Well drilling overflow early warning method and device

Abstract

The invention discloses a well drilling overflow early warning method and device. The method comprises the steps that a wavelet analysis and diagnosis model and an overflow diagnosis model corresponding to different well drilling working conditions are established in advance through historical logging data of multiple wells; acquiring drilling data in real time; determining a current drilling working condition according to the drilling data; determining a first overflow probability according to the drilling data and a wavelet analysis and diagnosis model; extracting overflow characteristic parameters from the drilling data, inputting the overflow characteristic parameters into an overflow diagnosis model corresponding to the current drilling working condition, and obtaining a second overflow probability according to the output of the overflow diagnosis model; and calculating an alarm evaluation index according to the first overflow probability and the second overflow probability. By means of the scheme, intelligent monitoring and early warning of well drilling overflow can be achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of oil and gas exploration and development, and in particular to a drilling overflow early warning method and device. Background Art

[0002] With the continuous exploitation of crude oil in my country, the field of oil and gas exploration and development is shifting from conventional oil and gas resources to deep and ultra-deep layers, facing challenges such as resource degradation and exploration diversification. The geological conditions in some oil and gas resource areas are complex, which is prone to gas invasion and overflow risk events. When overflow occurs, if it is not discovered in time or handled too late, it is likely to cause a blowout accident. The risk of overflow gas invasion will lead to problems such as threats to drilling safety and economic losses of personnel at the well site. Therefore, early identification of bottom hole pressure imbalance, discovery of overflow and proper handling are the difficulties in well control technology research.

[0003] Usually, the industry mostly uses pool volume and relative flow sensors to monitor overflows, and adopts a simple differential warning method. However, the measurement accuracy of the pool volume and relative flow sensors themselves is limited, so it is difficult to detect trace overflows in time, and there are also problems such as missed reports, late reports, and false reports. In recent years, the industry has also begun to study early overflow monitoring technology, and some monitoring equipment has been launched, such as the BH-EKLD early overflow monitor and the SK-EKD early well kick monitoring system. Although these technologies have begun to be initially tried on site, they are still in the exploratory stage and have not yet formed complete productivity. Summary of the invention

[0004] The present invention provides a drilling overflow early warning method and device to realize intelligent monitoring and early warning of drilling overflow.

[0005] To this end, the present invention provides the following technical solutions:

[0006] A drilling overflow early warning method, the method comprising:

[0007] The historical logging data of multiple wells are used to establish wavelet analysis diagnosis models and overflow diagnosis models corresponding to different drilling conditions;

[0008] Real-time collection of drilling data;

[0009] determining a current drilling condition according to the drilling data;

[0010] Determine a first overflow probability according to the drilling data and the wavelet analysis diagnostic model;

[0011] Extracting overflow characteristic parameters from the drilling data, inputting the overflow characteristic parameters into an overflow diagnosis model corresponding to the current drilling condition, and obtaining a second overflow probability according to an output of the overflow diagnosis model;

[0012] Calculate an alarm evaluation index based on the first overflow probability and the second overflow probability.

[0013] Optionally, the historical logging data includes: rotary table speed, casing pressure, riser pressure, bit position, bit speed, bit torque, and well depth.

[0014] Optionally, the pre-training of overflow diagnosis models for different drilling conditions using historical logging data from multiple wells includes:

[0015] Collect historical logging data from multiple wells;

[0016] Preprocess the historical logging data;

[0017] Use a drilling condition identification model to determine the drilling conditions corresponding to each historical logging data, and generate a training data set for each drilling condition;

[0018] Establish a wavelet analysis diagnosis model based on drilling pressure;

[0019] Use the training data set to train overflow diagnosis models for each drilling condition.

[0020] Optionally, the preprocessing of the historical logging data includes any one or more of the following: removing outliers, smoothing filtering, quantitative correlation analysis, non-parametric correlation test, data dimensionality reduction processing, and dimensionless processing.

[0021] Optionally, the establishment of the wavelet analysis diagnosis model based on drilling pressure includes:

[0022] Obtain the wellbore pressure wave response signal;

[0023] Decompose the pressure wave response signal using wavelet packet to obtain each decomposition curve;

[0024] Statistically analyze the ranges of the mean and variance of the decomposition curves in the overflow and non-overflow states;

[0025] Calculate the proportion of overflow data in different ranges;

[0026] Determine the wavelet analysis diagnosis model based on drilling pressure according to the calculation results.

