Intelligent data processing method for petroleum drilling

Abstract

The invention provides a petroleum drilling intelligent data processing method which comprises the following steps: acquiring underground data in real time, and transmitting the acquired data to a ground data processing system in real time for storage and analysis; the real-time data transmitted to the ground data processing system are analyzed, the time period passing through the stratum containing the high-concentration hydrogen sulfide gas is identified, and data segments in the time period are marked as risk segments; analyzing the underground data marked as the risk section, and predicting whether the hydrogen sulfide gas permeation can cause abnormal pressure fluctuation or not; when it is predicted that hydrogen sulfide gas permeation will cause abnormal pressure fluctuation, all signals generated in the risk section are analyzed, fluctuation signals caused by hydrogen sulfide gas permeation are recognized, and classification and marking are conducted; generating corresponding risk early warning and operation suggestions according to signal classification and marking results in the risk section; downhole data are continuously collected and updated, real-time analysis is carried out based on data changes, and risk marks and operation suggestions are dynamically adjusted.

Description

Technical Field

[0001] This invention relates to the field of oil drilling, and more specifically, to an intelligent data processing method for oil drilling. Background Technology

[0002] Oil drilling is the process of extracting oil and natural gas resources by inserting a drill bit deep into the ground using drilling rigs. During drilling, a large amount of complex real-time data needs to be processed, including downhole pressure, temperature, drilling speed, drill bit torque, mud flow rate, and wellbore trajectory. Real-time monitoring and processing of this data is crucial because the drilling environment is complex and constantly changing, and downhole conditions can change rapidly. Any abnormal parameters can trigger serious accidents such as lost circulation, well collapse, or blowouts. Therefore, real-time processing and analysis of this data are essential to ensure the safe, stable, and efficient operation of the drilling process.

[0003] Current oil drilling data processing technologies primarily rely on the collaborative work of downhole sensors, data acquisition systems (DAS), and data analysis software. First, downhole sensors (such as pressure, temperature, drilling speed, and torque sensors) collect various drilling parameters in real time and transmit the data to the surface via cable or wireless transmission systems. The surface data acquisition system receives and initially processes this raw data, performing filtering, noise reduction, and calibration operations to ensure basic data accuracy. Subsequently, the data analysis software inputs the pre-processed data into preset mathematical models or algorithms (such as drilling fluid dynamics models and downhole pressure calculation models), combining historical data and real-time monitoring results to predict and analyze potential risks during the drilling process. Specific processing steps typically include data cleaning, normalization, parameter fitting, and feature extraction to generate drilling parameter optimization suggestions and alerts. Finally, the analysis results are presented to operators through the monitoring system, allowing them to adjust operating parameters such as drilling pressure, pump pressure, and drilling speed according to the system's recommendations to ensure the safe and efficient operation of the drilling process.

[0004] During deep well drilling, especially when traversing formations containing high concentrations of hydrogen sulfide gas, the gas easily seeps into the drilling mud and gradually dissolves, causing minute and irregular fluctuations in downhole pressure data. These fluctuations typically manifest as intermittent, low-amplitude pressure changes. Particularly at low flow rates, the fluctuations caused by hydrogen sulfide gas infiltration are very similar to normal formation pressure changes, making them difficult to distinguish. Current technology cannot accurately identify whether these minute fluctuations originate from the chemical effects of hydrogen sulfide gas infiltration into the drilling mud or from normal formation pressure fluctuations. Therefore, the system often fails to promptly assess the risk of gas infiltration and fails to issue early warnings in the early stages of hydrogen sulfide escape. This causes operators to miss the optimal time for adjusting mud density or venting, potentially leading to well control failure and increasing the risk of serious accidents such as blowouts or well kicks. Summary of the Invention

[0005] The purpose of this invention is to address at least one of the aforementioned shortcomings of the prior art. For example, one objective of this invention is to provide a method for intelligent data processing in oil drilling caused by hydrogen sulfide gas permeation.

[0006] To achieve the above objectives, the present invention provides an intelligent data processing method for oil drilling.

[0007] The intelligent data processing method for oil drilling includes the following steps:

[0008] During the drilling process, downhole sensors are deployed to collect downhole data in real time, and the collected data is transmitted to the surface data processing system for storage and analysis.

[0009] The real-time data transmitted to the ground data processing system is analyzed to identify time periods passing through strata containing high concentrations of hydrogen sulfide gas, and the data segments within those time periods are marked as risk zones.

[0010] Analyze downhole data from areas marked as high-risk to predict whether hydrogen sulfide gas infiltration will cause abnormal pressure fluctuations.

[0011] Given the prediction that hydrogen sulfide gas infiltration would cause abnormal pressure fluctuations, all signals generated within the risk zone were analyzed to identify the fluctuation signals caused by hydrogen sulfide gas infiltration, and these signals were then classified and labeled.

[0012] Based on the signal classification and labeling results within the risk zone, corresponding risk warnings and operational suggestions are generated.

[0013] Continuously collect and update downhole data, analyze data changes in real time, and dynamically adjust risk markers and operational recommendations.

[0014] In an exemplary embodiment of the intelligent data processing method for oil drilling of the present invention, the step of analyzing the real-time data transmitted to the ground data processing system, identifying the time period passing through the formation containing high concentration of hydrogen sulfide gas, and marking the data segment within the time period as a risk segment may specifically include the following steps.

[0015] Preprocess the downhole data transmitted in real time to the ground data processing system.

[0016] Extract hydrogen sulfide gas concentration data from the preprocessed downhole data, and record the actual hydrogen sulfide gas concentration at each time point and the corresponding time period.

[0017] Determine the pre-set hydrogen sulfide gas concentration threshold.

[0018] The actual concentration of hydrogen sulfide gas at all times is compared with a pre-set hydrogen sulfide gas concentration threshold. The time period corresponding to the actual concentration of hydrogen sulfide gas that is greater than the pre-set hydrogen sulfide gas concentration threshold is marked as a risk zone.

[0019] In an exemplary embodiment of the intelligent data processing method for oil drilling of the present invention, the step of analyzing downhole data marked as risk zones to predict whether hydrogen sulfide gas infiltration will cause abnormal pressure fluctuations may specifically include the following steps:

[0020] Extract pressure fluctuation information and hydrogen sulfide gas diffusion information from downhole data marked as risk zones.

[0021] The extracted pressure fluctuation information and hydrogen sulfide gas diffusion information were analyzed to generate the permeation influence coefficient and the dissolution diffusion index, respectively.

[0022] An anomaly prediction model is constructed using the generated permeation influence coefficient and dissolution diffusion index to generate anomaly coefficients. These anomaly coefficients are then compared with a pre-set anomaly coefficient threshold. Based on the comparison results, it is predicted whether hydrogen sulfide gas permeation will cause abnormal pressure fluctuations.

[0023] In an exemplary embodiment of the intelligent data processing method for oil drilling of the present invention, the logic for obtaining the permeability influence coefficient and the dissolution-diffusion index can be as follows:

[0024] Extract pressure fluctuation information from downhole data marked as risk zones, including the actual pressure at the bottom of the well at different times over a period of time, and calibrate the actual pressure at the bottom of the well at different times over a period of time as P. i , i represents the number of the actual pressure at the bottom of the well at different times within a certain period of time, i = 1, 2, 3, ..., g, where g is a positive integer.

[0025] The actual pressure P at the bottom of the well at different times over a period of time i A set G can be constructed based on the time series, and can be denoted as G = {P1, P2, P3, P4, ..., P...}. g-1 P g}

[0026] The Fast Fourier Transform (FFT) is applied to perform frequency domain analysis on the actual pressure at the bottom of the well within a time period in set G, calculating the dominant frequency of the bottom-hole pressure fluctuation. The dominant frequency represents the frequency of change of the bottom-hole pressure per unit time. The specific formula is as follows:

[0027] f bd =FFT(G)=FFT({P1, P2, P3, P4,…,P g-1 P g})

[0028] In the above formula, fbd The dominant frequency of the bottom hole pressure fluctuation is denoted as , and FFT is a fast Fourier transform algorithm used to convert time-domain signals into frequency-domain signals.

[0029] The specific formula for calculating the fluctuation range of the actual pressure at the bottom of the well at different times over a period of time is as follows:

[0030]

[0031] In the above formula, A bd This represents the fluctuation range of the actual pressure at the bottom of the well at different times over a period of time.

[0032] The specific formula for calculating the permeability influence coefficient is as follows:

[0033] PIC = f bd *A bd

[0034] In the above formula, PIC is the permeability influence coefficient.

[0035] Extract hydrogen sulfide gas diffusion information from downhole data marked as risk zones, including hydrogen sulfide gas concentrations at different depths and corresponding depth values. Label the hydrogen sulfide gas concentrations and corresponding depth values ​​at different depths as C. j and X j j represents the hydrogen sulfide gas concentration at different depths and the corresponding depth value number, j = 1, 2, 3, ..., h, where h is a positive integer.

[0036] Calculate the hydrogen sulfide gas concentration gradient at different depths. The concentration gradient represents the rate at which hydrogen sulfide gas changes with depth. The specific formula is as follows:

[0037]

[0038] In the above formula, ΔCT represents the hydrogen sulfide gas concentration gradient at different depths, and X... j C represents the depth value at the j-th depth position. j This represents the hydrogen sulfide gas concentration at the j-th depth location.

[0039] Obtain the diffusion coefficient of hydrogen sulfide gas and the viscosity of the mud. Define the diffusion coefficient of hydrogen sulfide gas and the viscosity of the mud as D and θ, respectively. Calculate the diffusion rate of hydrogen sulfide gas in the mud. The specific calculation formula is as follows:

[0040]

[0041] In the above formula, V ks This represents the diffusion rate of hydrogen sulfide gas in the mud.

[0042] The specific formula for calculating the dissolution-diffusion index is as follows:

[0043]

[0044] In the above formula, DDI is the dissolution-diffusion index.

