Well drilling overflow early warning method, device, equipment, medium and product

Abstract

The embodiment of the invention provides an early warning method, device and equipment for well drilling overflow, a medium and a product, and is applied to the technical field of petroleum engineering. The method comprises the steps of generating second logging data in a second time period based on first logging data in a first time period; according to the first logging data and the second logging data, obtaining a risk identification result at each moment in the first time period and the second time period; based on the risk identification result of each moment, evidence construction is carried out according to the moment, and evidences of different time steps are obtained; based on the plurality of evidences and a preset fusion rule, performing evidence fusion to obtain a lowest probability value and a highest probability value of overflow in the second time period; when at least one of the lowest probability value and the highest probability value is greater than a preset threshold value, early warning information is sent out; wherein the early warning information represents that an overflow phenomenon can occur in a second time period in the future. The technical effect of improving the reliability of well drilling overflow early warning is achieved.

Description

Technical Field

[0001] This application relates to the field of petroleum engineering technology, and in particular to a method, device, equipment, medium and product for early warning of drilling overflow. Background Technology

[0002] In oil and gas drilling, a blowout is an early sign of a well runaway and requires timely detection and control to prevent casualties, environmental pollution, and equipment damage. As oil and gas resource development progresses, environmental uncertainties during drilling increase, leading to a greater risk of blowouts and necessitating further monitoring and detection of these risks.

[0003] Existing overflow detection methods mainly involve: real-time acquisition of overflow-related logging data generated during drilling; and prediction of whether an overflow will occur during drilling in future periods based on the logging data of the current period.

[0004] Because existing technologies mainly rely on real-time data to predict future drilling overflow situations, there are instances of inaccurate predictions, resulting in low reliability of drilling overflow early warning systems. Summary of the Invention

[0005] This application provides a method, apparatus, equipment, medium, and product for early warning of drilling overflow, in order to improve the technical effect of improving the reliability of drilling overflow early warning.

[0006] In a first aspect, embodiments of this application provide a method for early warning of drilling overflow, comprising:

[0007] Based on the first logging data in the first time period, the second logging data in the second time period is generated; wherein, the first logging data refers to the geological layer data and drilling equipment data when drilling is carried out in the first time period; the second logging data refers to the geological layer data and drilling equipment data if drilling is carried out in the future second time period;

[0008] Based on the first and second logging data, the risk identification results are obtained for each moment in the first and second time periods; wherein, the risk identification result at each moment represents the overflow risk at each moment.

[0009] Based on the risk identification results at each time point, evidence is constructed according to the time point to obtain evidence corresponding to different time steps; among them, evidence with more time steps contains evidence with fewer time steps, one time step corresponds to one time point, and each piece of evidence contains risk identification results corresponding to multiple time steps.

[0010] Based on multiple pieces of evidence and preset fusion rules, evidence fusion is performed to obtain the lowest and highest probability values ​​of overflow occurring in the second time period; and when at least one of the lowest and highest probability values ​​is greater than a preset threshold, an early warning message is issued; wherein, the early warning message indicates that an overflow phenomenon will occur in the future second time period.

[0011] In one possible implementation, based on the first logging data and the second logging data, the risk identification results for each moment within the first time period and the second time period are obtained, including:

[0012] Based on the first logging data and the second logging data, determine the sub-data corresponding to each moment in the first time period and the second time period;

[0013] Risk identification is performed on the sub-data corresponding to each time point to obtain the risk identification result corresponding to the sub-data at each time point;

[0014] Among them, sub-data represents the geological layer data and drilling equipment data during the drilling process at its corresponding time; risk identification results refer to the predicted risk type corresponding to the sub-data at each time.

[0015] In one possible implementation, evidence fusion is performed based on multiple pieces of evidence and a preset fusion rule to obtain the lowest and highest probability values ​​of overflow occurring in the second time period, including:

[0016] Based on the first and second logging data, determine the original risk type corresponding to the sub-data at each moment;

[0017] Based on the risk identification results corresponding to the sub-data at each time point, and the original risk type corresponding to the sub-data, the probability distribution of each original risk type in each risk identification result is calculated; where the probability distribution characterizes the reliability of the risk identification results under different original risk types in the first time period and the second time period.

[0018] Based on the mapping relationship between the original risk type and the overflow state, the distribution probability of each overflow state in each risk identification result is determined, and a distribution matrix composed of multiple distribution probabilities is obtained; where the distribution probability characterizes the reliability of the risk identification result under different overflow states in the first time period and the second time period.

[0019] Based on the distribution matrix corresponding to the overflow state, the minimum and maximum probability values ​​of overflow occurring in the second time period are calculated.

[0020] In one possible implementation, based on the distribution matrix corresponding to the overflow state, the minimum and maximum probability values ​​of overflow occurring in the second time period are calculated, including:

[0021] For each piece of evidence, the number of sub-data points corresponding to different risk identification results in the evidence is calculated; where each piece of evidence corresponds to a time period divided from the first time period and the second time period, and each piece of evidence contains risk identification results corresponding to multiple sub-data points;

[0022] Based on the number of sub-data corresponding to different risk identification results in each piece of evidence, a risk ratio vector corresponding to the evidence is generated. The distribution matrix is ​​then corrected based on the risk ratio vector to obtain the distribution probability vector corresponding to each piece of evidence. The distribution probability in the distribution probability vector represents the reliability of the risk identification result under different overflow states within the time period corresponding to the evidence.

[0023] Based on the probability distribution vectors corresponding to each piece of evidence, the fusion probability corresponding to different overflow states is obtained by synthesis calculation; the fusion probability represents the probability of different overflow states occurring in the second time period in the future.

[0024] Based on the fusion probabilities corresponding to different overflow states, the minimum and maximum probability values ​​of overflow occurring in the second time period are calculated.

[0025] In one possible implementation, the fusion probability corresponding to different overflow states is obtained by synthesizing the probability vectors of each piece of evidence, including:

[0026] Based on the probability distribution vector corresponding to each piece of evidence, determine the probability distribution for the same overflow state in each piece of evidence;

[0027] The fusion probability corresponding to different overflow states is calculated by multiplying the probability distributions of the same overflow state under different pieces of evidence.

[0028] In one possible implementation, based on the fusion probabilities corresponding to different overflow states, the minimum and maximum probability values ​​of overflow occurring in the second time period are calculated, including:

[0029] Based on the fusion probability corresponding to different overflow states, confidence and likelihood calculations are performed to obtain the lowest and highest probability values ​​of overflow occurring in the second time period.

