request, target well data of a target adjacent well is obtained, target steam injection data and target well data are input into a target identification model, and a target identification result of whether steam channeling occurs in the target adjacent well is obtained, wherein the target identification model is obtained by training an initial model with a positive sample data group and a negative sample data group, the positive sample data group and the negative sample data group are determined based on steam channeling parameter indicators and historical well logging data, the positive sample data group includes positive sample steam injection data, positive sample well data and positive sample identification results, the negative sample data group includes negative sample steam injection data, negative sample well data and negative sample identification results, the positive sample identification result is an identification result of steam channeling occurring in a positive sample adjacent well when steam is injected into a positive sample steam injection well using the positive sample steam injection data, and the negative sample identification result is an identification result of no steam channeling occurring in a negative sample adjacent well when steam is injected into a negative sample steam injection well using the negative sample steam injection data. Since the target recognition model is obtained by training the initial model with the positive sample data group and the negative sample data group, and the introduction of the negative sample data group allows the initial model to learn the characteristics and behaviors of the target neighboring wells when no steam channeling occurs. To a certain extent, this can help the target recognition model avoid misjudging the steam channeling phenomenon due to over-emphasis on a certain category, and can also increase the robustness of the target recognition model, improve the accuracy and reliability of model recognition, help to more accurately judge whether steam channeling occurs in the target neighboring wells, and provide more reliable recognition results, thereby solving the technical problem of low accuracy in identifying whether steam channeling occurs in adjacent wells when injecting steam into a steam injection well in related technologies. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0018] Figure 1 is a flow chart of a method for identifying steam crosstalk according to an embodiment of the present invention;

[0019] Figure 2 is a fluctuation diagram of the liquid production in different wheels provided by an optional fact mode of the present invention;

[0020] Figure 3 is a diagram of a classification system of operation time of pumping wells provided in an embodiment of the present invention;

[0021] Figure 4is a flow chart of extraction of pumping well operation events provided according to an optional embodiment of the present invention;

[0022] Figure 5 is a training graph of an initial model provided according to an optional implementation manner of the present invention;

[0023] Figure 6 is a flow chart of a target recognition model provided according to an optional implementation mode of the present invention;

[0024] Figure 7 4 is a structural block diagram of a device for identifying a steam crosstalk phenomenon according to an embodiment of the present invention. DETAILED DESCRIPTION

[0025] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0026] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0027] Example 1

[0028] According to an embodiment of the present invention, an embodiment of a method for identifying a steam crosstalk phenomenon is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0029] Figure 1 is a flow chart of a method for identifying steam crosstalk according to an embodiment of the present invention. Figure 1 As shown, the method comprises the following steps:

[0030] Step S102, receiving a steam channeling phenomenon identification request, wherein the steam channeling phenomenon identification request is used to identify whether steam channeling occurs in a target adjacent well when steam is injected into a target steam injection well using target steam injection data, the target adjacent well being a well whose distance from the target steam injection well is less than a predetermined threshold, and the identification request carries the target steam injection data;

[0031] In step S102 provided in the present application, the above-mentioned steam channeling phenomenon refers to the phenomenon that during the steam injection process, the injected steam rapidly diffuses in the formation to the adjacent well and enters the adjacent well; the above-mentioned target steam injection data refers to the data related to the target steam injection well recorded during the steam injection operation at the steam injection site, which may include steam injection parameters, steam injection time, steam injection well information, etc. The identification of the above-mentioned steam channeling phenomenon has an important impact on oil production.

[0032] It should be noted that, according to a predetermined threshold, a neighboring well whose distance from the target steam injection well is less than the threshold is found, and then the subsequent steam channeling phenomenon identification step can be performed through the acquired target neighboring well data to further determine whether the target neighboring well has steam channeling phenomenon. By analyzing and processing the data of the target steam injection well and the neighboring wells, it is possible to further identify whether the steam channeling phenomenon has occurred, thereby improving the accuracy of identifying the steam channeling phenomenon to a certain extent.

[0033] Step S104, in response to the steam crosstalk phenomenon identification request, acquiring target well data of the target adjacent well;

[0034] In step S104 provided in the present application, the target well data may include wellbore data, lithology data, geological data, etc. For the above steps, when responding to the steam channeling phenomenon identification request, it is necessary to obtain the target well data of the target adjacent well. The target well data can be obtained in different ways, including the analysis of logging data, field investigation and collection of literature, etc., and can be compared and analyzed with the target steam injection data to determine whether the target adjacent well has experienced steam channeling. The target well data of the target adjacent well obtained can provide important data basis for the subsequent steam channeling phenomenon identification step, and can improve the accuracy and reliability of steam channeling phenomenon identification to a certain extent.

[0035] Step S106, input the target steam injection data and the target well data into the target recognition model to obtain a target recognition result of whether steam crosstalk occurs in the target adjacent wells, wherein the target recognition model is obtained by training an initial model with a positive sample data group and a negative sample data group, the positive sample data group and the negative sample data group are determined based on steam crosstalk parameter indicators and historical well logging data, the positive sample data group includes positive sample steam injection data, positive sample well data and positive sample recognition results, the negative sample data group includes negative sample steam injection data, negative sample well data and negative sample recognition results, the positive sample recognition result is the recognition result of steam crosstalk occurring in the positive sample adjacent wells when steam is injected into the positive sample steam injection well with the positive sample steam injection data, and the negative sample recognition result is the recognition result of no steam crosstalk occurring in the negative sample adjacent wells when steam is injected into the negative sample steam injection well with the negative sample steam injection data.