[0045] In an exemplary embodiment of the intelligent data processing method for oil drilling of the present invention, an anomaly prediction model is constructed using the generated permeability influence coefficient (PIC) and dissolution-diffusion index (DDI). An anomaly coefficient (YC) is generated by weighted summation, and the generated anomaly coefficient (YC) is compared with a pre-set anomaly coefficient threshold (YC). yuzhi A comparison was conducted, and based on the comparison results, it was predicted whether hydrogen sulfide gas infiltration would cause abnormal pressure fluctuations. The specific analysis is as follows:

[0046] If YC≤YC yuzhi It is predicted that hydrogen sulfide gas infiltration will not cause abnormal pressure fluctuations.

[0047] If YC>YC yuzhi It is predicted that the infiltration of hydrogen sulfide gas will cause abnormal pressure fluctuations.

[0048] In an exemplary embodiment of the intelligent data processing method for oil drilling of the present invention, when it is predicted that abnormal pressure fluctuations will be caused by hydrogen sulfide gas infiltration, all signals generated in the risk zone are analyzed to identify the fluctuation signals caused by hydrogen sulfide gas infiltration, and these signals are classified and labeled, specifically including the following steps:

[0049] Given the prediction that hydrogen sulfide gas infiltration would cause abnormal pressure fluctuations, all signal data marked as risk zones were extracted, including the temporal characteristics of each signal.

[0050] The extracted temporal feature information of each signal is preprocessed.

[0051] The time-series characteristics of each preprocessed signal are analyzed to generate the signal consistency coefficient and fluctuation amplitude index for each signal.

[0052] A comprehensive analysis model is constructed using the signal consistency coefficient and fluctuation amplitude index of each generated signal to generate the fluctuation identification coefficient of each signal. The fluctuation identification coefficient of each generated signal is then compared with a pre-set fluctuation identification coefficient threshold. Based on the comparison results, the fluctuation signals caused by hydrogen sulfide gas infiltration are identified, and these signals are classified and labeled.

[0053] In an exemplary embodiment of the intelligent data processing method for oil drilling of the present invention, the logic for obtaining the signal consistency coefficient and fluctuation amplitude index of each signal is as follows:

[0054] Extract the temporal feature information of each preprocessed signal, including the amplitude of each signal at different times within a certain period and the corresponding time points, and then apply the function S to the time series. m Let S be an expression for time point (t), where t is a time interval [t1, t2]. m (t) represents the amplitude of the m-th signal at time t within a time period, where m = 1, 2, 3, ..., k, and k is a positive integer.

[0055] Calculate the rate of change of amplitude for each signal using the formula:

[0056]

[0057] In the above formula, V m (t) represents the amplitude change rate of the m-th signal.

[0058] The average rate of change of amplitude for each signal over a period of time is calculated using the following formula:

[0059]

[0060] In the above formula, Let be the average of the amplitude change rate of the m-th signal over a period of time.

[0061] The standard deviation of the rate of change of amplitude for each signal over a period of time is calculated using the following formula:

[0062]

[0063] In the above formula, Let be the standard deviation of the amplitude change rate of the m-th signal over a period of time.

[0064] The signal consistency coefficient for each signal is calculated using the following formula:

[0065]

[0066] In the above formula, CCON m Let be the signal consistency coefficient of the m-th signal.

[0067] The sum of squared amplitudes of each signal over the time interval [t1, t2] is calculated using the following formula:

[0068]

[0069] In the above formula, EM m Let be the sum of squares of the amplitude of the m-th signal during the time interval [t1, t2].

[0070] The sum of squared amplitudes of each signal over the entire time domain is calculated using the following formula:

[0071]

[0072] In the above formula, ET m Let m be the sum of squares of the amplitude of the m-th signal over the entire time domain.

[0073] The fluctuation amplitude exponent for each signal is calculated using the following formula:

[0074]

[0075] In the above formula, AAM m Let f be the fluctuation amplitude index of the m-th signal.

[0076] In an exemplary embodiment of the intelligent data processing method for oil drilling of the present invention, the signal consistency coefficient (CCON) of each generated signal is... m And volatility index AAM m A comprehensive analysis model is constructed, and the fluctuation identification coefficients (FICs) of each signal are generated through weighted summation. m And the fluctuation identification coefficients (FICs) of each generated signal. m Compared with the pre-set fluctuation identification coefficient threshold FIC yuzhi A comparison was performed, and the fluctuation signals caused by hydrogen sulfide gas permeation were identified based on the comparison results. These signals were then classified and labeled, and the specific analysis is as follows:

[0077] If FIC m <FIC yuzhi If the signal is a normal fluctuation signal, then the signal is classified as a normal signal and marked.

[0078] If FIC m ≥FIC yuzhi If the signal is a fluctuation caused by the permeation of hydrogen sulfide gas, then the signal is classified as an abnormal signal and marked.

[0079] In an exemplary embodiment of the intelligent data processing method for oil drilling of the present invention, based on the signal classification and labeling results within the risk zone, corresponding risk warnings and operational suggestions are generated, specifically including the following steps:

[0080] For signals marked as normal, generate a risk-free warning message based on the result of classifying them as normal, and keep the current drilling operation parameters unchanged.

[0081] For signals marked as anomalous, generate high-risk warning information based on the result of their classification as anomalous signals, and generate operation suggestions based on the characteristics of anomalous signals.

[0082] Another aspect of this invention provides an intelligent data processing method for oil drilling.

[0083] The intelligent data processing method for oil drilling includes the following steps:

[0084] Acquire real-time drilling data, which includes at least real-time hydrogen sulfide gas concentration data and the time period corresponding to the real-time hydrogen sulfide gas concentration data.

[0085] Based on the real-time drilling data, if a first time period exceeding a preset first hydrogen sulfide gas concentration is confirmed, the real-time drilling data acquired within the first time period will be used as the first data.

[0086] A first anomaly coefficient is determined based on the first data, and the first anomaly coefficient is used to characterize the degree of fluctuation of the first data.

[0087] When the first abnormal coefficient exceeds a preset first threshold, the first fluctuation identification coefficient of the first signal acquired within the first time period is confirmed. The first fluctuation coefficient is used to characterize the fluctuation degree of the first signal.

[0088] When the first fluctuation identification coefficient exceeds the preset second threshold, the first signal is regarded as an abnormal pressure fluctuation signal caused by hydrogen sulfide gas infiltration.

[0089] In an exemplary embodiment of the intelligent data processing method for oil drilling of the present invention, the step of confirming the first anomaly coefficient based on the first data may include:

[0090] Based on the first data, the permeability influence coefficient of downhole pressure fluctuations and the dissolution-diffusion index of downhole hydrogen sulfide gas diffusion were confirmed.

[0091] The first anomaly coefficient is obtained by weighted summation of the permeation influence coefficient and the dissolution diffusion index.

[0092] In an exemplary embodiment of the intelligent data processing method for oil drilling of the present invention, the permeability influence coefficient and the dissolution-diffusion index can be determined according to the permeability influence coefficient model and the dissolution-diffusion index model, respectively, wherein the permeability influence coefficient model may include:

[0093]

[0094] f bd =FFT(G),G={P1,P2,P3,P4,…,P g-1 P g}

[0095] In the above formula, PIC is the permeability influence coefficient; A bd The actual pressure P at the bottom of the well iThe fluctuation range; f bd The actual pressure P at the bottom of the well i The dominant frequency of the fluctuation; FFT() is the Fast Fourier Transform algorithm that converts a time-domain signal into a frequency-domain signal; G is the actual pressure P at the bottom of the well. i A collection constructed according to a time series; P i Let P be the actual pressure at the bottom of the well. i The time series numbering, i = 1, 2, 3, ..., g, where g is a positive integer.

[0096] The dissolution-diffusion index model may include:

[0097]

[0098] In the above formula, DDI is the dissolution-diffusion index; V ks denoted as ρ, where ρ is the diffusion rate of hydrogen sulfide gas in the mud; D is the diffusion coefficient of hydrogen sulfide gas; θ is the viscosity of the mud; ΔCT is the concentration gradient of hydrogen sulfide gas at different depths; and X is the diffusion velocity of hydrogen sulfide gas in the mud. j Let C be the depth value at the j-th depth location. j Let be the concentration of hydrogen sulfide gas at the j-th depth position; j is the depth position number, j = 1, 2, 3, ..., h, where h is a positive integer.

[0099] In an exemplary embodiment of the intelligent data processing method for oil drilling of the present invention, the confirmation of the first fluctuation identification coefficient of the first signal acquired within the first time period may include:

[0100] Confirm the consistency coefficient and fluctuation amplitude coefficient of the first signal.

[0101] The first fluctuation identification coefficient is obtained by weighted summation of the consistency coefficient and the fluctuation amplitude coefficient.

[0102] In an exemplary embodiment of the intelligent data processing method for oil drilling of the present invention, the consistency coefficient and the fluctuation amplitude coefficient are determined according to the consistency coefficient model and the fluctuation amplitude coefficient model, respectively, wherein the consistency coefficient model may include:

[0103]

[0104] In the above formula, CCON m Let m be the signal consistency coefficient of the m-th signal; Let be the standard deviation of the amplitude change rate of the m-th signal over the time interval [t1, t2]. V represents the mean of the amplitude change rate of the m-th signal over the time interval [t1, t2]. m (t) represents the rate of change of the amplitude of the m-th signal; S m(t) represents the amplitude of the m-th signal at time t within the time interval [t1, t2], where t is the time point within the time interval [t1, t2], and m = 1, 2, 3, ..., k, where k is a positive integer.

[0105] The fluctuation amplitude coefficient model may include:

[0106]

[0107] In the above formula, AAM m ET is the fluctuation amplitude exponent of the m-th signal; m EM is the sum of squared amplitudes of the m-th signal over the entire time domain; m Let be the sum of squares of the amplitude of the m-th signal within the time interval [t1, t2], where t is the time point within the time interval [t1, t2], and m = 1, 2, 3, ..., k, where k is a positive integer.

[0108] In an exemplary embodiment of the intelligent data processing method for oil drilling of the present invention, it may further include:

[0109] Update the real-time drilling data, and based on the real-time drilling data, confirm a second time period exceeding the preset second hydrogen sulfide gas concentration, and use the real-time drilling data acquired within the second time period as the second data.

[0110] The second anomaly coefficient is determined based on the second data, and the second anomaly coefficient is used to characterize the degree of fluctuation of the second data.