[0030] Among them, reliability calculation and likelihood calculation are used to transform the fusion probability into a probability value that can be used for decision-making. Reliability calculation is used to calculate the minimum probability value of overflow occurring in the second future time period, and likelihood calculation is used to calculate the maximum probability value of overflow occurring in the second future time period.

[0031] Secondly, embodiments of this application provide a drilling overflow early warning device, comprising:

[0032] The first processing module is used to generate second logging data for the second time period based on the first logging data in the first time period; wherein, the first logging data refers to the geological layer data and drilling equipment data when drilling is carried out in the first time period; the second logging data refers to the geological layer data and drilling equipment data if drilling is carried out in the future second time period;

[0033] The second processing module is used to obtain the risk identification result for each moment in the first time period and the second time period based on the first logging data and the second logging data; wherein the risk identification result for each moment represents the overflow risk at each moment.

[0034] The third processing module is used to construct evidence based on the risk identification results at each time point, and obtain evidence corresponding to different time steps. The evidence with more time steps contains evidence with fewer time steps. One time step corresponds to one time point, and each piece of evidence contains risk identification results corresponding to multiple time steps.

[0035] The fourth processing module is used to perform evidence fusion based on multiple pieces of evidence and preset fusion rules to obtain the lowest probability value and the highest probability value of overflow occurring in the second time period; and to issue a warning message when at least one of the lowest probability value and the highest probability value is greater than a preset threshold; wherein the warning message indicates that an overflow phenomenon will occur in the future second time period.

[0036] In one possible implementation, the second processing module is further used for

[0037] Based on the first logging data and the second logging data, determine the sub-data corresponding to each moment in the first time period and the second time period;

[0038] Risk identification is performed on the sub-data corresponding to each time point to obtain the risk identification result corresponding to the sub-data at each time point;

[0039] Among them, sub-data represents the geological layer data and drilling equipment data during the drilling process at its corresponding time; risk identification results refer to the predicted risk type corresponding to the sub-data at each time.

[0040] In one possible implementation, the fourth processing module is further configured to:

[0041] Based on the first and second logging data, determine the original risk type corresponding to the sub-data at each moment;

[0042] Based on the risk identification results corresponding to the sub-data at each time point, and the original risk type corresponding to the sub-data, the probability distribution of each original risk type in each risk identification result is calculated; where the probability distribution characterizes the reliability of the risk identification results under different original risk types in the first time period and the second time period.

[0043] Based on the mapping relationship between the original risk type and the overflow state, the distribution probability of each overflow state in each risk identification result is determined, and a distribution matrix composed of multiple distribution probabilities is obtained; where the distribution probability characterizes the reliability of the risk identification result under different overflow states in the first time period and the second time period.

[0044] Based on the distribution matrix corresponding to the overflow state, the minimum and maximum probability values ​​of overflow occurring in the second time period are calculated.

[0045] In one possible implementation, the fourth processing module is further configured to:

[0046] For each piece of evidence, the number of sub-data points corresponding to different risk identification results in the evidence is calculated; where each piece of evidence corresponds to a time period divided from the first time period and the second time period, and each piece of evidence contains multiple sub-data points;

[0047] Based on the number of sub-data corresponding to different risk identification results in each piece of evidence, a risk ratio vector corresponding to the evidence is generated. The distribution matrix is ​​then corrected based on the risk ratio vector to obtain the distribution probability vector corresponding to each piece of evidence. The distribution probability in the distribution probability vector represents the reliability of the risk identification result under different overflow states within the time period corresponding to the evidence.

[0048] Based on the probability distribution vectors corresponding to each piece of evidence, the fusion probability corresponding to different overflow states is obtained by synthesis calculation; the fusion probability represents the probability of different overflow states occurring in the second time period in the future.

[0049] Based on the fusion probabilities corresponding to different overflow states, the minimum and maximum probability values ​​of overflow occurring in the second time period are calculated.

[0050] In one possible implementation, the fourth processing module is also used for

[0051] Based on the probability distribution vector corresponding to each piece of evidence, determine the probability distribution for the same overflow state in each piece of evidence;

[0052] The fusion probability corresponding to different overflow states is calculated by multiplying the probability distributions of the same overflow state under different pieces of evidence.

[0053] In one possible implementation, the fourth processing module is also used for

[0054] Based on the fusion probability corresponding to different overflow states, confidence and likelihood calculations are performed to obtain the lowest and highest probability values ​​of overflow occurring in the second time period.

[0055] Among them, reliability calculation and likelihood calculation are used to transform the fusion probability into a probability value that can be used for decision-making. Reliability calculation is used to calculate the minimum probability value of overflow occurring in the second future time period, and likelihood calculation is used to calculate the maximum probability value of overflow occurring in the second future time period.

[0056] Thirdly, embodiments of this application provide an electronic device, including: a memory and a processor;

[0057] The memory stores the instructions that the computer executes;

[0058] The processor executes computer execution instructions stored in memory, causing the processor to perform the first aspect above and various possible implementations of the first aspect.

[0059] Fourthly, embodiments of this application provide a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the first aspect and various possible implementations thereof.

[0060] Fifthly, embodiments of this application provide a computer program product, including a computer program that, when executed by a processor, implements the first aspect and various possible implementations thereof.

[0061] This application provides a drilling spill early warning method, apparatus, equipment, medium, and product. The method generates second logging data for a future second time period based on first logging data; generates risk identification results for each moment in the first and second time periods based on the first and second logging data; divides the evidence according to the moment corresponding to the risk identification results, and determines the minimum and maximum probability values ​​for a spill occurring in the second time period based on the divided evidence; and issues an early warning message characterizing a spill occurring in the future second time period when at least one of the minimum and maximum probability values ​​is greater than a preset threshold. Compared with existing technologies, this application utilizes real-time acquired first logging data to predict second logging data for future periods; it uses the current first period and the future second period logging data to predict the probability of a blowout in the future second period, thereby predicting whether a blowout will occur during the drilling process in the future based on real-time and predicted data; it obtains evidence representing multiple time periods through evidence segmentation, and then uses the risk identification results corresponding to each piece of evidence to calculate the upper and lower limits of the probability of a blowout, further improving the accuracy of blowout prediction, thereby achieving the technical effect of improving the reliability of drilling blowout early warning. Attached Figure Description

[0062] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0063] Figure 1 Flowchart of the drilling blowout early warning method provided in this application Figure 1 ;

[0064] Figure 2 Flowchart of the drilling blowout early warning method provided in this application Figure 2 ;

[0065] Figure 3 Flowchart of the drilling blowout early warning method provided in this application Figure 3 ;

[0066] Figure 4 A flowchart illustrating the drilling overflow dynamic early warning method based on multi-source information fusion provided in this application;

[0067] Figure 5 A schematic diagram of the technical architecture for realizing dynamic early warning of drilling overflow based on multi-source information fusion, provided in this application;

[0068] Figure 6 This is a flowchart illustrating a DS evidence theory fusion method.