[0036] In step S106 provided in the present application, the target recognition model is obtained by training with a positive sample data group and a negative sample data group. The positive sample data group includes positive sample steam injection data, positive sample well data and positive sample recognition results, while the negative sample data group includes negative sample steam injection data, negative sample well data and negative sample recognition results. The positive sample recognition result refers to the result that the target recognition model determines that steam channeling has occurred in the adjacent wells of the positive sample when steam is injected into the positive sample steam injection well with the positive sample steam injection data. The negative sample recognition result refers to the result that the target recognition model determines that steam channeling has not occurred in the adjacent wells of the negative sample when steam is injected into the negative sample steam injection well with the negative sample steam injection data. The determination of positive samples and negative samples in the training data set of the target recognition model is based on the analysis and judgment of steam channeling parameter indicators and historical logging data. By extracting parameter indicators and historical logging data related to steam channeling, and taking the target steam injection data and target well data as input, the target recognition model can be trained and learned to determine whether steam channeling has occurred in the adjacent wells of the target.

[0037] It should be noted that by introducing the negative sample data group in the above steps, the initial model can learn the characteristics and behaviors of the target neighboring wells when no steam channeling occurs. To a certain extent, it can help the target recognition model avoid misjudging the steam channeling phenomenon due to over-emphasis on a certain category. It can also increase the robustness of the target recognition model, improve the accuracy and reliability of model recognition, and help to more accurately judge whether steam channeling occurs in the target neighboring wells and provide more reliable recognition results.

[0038] It should also be noted that by inputting the target steam injection data and the target well data into the trained target recognition model, the recognition result of whether the steam channeling phenomenon occurs in the target adjacent well can be obtained. In this step, the above-mentioned target recognition model can automatically analyze the characteristics of the target steam injection data and the target well data, and perform recognition based on the trained model, which reduces the need for manual judgment, improves the efficiency and accuracy of recognition, and provides data support for the solution of oil production, thereby effectively managing and controlling the impact of the steam channeling phenomenon on the liquid production.

[0039] Through the above steps S102-S106, a steam channeling phenomenon identification request is received, wherein the steam channeling phenomenon identification request is used to identify whether steam channeling occurs in a target adjacent well when steam is injected into a target steam injection well with target steam injection data, the target adjacent well being a well whose distance from the target steam injection well is less than a predetermined threshold, and the identification request carries the target steam injection data. In response to a steam channeling phenomenon identification request, target well data of a target adjacent well is obtained, target steam injection data and target well data are input into a target identification model, and a target identification result of whether steam channeling occurs in the target adjacent well is obtained, wherein the target identification model is obtained by training an initial model with a positive sample data group and a negative sample data group, the positive sample data group and the negative sample data group are determined based on steam channeling parameter indicators and historical well logging data, the positive sample data group includes positive sample steam injection data, positive sample well data and positive sample identification results, the negative sample data group includes negative sample steam injection data, negative sample well data and negative sample identification results, the positive sample identification result is an identification result of steam channeling occurring in a positive sample adjacent well when steam is injected into a positive sample steam injection well using the positive sample steam injection data, and the negative sample identification result is an identification result of no steam channeling occurring in a negative sample adjacent well when steam is injected into a negative sample steam injection well using the negative sample steam injection data. Since the target recognition model is obtained by training the initial model with the positive sample data group and the negative sample data group, and the introduction of the negative sample data group allows the initial model to learn the characteristics and behaviors of the target neighboring wells when no steam channeling occurs. To a certain extent, this can help the target recognition model avoid misjudging the steam channeling phenomenon due to over-emphasis on a certain category, and can also increase the robustness of the target recognition model, improve the accuracy and reliability of model recognition, help to more accurately judge whether steam channeling occurs in the target neighboring wells, and provide more reliable recognition results, thereby solving the technical problem of low accuracy in identifying whether steam channeling occurs in adjacent wells when injecting steam into a steam injection well in related technologies.

[0040] The above method of this embodiment is further introduced below.

[0041] As an optional embodiment, before inputting the target steam injection data and the target well data into the target recognition model to obtain the target recognition result of whether steam channeling occurs in the target adjacent wells, it also includes: obtaining historical logging data, wherein the historical logging data includes logging data related to the sample steam injection well and logging data related to the sample adjacent wells; determining the steam channeling parameter value corresponding to the steam channeling parameter index from the historical logging data; determining the sample result of whether steam channeling occurs in the sample adjacent wells based on the steam channeling parameter value; based on the sample result, determining the positive sample steam injection data and the positive sample well data, as well as the negative sample steam injection data and the negative sample well data from the historical logging data; determining the positive sample recognition result in the sample result, the positive sample steam injection data and the positive sample well data are the positive sample data group, and determining the negative sample recognition result in the sample result, the negative sample steam injection data and the negative sample well data are the negative sample data group.

[0042] In this embodiment, before the target recognition model is constructed, it is also necessary to obtain the logging data related to the sample steam injection well and the logging data related to the sample adjacent well. From the historical logging data obtained, the logging curve related to it is identified according to the steam channeling parameter index, and the corresponding steam channeling parameter value is extracted. The steam channeling parameter index may include fluid pressure, fluid saturation, permeability and other indicators, which are not limited here and can be customized according to specific scenarios and applications. Based on the steam channeling parameter value, the sample result of whether the sample adjacent well has steam channeling is determined. According to the determined sample result, the positive sample steam injection data and the positive sample well data, as well as the negative sample steam injection data and the negative sample well data are selected from the historical logging data. The positive sample group includes the positive sample steam injection data, the positive sample well data and the positive sample recognition result (the sample adjacent well has steam channeling), and the negative sample group includes the negative sample steam injection data, the negative sample well data and the negative sample recognition result (the sample adjacent well has no steam channeling). According to the determined sample results, positive sample steam injection data and positive sample well data, as well as negative sample steam injection data and negative sample well data are selected from the historical well logging data. The positive sample group includes positive sample steam injection data, positive sample well data, and positive sample identification results (steam channeling occurs in the sample's adjacent wells), and the negative sample group includes negative sample steam injection data, negative sample well data, and negative sample identification results (steam channeling does not occur in the sample's adjacent wells).