[0111] When the second anomaly coefficient exceeds a preset third threshold, the second fluctuation identification coefficient of the second signal acquired during the second time period is confirmed. The second fluctuation coefficient is used to characterize the fluctuation degree of the second signal.

[0112] When the second fluctuation identification coefficient exceeds the preset fourth threshold, the second signal is regarded as an abnormal pressure fluctuation signal caused by hydrogen sulfide gas infiltration.

[0113] In an exemplary embodiment of the intelligent data processing method for oil drilling of the present invention, when it is confirmed that the first signal is an abnormal pressure fluctuation signal caused by hydrogen sulfide gas infiltration, a high-risk warning is issued and operation suggestions corresponding to the first signal are provided.

[0114] In another aspect, the present invention provides an intelligent data processing device for oil drilling.

[0115] The intelligent data processing device for oil drilling includes a real-time drilling data acquisition module, a first data determination module, a first anomaly coefficient confirmation module, a first fluctuation identification coefficient determination module, and an abnormal pressure fluctuation signal confirmation module, which are connected in sequence.

[0116] The drilling real-time data acquisition module is configured to acquire drilling real-time data, which includes at least real-time hydrogen sulfide gas concentration data and the time period corresponding to the real-time hydrogen sulfide gas concentration data.

[0117] The first data determination module is configured to determine a first time period exceeding a preset first hydrogen sulfide gas concentration based on the real-time drilling data, and to use the real-time drilling data acquired within the first time period as the first data.

[0118] The first anomaly coefficient confirmation module is configured to confirm a first anomaly coefficient based on the first data, wherein the first anomaly coefficient is used to characterize the degree of fluctuation of the first data.

[0119] The first fluctuation identification coefficient module is configured to confirm the first fluctuation identification coefficient of the first signal acquired within the first time period when the first abnormal coefficient exceeds a preset first threshold. The first fluctuation coefficient is used to characterize the fluctuation degree of the first signal.

[0120] The abnormal pressure fluctuation signal confirmation module is configured to recognize the first signal as an abnormal pressure fluctuation signal caused by hydrogen sulfide gas infiltration when the first fluctuation recognition coefficient exceeds a preset second threshold.

[0121] In another aspect, the present invention provides a computer device, the computer device comprising:

[0122] Processor; memory storing a computer program that, when executed by the processor, implements the intelligent data processing method for oil drilling as described above.

[0123] In another aspect, the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the intelligent data processing method for oil drilling as described above.

[0124] Compared with the prior art, the beneficial effects of the present invention include at least one of the following:

[0125] (1) This invention utilizes downhole sensors to collect key downhole data such as pressure and hydrogen sulfide gas concentration in real time during deep well drilling. Comprehensive preprocessing and analysis of this data allows for the accurate identification of time periods passing through formations containing high concentrations of hydrogen sulfide gas, marking these as risk zones. Further analysis of this data generates permeability influence coefficients and dissolution-diffusion indices, and constructs anomaly prediction models. These technical steps significantly improve the ability to identify minute and irregular pressure fluctuations. Unlike traditional technologies, this approach effectively distinguishes between chemical effects caused by hydrogen sulfide gas permeation and normal formation pressure fluctuations, greatly reducing the risk of misjudgment.

[0126] (2) This invention acquires and processes all signal data within the risk zone, extracts the temporal and frequency domain characteristics of each signal, and generates a signal consistency coefficient and fluctuation amplitude index to further identify abnormal signals caused by hydrogen sulfide gas infiltration. This signal classification and labeling process, combined with a comprehensive analysis model and a weighted summation strategy, enables the system to accurately determine which signals are related to hydrogen sulfide gas infiltration, thereby issuing precise risk warnings and operational recommendations. This design avoids erroneous judgments caused by signal interference, improving the reliability and accuracy of the system.

[0127] (3) By continuously collecting and dynamically updating downhole data, the system can automatically adjust risk markers and operational recommendations based on changes in real-time data. This ensures that in complex and constantly changing deep well environments, the system can respond promptly to the risks posed by hydrogen sulfide gas infiltration and take effective preventative measures. This dynamic adjustment mechanism enables the system to provide efficient and accurate decision support during deep well drilling, thereby significantly reducing the probability of serious accidents such as well control failure, blowouts, or well kicks, and ensuring the safety and stability of drilling operations. Attached Figure Description

[0128] The above and other objects and / or features of the present invention will become clearer from the following description taken in conjunction with the accompanying drawings, in which:

[0129] Figure 1 A flowchart illustrating an embodiment of the intelligent data processing method for oil drilling according to the present invention is shown. Detailed Implementation

[0130] In the following sections, an intelligent data processing method for oil drilling according to the present invention will be described in detail with reference to exemplary embodiments.

[0131] It should be noted that terms such as "first," "second," and "third" are used merely for ease of description and distinction, and should not be interpreted as indicating or implying relative importance. The terms "S1," "S2," and "S3" are used to distinguish similar objects, not necessarily to describe a specific order or sequence.

[0132] Example 1

[0133] In this embodiment, as Figure 1 As shown, the intelligent data processing method for oil drilling specifically includes the following steps:

[0134] During the drilling process, downhole sensors are deployed to collect downhole data in real time, and the collected data is transmitted to the surface data processing system for storage and analysis.

[0135] During drilling, a series of downhole sensors can be deployed, including pressure sensors, hydrogen sulfide gas sensors, and mud flow sensors. These sensors can be installed near the drill string, drill bit, and key locations in the wellbore to collect real-time data on downhole pressure changes, hydrogen sulfide gas concentration, and mud flow. Pressure sensors monitor downhole pressure fluctuations, especially minute anomalies; hydrogen sulfide gas sensors detect the presence of harmful gas infiltration; and mud flow sensors monitor mud flow. This data collection aims to identify potential anomalies when traversing formations containing high concentrations of hydrogen sulfide gas, ensuring timely identification of hydrogen sulfide gas infiltration risks during drilling and preventing serious accidents such as well control failure.

[0136] The collected downhole data can be transmitted to the surface data processing system in real time via wireless or wired transmission. Wireless transmission typically employs radio signal transmission technology between downhole sensors and surface receiving devices to ensure stable data transmission in harsh environments. Wired transmission usually relies on fiber optic cables or cables laid inside the drill string to transmit data to the surface in digital signal form. The transmission system incorporates data buffering and compression modules to perform preliminary data processing, ensuring the stability and integrity of large-scale data transmission. Data transmitted to the surface is centralized at the receiving station and connected to the data processing system via a high-speed network for real-time analysis.

[0137] Ground-based data processing systems typically consist of integrated hardware and software platforms, including modules for data reception, storage, analysis, and visualization. The data reception module acquires real-time data transmitted from downhole via sensor links and stores this data in a high-performance database, storing it in time-series format for easy subsequent analysis. The system's data processing software preprocesses, cleans, and normalizes multi-dimensional data such as pressure, gas concentration, and flow rate, providing a high-quality data foundation for subsequent risk assessment and anomaly identification. The data processing system automatically updates and analyzes new data, generating real-time charts and alerts to provide timely decision support for operators.

[0138] The real-time data transmitted to the ground data processing system is analyzed to identify time periods passing through strata containing high concentrations of hydrogen sulfide gas, and the data segments within those time periods are marked as risk zones.

[0139] In this embodiment, the real-time data transmitted to the ground data processing system is analyzed to identify time periods passing through strata containing high concentrations of hydrogen sulfide gas, and data segments within these time periods are marked as risk zones. The specific steps include:

[0140] Preprocess the downhole data transmitted in real time to the ground data processing system.

[0141] When preprocessing downhole data transmitted in real time to the surface data processing system, preprocessing can be performed through data cleaning, outlier filtering, and data normalization. First, data cleaning removes noise and invalid data generated during acquisition, ensuring the quality of data transmitted to the system. This is accomplished by detecting and correcting null values, duplicate values, and erroneous data. Second, outlier filtering identifies and removes data exceeding reasonable ranges by setting upper and lower thresholds or using statistical methods (such as standard deviation and IQR) to prevent the impact of outliers on the analysis results. Finally, data normalization unifies the processing of data from different sensors. Common normalization algorithms (such as Min-Max normalization) are used to convert data from different dimensions into a uniform range, improving the accuracy of data analysis. These preprocessing steps are automatically completed by the software modules of the surface data processing system, providing an accurate and reliable data foundation for subsequent analysis of hydrogen sulfide concentration, identification of pressure fluctuations, and marking of risk zones.

[0142] Extract hydrogen sulfide gas concentration data from the preprocessed downhole data, and record the actual hydrogen sulfide gas concentration at each time point and the corresponding time period.

[0143] When extracting hydrogen sulfide gas concentration data from preprocessed downhole data, the system first filters out relevant data from the sensor dataset. This data typically includes the actual hydrogen sulfide gas concentration at each time point and the corresponding timestamp. The system then records the hydrogen sulfide concentration for each time period based on the time series, mapping it to the time period to form a dataset that includes a "time period - hydrogen sulfide concentration" correlation. This dataset is then used to compare with preset hydrogen sulfide concentration thresholds to identify risk areas.

[0144] In the specific extraction process, common methods include keyword- or data tag-based extraction and time-series-based filtering. First, the software system filters data related to hydrogen sulfide gas from the original dataset using predefined data tags or field names (such as "hydrogen sulfide concentration"). This step typically relies on data querying or data segmentation techniques. Second, the system arranges this data chronologically and extracts the hydrogen sulfide concentration data according to time periods, forming a structured dataset of "time period - hydrogen sulfide concentration". The entire process is automated through the data processing software's module to ensure the accuracy and real-time nature of the data extraction.

[0145] Determine the pre-set hydrogen sulfide gas concentration threshold.

[0146] Determining the pre-set hydrogen sulfide gas concentration threshold can be done through various methods, including historical data analysis, industry standards, and expert experience. First, the system can analyze hydrogen sulfide concentration data collected during previous drilling operations. Based on the relationship between risk events and hydrogen sulfide concentration in historical data, a threshold applicable to the current drilling environment can be determined. Additionally, a basic safety threshold can be set based on the safe concentration range specified in industry standards or regulations. Building upon this, and considering factors such as formation conditions, well depth, and mud characteristics, experts refine the final threshold range by inputting adjustment values ​​through software. The final hydrogen sulfide concentration threshold is set through the configuration module of the data processing system, which uses this threshold as a comparison standard to subsequently determine whether the hydrogen sulfide concentration exceeds the safe range.