[0069] Figure 7Comparison of early warning effects provided for this application Figure 1 ;

[0070] Figure 8 Comparison of early warning effects provided for this application Figure 2 ;

[0071] Figure 9 A schematic diagram of the structure of the drilling overflow early warning device provided in this application;

[0072] Figure 10 A schematic diagram of the structure of the electronic device provided in this application.

[0073] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0074] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0075] First, let me explain the terms used in this application:

[0076] Surface Pressure of Annular Average (SPPA): This refers to the average pressure of drilling fluid in the annular space between the drill pipe and the wellbore at the wellhead surface during the drilling process. It directly reflects the pressure stability of the wellhead annulus.

[0077] True Vertical Correction Amount (TVCA): This refers to the correction amount for the deviation between the actual length of the wellbore and the true vertical depth. It is used to accurately calculate the vertical position of the formation.

[0078] Gas Analysis Sample Average (GASA): This refers to the average gas concentration in the well per unit time, reflecting the intensity of hydrocarbon gas spillover from the formation. It is used to identify oil and gas-bearing formations and determine the gas abundance and type of the formation.

[0079] Mud Flow & Pressure Composite Parameter (MFOP): This refers to a variable that comprehensively reflects the coupling relationship between "flow rate" and "pump pressure" in the drilling fluid circulation system, and is used to evaluate mud circulation efficiency and wellbore cleanliness.

[0080] In existing technologies, the main methods for detecting wellbore overflows in oil and gas drilling are as follows: by real-time monitoring of geological layer data and drilling equipment-related data during the drilling process, these are identified as real-time logging data during the drilling process; based on the real-time logging data and a pre-trained prediction model, the probability of wellbore overflows occurring in the future is predicted.

[0081] However, existing technologies for overflow detection use real-time drilling data. When making predictions, relying solely on this data lacks consideration for sudden environmental changes during long-term drilling operations. Furthermore, to improve prediction accuracy, existing technologies predict future drilling data and use this data for further risk assessment. However, this method ignores the inherent errors between predicted and actual data, leading to significant discrepancies between the final overflow detection results and the actual overflow occurrence. Therefore, existing technologies suffer from low reliability in drilling overflow early warning systems.

[0082] To address the aforementioned technical issues, this application proposes the following technical concept: First, first logging data for the current first time period is collected; second logging data for future time periods is generated based on the first logging data, taking into account potential environmental changes during long-term drilling. Based on the time steps corresponding to the first and second logging data, the risk identification results corresponding to the two sets of logging data are divided into evidence segments, obtaining evidence corresponding to different time steps. Different time steps represent the number of risk identification results covered by this evidence. The minimum and maximum probability values ​​of a spill occurring in the second time period are obtained using multiple pieces of evidence and preset fusion rules. In generating probability values ​​for the evidence, because the time steps corresponding to the evidence are inconsistent, time steps closer to the current time occupy a higher proportion of proof, while time steps farther from the current time occupy a lower proportion of proof; the longer the prediction step, the lower the impact of the logging data corresponding to that step on the final generated probability value, taking into account the inherent error between predicted and actual data, thereby achieving the technical effect of improving the reliability of drilling spill early warning.

[0083] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.

[0084] Figure 1 Flowchart of the drilling blowout early warning method provided in this application Figure 1 ,like Figure 1 As shown, the method includes:

[0085] S101. Based on the first logging data in the first time period, generate the second logging data in the second time period.

[0086] In this step, before generating the second logging data, it is necessary to acquire the first logging data during the drilling process in the first time period. The first logging data refers to the geological layer data and drilling equipment data during drilling in the first time period; the second logging data refers to the geological layer data and drilling equipment data if drilling were to take place in the future second time period. The first time period represents the current time period including the current moment, and the second time period represents the future time period used for logging data prediction.

[0087] For example, the first logging data collected mainly includes data tags and data content. Data content refers to parameters used to reflect key physical quantities or formation characteristics during the drilling process. The data content consists of four parameters: SPPA, TVCA, GASA, and MFOP. Data tags refer to the original risk type corresponding to each piece of data.

[0088] In this step, the core of logging data refers to inferring bottom characteristics, wellbore status, and drilling safety by monitoring physical quantities during the drilling process in real time. The first logging data refers to multi-dimensional time-series data. "Multi-dimensional" refers to data collected from multiple monitoring dimensions of the drilling process, and "time-series data" refers to time-series data containing timestamps collected within the first time period. Data labels in the first logging data can be manually labeled, or determined using preset labeling rules or a pre-trained labeling model. Each sub-data point in the first logging data also has its corresponding data acquisition timestamp and data label.

[0089] It should be noted that the data structure of the second logging data is the same as that of the first logging data. The second logging data is the prediction data obtained based on the first logging data.

[0090] For example, the second logging data can be generated as follows: the first time period is 12:00-12:30; data is recorded every 30 seconds, resulting in first logging data consisting of 60 sub-data points. The first logging data includes drilling equipment data and geological layer data collected from multiple dimensions. A pre-trained time-series prediction model is used to analyze the data change trend of the first time period, generating the second logging data for the second time period, 12:30-13:00. The parameter types and time granularity of the second logging data are completely identical to those of the first logging data; only the parameter values ​​and data labels are results obtained based on trend prediction.

[0091] It should be noted that the parameters in the first and second logging data refer to key characterization data related to drilling blowouts. When selecting and collecting multi-dimensional data, the parameters should be chosen based on the actual drilling blowout warning situation to ensure that the selected parameters can be used for accurate blowout warnings.

[0092] For example, the data content corresponding to a single moment is the four-dimensional parameter values ​​of SPPA, TVCA, GASA, and MFOP. Low overflow risk in the risk identification results means that all four-dimensional parameters are within the safety threshold; medium overflow risk means that some of the four-dimensional parameters are close to the safety threshold; and high overflow risk means that the four-dimensional parameters exceed the safety threshold.

[0093] S102. Based on the first logging data and the second logging data, obtain the risk identification results for each moment in the first time period and the second time period.

[0094] In this step, the risk identification result at each time step represents the overflow risk at that time step. The risk identification result refers to the predicted risk type output after analyzing the sub-data at each time step through a pre-trained risk identification model. The predicted risk type can be divided into: low overflow risk, medium overflow risk, and high overflow risk, which are used to characterize the probability level of overflow occurring at that time step.

[0095] Each moment refers to the smallest unit of time within the first and second time periods, divided according to a fixed time interval, such as 30 seconds per moment. The division of moments is consistent with the frequency of data collection.