[0043] It should be noted that by extracting the steam channeling parameter values ​​corresponding to the steam channeling parameter indicators related to the steam channeling phenomenon as features, the expressive ability of the target recognition model can be enhanced, and the key factors that may cause the steam channeling phenomenon during the steam injection process can be captured. Different steam channeling parameter values ​​have different degrees of influence on the steam channeling phenomenon, which can provide the model with more discriminative features to a certain extent, thereby improving the accuracy of the target recognition model.

[0044] As an optional embodiment, determining the sample result of whether steam channeling occurs in the sample adjacent wells based on the steam channeling parameter value includes: determining a steam channeling parameter threshold corresponding to the steam channeling parameter index; and determining the sample result of whether steam channeling occurs in the sample adjacent wells based on the steam channeling parameter value and the steam channeling parameter threshold.

[0045] In this embodiment, the steam channeling parameter threshold corresponding to the steam channeling parameter index can be determined by analyzing the historical logging data. The determination of the steam channeling parameter threshold is not limited here and can be determined based on experience and other methods, and can be used to determine whether the steam channeling parameter value exceeds a preset critical value.

[0046] It should be noted that, for the above steps, the obtained steam channeling parameter value is compared with the steam channeling parameter threshold. There are many types of steam channeling parameter indicators, and different types of indicators represent different steam channeling parameter values. Therefore, the size of the steam channeling parameter value and the steam channeling parameter threshold can be judged according to the different types of steam channeling parameter indicators. For example, the obtained steam channeling parameter indicators are of two types, A and B. The steam channeling parameter value corresponding to A is a, and the steam channeling parameter value corresponding to B is c. Then the steam channeling parameter thresholds corresponding to a and b are not exactly the same, and can be p or q. If a exceeds its corresponding steam channeling parameter threshold p, it can be considered that steam channeling has occurred in the sample's neighboring wells, and if b exceeds its corresponding steam channeling parameter threshold q, it can be considered that steam channeling has not occurred in the sample's neighboring wells. It is determined according to the type and characteristics of the steam channeling parameter indicators. Therefore, the sample results of whether steam channeling occurs in the sample's neighboring wells can be determined based on the threshold of the steam channeling parameter. By determining the steam channeling parameter threshold corresponding to its channeling parameter indicator, in the model training and target recognition process, by comparing with the threshold, the steam channeling parameter values ​​that exceed or fall below the expected range can be captured, and the steam channeling parameter values ​​can be divided into two categories: steam channeling occurs and steam channeling does not occur. This classification method can help accurately determine whether steam channeling occurs in the sample's neighboring wells, provide an effective decision-making basis, and to a certain extent enhance the generalization ability of the model, making it better adaptable to new samples.

[0047] As an optional embodiment, obtaining historical logging data includes: obtaining historical initial logging data; filtering out abnormal logging data that meets predetermined conditions from the historical initial logging data to obtain historical logging data, wherein the predetermined conditions include at least one of the following: the number of dates between the steam channeling date and the steam injection date exceeds a first threshold, and a fluctuation index corresponding to the steam channeling parameter value exceeds a second threshold.

[0048] In this embodiment, first, historical initial logging data are obtained from an available database. These data may include information such as wellhead temperature, wellhead pressure, and liquid production. In the historical initial logging data, abnormal logging data are screened out according to predetermined conditions. The predetermined conditions include at least one of the following two:

[0049] 1) The number of days between the steam breakthrough date and the steam injection date exceeds a first threshold: Based on the known steam injection date and the steam breakthrough date, the number of days between the dates is calculated. If the number of days exceeds a given first threshold, the well logging data is classified as abnormal well logging data.

[0050] 2) The fluctuation index corresponding to the steam channeling parameter value exceeds the second threshold: For each well logging data, the fluctuation index is calculated according to the fluctuation of its steam channeling parameter value. If the fluctuation index exceeds the given second threshold, the well logging data is classified as abnormal well logging data.

[0051] It should be noted that the non-abnormal logging data obtained after the screening step is further used as the result of historical logging data, which includes logging data related to the sample steam injection well and logging data related to the sample adjacent wells. By setting predetermined conditions, most normal logging data can be screened out, unnecessary analysis and processing work can be reduced, time and resources can be saved, and the analysis focusing on abnormal data can obtain accurate results and conclusions more quickly, improving work efficiency.

[0052] As an optional embodiment, the target steam injection data and the target well data are input into the target recognition model to obtain a target recognition result of whether steam channeling occurs in the target adjacent wells, including: obtaining a steam channeling database, wherein the steam channeling database is a database that records the recognition results of whether steam channeling occurs in predetermined adjacent wells of the predetermined well data when steam is injected into the target steam injection well with predetermined steam injection data, the predetermined adjacent wells are wells whose distance from the target steam injection well is less than a predetermined threshold, and the predetermined adjacent wells include the target adjacent wells; the steam channeling database, the target steam injection data and the target well data are input into the target recognition model to obtain a target recognition result of whether steam channeling occurs in the target adjacent wells.

[0053] In this embodiment, the steam channeling database is a database that records the identification results of whether steam channeling occurs in the predetermined adjacent wells of the predetermined well data when the target steam injection well is injected with steam using the predetermined steam injection data. According to the predetermined threshold setting, a well that is closer to the target steam injection well is selected, and the steam channeling identification results of the well and the target steam injection well are recorded in the steam channeling database. The steam channeling database, the target steam injection data, and the target well data are input into the target identification model for identification, and the target identification results of whether steam channeling occurs in the target adjacent wells are further obtained.