[0147] The actual concentration of hydrogen sulfide gas at all times is compared with a pre-set hydrogen sulfide gas concentration threshold. The time period corresponding to the actual concentration of hydrogen sulfide gas that is greater than the pre-set hydrogen sulfide gas concentration threshold is marked as a risk zone.

[0148] When comparing the actual hydrogen sulfide gas concentration at all times with the preset hydrogen sulfide gas concentration threshold, the system first compares the hydrogen sulfide concentration value at each time point with the set threshold in chronological order. If the actual concentration at a certain time point exceeds the preset threshold, the system will determine that the corresponding time period is a potential risk zone. This comparison process is implemented through the software's built-in automated calculation module. The system processes and judges the hydrogen sulfide concentration data in real time, ensuring rapid identification of times when the concentration exceeds the standard, and including the data of the corresponding time period in the further analysis.

[0149] Common methods for labeling include time-based labeling and data tagging. After the system identifies a time period where hydrogen sulfide concentration exceeds the standard, it automatically labels that time period as a "risk zone." Specifically, the data processing module categorizes eligible time-based data and attaches a "risk zone" tag or identifier. This tag information is typically stored in the database along with timestamps and stratigraphic information. This labeling method allows subsequent analysis and processing to quickly locate these high-risk zones. All labeling operations are automated through data management software to ensure the accuracy and real-time nature of the labeling process.

[0150] Analyze downhole data from areas marked as high-risk to predict whether hydrogen sulfide gas infiltration will cause abnormal pressure fluctuations.

[0151] In this embodiment, downhole data marked as risk zones are analyzed to predict whether hydrogen sulfide gas infiltration will cause abnormal pressure fluctuations. The specific steps include:

[0152] Extract pressure fluctuation information and hydrogen sulfide gas diffusion information from downhole data marked as risk zones.

[0153] When extracting pressure fluctuation and hydrogen sulfide gas diffusion information from downhole data marked as risk zones, the system first filters data related to these parameters using data tags or field names. Specifically, within the dataset marked as risk zones, the software extracts key fields based on pre-defined "pressure fluctuation" or "hydrogen sulfide diffusion" tags. For pressure fluctuation information, the system extracts pressure values ​​from the time series and automatically calculates the fluctuation amplitude and frequency. For hydrogen sulfide gas diffusion information, the system extracts hydrogen sulfide concentration data at different depths and generates diffusion characteristic data by calculating the concentration gradient and diffusion rate. The entire extraction process relies on the software's built-in data query, segmentation, and calculation modules, automatically filtering, extracting, and processing relevant information from large amounts of data to ensure accurate extraction results that meet subsequent analysis requirements.

[0154] The extracted pressure fluctuation information and hydrogen sulfide gas diffusion information were analyzed to generate the permeation influence coefficient and the dissolution diffusion index, respectively.

[0155] An anomaly prediction model is constructed using the generated permeation influence coefficient and dissolution diffusion index to generate anomaly coefficients. These anomaly coefficients are then compared with a pre-set anomaly coefficient threshold. Based on the comparison results, it is predicted whether hydrogen sulfide gas permeation will cause abnormal pressure fluctuations.

[0156] When determining the pre-set anomaly coefficient threshold, various methods can be used, including historical data analysis, expert experience setting, and machine learning algorithm optimization. First, the system analyzes the permeability influence coefficient and dissolution-diffusion index collected from similar drilling operations in the past. Based on known anomalies in this historical data, it statistically analyzes the range of coefficients related to abnormal pressure fluctuations, initially determining a basic threshold. Then, experts can fine-tune the threshold through the software configuration interface according to specific working conditions (such as well depth and formation type), setting an initial threshold more suitable for the current operation. Furthermore, the system can utilize machine learning algorithms to continuously optimize and update the threshold by analyzing real-time collected data and actual downhole conditions, making it more accurate. Finally, the set anomaly coefficient threshold is stored in the system and used as a comparison standard during the prediction process to determine whether hydrogen sulfide gas permeation may cause abnormal pressure fluctuations.

[0157] In this embodiment, the logic for obtaining the permeation influence coefficient and the dissolution-diffusion index is as follows:

[0158] Extract pressure fluctuation information from downhole data marked as risk zones, including the actual pressure at the bottom of the well at different times over a period of time, and calibrate the actual pressure at the bottom of the well at different times over a period of time as P.i , i represents the number of the actual pressure at the bottom of the well at different times within a certain period of time, i = 1, 2, 3, ..., g, where g is a positive integer.

[0159] When extracting pressure fluctuation information from downhole data marked as risk zones, the first step is to filter pressure-related data using data labels or field names. Specifically, the software system extracts relevant fields from the dataset marked as risk zones, according to predefined labels such as "Pressure" or "Bottomhole Pressure." These fields contain bottomhole pressure values ​​at different time points. After extraction, the system sorts this data by time series to ensure the pressure data is arranged chronologically for subsequent frequency and amplitude analysis. The entire process relies on the automated query and filtering functions of the data processing software, achieving accurate data extraction and organization through the setting of conditions and rules, thereby providing the foundational data for further pressure fluctuation analysis.

[0160] The actual pressure P at the bottom of the well at different times over a period of time i Construct a set G based on the time series, denoted as G = {P1, P2, P3, P4, ..., P...} g-1 P g}

[0161] The Fast Fourier Transform (FFT) is applied to perform frequency domain analysis on the actual pressure at the bottom of the well within a time period in set G, calculating the dominant frequency of the bottom-hole pressure fluctuation. The dominant frequency represents the frequency of the bottom-hole pressure change per unit time, and the specific formula is as follows:

[0162] f bd =FFT(G)=FFT({P1, P2, P3, P4,…,P g-1 P g}).

[0163] In the above formula, f bd The dominant frequency of the bottom hole pressure fluctuation is denoted as , and FFT is a fast Fourier transform algorithm used to convert time-domain signals into frequency-domain signals.

[0164] Frequency domain analysis of actual bottom hole pressure data at different time points over a period of time is performed using Fast Fourier Transform (FFT) to extract the dominant frequency of bottom hole pressure fluctuations. Specifically, the pressure data, as a time-series signal, is originally located in the time domain (i.e., arranged chronologically). To analyze the frequency characteristics of the pressure signal, it is necessary to convert the time-domain signal into a frequency-domain signal. FFT is an efficient algorithm that can quickly convert signals in a time series into frequency components. This is achieved by using a set G = {P1, P2, P3, P4, ..., P...} g-1 P g Perform an FFT operation to obtain the amplitude corresponding to each frequency, where the frequency with the largest amplitude is the dominant frequency f.bd This represents the most significant periodic change in the pressure signal. The magnitude of the dominant frequency indicates the frequency of change in bottom hole pressure per unit time, thus reflecting the main characteristics of bottom hole pressure fluctuations. This process provides fundamental data for subsequent analysis of the dynamic changes in the bottom hole environment.

[0165] The specific formula for calculating the fluctuation range of the actual pressure at the bottom of the well at different times over a period of time is as follows:

[0166]

[0167] In the above formula, A bd This represents the fluctuation range of the actual pressure at the bottom of the well at different times over a period of time.

[0168] The specific formula for calculating the permeability influence coefficient is as follows:

[0169] PIC = f bd *A bd

[0170] In the above formula, PIC is the permeability influence coefficient.

[0171] The magnitude of the permeability influence coefficient directly reflects the comprehensive impact of hydrogen sulfide gas permeation on bottom hole pressure fluctuations. A large permeability influence coefficient typically indicates a significant increase in both the frequency and amplitude of bottom hole pressure fluctuations, suggesting that hydrogen sulfide gas permeation may be triggering abnormal pressure fluctuations. These abnormal fluctuations may be due to pressure instability caused by chemical reactions or physical diffusion resulting from hydrogen sulfide gas permeation. Therefore, the magnitude of the permeability influence coefficient can serve as an important indicator for determining whether hydrogen sulfide gas permeation will cause abnormal pressure fluctuations. When this coefficient exceeds a preset safety threshold, the system will issue a warning, indicating the potential for abnormal pressure fluctuations and helping operators take timely countermeasures.

[0172] Extract hydrogen sulfide gas diffusion information from downhole data marked as risk zones, including hydrogen sulfide gas concentrations at different depths and corresponding depth values. Label the hydrogen sulfide gas concentrations and corresponding depth values ​​at different depths as C. j and X j j represents the hydrogen sulfide gas concentration at different depths and the corresponding depth value number, j = 1, 2, 3, ..., h, where h is a positive integer.

[0173] When extracting hydrogen sulfide gas diffusion information from downhole data marked as risk zones, the first step is to use the software's data query and filtering functions to select data related to hydrogen sulfide concentration and depth location. Specifically, the system extracts data within the risk zone based on the labels or fields of "hydrogen sulfide concentration" and "depth location." The extracted data includes hydrogen sulfide concentration values ​​at different depths and their corresponding depth coordinates. The software then arranges this data in depth order for subsequent calculations of the hydrogen sulfide gas concentration gradient and diffusion rate. This entire process relies on the automated data extraction and sorting functions of the data processing software. By setting rules, accurate hydrogen sulfide diffusion information at different depths in the well can be obtained, ensuring the accuracy and completeness of the data.

[0174] Calculate the hydrogen sulfide gas concentration gradient at different depths. The concentration gradient represents the rate at which hydrogen sulfide gas changes with depth. The specific formula is as follows:

[0175]

[0176] In the above formula, ΔCT represents the hydrogen sulfide gas concentration gradient at different depths, and X... j C represents the depth value at the j-th depth position. j This represents the hydrogen sulfide gas concentration at the j-th depth location.