[0096] Alternatively, one possible way to obtain the risk identification results at each time step is as follows:

[0097] S1021. Based on the first logging data and the second logging data, determine the sub-data corresponding to each moment in the first time period and the second time period.

[0098] In this step, sub-data refers to the logging data corresponding to a single moment, which is the smallest unit after the first logging data and the second logging data are separated by time difference.

[0099] S1022. Perform risk identification on the sub-data corresponding to each time moment to obtain the risk identification result corresponding to the sub-data at each time moment.

[0100] In this step, the sub-data represents the geological layer data and drilling equipment data at the corresponding time point during the drilling process; the risk identification result refers to the predicted risk type corresponding to the sub-data at each time point.

[0101] For example, the data for the first time period at 14:00:00 is: SPPA=0.2, TVCA=0.1. The corresponding model output is: low overflow risk; the data for the second time period at 14:40:00 is: SPPA=0.9, TVCA=1.1, GASA=1.0, MFOP=9, and the corresponding model output is: medium overflow risk.

[0102] It should be noted that the risk identification model used is based on a large amount of historical overflow data trained on the model.

[0103] S103. Based on the risk identification results at each time point, evidence is constructed according to the time point to obtain evidence corresponding to different time steps.

[0104] In this step, evidence with more time steps contains evidence with fewer time steps. One time step corresponds to one moment, and each piece of evidence contains risk identification results corresponding to multiple time steps. Each piece of evidence corresponds to a time period, which refers to the time period obtained by dividing the time steps based on the first and second time periods. Specifically, the period and time steps are divided based on the current moment in the first time period and all moments in the second time period.

[0105] For example, if we set the current time in the first time period to 0, and the duration of the second time period to 29, then the time intervals from 0 to 9 can be defined as the first time cycle. Each time interval corresponds to a time step, so the first time cycle contains 10 time steps. The risk identification result corresponding to each time interval in the first time cycle is used as the evidence for that time cycle. In other words, the risk identification results corresponding to each of the 10 time steps in the first time cycle constitute the evidence for the first cycle. If we define the time intervals from 0 to 19 as the second time cycle, and the time intervals from 0 to 29 as the third time cycle, then the evidence for the second time cycle contains the risk identification results for 20 time steps; the evidence for the third time cycle contains the risk identification results for 30 time steps. Furthermore, according to the order of the time intervals, it can be determined that the evidence for the second time cycle includes the evidence for the first time cycle; and the evidence for the third time cycle includes the evidence for the second time cycle.

[0106] S104. Based on multiple pieces of evidence and preset fusion rules, perform evidence fusion to obtain the lowest probability value and the highest probability value of overflow occurring in the second time period; and issue a warning message when at least one of the lowest probability value and the highest probability value is greater than a preset threshold.

[0107] In this step, the warning information indicates that an overflow will occur in the second future time period. The minimum probability value refers to the lowest level of confidence that clearly supports the occurrence of the overflow, calculated by combining all risk identification results from the first and second time periods. The maximum probability value refers to the highest probability that does not oppose the occurrence of the overflow, calculated by combining all risk identification results from the first and second time periods.

[0108] For example, the preset threshold is set to 0.6. When the highest probability value is below 0.6, it is determined that no overflow will occur in the second time period in the future. When the highest probability value is greater than 0.6 and the lowest probability value is less than 0.6, it is determined that an overflow may occur in the second time period in the future, and a yellow warning is required to remind people to pay attention to overflow prevention and equipment inspection and maintenance. When the lowest probability value is greater than 0.6, it is determined that an overflow will occur in the second time period in the future, triggering a red warning. A red warning message is pushed to the client and a red warning signal is issued on site.

[0109] It should be noted that the calculation of the highest and lowest probability values ​​in this step is as follows: Figure 2 and Figure 3 Further explanation will be provided in the embodiments shown, and will not be repeated here.

[0110] The drilling spill early warning method provided in this application generates second logging data for a future second time period based on first logging data; generates risk identification results corresponding to each moment in the first and second time periods based on the first and second logging data; divides the evidence according to the moment corresponding to the risk identification results, and determines the minimum and maximum probability values ​​of a spill occurring in the second time period based on the divided evidence; and issues an early warning message characterizing a spill phenomenon occurring in the future second time period when at least one of the minimum and maximum probability values ​​is greater than a preset threshold. Compared with the prior art, this application uses real-time acquired first logging data to predict the second logging data for future periods; it uses the logging data of the current first time period and the future second time period to predict the probability of a spill occurring in the future second time period, thereby predicting whether a spill will occur in the drilling process in the future period based on real-time data and predicted data; it uses evidence division to obtain evidence characterizing multiple time periods; and it uses the risk identification results corresponding to multiple pieces of evidence to calculate the upper and lower limits of the probability of a spill, further improving the accuracy of spill prediction, thereby achieving the technical effect of improving the reliability of drilling spill early warning.

[0111] Figure 2 Flowchart of the drilling blowout early warning method provided in this application Figure 2 ,like Figure 2 As shown, the method includes:

[0112] S201. Based on the first logging data and the second logging data, determine the original risk type corresponding to the sub-data at each moment.

[0113] In this step, the original risk type refers to the actual risk level corresponding to the sub-data, which is an objective label of whether a spill risk actually exists during the drilling process, rather than a model prediction result. It is the benchmark label for evaluating the reliability of the model prediction. Specifically, the original risk type of the sub-data in the first time period can be determined by combining historical records with on-site verification; since the sub-data in the second time period has not yet occurred, its original risk type needs to be preset by analogy with similar historical scenarios.

[0114] For example, for the sub-data in the first time period, the original risk type corresponding to each sub-data can be determined by means of on-site drilling logs, post-overflow verification reports, or expert annotations; for the sub-data in the second time period, the preset original risk type needs to be determined based on historical similarity parameters.

[0115] S202. Based on the risk identification results corresponding to the sub-data at each time point and the original risk type corresponding to the sub-data, calculate the distribution probability of each original risk type in each risk identification result.

[0116] In this step, the probability distribution characterizes the reliability of the risk identification results under different original risk types in the first and second time periods. That is, it represents the proportion of actual original risk types that are low / medium / high when the model predicts a certain risk type, used to measure the reliability of the model's prediction results. For example, if 97% of the predicted low risk is actually an original low risk, it indicates that the model's prediction of low risk is relatively reliable.

[0117] For example, the probability distribution can be calculated by: constructing a confusion matrix of the risk identification result relative to the original risk type, that is, calculating the data of the original risk type under each prediction result; normalizing each row of the confusion matrix to obtain the probability distribution.