[0054] It should be noted that the steam channeling database established records whether the steam channeling phenomenon occurs in the predetermined adjacent wells near the target steam injection well, and uses the target recognition model to identify the steam channeling phenomenon in the target adjacent wells. By establishing a steam channeling database, the steam channeling phenomenon identification results between different time periods and different wells can be integrated and recorded in the same database. On the one hand, it is convenient to query and obtain relevant steam channeling data, providing convenient data support for subsequent analysis, evaluation and decision-making. On the other hand, the steam channeling database will accumulate more and more historical data, which will help mining personnel to more accurately understand the changing trend of steam channeling phenomena, and enable mining personnel to more effectively prevent and control the occurrence of steam channeling phenomena.

[0055] As an optional embodiment, after the steam channeling database, target steam injection data and target well data are input into the target recognition model to obtain the target recognition result of whether steam channeling occurs in the target adjacent wells, it also includes: when the target recognition result is the recognition result of the steam channeling phenomenon occurring in the target well data, determining whether there is a predetermined recognition result corresponding to the target adjacent well of the target well data when steam is injected into the target steam injection well with the target steam injection data in the steam channeling database, and obtaining an existence result; when the existence result is that the predetermined recognition result does not exist in the steam channeling database, the target recognition result of the steam channeling phenomenon sent by the target adjacent well of the target well data when steam is injected into the target steam injection well with the target steam injection data is added to the steam channeling database, and an updated steam channeling database is obtained.

[0056] In this embodiment, the predetermined identification result corresponding to the target neighboring well of the target well data when the target steam injection well is injected with the target steam injection data is queried in the steam channeling database. If the predetermined identification result exists, the existence result is obtained and corresponding processing or recording is performed. If the predetermined identification result does not exist, that is, the identification result of the target neighboring well of the target well data when the target steam injection well is injected with the target steam injection data is not recorded in the predetermined database, then the identification result is added to the steam channeling database, and after the steam channeling database is updated, the new identification result will be included, so that the steam channeling database is more comprehensive and real-time.

[0057] It should be noted that by updating the steam channeling database, on the one hand, it can provide training data for the optimization and improvement of the target recognition model. The newly added target recognition results can be used for model retraining to improve the accuracy and performance of the model and better identify whether steam channeling occurs in the target adjacent wells. On the other hand, by updating the steam channeling database, the integrity and accuracy of the data can be guaranteed, the real-time update of the data can be realized, and support can be provided for data accumulation and experience summary, which is helpful to improve the management and application of steam channeling data.

[0058] As an optional embodiment, after the target steam injection data and the target well data are input into the target recognition model to obtain the target recognition result of whether steam channeling occurs in the target adjacent wells, it also includes: when the target recognition result is an recognition result that steam channeling does not occur in the target adjacent wells, the target steam injection data, the target well data, and the target recognition result are added to the negative sample data group to obtain an updated negative sample data group.

[0059] In this embodiment, after obtaining the target identification result of whether steam channeling occurs in the target adjacent wells, the identification result that steam channeling does not occur in the target adjacent wells can be added to the negative sample data group. Further, the target steam injection data, target well data, and target identification result are added to the negative sample data group to obtain an updated negative sample data group.

[0060] It should be noted that by adding the recognition results of no steam channeling to the negative sample data group, on the one hand, the diversity of negative samples can be expanded, so that the target recognition model will be able to more comprehensively learn and understand the characteristics of the data without steam channeling, thereby improving the judgment and classification performance of the model in the scenario where no steam channeling occurs; on the other hand, the updated negative sample data group can keep the negative sample data group consistent with the latest situation of the target adjacent wells, thereby better reflecting the current state of the steam channeling phenomenon, which is helpful to the training and iteration of the target recognition model to a certain extent, improves the generalization ability and accuracy of the model, and provides a more accurate decision-making basis for steam channeling prevention and control.

[0061] Based on the above embodiments and optional embodiments, an optional implementation is provided, which is described in detail below.

[0062] In the related art, when injecting steam into a steam injection well, there is a problem of low accuracy in identifying whether steam cross-talk occurs in an adjacent well.

[0063] Introduction to the research background of related technologies: The current technical work is mainly focused on three aspects: 1) Research on how to effectively control steam channeling, reduce production losses, improve thermal energy efficiency, and increase heavy oil production. 2) The measurement of steam channeling can be determined by calculating the steam channeling factor. 3) Optimization of steam injection strategy. In the steam injection plan, well groups with serious steam channeling adopt combined steam injection or group steam injection strategies. For example, some operating specifications, based on the information of steam channeling relationship, consider the relationship between geology, well spacing and manifold, and improve the overall production effect through group steam injection or joint steam injection. Steam channeling identification currently relies mainly on field experience and reports. It is mainly based on the typical characteristics of steam channeling: when adjacent wells are injected with steam, the liquid production of the production well will increase accordingly, the water content will continue to rise, the wellhead temperature will continue to rise, and sometimes it will be accompanied by a certain amount of steam. This is a typical feature of oil well steam channeling. In Xinjiang Oilfield, these field information are usually recorded in the notes of the daily production report.

[0064] In view of this, a method for identifying steam crosstalk is provided in an optional implementation manner of the present invention, and the optional implementation manner of the present application is described in detail below.

[0065] S1. Receive a steam channeling phenomenon identification request, wherein the steam channeling phenomenon identification request is used to identify whether steam channeling occurs in a target adjacent well when steam is injected into a target steam injection well using target steam injection data, wherein the target adjacent well is a well whose distance from the target steam injection well is less than a predetermined threshold, and the identification request carries the target steam injection data;

[0066] S2, in response to a steam channeling phenomenon identification request, acquiring target well data of a target adjacent well;

[0067] S3, acquiring historical initial well logging data, filtering out abnormal well logging data that meets predetermined conditions from the historical initial well logging data, and obtaining historical well logging data, wherein the predetermined conditions include at least one of the following: the number of dates between the steam channeling date and the steam injection date exceeds a first threshold, and a fluctuation index corresponding to the steam channeling parameter value exceeds a second threshold;