[0177] Calculating the hydrogen sulfide gas concentration gradient at different depths is to determine the diffusion rate of hydrogen sulfide gas downhole. The concentration gradient ΔCT represents the rate at which the hydrogen sulfide gas concentration changes with depth. Specifically, the calculation method involves comparing the hydrogen sulfide gas concentration C at two adjacent depth locations. j+1 and C j Perform the difference calculation and divide it by the corresponding depth difference X. j+1 -X j The concentration gradient is obtained. This formula reflects the rate of change of hydrogen sulfide concentration between different depths. A larger concentration gradient indicates that hydrogen sulfide gas diffuses faster downhole, which is crucial for assessing the impact of hydrogen sulfide permeation on the downhole environment. The calculation results of this step provide key parameters for subsequent diffusion rate analysis and dissolution-diffusion index calculation.

[0178] Obtain the diffusion coefficient of hydrogen sulfide gas and the viscosity of the mud. Define the diffusion coefficient of hydrogen sulfide gas and the viscosity of the mud as D and θ, respectively. Calculate the diffusion rate of hydrogen sulfide gas in the mud. The specific calculation formula is as follows:

[0179]

[0180] In the above formula, V ks This represents the diffusion rate of hydrogen sulfide gas in the mud.

[0181] Calculate the diffusion rate V of hydrogen sulfide gas in the mud. ks This is to assess the diffusion rate of hydrogen sulfide gas during downhole penetration. The diffusion rate is determined by the diffusion coefficient D of the hydrogen sulfide gas and the viscosity θ of the drilling mud, as shown in the formula: The diffusion coefficient D represents the diffusion capacity of hydrogen sulfide gas in the mud, while the viscosity θ reflects the resistance of the mud to gas diffusion. This formula shows that the diffusion rate is directly proportional to the diffusion coefficient and inversely proportional to the viscosity; that is, the larger the diffusion coefficient and the smaller the viscosity, the faster the gas diffuses in the mud.

[0182] The diffusion coefficient D of hydrogen sulfide gas and the viscosity θ of the drilling mud can be obtained in the following ways: The diffusion coefficient D is usually obtained through experimental data and existing literature. It can be found in the physical property database of the software or calculated using empirical formulas based on the downhole temperature, pressure, and molecular characteristics of hydrogen sulfide. The viscosity θ of the drilling mud can be obtained by real-time monitoring of the downhole drilling mud rheological properties. The software system will collect rheological data from downhole sensors and calculate the relationship between shear stress and shear rate based on this data to obtain the viscosity parameter.

[0183] The diffusion coefficient D of hydrogen sulfide gas represents the ability of gas molecules to diffuse in drilling mud, and is affected by downhole temperature, pressure, and the physical properties of the mud. The diffusion coefficient is usually obtained from experiments or physical property databases and is a key parameter reflecting the diffusion characteristics of gases in liquids. The viscosity θ of the drilling mud is a physical quantity reflecting the internal flow resistance of the fluid, usually obtained through rheological testing. The software collects downhole mud flow data through sensors and analyzes it using a built-in calculation module to calculate the viscosity value. Higher viscosity means greater fluid resistance and slower diffusion rate; therefore, viscosity is an important parameter for calculating diffusion rate.

[0184] The specific formula for calculating the dissolution-diffusion index is as follows:

[0185]

[0186] In the above formula, DDI is the dissolution-diffusion index.

[0187] The dissolution-diffusion index (DDI) reflects the diffusion rate and concentration trend of hydrogen sulfide gas in drilling mud, and is closely related to predicting whether hydrogen sulfide gas infiltration will cause abnormal pressure fluctuations. A larger DDI indicates that hydrogen sulfide gas diffuses faster and has a steeper concentration gradient in the drilling mud, which may lead to rapid fluctuations in bottom hole pressure, thus increasing the risk of abnormal pressure fluctuations. Conversely, a smaller DDI indicates slower gas diffusion, more stable concentration changes, smaller pressure fluctuations, and a relatively lower risk of abnormal fluctuations. Therefore, by monitoring the DDI, the potential impact of hydrogen sulfide gas infiltration on bottom hole pressure fluctuations can be effectively assessed, providing a basis for early warning.

[0188] In this embodiment, the generated permeability influence coefficient (PIC) and dissolution-diffusion index (DDI) are used to construct an anomaly prediction model. An anomaly coefficient (YC) is generated by weighted summation, and the generated anomaly coefficient (YC) is compared with a pre-set anomaly coefficient threshold (YC). yuzhi A comparison was conducted, and based on the comparison results, it was predicted whether hydrogen sulfide gas infiltration would cause abnormal pressure fluctuations. The specific analysis is as follows:

[0189] If YC≤YC yuzhi The prediction that hydrogen sulfide gas infiltration will not cause abnormal pressure fluctuations means that in the current downhole environment, the degree of hydrogen sulfide gas infiltration is low, and the diffusion rate and the pressure fluctuation amplitude under equilibrium conditions are within a controllable range. The values ​​of the infiltration influence coefficient and the dissolution-diffusion index are relatively small, indicating that the interference of hydrogen sulfide gas infiltration on the bottom hole pressure is limited and has not reached the critical point for triggering an early warning. Under these circumstances, the pressure change trend at the bottom hole remains stable and there will be no drastic fluctuations. For drilling operations, this means that operations can continue as planned without the need for additional well control measures. The operating environment is relatively safe, and the system will continue to perform routine monitoring to ensure the continued stability of downhole conditions.

[0190] If YC>YC yuzhi The prediction indicates that hydrogen sulfide gas infiltration will cause abnormal pressure fluctuations, suggesting that the infiltration has reached a level that may trigger abnormal pressure fluctuations at the bottom of the well. This abnormality may be due to increased pressure fluctuations caused by hydrogen sulfide gas seeping into the drilling mud, which may lead to downhole instability, such as well control risks like blowouts and well kicks. In this case, the system will issue an early warning, prompting operators to quickly assess the current situation and take countermeasures according to the system's recommendations, such as adjusting the mud density, reducing the drilling rate, or implementing venting. This situation has a significant impact, and if not handled in time, it may lead to serious safety accidents. Therefore, operators need to pay close attention and respond quickly.

[0191] When constructing anomaly prediction models, the first step is to combine the Permeability Influence Coefficient (PIC) and the Dissolution-Diffusion Index (DDI) and generate the anomaly coefficient YC through weighted summation. Specifically, PIC and DDI are assigned different weight coefficients ω1 and ω2, and adjusted according to their relative contributions to abnormal pressure fluctuations. The weight coefficients are typically determined based on historical data and actual operating environments, optimized by comparing pressure fluctuations under different conditions. For example, if the PIC has a more significant impact on abnormal pressure fluctuations, its weight ω1 will be relatively large, while the DDI's weight ω2 will be relatively small. The final anomaly coefficient calculation formula is: YC = ω1 * PIC + ω2 * DDI. Through weighted summation, the model can more accurately reflect the contributions of permeability and diffusion effects to abnormal pressure fluctuations. Adjusting the weight coefficients allows for dynamic optimization based on different operating scenarios, ensuring the accuracy of the prediction model.

[0192] Given the prediction that hydrogen sulfide gas infiltration would cause abnormal pressure fluctuations, all signals generated within the risk zone were analyzed to identify the fluctuation signals caused by hydrogen sulfide gas infiltration, and these signals were then classified and labeled.

[0193] In this embodiment, when it is predicted that hydrogen sulfide gas infiltration will cause abnormal pressure fluctuations, all signals generated in the risk area are analyzed to identify the fluctuation signals caused by hydrogen sulfide gas infiltration, and these signals are classified and labeled. Specifically, this includes the following steps:

[0194] Given the prediction that hydrogen sulfide gas infiltration would cause abnormal pressure fluctuations, all signal data marked as risk zones were extracted, including the temporal characteristics of each signal.

[0195] When abnormal pressure fluctuations are predicted to be caused by hydrogen sulfide gas infiltration, all signal data marked as risk zones are extracted, including the time-series characteristics of each signal. This is typically achieved through automated extraction functions in data processing software. In practice, the software uses conditional filtering based on predefined risk zone labels to select relevant multidimensional signal data. This signal data includes time-series data such as downhole pressure, temperature, flow rate, and hydrogen sulfide concentration. The extraction process mainly relies on database queries and data mining tools. The software retrieves and extracts signal data line by line within a specified time period based on set labels and fields, and automatically categorizes and stores this time-series data according to signal type. After extraction, the data is stored in a structured format within the system for subsequent analysis and processing.

[0196] The extracted temporal feature information of each signal is preprocessed.

[0197] Before analyzing the time-series features of the extracted signals, preprocessing is essential to improve the accuracy, completeness, and usability of the data, ensuring the reliability of subsequent analysis results. Preprocessing typically includes steps such as noise filtering, data smoothing, missing data completion, and data normalization. First, noise filtering removes high-frequency interference and abnormal fluctuations from the signal using filters, ensuring signal purity. Second, data smoothing eliminates short-term fluctuations in the signal using methods such as moving averages, making the data trend more stable. For missing data, interpolation algorithms can be used to complete the data, avoiding errors caused by incomplete data. Finally, data normalization standardizes data of different dimensions, allowing data with different characteristics to be compared on the same scale. The entire preprocessing process is automatically implemented through a series of algorithm modules in the data processing software. The software sequentially performs filtering, smoothing, completion, and normalization operations according to preset rules, thereby generating high-quality signal data and providing a reliable foundation for subsequent comprehensive analysis.

[0198] The time-series characteristics of each preprocessed signal are analyzed to generate the signal consistency coefficient and fluctuation amplitude index for each signal.

[0199] A comprehensive analysis model is constructed using the signal consistency coefficient and fluctuation amplitude index of each generated signal to generate the fluctuation identification coefficient of each signal. The fluctuation identification coefficient of each generated signal is then compared with a pre-set fluctuation identification coefficient threshold. Based on the comparison results, the fluctuation signals caused by hydrogen sulfide gas infiltration are identified, and these signals are classified and labeled.

[0200] The pre-set fluctuation identification coefficient threshold is typically determined by analyzing historical and experimental data, and optimized using multi-dimensional statistical analysis. Specifically, the determination process begins with collecting historical signal data under different operating conditions and geological formations. This data includes signal characteristics under normal operating conditions and abnormal hydrogen sulfide permeation conditions. Next, machine learning algorithms in data processing software are used to classify and train these signals, generating fluctuation identification coefficient distribution models for different operating conditions. The software then uses these models for cluster analysis to find the boundary point between normal and abnormal fluctuations; this boundary point is the initially determined threshold. Further sensitivity analysis and simulation testing are then used to adjust and optimize the threshold, ensuring accurate differentiation of abnormal fluctuations caused by hydrogen sulfide permeation in practical applications. Finally, after multiple iterations and verifications, an optimal identification threshold is determined and used as a fixed parameter within the system for subsequent analysis and comparison.