[0118] Assuming that the first and second time periods each contain 1000 time-based sub-data points, then their corresponding confusion matrix is:

[0119] [Risk identification result is low: (Original risk type: low: 702; Original risk type: medium: 15; Original risk type: high: 6)]

[0120] Risk identification results are as follows: (Original risk type: Low: 7; Original risk type: Medium: 848; Original risk type: High: 22)

[0121] Risk identification result is high: (original risk type is low: 0; original risk type is medium: 20; original risk type is high: 132)).

[0122] The probability distribution is calculated based on the basic trust assignment function, as shown in Formula 1:

[0123]

[0124] in, The representative identification result is The actual situation in China is Data percentage For the true result of the i-th data, Let m be the recognition result of the i-th data point, and m be the number of data points. This is an indicator function that takes the value 1 if the condition is met, and 0 otherwise.

[0125] The probability distribution calculated based on the above confusion matrix is:

[0126] [Risk identification result is low: (Probability of original risk type being low: 0.97; Probability of original risk type being medium: 0.02; Probability of original risk type being high: 0.01)]

[0127] The risk identification result is medium: (probability of original risk type being low: 0.005; probability of original risk type being medium: 0.97; probability of original risk type being high: 0.025).

[0128] Risk identification result is high: (Probability of original risk type being low: 0; Probability of original risk type being medium: 0.13; Probability of original risk type being high: 0.87)).

[0129] S203. Based on the mapping relationship between the original risk type and the overflow state, determine the distribution probability of each overflow state in each risk identification result, and obtain a distribution matrix composed of multiple distribution probabilities.

[0130] In this step, the probability distribution characterizes the reliability of the risk identification results under different overflow states in the first and second time periods.

[0131] For example, based on whether an overflow occurs in the identification problem, the overflow state is determined, which can be one of three cases: safe, overflow, or uncertain. Safe is set to Sa, and overflow is set to Ov. By constructing the corresponding non-empty set with a finite number of elements, the resulting set is: { ,{Sa},{Ov},{Sa,Ov}}. This represents three overflow states, among which, {Sa, Ov} is an empty set, and {Sa, Ov} is the union of Sa and Ov. The corresponding overflow state is uncertain, that is, the result is either Sa or Ov.

[0132] The mapping relationship between overflow status and original risk type is as follows: Sa corresponds to low overflow risk in the original overflow type, {Sa,Ov} corresponds to medium overflow risk in the original overflow risk type, and Ov corresponds to high overflow risk in the original overflow risk type.

[0133] For example, based on the example in step S202, the obtained distribution matrix can be:

[0134] [Risk identification result is low: (Sa: 0.97; Sa, Ov: 0.02; Ov: 0.01)]

[0135] The risk identification result is medium: (Sa: 0.005; Sa, Ov: 0.97; Ov: 0.025).

[0136] The risk identification result is high: (Sa: 0; Sa, Ov: 0.13; Ov: 0.87)).

[0137] S204. Based on the distribution matrix corresponding to the overflow state, calculate the minimum and maximum probability values ​​of overflow occurring in the second time period.

[0138] In this step, when calculating the minimum and maximum probability values, it is necessary to adjust the size of the parameters in the distribution matrix by dividing the evidence, and combine different pieces of evidence to obtain the probability distribution vector corresponding to each piece of evidence. The final minimum and maximum probability values ​​are then calculated using the probability distribution vectors corresponding to different pieces of evidence.

[0139] It should be noted that the method of calculating probability values ​​through evidence division is as follows: Figure 3 Further explanation will be provided in the embodiments shown, and will not be repeated here.

[0140] Figure 3 Flowchart of the drilling blowout early warning method provided in this application Figure 3 ,like Figure 3 As shown, the method includes:

[0141] S301. For each piece of evidence, calculate the number of sub-data corresponding to different risk identification results in the evidence.

[0142] In this step, each piece of evidence corresponds to a time period divided from the first and second time periods, and each piece of evidence contains multiple sub-data.

[0143] For example, suppose the data collected in the first time period is the first logging data from 14:00 to 14:30, and the data generated in the future second time period is the second logging data from 14:30 to 16:00. Then, the current time of the first time period can be considered as 14:30. Based on the current time and the second time period, the data is divided into three time periods, with 14:30-15:00 as the first period, 14:30-15:30 as the second period, and 14:30-16:00 as the third period. The purpose of this time period division is to obtain three types of evidence. The risk identification results in each type of evidence are not uniformly distributed, reflecting changes in risk identification prediction over time. According to the time period division method, the first and second logging data can be divided into near-term evidence, medium-term evidence, and long-term evidence. Following a 5-second time step, i.e., each 5 seconds corresponds to a sub-data point, the number of sub-data points in the first period is 360, the number in the second period is 720, and the number in the third period is 1080.

[0144] The purpose of calculating the number of sub-data points corresponding to different risk identification results in each piece of evidence is to statistically analyze the distribution of different risk identification results in different pieces of evidence.

[0145] For example, there are three time periods corresponding to the evidence: recent evidence, medium-term evidence, and long-term evidence. The recent evidence contains 20 data points, of which 18 are low-risk, 2 are medium-risk, and 0 are high-risk, as determined by risk identification. The medium-term evidence contains 40 data points, of which 23 are low-risk, 12 are medium-risk, and 5 are high-risk. The long-term evidence contains 60 data points, of which 23 are low-risk, 15 are medium-risk, and 23 are high-risk. The distribution of risk identification results in each piece of evidence can be determined based on the number of data points corresponding to different risk identification results in each piece of evidence.

[0146] S302. Based on the number of sub-data corresponding to different risk identification results in each piece of evidence, generate a risk ratio vector corresponding to the evidence. Based on the risk ratio vector, correct the distribution matrix to obtain the distribution probability vector corresponding to the evidence.

[0147] In this step, the probability distribution in the probability distribution vector represents the reliability of the risk identification result under different overflow states within the time period corresponding to the evidence.

[0148] For example, the probability distribution vector can be calculated as follows:

[0149] Based on the number of sub-data points corresponding to different risk identification results, the risk ratio vector R corresponding to each type of evidence is calculated. E =[r L ,r M ,r H ], where r L r M r H These represent the proportions of low, medium, and high spillover risk outcomes among all risk information within Evidence E. There are three pieces of evidence: the first period evidence, the second period evidence, and the third period evidence. The risk ratio vector for the first period evidence is (0.9, 0.1, 0); the risk ratio vector for the second period evidence is (0.575, 0.3, 0.125); and the risk ratio vector for the third period evidence is (0.383, 0.25, 0.367). The corresponding distribution matrix is ​​as follows:

[0150] [Risk identification result is low: (Sa: 0.97; Sa, Ov: 0.02; Ov: 0.00)]

[0151] The risk identification result is medium: (Sa: 0.02; Sa, Ov: 0.96; Ov: 0.02).