[0068] It should be noted that after determining the historical well logging data, the steam channeling parameter value corresponding to the steam channeling parameter index is determined from the historical well logging data, and based on the steam channeling parameter value, the sample results of whether steam channeling occurs in the adjacent wells of the sample are determined. Based on the sample results, the positive sample steam injection data and the positive sample well data, as well as the negative sample steam injection data and the negative sample well data are determined from the historical well logging data, and the positive sample recognition results in the sample results are determined, and the positive sample steam injection data and the positive sample well data are the positive sample data group, and the negative sample recognition results in the sample results are determined, and the negative sample steam injection data and the negative sample well data are the negative sample data group. Figure 2 The performance fluctuation diagram of the liquid production in different runners provided by the optional fact mode of the present invention (taking FHW34119 as an example) is used in this step. In order to make the judgment more effective but not missed at the same time, this subcomponent identifies the distribution range (maximum and minimum values) of each parameter, including steam channeling time delay (the number of days from the steam channeling occurrence date minus the steam injection end date), production index jump parameters (for example, the increase rate of liquid production relative to the maximum value of the past 3 days, the increase rate of liquid production in 5 days, the difference between liquid production and yesterday, etc.). In actual use, the maximum range of the above predetermined conditions is selected.

[0069] It should also be noted that the above historical logging data extracts events from the remark information (remark field) of the daily data through text analysis tools, and according to the event type, the event association details are viewed, such as: well maintenance information, measures information, etc. Support the classification management of the remark of the daily data, business personnel can view and modify it, and provide a flexible extension interface for geologists to work locally. The remark information is very short, and the writing style varies greatly, with a high degree of professionalism (involving multiple professional fields such as geology, technology, and oil production engineering). Although the repetition is quite high, the number of unique phrases in the more than 13 million daily reports in the past 10 years is still as high as more than 330,000. It is necessary to use text analysis and knowledge graph technology to further refine the event classification. And realize the extraction of operation event entities, structural association with entities such as wells, blocks, and measures, and support the deduction analysis of the impact of operation events, as well as the traceability analysis of production fluctuations. The basic professional word segmentation is constructed through data dictionaries such as EPDM to solve the problem of short sentences. Based on the ontology relationship of the domain model, guide the construction of entity relationships. Figure 3 The present invention is an implementable embodiment of the oil well operation time classification system diagram, in which the operation events can be extracted into more than 2,000 entities, classified into 6 major categories, 20 medium categories, 62 small categories, and 406 items of the hierarchy. Figure 4 The flow chart of extraction of pumping well operation events provided for the optional implementation mode of the present invention, in the offline modeling stage, uses the word segmentation engine to construct the domain dictionary, based on frequency statistics, discovers common phrase patterns, forms regular expressions, summarizes and summarizes the typical event types of heavy oil production, including abnormal events (as shown in the following table), production management events (such as parameter adjustment), new measures events (such as microbial measures), daily well operation and maintenance events (such as new well maintenance), heavy oil production layer switching and cycle conversion events, external interference events (such as power outages), and forms a complete operation event classification tree system after business discussions. During online processing, according to the offline trained domain dictionary, phrase pattern library, and operation event classification tree, the remark information in the incremental data (daily production, daily steam injection) is extracted, and the events are merged and summarized according to the wheel cycle.

[0070] S4, inputting the target steam injection data and the target well data into the target recognition model to obtain a target recognition result of whether steam channeling occurs in the target adjacent wells;

[0071] It should be noted that the target recognition model is obtained by training the initial model with the positive sample data group and the negative sample data group. The positive sample data group and the negative sample data group are determined based on the steam channeling parameter index and the historical well logging data. The positive sample data group includes the positive sample steam injection data, the positive sample well data and the positive sample recognition result. The negative sample data group includes the negative sample steam injection data, the negative sample well data and the negative sample recognition result. The positive sample recognition result is the recognition result of the steam channeling phenomenon occurring in the adjacent wells of the positive sample when the positive sample steam injection data is used to inject steam into the positive sample steam injection well. The negative sample recognition result is the recognition result of the steam channeling phenomenon not occurring in the adjacent wells of the negative sample when the negative sample steam injection data is used to inject steam into the negative sample steam injection well. Figure 5 The training diagram of the initial model provided by the optional implementation mode of the present invention, specifically the process of obtaining the target recognition model by training the initial model with the positive sample data group and the negative sample data group is as follows Figure 5 shown.

[0072] The construction of the above negative sample data set is mainly based on the following assumption: the on-site identification of steam channeling relationships is timely with a high probability. This assumption can be guaranteed under normal production management levels. Based on this assumption, two negative sample search rules can be derived: There is no steam channeling (at least from a probability perspective) in the two steam channeling wells before the identification date (the earliest date that is considered to have a steam channeling relationship). If there is a situation that satisfies the rule model before the identification date, it should be a negative sample. There is currently no adjacent well pair that has been identified as a steam channeling well (at least from a probability perspective). If there is a situation that satisfies the rule model, it should be a negative sample. Within the above two time and well pair combinations, the steam injection event is used to drive (so that the model does not need to be run frequently). In the adjacent production wells of the steam injection wells, according to the rule model, the time series segments with index jumps are found, and the time series features are extracted to form negative samples.

[0073] Figure 6 The flowchart of the target recognition model provided by the first optional implementation mode of the present invention is as follows: Figure 5 After the initial model is trained as shown, a target recognition model is obtained, wherein the steam channeling relationship library (the same as the steam channeling database mentioned above) is a database for recording the recognition results of whether steam channeling occurs in a predetermined adjacent well of the predetermined well data when steam is injected into a target steam injection well with predetermined steam injection data. Then, the steam channeling database, the target steam injection data and the target well data can be input into the target recognition model to obtain a target recognition result of whether steam channeling occurs in the target adjacent well.