[0201] In this embodiment, the logic for obtaining the signal consistency coefficient and fluctuation amplitude index of each signal is as follows:

[0202] Extract the temporal feature information of each preprocessed signal, including the amplitude of each signal at different times within a certain period and the corresponding time points, and then apply the function S to the time series. m Let S be an expression for time point (t), where t is a time interval [t1, t2]. m (t) represents the amplitude of the m-th signal at time t within a time period, where m = 1, 2, 3, ..., k, and k is a positive integer.

[0203] To extract the temporal characteristics of each preprocessed signal, including the amplitude of each signal at different times within a given period and the corresponding time points, the extraction can be performed as follows: First, after preprocessing the original signal, the system samples the signal and generates sampling points according to a preset time interval. Each sampling point records the signal's amplitude information and time point at that moment. Specifically, data processing software segments the data from these sampling points to obtain the signal amplitude S within a specific time range (e.g., from t1 to t2). m (t) and the corresponding time point t. Extraction methods typically involve traversing the sampled data table or querying relevant data from a database by time series to generate the signal amplitude sequence and time point sequence. This process can be automated and is suitable for batch processing of large-scale time series data, thus providing a foundation for subsequent analysis.

[0204] Calculate the rate of change of amplitude for each signal using the formula:

[0205]

[0206] In the above formula, V m (t) represents the amplitude change rate of the m-th signal.

[0207] The average rate of change of amplitude for each signal over a period of time is calculated using the following formula:

[0208]

[0209] In the above formula, Let be the average of the amplitude change rate of the m-th signal over a period of time.

[0210] The standard deviation of the rate of change of amplitude for each signal over a period of time is calculated using the following formula:

[0211]

[0212] In the above formula, Let be the standard deviation of the amplitude change rate of the m-th signal over a period of time.

[0213] The signal consistency coefficient for each signal is calculated using the following formula:

[0214]

[0215] In the above formula, CCON m Let be the signal consistency coefficient of the m-th signal.

[0216] The signal consistency coefficient CCON of the m-th signal m Used to measure the stability and consistency of the rate of change of signal amplitude over a period of time. When CCON m A larger CCON value indicates that the signal amplitude is relatively stable and the deviation is small. This usually means that the signal fluctuation is greatly affected by other factors (such as normal changes in formation pressure), and belongs to the background signal. m A small signal consistency coefficient indicates significant and inconsistent fluctuations in the signal amplitude. This inconsistency is more likely caused by sudden factors, such as abnormal fluctuations due to hydrogen sulfide gas infiltration. Therefore, a small signal consistency coefficient can serve as an important reference indicator for identifying fluctuating signals caused by hydrogen sulfide gas infiltration, helping to detect the risk of hydrogen sulfide gas infiltration in its early stages and take timely measures.

[0217] The sum of squared amplitudes of each signal over the time interval [t1, t2] is calculated using the following formula:

[0218]

[0219] In the above formula, EM m Let be the sum of squares of the amplitude of the m-th signal during the time interval [t1, t2].

[0220] The sum of squared amplitudes of each signal over the entire time domain is calculated using the following formula:

[0221]

[0222] In the above formula, ET m Let m be the sum of squares of the amplitude of the m-th signal over the entire time domain.

[0223] The fluctuation amplitude exponent for each signal is calculated using the following formula:

[0224]

[0225] In the above formula, AAM m Let f be the fluctuation amplitude index of the m-th signal.

[0226] The fluctuation amplitude index AAM of the m-th signal m Used to quantify the energy concentration of a signal within a specific time period. When AAM... mA larger value indicates that the signal energy is concentrated within a specific frequency range, which usually reflects significant fluctuations in the signal. These fluctuations are more likely to be caused by anomalous factors (such as hydrogen sulfide gas infiltration). Therefore, a higher fluctuation amplitude index can indicate potential anomalous fluctuation signals, distinguishing them from normal formation pressure changes. This can be achieved by evaluating AAM. m The size of the signal can effectively identify fluctuation signals caused by hydrogen sulfide gas infiltration, thereby issuing timely warnings during well control and preventing further accident risks.

[0227] In this embodiment, the signal consistency coefficient CCON of each generated signal is... m And volatility index AAM m A comprehensive analysis model is constructed, and the fluctuation identification coefficients (FICs) of each signal are generated through weighted summation. m And the fluctuation identification coefficients (FICs) of each generated signal. m Compared with the pre-set fluctuation identification coefficient threshold FIC yuzhi A comparison was performed, and the fluctuation signals caused by hydrogen sulfide gas permeation were identified based on the comparison results. These signals were then classified and labeled, and the specific analysis is as follows:

[0228] If FIC m <FIC yuzhi If the signal is a normal fluctuation signal, it is classified as a normal signal and marked, indicating that the fluctuation characteristics of the signal are consistent with the formation pressure changes under normal operating conditions and are not disturbed by abnormal factors such as hydrogen sulfide gas infiltration. Such signals usually reflect stable formation pressure or downhole fluid conditions, so no further intervention or processing is required. For the marking of such signals, the software can automatically classify them as "normal signals" and use low priority or not process them. The specific marking method can be to assign a status label to each signal, such as "Normal" or "Standard", and store this label in the data logging or signal processing system. After marking, these signals can be ignored or archived for centralized processing of higher priority abnormal signals.

[0229] If FIC m ≥FIC yuzhiIf the signal is a fluctuation caused by hydrogen sulfide gas infiltration, it is classified as an abnormal signal and marked. This means that the fluctuation characteristics of the signal are abnormal and may be caused by hydrogen sulfide gas infiltration. Such signals are often accompanied by irregular pressure fluctuations and abnormal amplitude changes, which require attention and corresponding measures. In this case, the signal can be automatically marked as an "abnormal signal" by software and given a higher priority so as to trigger an early warning or initiate well control operations. The specific marking method can be to assign the signal the "Anomalous" or "Alert" label and generate an alarm message to be sent to the monitoring system. In addition, these marked signals can be centrally stored in an anomaly database for further analysis and subsequent decision-making.

[0230] When constructing a comprehensive analysis model and generating fluctuation identification coefficients for each signal, the first step is to calculate the signal consistency coefficient (CCON) for each signal. m And volatility index AAM m As input parameters. Based on the preceding analysis, the volatility index AAM m It plays a more important role in identifying fluctuation signals caused by hydrogen sulfide gas permeation, therefore AAM should be assigned during weighted summation. m A larger weighting coefficient ω4, while the signal consistency coefficient CCON m The weighting coefficient ω3 is relatively small. The calculation formula for the comprehensive analysis model is FIC. m =ω_3*CCON m +ω4*AAM m The weighting coefficients ω3 and ω4 must satisfy ω3 + ω4 = 1, and ω4 > ω3. For example, ω3 = 0.3 and ω4 = 0.7 can typically be set to highlight the importance of the fluctuation amplitude index in anomaly signal identification. Through this weighted summation method, the model can more accurately calculate the fluctuation identification coefficient (FIC) for each signal. m This is used to further determine whether the signal is caused by the permeation of hydrogen sulfide gas, and to provide a basis for subsequent signal classification and labeling.

[0231] Based on the signal classification and labeling results within the risk zone, corresponding risk warnings and operational suggestions are generated.

[0232] In this embodiment, based on the signal classification and labeling results within the risk zone, corresponding risk warnings and operational suggestions are generated, specifically including the following steps:

[0233] For signals marked as normal, generate a risk-free warning message based on the result of classifying them as normal, and keep the current drilling operation parameters unchanged.

[0234] For signals marked as normal, the system can determine the signal classification result as "normal" based on preset logical conditions. Once the signal is confirmed as normal, the system will generate a no-risk warning message and maintain the current operating parameters. Specifically, this involves setting classification labels in the signal processing software, classifying "normal" signals as low-priority signals, and simultaneously displaying a "no anomaly" message on the operating interface. Upon receiving such signals, the system will not issue warnings or adjust operating parameters, but will maintain existing drilling speed, mud density, and other operating conditions to ensure continuous operation. This process can be automated without manual intervention.

[0235] For signals marked as anomalous, a high-risk warning message is generated based on the result of their classification as anomalous signals, and operational suggestions are generated based on the characteristics of the anomalous signals, including operational measures such as adjusting mud density, reducing drilling speed, or increasing the frequency of air venting.

[0236] For signals marked as abnormal, the system will generate risk warnings and specific operational suggestions based on the characteristics of the abnormal signal. Specifically, the system first classifies the signal as a "high-priority" signal and triggers the early warning module to generate a high-risk warning. When generating operational suggestions, the system will select appropriate response strategies based on the signal characteristics, such as automatically generating suggestions to adjust mud density, slow down drilling speed, or increase the frequency of air venting through built-in rules or algorithms. After the operational suggestions are generated, the system will display the specific operational steps on the monitoring interface and simultaneously send these suggestions to the control system for operators to refer to or directly execute.

[0237] Continuously collect and update downhole data, analyze data changes in real time, and dynamically adjust risk markers and operational recommendations.

[0238] The goal of "continuously collecting and updating downhole data, analyzing data changes in real time, and dynamically adjusting risk markers and operational recommendations" can be achieved through the following methods:

[0239] First, the system can collect key parameters such as downhole pressure, hydrogen sulfide gas concentration, and mud flow rate in real time using deployed downhole sensors, and continuously transmit this data to the surface data processing system. The surface system stores this data and dynamically updates the risk assessment model based on the latest downhole data. Specifically, the system uses cyclic sampling and continuous monitoring to continuously collect new data and batch process and analyze this data at regular intervals (e.g., every second or every minute). By setting data update trigger conditions, if a key parameter changes significantly, the system will automatically trigger real-time analysis and reassess the risk level under the current operating conditions. In this way, the system can adjust risk markers at any time based on the latest downhole data, ensuring timely detection of potential anomalies.