[0152] The risk identification result is high: (Sa: 0.04; Sa, Ov: 0.14; Ov: 0.82)).

[0153] For each piece of evidence, its corresponding risk ratio vector and distribution matrix are multiplied together to obtain the probability distribution vectors for each piece of evidence as follows: first period evidence [0.875, 0.114, 0.002]; second period evidence [0.569, 0.317, 0.109]; third period evidence [0.391, 0.299, 0.306].

[0154] S303. Based on the probability distribution vectors corresponding to each piece of evidence, perform synthesis calculations to obtain the fusion probability corresponding to different overflow states.

[0155] In this step, the fusion probability represents the probability of different overflow states occurring in the second time period in the future.

[0156] Alternatively, one possible implementation for calculating the fusion probability is as follows:

[0157] S3031. Based on the probability distribution vector corresponding to each piece of evidence, determine the probability distribution for the same overflow state in each piece of evidence.

[0158] For example, based on the probability distribution vector corresponding to each piece of evidence in S303, the probability distribution for the same overflow state in different pieces of evidence is obtained. The probability distribution for overflow state Sa is (0.875, 0.569, 0.391), the probability distribution for overflow state {Sa, Ov} is (0.105, 0.311, 0.295), and the probability distribution for overflow state Ov is (0.002, 0.109, 0.306).

[0159] S3032. Based on the product of the probability distributions corresponding to the same overflow state under each piece of evidence, the fusion probability corresponding to different overflow states is calculated.

[0160] In this step, the fusion probability is calculated as shown in Formula 2:

[0161]

[0162] in, The intersection of the representatives is All overflow states , It is the basic trust assignment function for the e-th piece of evidence. It is the result of fusing all evidence to determine any overflow state within the overflow state domain. The basic trust allocation function is given by formula 3, where n refers to the total number of pieces of evidence and K is the conflict factor.

[0163]

[0164] S304. Based on the fusion probability corresponding to different overflow states, calculate the minimum and maximum probability values ​​of overflow occurring in the second time period.

[0165] Optionally, one possible way to calculate the minimum and maximum probability values ​​is to perform confidence and likelihood calculations based on the fusion probabilities corresponding to different overflow states, and obtain the minimum and maximum probability values ​​of overflow occurring in the second time period.

[0166] In this step, reliability calculation and likelihood calculation are used to transform the fusion probability into a probability value that can be used for decision-making. Reliability calculation is used to calculate the minimum probability value of overflow occurring in the second future time period, and likelihood calculation is used to calculate the maximum probability value of overflow occurring in the second future time period.

[0167] For example, the reliability function is shown in Formula 4:

[0168]

[0169] The likelihood function is shown in Equation 5:

[0170]

[0171] in, Overflow state The confidence function, Overflow state The likelihood function is given, and the remaining parameters are explained in Formulas 2 and 3 above.

[0172] Based on the above embodiments, this application provides a dynamic early warning method for drilling overflow based on multi-source information fusion. Figure 4 A flowchart illustrating the drilling overflow dynamic early warning method based on multi-source information fusion provided in this application is shown below. Figure 4 As shown, the method includes:

[0173] A1. Collect real-time logging data during the drilling process, perform multi-step predictions on the real-time logging data, and obtain predicted logging data for the next 30 steps.

[0174] A2. Risk identification is performed on real-time logging data and predictive logging data to obtain the risk identification results corresponding to the logging data at each moment.

[0175] A3. Using the risk identification results and the original risk types corresponding to the logging data, calculate the minimum and maximum probability values ​​of future overflows; and classify and warn of future overflow possibilities based on the minimum and maximum probability values.

[0176] exist Figure 4 Based on the embodiments shown, this application also provides a technical architecture for realizing dynamic early warning of drilling overflow based on multi-source information fusion. Figure 5 This application provides a schematic diagram of the technical architecture for implementing dynamic early warning of drilling overflow based on multi-source information fusion. Figure 5 As shown, this technical approach includes:

[0177] The data generation module 501 is used to predict real-time logging data using a pre-trained multi-dimensional, multi-step time-series prediction model, obtaining predicted logging data for the next 30 steps. Both the real-time and predicted logging data contain key characterization parameters of the overflow anomaly.

[0178] The risk identification module 502 is used to input real-time logging data and predicted logging data from the multi-source data matrix into a pre-trained intelligent overflow identification model to obtain multi-source overflow risk information. This multi-source overflow risk information includes the current overflow risk for the real-time logging data and the future overflow risk for the predicted time-series data; the overflow risk is categorized as low overflow risk, medium overflow risk, and high overflow risk.

[0179] The risk quantification module 503 is used to calculate the proportion of data representing low, medium, and high spillover risks in different spillover risks based on the basic trust allocation function, and to construct the basic trust allocation function, as shown in Formula 1 above. By dividing current spillover risks and future spillover risks, near-term, medium-term, and long-term evidence are obtained. Using the distribution of different spillover risk types in the near-term, medium-term, and long-term evidence, the trust allocation function corresponding to the data in each period is adjusted, and the corresponding function value is calculated.

[0180] The overflow early warning module 504 is used to perform quantitative risk fusion based on the function values ​​corresponding to recent, medium-term, and long-term evidence to obtain the minimum and maximum probability values ​​for overflows occurring in future periods; and to perform graded overflow early warning based on the minimum and maximum probability values. The quantitative risk fusion method can refer to the implementation process of formulas 2 to 5 in the above embodiments.

[0181] exist Figure 5 Based on the illustrated embodiment, the DS evidence theory fusion method is used to achieve quantitative risk fusion. Figure 6 This is a flowchart illustrating a DS evidence theory fusion method, such as... Figure 6 As shown, the method includes:

[0182] B1. Determine the recognition framework based on the recognition target.

[0183] In this step, the targets to be identified are overflow Ov and safety Sa, and the corresponding identification framework is: Θ={Sa, Ov}, where Θ refers to the complete set of identification frameworks.

[0184] B2. Determine a subset of the recognition framework.

[0185] In this step, the subset of the identified frames includes {{ }, {Sa}, {Sa, Ov}, {Ov}}.

[0186] B3. Calculate the basic trust assignment probability based on a subset of the recognition framework. Figure 5 The implementation process of the risk quantification module 503 shown is consistent.

[0187] B4. Fusion calculation of multi-source evidence based on periodic division. For example... Figure 5 The implementation process of the overflow warning module 504 in the illustrated embodiment.