[0027] Optionally, the calculation of the alarm evaluation index based on the first overflow probability and the second overflow probability includes:

[0028] Perform weighted calculation on the first overflow probability and the second overflow probability to obtain the alarm evaluation index.

[0029] Optionally, the method further includes: when the alarm evaluation index exceeds a set threshold, an overflow alarm is issued.

[0030] Optionally, the method further includes:

[0031] Presetting a plurality of alarm levels, where the alarm evaluation index ranges corresponding to different alarm levels are different;

[0032] Determining the current alarm level according to the alarm evaluation index;

[0033] Performing an overflow alarm according to the current alarm level.

[0034] A drilling overflow early warning device, the device includes:

[0035] A data acquisition module, configured to collect drilling data in real time;

[0036] A working condition determination module, configured to determine the current working condition according to the drilling data;

[0037] A first analysis module, configured to determine a first overflow probability according to the drilling data and a pre-established wavelet analysis diagnosis model;

[0038] A second analysis module, configured to extract overflow characteristic parameters from the drilling data, input the overflow characteristic parameters into an overflow diagnosis model corresponding to the current working condition, and obtain a second overflow probability according to the output of the overflow diagnosis model;

[0039] A calculation module, configured to calculate an alarm evaluation index according to the first overflow probability and the second overflow probability.

[0040] Optionally, the device further includes:

[0041] An alarm module, configured to perform an alarm according to the alarm evaluation index.

[0042] For the drilling overflow early warning method and device provided by the present invention, on the one hand, a wavelet analysis diagnosis model based on drilling pressure is used, and the wavelet analysis pressure signal is used to constrain the result to realize the determination of whether there is an overflow; on the other hand, an overflow early warning model established based on historical logging data is used to perform real-time diagnosis based on on-site real-time data. The alarm evaluation index is calculated based on the overflow probabilities obtained by combining the two methods, so as to realize the accurate judgment of early drilling overflow.

[0043] Furthermore, an alarm is performed according to the alarm evaluation index, and an overflow signal is timely provided for subsequent overflow kill well control, so that the problem can be processed as soon as possible, avoiding blowout accidents and ensuring the safety of wellsite personnel. Description of the Drawings

[0044] Figure 1 It is a flowchart of establishing a wavelet analysis diagnosis model and an overflow diagnosis model in the method of the present invention;

[0045] Figure 2 is a flowchart of a drilling overflow warning method provided by the present invention;

[0046] Figure 3 is a schematic structural diagram of a drilling overflow warning device provided by the present invention. Detailed implementation manners

[0047] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0048] The present invention will be described in detail below in conjunction with the drawings and specific implementation manners. The implementation manners cannot be described in detail one by one here, but the implementation manners of the present invention are not limited to the following implementation manners.

[0049] The present invention provides a drilling overflow warning method and device, and respectively establishes an overflow diagnosis model of a wavelet analysis diagnosis model corresponding to different drilling conditions in advance by using the historical logging data of multiple wells. During the drilling process, drilling data is collected in real time, and judgments are respectively made based on the drilling data by using the wavelet analysis diagnosis model and the overflow diagnosis model, and then an alarm evaluation index is calculated by synthesizing the judgment results obtained by the two models.

[0050] The following first describes the process of establishing the wavelet analysis diagnosis model and the overflow diagnosis model in the embodiments of the present invention.

[0051] As Figure 1 shown, it is a flowchart of establishing the wavelet analysis diagnosis model and the overflow diagnosis model in the method of the present invention, including the following steps:

[0052] Step 101, collect the historical logging data of multiple wells.

[0053] The historical logging data includes but is not limited to: rotary table speed, casing pressure, riser pressure, bit position, bit speed, bit torque, well depth, etc.

[0054] Step 102, preprocess the historical logging data.

[0055] For the collected historical logging data, the historical logging data can be analyzed and saved according to the WITS (Wireless Information Transmission System) protocol.

[0056] Furthermore, the historical logging data can be classified and visually displayed. For example, the logging data can be divided into five groups: drilling parameters, circulation parameters, inlet and outlet parameters, gas logging parameters, and depth parameters, which facilitates a comprehensive analysis of the drilling situation in the follow-up.