[0240] Building upon the dynamic adjustment of risk markers, the system also adjusts operational recommendations based on data analysis results. Specifically, the system continuously assesses the changing trends of the downhole environment, combining real-time updated risk markers and historical data to dynamically adjust previously generated operational recommendations. For example, if the system detects a gradual increase in hydrogen sulfide concentration with abnormal fluctuations, it may adjust previous operational recommendations, suggesting further reductions in drilling speed or increases in venting frequency. Through this continuous updating and adjustment, the system ensures that operational recommendations always match the current actual situation, thereby minimizing risks such as blowouts and well kicks. The aim is to make the decision-making process more adaptive and to respond to sudden changes in the downhole environment, ensuring the safe and efficient operation of the drilling process.

[0241] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0242] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0243] Example 2

[0244] In this embodiment, the intelligent data processing method for oil drilling includes the following steps:

[0245] S1. Obtain real-time drilling data, which includes at least the time period corresponding to the real-time concentration data of hydrogen sulfide gas and the real-time concentration data of hydrogen sulfide body.

[0246] Optionally, during the drilling process, downhole sensors can be deployed to acquire real-time drilling data, which can then be transmitted in real-time to a surface data processing system for storage and further analysis.

[0247] S2. Based on the real-time drilling data, confirm the first time period exceeding the preset first hydrogen sulfide gas concentration, and use the real-time drilling data obtained within the first time period as the first data.

[0248] S3. Confirm the first anomaly coefficient based on the first data. The first anomaly coefficient is used to characterize the degree of fluctuation of the first data.

[0249] S31. Based on the first data, confirm the permeability influence coefficient of downhole pressure fluctuation and the dissolution-diffusion index of downhole hydrogen sulfide gas diffusion.

[0250] The permeability influence coefficient and the dissolution-diffusion index can be determined based on the permeability influence coefficient model and the dissolution-diffusion index model, respectively. The permeability influence coefficient model may include:

[0251]

[0252] f bd =FFT(G),G={P1,P2,P3,P4,…,P g-1 P g}

[0253] In the above formula, PIC is the permeability influence coefficient; A bd The actual pressure P at the bottom of the well i The fluctuation range; f bd The actual pressure P at the bottom of the well i The dominant frequency of the fluctuation; FFT() is the Fast Fourier Transform algorithm that converts a time-domain signal into a frequency-domain signal; G is the actual pressure P at the bottom of the well. i A collection constructed according to a time series; P i Let P be the actual pressure at the bottom of the well. i The time series numbering, i = 1, 2, 3, ..., g, where g is a positive integer.

[0254] Dissolution-diffusion index models may include:

[0255]

[0256] In the above formula, DDI is the dissolution-diffusion index; V ks denoted as ρ, where ρ is the diffusion rate of hydrogen sulfide gas in the mud; D is the diffusion coefficient of hydrogen sulfide gas; θ is the viscosity of the mud; ΔCT is the concentration gradient of hydrogen sulfide gas at different depths; and X is the diffusion velocity of hydrogen sulfide gas in the mud. jLet C be the depth value at the j-th depth location. j Let be the concentration of hydrogen sulfide gas at the j-th depth position; j is the depth position number, j = 1, 2, 3, ..., h, where h is a positive integer.

[0257] S32. The first anomaly coefficient is obtained by weighted summation of the permeation influence coefficient and the dissolution diffusion index.

[0258] S4. When the first abnormal coefficient exceeds the preset first threshold, confirm the first fluctuation identification coefficient of the first signal acquired in the first time period. The first fluctuation coefficient is used to characterize the fluctuation degree of the first signal.

[0259] S41. Confirm the consistency coefficient and fluctuation amplitude coefficient of the first signal.

[0260] The consistency coefficient and volatility coefficient are determined based on the consistency coefficient model and volatility coefficient model, respectively. The consistency coefficient model may include:

[0261]

[0262] In the above formula, CCON m Let m be the signal consistency coefficient of the m-th signal; Let be the standard deviation of the amplitude change rate of the m-th signal over the time interval [t1, t2]. V represents the mean of the amplitude change rate of the m-th signal over the time interval [t1, t2]. m (t) represents the rate of change of the amplitude of the m-th signal; S m (t) represents the amplitude of the m-th signal at time t within the time interval [t1, t2], where t is the time point within the time interval [t1, t2], and m = 1, 2, 3, ..., k, where k is a positive integer.

[0263] The volatility coefficient model may include:

[0264]

[0265] In the above formula, AAM m ET is the fluctuation amplitude exponent of the m-th signal; m EM is the sum of squared amplitudes of the m-th signal over the entire time domain; m Let be the sum of squares of the amplitude of the m-th signal within the time interval [t1, t2], where t is the time point within the time interval [t1, t2], and m = 1, 2, 3, ..., k, where k is a positive integer.

[0266] S42. The consistency coefficient and the fluctuation amplitude coefficient are weighted and summed to obtain the first fluctuation identification coefficient.

[0267] S5. When the first fluctuation identification coefficient exceeds the preset second threshold, the first signal is taken as an abnormal pressure fluctuation signal caused by hydrogen sulfide gas infiltration.

[0268] Optionally, the intelligent data processing method for oil drilling may also include the following steps:

[0269] S6. Update the real-time drilling data. Based on the real-time drilling data, confirm the second time period exceeding the preset second hydrogen sulfide gas concentration, and use the real-time drilling data obtained within the second time period as the second data.

[0270] S7. Confirm the second anomaly coefficient based on the second data. The second anomaly coefficient is used to characterize the degree of fluctuation of the second data.

[0271] S8. When the second abnormal coefficient exceeds the preset third threshold, confirm the second fluctuation identification coefficient of the second signal acquired in the second time period. The second fluctuation coefficient is used to characterize the fluctuation degree of the second signal.

[0272] S9. When the second fluctuation identification coefficient exceeds the preset fourth threshold, the second signal is taken as an abnormal pressure fluctuation signal caused by hydrogen sulfide gas infiltration.

[0273] Optionally, when the first signal is confirmed to be an abnormal pressure fluctuation signal caused by hydrogen sulfide gas infiltration, a high-risk warning can be issued, and corresponding operational suggestions can be provided.

[0274] More specifically, the first signal can be categorized or labeled to generate corresponding risk warnings and operational suggestions.

[0275] Optionally, when the first signal is confirmed to be a normal signal, a risk-free warning message can be generated, and the current drilling operation parameters can be kept unchanged.

[0276] Example 3

[0277] This embodiment provides an intelligent data processing device for oil drilling. The intelligent data processing device for oil drilling includes a real-time drilling data acquisition module, a first data determination module, a first anomaly coefficient confirmation module, a first fluctuation identification coefficient determination module, and an abnormal pressure fluctuation signal confirmation module, which are connected in sequence.

[0278] The drilling real-time data acquisition module is configured to acquire drilling real-time data, which includes at least the time period corresponding to the real-time concentration data of hydrogen sulfide gas and the real-time concentration data of hydrogen sulfide body.

[0279] The first data determination module is configured to identify a first time period exceeding a preset first hydrogen sulfide gas concentration based on real-time drilling data, and to use the real-time drilling data acquired within the first time period as the first data.

[0280] The first anomaly coefficient confirmation module is configured to confirm the first anomaly coefficient based on the first data. The first anomaly coefficient is used to characterize the degree of fluctuation of the first data.

[0281] The first fluctuation identification coefficient module is configured to confirm the first fluctuation identification coefficient of the first signal acquired in the first time period when the first abnormal coefficient exceeds the preset first threshold. The first fluctuation coefficient is used to characterize the fluctuation degree of the first signal.

[0282] The abnormal pressure fluctuation signal confirmation module is configured to recognize the first signal as an abnormal pressure fluctuation signal caused by hydrogen sulfide gas infiltration when the first fluctuation recognition coefficient exceeds the preset second threshold.

[0283] Example 4

[0284] This embodiment provides a computer device. The computer device includes a processor and a memory. The memory stores computer programs. The computer programs are executed by the processor, causing the processor to execute the computer program of the intelligent data processing method for oil drilling according to the present invention.

[0285] Example 5

[0286] This embodiment provides a computer-readable storage medium storing a computer program. The computer-readable storage medium stores a computer program that, when executed by a processor, causes the processor to execute the intelligent data processing method for oil drilling according to the present invention. The computer-readable recording medium is any data storage device capable of storing data read by a computer system. Examples of computer-readable recording media include: read-only memory, random access memory, read-only optical disc, magnetic tape, floppy disk, optical data storage device, and carrier waves (such as data transmission via the Internet through wired or wireless transmission paths).

[0287] It should be understood that, in the various embodiments of the present invention, the order of the above-mentioned process numbers does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0288] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0289] In the several embodiments provided in this application, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and 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 an indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0290] Although the present invention has been described above in conjunction with exemplary embodiments and accompanying drawings, those skilled in the art should understand that various modifications can be made to the above embodiments without departing from the spirit and scope of the claims.

Claims

1. A method for intelligent data processing in oil drilling, characterized in that, The intelligent data processing method for oil drilling includes the following steps: During the drilling process, downhole sensors are deployed to collect downhole data in real time, and the collected data is transmitted to the surface data processing system for storage and analysis in real time. The real-time data transmitted to the ground data processing system is analyzed to identify time periods passing through strata containing high concentrations of hydrogen sulfide gas, and the data segments within those time periods are marked as risk zones. Analyze downhole data in areas marked as risk zones to predict whether hydrogen sulfide gas infiltration will cause abnormal pressure fluctuations; Given the prediction that hydrogen sulfide gas infiltration would cause abnormal pressure fluctuations, all signals generated in the risk zone were analyzed to identify the fluctuation signals caused by hydrogen sulfide gas infiltration, and these signals were classified and labeled. Based on the signal classification and labeling results within the risk zone, corresponding risk warnings and operational suggestions are generated; Continuously collect and update downhole data, analyze data changes in real time, and dynamically adjust risk markers and operational recommendations.