[0188] Figure 7 and Figure 8 A comparison chart of the early warning effects provided in this application. (For example...) Figure 7As shown, the comparison shows the lead time for a yellow alert based solely on current information versus one based on multi-source information. When issuing a yellow alert, the lead time for medium overflow risk based solely on current information is negative, indicating that the alert is triggered only after an overflow occurs on-site. The lead time for medium overflow risk based on multi-source information is 10 seconds. The lead time for high overflow risk based solely on current information is 97 seconds, and the lead time for medium overflow risk based on multi-source information is 120 seconds. Figure 8 As shown, when a red alert is issued, only current information is used to issue the alert, without providing advance warning of the overflow risk in the middle, indicating that using only current information cannot achieve advance warning of the overflow risk in the middle. The advance warning time for the overflow risk in the middle based on multi-source information is 6 seconds. The advance warning time for the high overflow risk based solely on current information is 16 seconds, and the advance warning time for the overflow risk in the middle based on multi-source information is 70 seconds.

[0189] Figure 9 A schematic diagram of the structure of the drilling overflow early warning device provided in this application is shown below. Figure 9 As shown, the drilling overflow early warning device provided in this embodiment includes:

[0190] The first processing module 901 is used to generate second logging data for the second time period based on the first logging data in the first time period; wherein, the first logging data refers to the geological layer data and drilling equipment data when drilling is carried out in the first time period; the second logging data refers to the geological layer data and drilling equipment data if drilling is carried out in the future second time period.

[0191] The second processing module 902 is used to obtain the risk identification result for each moment in the first time period and the second time period based on the first logging data and the second logging data; wherein the risk identification result for each moment represents the overflow risk at each moment.

[0192] The third processing module 903 is used to construct evidence according to time based on the risk identification results at each time, and obtain evidence corresponding to different time steps; among them, evidence with more time steps contains evidence with fewer time steps, one time step corresponds to one time, and each piece of evidence contains risk identification results corresponding to multiple time steps.

[0193] The fourth processing module 904 is used to perform evidence fusion based on multiple pieces of evidence and preset fusion rules to obtain the lowest probability value and the highest probability value of overflow occurring in the second time period; and to issue a warning message when at least one of the lowest probability value and the highest probability value is greater than a preset threshold; wherein the warning message indicates that an overflow phenomenon will occur in the future second time period.

[0194] In one possible implementation, the second processing module 902 is further configured to:

[0195] Based on the first logging data and the second logging data, the sub-data corresponding to each moment in the first time period and the second time period is determined.

[0196] Risk identification is performed on the sub-data corresponding to each time point to obtain the risk identification result corresponding to the sub-data at each time point.

[0197] Among them, sub-data represents the geological layer data and drilling equipment data during the drilling process at its corresponding time; risk identification results refer to the predicted risk type corresponding to the sub-data at each time.

[0198] In one possible implementation, the fourth processing module 904 is further configured to:

[0199] Based on the first and second logging data, the original risk type corresponding to the sub-data at each time point is determined.

[0200] Based on the risk identification results corresponding to the sub-data at each time point, and the original risk type corresponding to the sub-data, the distribution probability of each original risk type in each risk identification result is calculated; where the distribution probability represents the reliability of the risk identification results under different original risk types in the first time period and the second time period.

[0201] Based on the mapping relationship between the original risk type and the overflow state, the distribution probability of each overflow state in each risk identification result is determined, and a distribution matrix composed of multiple distribution probabilities is obtained; where the distribution probability characterizes the reliability of the risk identification result under different overflow states in the first time period and the second time period.

[0202] Based on the distribution matrix corresponding to the overflow state, the minimum and maximum probability values ​​of overflow occurring in the second time period are calculated.

[0203] In one possible implementation, the fourth processing module 904 is further configured to:

[0204] For each piece of evidence, the number of sub-data points corresponding to different risk identification results in the evidence is calculated; where each piece of evidence corresponds to a time period divided from the first time period and the second time period, and each piece of evidence contains multiple sub-data points.

[0205] Based on the number of sub-data corresponding to different risk identification results in each piece of evidence, a risk ratio vector corresponding to the evidence is generated. The distribution matrix is ​​then corrected based on the risk ratio vector to obtain the distribution probability vector corresponding to each piece of evidence. The distribution probability in the distribution probability vector represents the reliability of the risk identification result under different overflow states within the time period corresponding to the evidence.

[0206] The fusion probability corresponding to different overflow states is obtained by synthesizing the probability vectors of the distribution of each piece of evidence; the fusion probability represents the probability of different overflow states occurring in the second time period in the future.

[0207] Based on the fusion probabilities corresponding to different overflow states, the minimum and maximum probability values ​​of overflow occurring in the second time period are calculated.

[0208] In one possible implementation, the fourth processing module 904 is further used for

[0209] Based on the probability distribution vector corresponding to each piece of evidence, determine the probability distribution for the same overflow state in each piece of evidence.

[0210] The fusion probability corresponding to different overflow states is calculated by multiplying the probability distributions of the same overflow state under different pieces of evidence.

[0211] In one possible implementation, the fourth processing module 904 is further used for

[0212] Based on the fusion probabilities corresponding to different overflow states, confidence and likelihood calculations are performed to obtain the lowest and highest probability values ​​of overflow occurring in the second time period.

[0213] Among them, reliability calculation and likelihood calculation are used to transform the fusion probability into a probability value that can be used for decision-making. Reliability calculation is used to calculate the minimum probability value of overflow occurring in the second future time period, and likelihood calculation is used to calculate the maximum probability value of overflow occurring in the second future time period.

[0214] The apparatus provided in this embodiment can execute the method provided in the above method embodiment. Its implementation principle and technical effect are similar, and will not be described in detail here.

[0215] Figure 10 A schematic diagram of the structure of the electronic device provided in this application. Figure 10 As shown, the electronic device provided in this embodiment includes at least one processor 1001 and a memory 1002. Optionally, the device further includes a communication component 1003. The processor 1001, memory 1002, and communication component 1003 are connected via a bus 1004.

[0216] In the specific implementation process, at least one processor 1001 executes computer execution instructions stored in memory 1002, causing at least one processor 1001 to execute the aforementioned drilling overflow early warning method or method.

[0217] The specific implementation process of processor 1001 can be found in the above method embodiments, and its implementation principle and technical effect are similar. It will not be repeated here.

[0218] In the above embodiments, it should be understood that the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor.

[0219] The memory may include random access memory (RAM) and may also include non-volatile memory (NVM), such as at least one disk storage device.

[0220] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.

[0221] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method.

[0222] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the above-described method.