[0074] For the above steps, "the wellhead temperature, wellhead pressure, and liquid production of adjacent wells jump during steam injection" is the basic rule for judging steam channeling, but this is only a qualitative description and cannot be directly applied to achieve automatic identification. Of course, unsupervised learning methods can be used to obtain steam channeling events (positive samples) based on the remark information, and the range of time lag and the description parameters of each indicator jump can be obtained through statistics. However, this is an unsupervised learning process, and there may be many counterexamples (in normal production, when there is no steam channeling, many indicators also jump). It is best to include indicators that have jumped but are not steam channeling (negative samples) to form a binary classification problem, so that the steam channeling model constructed in this way is more robust. In this step, the main idea is to obtain the distribution range of the parameters through the parameters of the positive sample set, and according to these parameter ranges, select negative samples that meet the conditions on the historical data, and use these positive / negative samples to drive model training.

[0075] When applying the model, the steam injection event is used to drive the model, so that the model does not need to be run frequently, which improves the operating efficiency. In the production wells adjacent to the steam injection wells, the time series segments where the indicators jump up are found according to the rule model, the features are extracted, and the steam channeling relationship recognition model is used to determine whether it is a new steam channeling relationship. Even if it is judged not to be a steam channeling relationship, the sample can be stored in the negative sample library to continuously enrich the number of samples and support experts to further identify and mark negative samples.

[0076] Through the above steps, after obtaining the target recognition result of whether the steam channeling phenomenon occurs in the target adjacent well, it is determined that when the target recognition result is the recognition result that the steam channeling phenomenon occurs in the target well data, whether there is a predetermined recognition result corresponding to the target adjacent well of the target well data when the target steam injection data is used to inject steam into the target steam injection well in the steam channeling database, and the existence result is obtained; when the existence result is that the predetermined recognition result does not exist in the steam channeling database, the target recognition result of the target adjacent well of the target well data sending the steam channeling phenomenon when the target steam injection data is used to inject steam into the target steam injection well is added to the steam channeling database to obtain an updated steam channeling database. In addition, the target steam injection data, the target well data, and the target recognition result can be added to the negative sample data group when the target recognition result is the recognition result that the steam channeling phenomenon does not occur in the target adjacent well, to obtain an updated negative sample data group.

[0077] For the above steps S1-S4, we can take the steam channeling process of the 7th round of well FHW34119 as an example, and use the data of more than 3,800 heavy oil wells in the Fengcheng operating area of ​​a certain oil field for the past 10 years (since January 1, 2011) for testing (the maximum number of runner cycles is the 45th round), with a total of more than 11.8 million daily production records and more than 1.2 million daily steam injection records, and more than 70,000 runner cycles were designed. In the daily production report, tens of thousands of steam channeling records were extracted.

[0078] Take the steam channeling process of the 7th round of well FHW34119 as an example. There is serious data missing in the 7th round. From the 30th day, the note has been "pending injection". On November 1, 2014, at 12:00, the steam channeling relationship was affected by the steam injection (serial pump) of (FHW34062), and this impact lasted until January 12. Among them, steam channeling is a development process. The steam injection of FHW34062 well started on October 26, but the steam channeling was only discovered on November 1 (the affected well was discovered after 5 days of production). The end date of steam injection was November 6, but the steam channeling impact lasted until November 12. The impact of the steam channeling on the liquid production is very obvious. The liquid production on November 1 is more than twice that of the previous day.

[0079] The 14th round lasted 512 days, and the number of days of interference with self-flow and steam channeling caused by other wells was as high as 207 days. Five wells were interfered with successively. Table 1 is the relationship between the steam injection date and the occurrence of steam channeling, as shown in Table 1 below.

[0080] Table 1

[0081]

[0082] From the above examples, we can see that when there is no steam channeling, the fluctuation of liquid production is also quite large. Therefore, we use interquartile range (IQR), three times variance and other methods to detect the cluster points. The distance of the wells can be directly calculated using the wellhead coordinates. Based on these characteristic variables, well spacing, etc., according to the above method, the model is trained on 70% of the wells and tested on the remaining 30% of the wells.

[0083] Applying the trained model to well FHW34119 as an example, it has 45 clear notes of the well number of steam channeling sources in 26 rotors. In this example, 40 steam channelings were identified using the above target recognition model. Among them, the notes often use the habit of yesterday, while the target recognition model only reports the first time. For example, on May 26, 2014, the target recognition model and the notes marked steam channeling, but on May 27, the liquid production had dropped to normal levels. The above description method would not report steam channeling again, but the notes continued to use yesterday, with steam channeling marks; and for some small jumps, the target recognition model would not report (these jumps are not significantly different from the jumps in normal production), but the notes considered them to be steam channeling, for example, the liquid production on August 9, 2014 was about 30 tons, which was not much different from the 22.4 tons in the previous three days. In addition, there are omissions caused by missing production data, such as the 7th round. However, this rotor model can also identify some steam crosstalk relationships that are not marked in the notes. For example, on March 15, 2014, its liquid production was much higher than the production in the past three days. After analysis by business experts, it was confirmed to be a typical steam crosstalk event.

[0084] Through the above optional implementation, at least the following beneficial effects can be achieved:

[0085] (1) Since the target recognition model is obtained by training the initial model with the positive sample data set and the negative sample data set, the introduction of the negative sample data set allows the initial model to learn the characteristics and behaviors of the target neighboring wells when no gas channeling occurs. To a certain extent, this can help the target recognition model avoid misjudging the gas channeling phenomenon caused by over-emphasis on a certain category, and can also increase the robustness of the target recognition model, improve the accuracy and reliability of the model recognition, and help to more accurately determine whether the target neighboring wells have gas channeling, and provide more reliable recognition results.