2. The intelligent data processing method for oil drilling according to claim 1, characterized in that, The process of analyzing real-time data transmitted to the ground data processing system, identifying time periods passing through strata containing high concentrations of hydrogen sulfide gas, and marking data segments within those time periods as risk zones specifically includes the following steps: Preprocess the downhole data transmitted in real time to the ground data processing system; Extract hydrogen sulfide gas concentration data from the preprocessed downhole data, and record the actual hydrogen sulfide gas concentration at each time point and the corresponding time period. Determine the pre-set hydrogen sulfide gas concentration threshold; The actual concentration of hydrogen sulfide gas at all times is compared with a pre-set hydrogen sulfide gas concentration threshold. The time period corresponding to the actual concentration of hydrogen sulfide gas that is greater than the pre-set hydrogen sulfide gas concentration threshold is marked as a risk zone.

3. The intelligent data processing method for oil drilling according to claim 1, characterized in that, The analysis of downhole data marked as risk zones to predict whether hydrogen sulfide gas infiltration will cause abnormal pressure fluctuations specifically includes the following steps: Extract pressure fluctuation information and hydrogen sulfide gas diffusion information from downhole data marked as risk zones; The extracted pressure fluctuation information and hydrogen sulfide gas diffusion information were analyzed to generate the permeation influence coefficient and the dissolution diffusion index, respectively. An anomaly prediction model is constructed using the generated permeation influence coefficient and dissolution diffusion index to generate anomaly coefficients. These anomaly coefficients are then compared with a pre-set anomaly coefficient threshold. Based on the comparison results, it is predicted whether hydrogen sulfide gas permeation will cause abnormal pressure fluctuations.

4. The intelligent data processing method for oil drilling according to claim 3, characterized in that, The logic for obtaining the permeation influence coefficient and the dissolution-diffusion index is as follows: Extract pressure fluctuation information from downhole data marked as risk zones, including the actual pressure at the bottom of the well at different times over a period of time, and calibrate the actual pressure at the bottom of the well at different times over a period of time as P. i , i represents the number of the actual pressure at the bottom of the well at different times within a certain period of time, i = 1, 2, 3, ..., g, where g is a positive integer; The actual pressure P at the bottom of the well at different times over a period of time i Construct a set G based on the time series, denoted as G = {P1, P2, P3, P4, ..., P...} g-1 P g }; The Fast Fourier Transform (FFT) is applied to perform frequency domain analysis on the actual pressure at the bottom of the well within a time period in set G, calculating the dominant frequency of the bottom-hole pressure fluctuation. The dominant frequency represents the frequency of the bottom-hole pressure change per unit time, and the specific formula is as follows: f bd =FFT(G)=FFT({P1,P2,P3,P4,…,P g-1 ,P g }) In the above formula, f bd The dominant frequency of the bottom hole pressure fluctuation is denoted as FFT, which is a fast Fourier transform algorithm used to convert time-domain signals into frequency-domain signals. The specific formula for calculating the fluctuation range of the actual pressure at the bottom of the well at different times over a period of time is as follows: In the above formula, A bd This represents the fluctuation range of the actual pressure at the bottom of the well at different times over a period of time. The specific formula for calculating the permeability influence coefficient is as follows: PIC=f bd *A bd In the above formula, PIC is the permeability influence coefficient; Extract hydrogen sulfide gas diffusion information from downhole data marked as risk zones, including hydrogen sulfide gas concentrations at different depths and corresponding depth values. Label the hydrogen sulfide gas concentrations and corresponding depth values ​​at different depths as C. j and X j j represents the hydrogen sulfide gas concentration at different depths and the corresponding depth value number, j = 1, 2, 3, ..., h, where h is a positive integer; Calculate the hydrogen sulfide gas concentration gradient at different depths. The concentration gradient represents the rate at which hydrogen sulfide gas changes with depth. The specific formula is as follows: In the above formula, ΔCT represents the hydrogen sulfide gas concentration gradient at different depths, and X... j C represents the depth value at the j-th depth position. j This represents the hydrogen sulfide gas concentration at the j-th depth location; Obtain the diffusion coefficient of hydrogen sulfide gas and the viscosity of the mud. Define the diffusion coefficient of hydrogen sulfide gas and the viscosity of the mud as D and θ, respectively. Calculate the diffusion rate of hydrogen sulfide gas in the mud. The specific calculation formula is as follows: In the above formula, V ks This represents the diffusion rate of hydrogen sulfide gas in the mud. The specific formula for calculating the dissolution-diffusion index is as follows: In the above formula, DDI is the dissolution-diffusion index.

5. The intelligent data processing method for oil drilling according to claim 4, characterized in that, An anomaly prediction model is constructed using the generated permeability influence coefficient (PIC) and dissolution-diffusion index (DDI). An anomaly coefficient (YC) is generated through weighted summation, and this anomaly coefficient (YC) is compared with a pre-set anomaly coefficient threshold (YC). yuzhi A comparison was conducted, and based on the comparison results, it was predicted whether hydrogen sulfide gas infiltration would cause abnormal pressure fluctuations. The specific analysis is as follows: If YC≤YC yuzhi It is predicted that hydrogen sulfide gas infiltration will not cause abnormal pressure fluctuations; If YC>YC yuzhi It is predicted that the infiltration of hydrogen sulfide gas will cause abnormal pressure fluctuations.

6. The intelligent data processing method for oil drilling according to claim 1, characterized in that, Given the prediction that hydrogen sulfide gas infiltration will cause abnormal pressure fluctuations, all signals generated within the risk zone are analyzed to identify the fluctuation signals caused by hydrogen sulfide gas infiltration. These signals are then classified and labeled, specifically including the following steps: Given the prediction that hydrogen sulfide gas infiltration would cause abnormal pressure fluctuations, all signal data marked as risk zones were extracted, including the time-series characteristics of each signal. The temporal feature information of each extracted signal is preprocessed; The time-series characteristics of each preprocessed signal are analyzed to generate the signal consistency coefficient and fluctuation amplitude index for each signal. A comprehensive analysis model is constructed using the signal consistency coefficient and fluctuation amplitude index of each generated signal to generate the fluctuation identification coefficient of each signal. The fluctuation identification coefficient of each generated signal is then compared with a pre-set fluctuation identification coefficient threshold. Based on the comparison results, the fluctuation signals caused by hydrogen sulfide gas infiltration are identified, and these signals are classified and labeled.

7. The intelligent data processing method for oil drilling according to claim 6, characterized in that, The logic for obtaining the signal consistency coefficient and fluctuation amplitude index of each signal is as follows: Extract the temporal feature information of each preprocessed signal, including the amplitude of each signal at different times within a certain period and the corresponding time points, and then apply the function S to the time series. m Let S be an expression for time point (t), where t is a time interval [t1, t2]. m (t) represents the amplitude of the m-th signal at time t within a time period, where m = 1, 2, 3, ..., k, and k is a positive integer; Calculate the rate of change of amplitude for each signal using the formula: In the above formula, V m (t) represents the rate of change of the amplitude of the m-th signal; The average rate of change of amplitude for each signal over a period of time is calculated using the following formula: In the above formula, Let be the average of the rate of change of the amplitude of the m-th signal over a period of time; The standard deviation of the rate of change of amplitude for each signal over a period of time is calculated using the following formula: In the above formula, Let be the standard deviation of the rate of change of the amplitude of the m-th signal over a period of time; The signal consistency coefficient for each signal is calculated using the following formula: In the above formula, CCON m Let m be the signal consistency coefficient of the m-th signal; The sum of squared amplitudes of each signal over the time interval [t1, t2] is calculated using the following formula: In the above formula, EM m Let m be the sum of squares of the amplitude of the m-th signal over the time interval [t1, t2]. The sum of squared amplitudes of each signal over the entire time domain is calculated using the following formula: In the above formula, ET m Let m be the sum of squares of the amplitudes of the m-th signal over the entire time domain; The fluctuation amplitude exponent for each signal is calculated using the following formula: In the above formula, AAM m Let f be the fluctuation amplitude index of the m-th signal.

8. The intelligent data processing method for oil drilling according to claim 7, characterized in that, The signal consistency coefficient CCON of each generated signal m And volatility index AAM m A comprehensive analysis model is constructed, and the fluctuation identification coefficients (FICs) of each signal are generated through weighted summation. m And the fluctuation identification coefficients (FICs) of each generated signal. m Compared with the pre-set fluctuation identification coefficient threshold FIC yuzhi A comparison was performed, and the fluctuation signals caused by hydrogen sulfide gas permeation were identified based on the comparison results. These signals were then classified and labeled, and the specific analysis is as follows: If FIC m <FIC yuzhi If the signal is a normal fluctuation signal, then the signal is classified as a normal signal and marked. If FIC m ≥FIC yuzhi If the signal is a fluctuation caused by the permeation of hydrogen sulfide gas, then the signal is classified as an abnormal signal and marked.

9. The intelligent data processing method for oil drilling according to claim 1, characterized in that, Based on the signal classification and labeling results within the risk zone, corresponding risk warnings and operational suggestions are generated, specifically including the following steps: For signals marked as normal, generate a risk-free warning message based on the result of their classification as normal signals, and keep the current drilling operation parameters unchanged; For signals marked as anomalous, generate high-risk warning information based on the result of their classification as anomalous signals, and generate operation suggestions based on the characteristics of anomalous signals.

10. A method for intelligent data processing in oil drilling, characterized in that, The intelligent data processing method for oil drilling includes the following steps: Acquire real-time drilling data, which includes at least real-time hydrogen sulfide gas concentration data and the time period corresponding to the real-time hydrogen sulfide gas concentration data; Based on the real-time drilling data, if a first time period exceeding a preset first hydrogen sulfide gas concentration is confirmed, the real-time drilling data obtained within the first time period will be used as the first data. A first anomaly coefficient is determined based on the first data, and the first anomaly coefficient is used to characterize the degree of fluctuation of the first data. When the first anomaly coefficient exceeds a preset first threshold, the first fluctuation identification coefficient of the first signal acquired within the first time period is confirmed, wherein the first fluctuation coefficient is used to characterize the degree of fluctuation of the first signal; and When the first fluctuation identification coefficient exceeds the preset second threshold, the first signal is regarded as an abnormal pressure fluctuation signal caused by hydrogen sulfide gas infiltration.

Images