[0057] The preprocessing of the historical logging data includes but is not limited to any one or more of the following: outlier removal, smoothing filtering, quantitative correlation analysis, non-parametric correlation test, data dimensionality reduction processing, dimensionless processing, etc.

[0058] The following briefly describes the above various preprocessing methods.

[0059] 1) Outlier processing of data: The outliers are comprehensively judged through the mechanism and data fusion method to further improve the data validity. Among them, the mechanism method refers to the comparative analysis of the average mechanical drilling rate calculated by the well depth change and the mechanical drilling rate of the actual logging parameters, and a 5% error threshold is set. If it exceeds the threshold, the data is regarded as abnormal and deleted; the data method mainly uses the data statistics theory and uses methods such as box plots to count the distance of each point from the median and 1.5 times the interquartile range, and deletes the outlier data.

[0060] 2) Data smoothing filtering processing: A fixed window length is set, polynomial fitting is completed for the data within the window, and the data points are restored. While retaining the data change trend, the data noise is reduced and the data quality is improved.

[0061] 3) Quantitative correlation analysis: An analysis and evaluation table of variance threshold and correlation coefficient threshold is set, as shown in Table 1. Based on the characteristic variance and correlation coefficient values and the analysis and evaluation table, the features are initially screened.

[0062] Table 1

[0063]

[0064] 4) Non-parametric correlation test: Based on the chi-square test and mutual information test methods, the correlation between the above parameters and the overflow is determined. For example, based on the overflow characterization parameter table shown in Table 2, the optimal parameters are selected.

[0065] Table 2

[0066]

[0067]

[0068] It should be noted that the selection of the optimal parameters can be based on the parameters in the overflow characterization parameters shown in Table 2 above and determined according to the quantitative relationship (see Table 1). (Strong correlation leads to duplicate data, so it is not applicable; small variance represents less information content of the parameter, so it is not used)

[0069] For example, for the drilling time and drilling speed, both are shown in Table 2 and they have a strong correlation or connection. In such a case, generally either one can be deleted because strong correlation may lead to duplicate data.

[0070] Another example: for the inlet flow rate parameter, it is not in Table 2, and this parameter has no correlation with other parameters, but it can form a flow difference with the outlet flow rate, so it will be retained.

[0071] 4) Data dimensionality reduction processing and analysis: Use the principal component analysis method to reduce the data dimension to improve the model calculation efficiency.

[0072] 5) Dimensionless processing: Set the minimum-maximum value method and the standardization method to convert all feature dimensions into the same order of magnitude to reduce the error caused by data units.

[0073] Based on the preprocessed data above, build a basic overflow database to provide a data basis for building the model.

[0074] Step 103, use the drilling condition identification model to determine the drilling conditions corresponding to each historical logging data, and generate a training data set corresponding to each drilling condition.

[0075] Specifically, a drilling condition identification model can be established based on drilling parameter judgment and logical reasoning.

[0076] Considering the influence of on-site noise interference, the hook position, hook load, bit position, and well depth change trend can be calculated based on the least squares method, and the parameter change situation can be judged according to the slope and the set threshold. For example, set the drilling parameter trend judgment criteria as shown in Table 3.

[0077] Table 3

[0078]

[0079] Through 9 parameters including real-time logging data such as well depth, hook position, hook load, rotation speed, torque, pump pressure, bit position, and the processed bit position change trend and well depth change trend, combined with expert experience, build a logical rule inference model to judge 8 drilling conditions including drilling, tripping in and out, etc. The specific rules are the automatic drilling condition identification rules shown in Table 4.

[0080] Table 4

[0081]

[0082] In the above Table 4, ↑ represents an increasing trend, ↓ represents a decreasing trend, - represents a constant value, ↑↓ represents a trend fluctuation; > represents a value greater than, < represents a value less than.

[0083] The above rules can be programmed using Python to enable real-time determination of the drilling conditions by inputting the above key parameters.

[0084] Step 104: Establish a wavelet analysis diagnosis model based on drilling pressure.

[0085] In the embodiment of the present invention, based on wavelet packet analysis, db6 can be preferably selected as the mother wavelet function; the wellbore pressure wave response frequency can be analyzed according to the water hammer effect. For example, the pressure signal with a frequency below 1 Hz at a well depth of more than 3000 m can be selected as the analysis basis; the pressure (standpipe pressure / casing pressure) signal of the initial 1 Hz is decomposed by wavelet packet, and the signal below 0.5 Hz is extracted, and early warning of overflow is carried out by setting thresholds for the mean and variance of the extracted signal. Among them, the wellbore pressure wave specifically refers to the additional wellbore pressure wave caused by overflow, rather than the wellbore pressure during circulating drilling fluid.