[0086] (2) Since the steam channeling parameter values ​​are determined from historical logging data, that is, by extracting the steam channeling parameter values ​​corresponding to the steam channeling parameter indicators related to the steam channeling phenomenon as features, the expression ability of the target recognition model can be enhanced, and the key factors that may cause the steam channeling phenomenon during the steam injection process can be captured. Different steam channeling parameter values ​​have different degrees of influence on the steam channeling phenomenon, which can provide the model with more discriminative features to a certain extent, thereby improving the accuracy of the target recognition model;

[0087] (3) Since the identification results of no steam channeling are added to the negative sample data set, an updated negative sample data set is obtained. That is, on the one hand, the diversity of negative samples can be expanded, so that the target recognition model will be able to more comprehensively learn and understand the characteristics of data without steam channeling, thereby improving the judgment and classification performance of the model in the scenario where no steam channeling occurs; on the other hand, the updated negative sample data set can keep the negative sample data set consistent with the latest situation of the target neighboring wells, thereby better reflecting the current state of steam channeling, which is helpful to the training and iteration of the target recognition model to a certain extent, improves the generalization ability and accuracy of the model, and provides a more accurate decision-making basis for steam channeling prevention and control;

[0088] (4) The steam channeling database is updated by adding the target steam injection data, target well data, and target identification results of the target wells where steam channeling occurs to the steam channeling database. On the one hand, it can provide valuable training data for the optimization and improvement of the target identification model. The newly added target identification results can be used for retraining the model to improve the accuracy and performance of the model and better identify whether steam channeling occurs in the target adjacent wells. On the other hand, by updating the steam channeling database, the integrity and accuracy of the data can be guaranteed, the real-time update of the data can be realized, and support can be provided for data accumulation and experience summary, which is helpful to improve the management and application of steam channeling data.

[0089] It should be noted that, for the above-mentioned method embodiments, for the sake of simplicity, they are all described as a series of action combinations, but those skilled in the art should know that the present invention is not limited by the described action sequence, because according to the present invention, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present invention.

[0090] Through the description of the above implementation methods, those skilled in the art can clearly understand that the method according to the above embodiment can be implemented by means of software plus a necessary general hardware platform, and of course can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, a magnetic disk, or an optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, a computer, a server, or a network device, etc.) to execute the methods of various embodiments of the present invention.

[0091] Example 2

[0092] According to an embodiment of the present invention, a device for implementing the above-mentioned steam crosstalk phenomenon identification method is also provided. Figure 7 is a structural block diagram of a steam channeling phenomenon identification device according to an embodiment of the present invention. Figure 7 As shown, the device includes: a receiving module 702, an acquisition module 704 and an input module 706. The device is described in detail below.

[0093] A receiving module 702 is used to receive a steam channeling phenomenon identification request, wherein the steam channeling phenomenon identification request is used to identify whether steam channeling occurs in a target adjacent well when steam is injected into a target steam injection well using target steam injection data, the target adjacent well being a well whose distance from the target steam injection well is less than a predetermined threshold, and the identification request carries the target steam injection data;

[0094] An acquisition module 704, connected to the acquisition module 702, is used to respond to a steam channeling phenomenon identification request and acquire target well data of a target adjacent well;

[0095] Input module 706, connected to the above-mentioned input module 704, is used to input the target steam injection data and the target well data into the target recognition model to obtain a target recognition result of whether steam channeling occurs in the target adjacent wells, wherein the target recognition model is obtained by training an initial model with a positive sample data group and a negative sample data group, and the positive sample data group and the negative sample data group are determined based on steam channeling parameter indicators and historical well logging data. The positive sample data group includes positive sample steam injection data, positive sample well data and positive sample recognition results, and the negative sample data group includes negative sample steam injection data, negative sample well data and negative sample recognition results. The positive sample recognition result is the recognition result of steam channeling occurring in the positive sample adjacent wells when steam is injected into the positive sample steam injection well with the positive sample steam injection data, and the negative sample recognition result is the recognition result of no steam channeling occurring in the negative sample adjacent wells when steam is injected into the negative sample steam injection well with the negative sample steam injection data.

[0096] It should be noted here that the above-mentioned receiving module 702, acquisition module 704 and input module 706 correspond to steps S102 to S106 in implementing the steam crosstalk phenomenon identification method, and the examples and application scenarios implemented by the multiple modules and the corresponding steps are the same, but are not limited to the contents disclosed in the above-mentioned embodiment 1.

[0097] Example 3

[0098] According to another aspect of an embodiment of the present invention, there is also provided an electronic device, comprising: a processor; and a memory for storing instructions executable by the processor, wherein the processor is configured to execute the instructions to implement any of the above-mentioned methods for identifying steam crosstalk.

[0099] Example 4

[0100] According to another aspect of an embodiment of the present invention, a computer-readable storage medium is provided. When instructions in the computer-readable storage medium are executed by a processor of an electronic device, the electronic device can execute any of the above-mentioned methods for identifying steam crosstalk.

[0101] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0102] In the above embodiments of the present invention, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0103] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the units can be a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0104] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0105] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0106] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, server or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, disk or optical disk and other media that can store program codes.

[0107] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for identifying steam channeling, characterized in that: include: receiving a steam channeling phenomenon identification request, wherein the steam channeling phenomenon identification request is used to identify whether steam channeling occurs in a target adjacent well when steam is injected into a target steam injection well using target steam injection data, the target adjacent well being a well whose distance from the target steam injection well is less than a predetermined threshold, and the identification request carries the target steam injection data; In response to the steam crosstalk phenomenon identification request, acquiring target well data of the target adjacent well; The target steam injection data and the target well data are input into a target recognition model to obtain a target recognition result of whether the steam crosstalk phenomenon occurs in the target adjacent wells, wherein the target recognition model is obtained by training an initial model with a positive sample data group and a negative sample data group, and the positive sample data group and the negative sample data group are determined based on steam crosstalk parameter indicators and historical well logging data. The positive sample data group includes positive sample steam injection data, positive sample well data and positive sample recognition results, and the negative sample data group includes negative sample steam injection data, negative sample well data and negative sample recognition results. The positive sample recognition result is a recognition result of the occurrence of the steam crosstalk phenomenon in the positive sample adjacent wells when steam is injected into the positive sample steam injection well with the positive sample steam injection data, and the negative sample recognition result is a recognition result of the non-occurrence of the steam crosstalk phenomenon in the negative sample adjacent wells when steam is injected into the negative sample steam injection well with the negative sample steam injection data.