[0223] The aforementioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The readable storage medium can be any available medium accessible to a general-purpose or special-purpose computer.

[0224] An exemplary readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can reside in an Application Specific Integrated Circuit (ASIC). Alternatively, the processor and the readable storage medium can exist as discrete components in the device.

[0225] The division of units is merely a logical functional division; in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0226] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0227] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0228] If a function is implemented as 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 this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0229] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0230] Finally, it should be noted that other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein, and is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of the invention is limited only by the appended claims.

Claims

1. A method for early warning of drilling overflow, characterized in that, include: Based on the first logging data in the first time period, the second logging data in the second time period is generated; wherein, the first logging data refers to the geological layer data and drilling equipment data when drilling is carried out in the first time period; the second logging data refers to the geological layer data and drilling equipment data if drilling is carried out in the future second time period; Based on the first logging data and the second logging data, risk identification results are obtained at each moment within the first time period and the second time period; wherein, the risk identification result at each moment represents the overflow risk at each moment; Based on the risk identification results at each time point, evidence is constructed according to the time point to obtain evidence corresponding to different time steps; among them, evidence with more time steps contains evidence with fewer time steps, one time step corresponds to one time point, and each piece of evidence contains risk identification results corresponding to multiple time steps; Based on multiple pieces of evidence and preset fusion rules, evidence fusion is performed to obtain the lowest probability value and the highest probability value of overflow occurring in the second time period; and when at least one of the lowest probability value and the highest probability value is greater than a preset threshold, a warning message is issued; wherein, the warning message indicates that an overflow phenomenon will occur in the future second time period.

2. The method according to claim 1, characterized in that, Based on the first logging data and the second logging data, risk identification results are obtained for each moment within the first and second time periods, including: Based on the first logging data and the second logging data, determine the sub-data corresponding to each moment in the first time period and the second time period; Risk identification is performed on the sub-data corresponding to each time moment to obtain the risk identification result corresponding to the sub-data at each time moment; The sub-data represents the geological layer data and drilling equipment data during the drilling process at its corresponding time; the risk identification result refers to the predicted risk type corresponding to the sub-data at each time point.

3. The method according to claim 1, characterized in that, The process of fusing evidence based on multiple pieces of evidence and preset fusion rules to obtain the lowest and highest probability values ​​of overflow occurring in the second time period includes: Based on the first logging data and the second logging data, determine the original risk type corresponding to the sub-data at each moment; Based on the risk identification result corresponding to the sub-data at each time point, and the original risk type corresponding to the sub-data, the distribution probability of each original risk type in each risk identification result is calculated; wherein, the distribution probability represents the reliability of the risk identification result under different original risk types in the first time period and the second time period; Based on the mapping relationship between the original risk type and the overflow state, the distribution probability of each overflow state in each risk identification result is determined, and a distribution matrix composed of multiple distribution probabilities is obtained; wherein, the distribution probability characterizes the reliability of the risk identification result under different overflow states in the first time period and the second time period; Based on the distribution matrix corresponding to the overflow state, the minimum and maximum probability values ​​of overflow occurring during the second time period are calculated.

4. The method according to claim 3, characterized in that, The calculation of the minimum and maximum probability values ​​of overflow occurring during the second time period based on the distribution matrix corresponding to the overflow state includes: For each piece of evidence, the number of sub-data corresponding to different risk identification results in the evidence is calculated; wherein, each piece of evidence corresponds to a time period divided from the first time period and the second time period, and each piece of evidence contains risk identification results corresponding to multiple sub-data; Based on the number of sub-data corresponding to different risk identification results in each piece of evidence, a risk ratio vector corresponding to the evidence is generated. The distribution matrix is ​​then corrected based on the risk ratio vector to obtain a distribution probability vector corresponding to each piece of evidence. The distribution probability in the distribution probability vector represents the reliability of the risk identification result under different overflow states within the time period corresponding to the evidence. Based on the probability distribution vectors corresponding to each piece of evidence, a synthesis calculation is performed to obtain the fusion probability corresponding to different overflow states; the fusion probability represents the probability of the different overflow states occurring in the future second time period. Based on the fusion probabilities corresponding to the different overflow states, the lowest and highest probability values ​​of overflow occurring during the second time period are calculated.

5. The method according to claim 4, characterized in that, The synthesis calculation based on the probability distribution vectors corresponding to each piece of evidence to obtain the fusion probability corresponding to different overflow states includes: Based on the probability distribution vector corresponding to each piece of evidence, determine the probability distribution for the same overflow state in each piece of evidence; The fusion probability corresponding to the different overflow states is calculated by multiplying the probability distributions of the same overflow state under each piece of evidence.

6. The method according to claim 4, characterized in that, The calculation of the minimum and maximum probability values ​​of overflow occurring during the second time period based on the fusion probabilities corresponding to the different overflow states includes: Based on the fusion probabilities corresponding to the different overflow states, confidence and likelihood calculations are performed to obtain the lowest and highest probability values ​​of overflow occurring in the second time period. The confidence calculation and the likelihood calculation are used to convert the fusion probability into a probability value that can be used for decision-making. The confidence calculation is used to calculate the minimum probability value of overflow occurring in the second future time period, and the likelihood calculation is used to calculate the maximum probability value of overflow occurring in the second future time period.

7. A drilling overflow early warning device, characterized in that, include: The first processing module is used to generate second logging data for a second time period based on the first logging data in the first time period; wherein, the first logging data refers to the geological layer data and drilling equipment data when drilling is carried out in the first time period; and the second logging data refers to the geological layer data and drilling equipment data if drilling is carried out in the future second time period. The second processing module is used to obtain the risk identification result at each moment within the first time period and the second time period based on the first logging data and the second logging data; wherein the risk identification result at each moment represents the overflow risk at each moment; The third processing module is used to construct evidence according to time based on the risk identification results at each time point, and obtain evidence corresponding to different time steps; wherein the evidence with more time steps contains evidence with fewer time steps, one time step corresponds to one time point, and each piece of evidence contains risk identification results corresponding to multiple time steps. The fourth processing module is used to perform evidence fusion based on multiple pieces of evidence and preset fusion rules to obtain the lowest probability value and the highest probability value of overflow occurring in the second time period; and to issue a warning message when at least one of the lowest probability value and the highest probability value is greater than a preset threshold; wherein the warning message indicates that an overflow phenomenon will occur in the future second time period.

8. An electronic device, characterized in that, include: Memory, processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory, causing the processor to perform the method as described in any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method as described in any one of claims 1-6.

10. A computer program product, characterized in that, Includes a computer program that, when executed by a processor, implements the method described in any one of claims 1-6.

Images