2. The method according to claim 1, characterized in that Before inputting the target steam injection data and the target well data into the target recognition model to obtain the target recognition result of whether the steam cross-talk occurs in the target adjacent well, the method further includes: Acquiring the historical well logging data, wherein the historical well logging data includes well logging data related to the sample steam injection well and well logging data related to the sample adjacent wells; Determining a steam channeling parameter value corresponding to the steam channeling parameter index from the historical well logging data; Determine, based on the steam channeling parameter value, whether the sample adjacent well has steam channeling phenomenon; Based on the sample results, determining the positive sample steam injection data and the positive sample well data, and the negative sample steam injection data and the negative sample well data from the historical well logging data; The positive sample identification result in the sample result is determined, the positive sample steam injection data and the positive sample well data are the positive sample data group, and the negative sample identification result in the sample result is determined, the negative sample steam injection data and the negative sample well data are the negative sample data group.

3. The method according to claim 2, characterized in that The sample result of determining whether steam channeling occurs in the sample adjacent wells according to the steam channeling parameter value includes: Determining a steam blowby parameter threshold value corresponding to the steam blowby parameter indicator; The sample result of whether steam channeling occurs in the sample adjacent well is determined according to the steam channeling parameter value and the steam channeling parameter threshold.

4. The method according to claim 2, characterized in that: The obtaining of the historical logging data comprises: Obtain historical initial logging data; Abnormal logging data meeting predetermined conditions are screened out from the historical initial logging data to obtain the historical logging data, wherein the predetermined conditions include at least one of the following: the number of dates between the steam crosstalk date and the steam injection date exceeds a first threshold, and a fluctuation index corresponding to a steam crosstalk parameter value exceeds a second threshold.

5. The method according to claim 1, characterized in that The step of inputting the target steam injection data and the target well data into a target recognition model to obtain a target recognition result of whether the steam cross-talk occurs in the target adjacent wells includes: Acquire a steam channeling database, wherein the steam channeling database is a database recording an identification result of whether the steam channeling phenomenon occurs in a predetermined adjacent well of the predetermined well data when steam is injected into the target steam injection well with the predetermined steam injection data, the predetermined adjacent well being a well whose distance from the target steam injection well is less than the predetermined threshold, and the target adjacent well is included in the predetermined adjacent well; The steam crosstalk database, the target steam injection data and the target well data are input into the target recognition model to obtain the target recognition result of whether the steam crosstalk phenomenon occurs in the target adjacent well.

6. The method according to claim 5, characterized in that After inputting the steam channeling database, the target steam injection data and the target well data into the target recognition model to obtain the target recognition result of whether the steam channeling phenomenon occurs in the target adjacent well, the method further includes: In the case where the target recognition result is a recognition result that the steam crosstalk phenomenon occurs in the target well data, determining whether there is a predetermined recognition result corresponding to a target adjacent well of the target well data when steam is injected into the target steam injection well using the target steam injection data in the steam crosstalk database, and obtaining an existence result; In the case where the existence result is that the predetermined identification result does not exist in the steam channeling database, when the target steam injection well is injected with the target steam injection data, the target identification result of the steam channeling phenomenon sent by the target adjacent well of the target well data is added to the steam channeling database to obtain an updated steam channeling database.

7. The method according to any one of claims 1 to 6, characterized in that After inputting the target steam injection data and the target well data into the target recognition model to obtain a target recognition result of whether the steam cross-talk occurs in the target adjacent well, the method further includes: When the target identification result is an identification result that the steam crosstalk phenomenon does not occur in the target adjacent well, the target steam injection data, the target well data, and the target identification result are added to the negative sample data group to obtain an updated negative sample data group.

8. A device for identifying steam channeling, characterized in that: include: a receiving module, configured to receive a steam channeling phenomenon identification request, wherein the steam channeling phenomenon identification request is used to identify whether steam channeling occurs in a target adjacent well when steam is injected into a target steam injection well using target steam injection data, wherein the target adjacent well is a well whose distance from the target steam injection well is less than a predetermined threshold, and the identification request carries the target steam injection data; An acquisition module, configured to acquire target well data of the target adjacent well in response to the steam crosstalk phenomenon identification request; An input module is used to input the target steam injection data and the target well data into a target recognition model to obtain a target recognition result of whether the steam crosstalk phenomenon occurs in the target adjacent wells, wherein the target recognition model is obtained by training an initial model with a positive sample data group and a negative sample data group, the positive sample data group and the negative sample data group are determined based on steam crosstalk parameter indicators and historical well logging data, the positive sample data group includes positive sample steam injection data, positive sample well data and positive sample recognition results, the negative sample data group includes negative sample steam injection data, negative sample well data and negative sample recognition results, the positive sample recognition result is a recognition result of the occurrence of the steam crosstalk phenomenon in the positive sample adjacent wells when steam is injected into the positive sample steam injection well with the positive sample steam injection data, and the negative sample recognition result is a recognition result of the non-occurrence of the steam crosstalk phenomenon in the negative sample adjacent wells when steam is injected into the negative sample steam injection well with the negative sample steam injection data.

9. An electronic device, characterized in that: include: processor; a memory for storing instructions executable by the processor; The processor is configured to execute the instructions to implement the method for identifying the steam blowby phenomenon according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: When the instructions in the computer-readable storage medium are executed by a processor of an electronic device, the electronic device is enabled to execute the method for identifying the gas blowby phenomenon as claimed in any one of claims 1 to 7.

Images