[0086] Specifically, the wavelet analysis diagnosis model can be established in the following manner:

[0087] First, obtain the wellbore pressure wave (standpipe pressure or casing pressure) response signal; then decompose the pressure wave response signal by wavelet packet to obtain each decomposition curve; statistically analyze the ranges of the mean and variance of the decomposition curves in the overflow and non-overflow states; calculate the proportion of overflow data in different ranges; establish a wavelet analysis diagnosis model based on drilling pressure according to the calculation results.

[0088] The wavelet analysis diagnosis model corresponds to the overflow risk index for different ranges of mean and variance.

[0089] For example, a wavelet analysis diagnosis model as shown in Table 5 can be established. Among them, the greater the overflow risk index, the higher the possibility of overflow.

[0090] Table 5

[0091]

[0092] It should be noted that the risk index values corresponding to different means and variances in Table 5 above are only exemplary descriptions, and can be corrected according to the actual well site conditions in specific applications.

[0093] Step 105: Use the training data set to train an overflow diagnosis model corresponding to each drilling condition.

[0094] The overflow diagnosis model can adopt a neural network model. The model input characteristic parameters (also referred to as overflow characteristic parameters in the following description) under different working conditions of overflow can be as shown in Table 6 below. The output of the model is the overflow probability.

[0095] Table 6

[0096]

[0097] The training method of the overflow diagnosis model can adopt the corresponding conventional training method according to the model structure, and the embodiments of the present invention do not make any limitations in this regard.

[0098] Based on the above-mentioned wavelet analysis diagnosis model and overflow diagnosis model, the drilling overflow warning method provided by the present invention is as Figure 2 shown.

[0099] Referring to Figure 2 , which is a flowchart of a drilling overflow warning method provided by an embodiment of the present invention, including the following steps:

[0100] Step 201, collect drilling data in real time.

[0101] The drilling data may include, but is not limited to: rotary table speed, casing pressure, riser pressure, bit position, bit speed, bit torque, well depth and other data.

[0102] Furthermore, the drilling data can be preprocessed. For example, the preprocessing process is as follows: comprehensively judge outliers through the mechanism and data fusion method, delete mechanism outliers exceeding 5% of the error by using the mechanical drilling rate parameter comparison, use the box plot method to statistically calculate the distance of each point from the median and 1.5 times the interquartile range, delete outlier data, and collect the normal value range and interquartile range; use data smoothing filtering, preferably with a fixed window length of 15 and a polynomial fitting level of 3; through quantitative correlation analysis and non-parametric correlation test.

[0103] Furthermore, the principal component analysis method can be used to reduce the data dimension; the standardization method can be used to normalize the data and collect the normalization parameters.

[0104] Step 202, determine the current drilling working condition according to the drilling data.

[0105] Specifically, the drilling working condition identification model described above can be used to determine the current drilling working condition.

[0106] For example, calculate the change trends of the hook position, hook load, bit position and well depth by the least squares method, and input the original values and trends of the four variables, bit weight on bit, rotary table speed, rotary table torque and pump pressure into the drilling working condition identification model to judge the drilling working condition.

[0107] It should be noted that the preprocessing process of the drilling data described above can also be carried out after the above step 202, and the present invention does not make any limitations in this regard.

[0108] Step 203, determine the first overflow probability R1 according to the drilling data and the wavelet analysis diagnosis model, and its value range is 0-1.

[0109] Specifically, the original standpipe pressure / casing pressure parameters can be input into the wavelet analysis diagnosis model to obtain the first overflow probability R1.

[0110] Step 204: Extract the overflow characteristic parameters from the drilling data, input the overflow characteristic parameters into the overflow diagnosis model corresponding to the current drilling condition, and obtain the second overflow probability R2 according to the output of the overflow diagnosis model, where the value range of R2 is 0 to 1.

[0111] The overflow characteristic parameters can be as shown in Table 6, that is, the input characteristic parameters in Table 6. For different drilling conditions, there can be different overflow characteristic parameters.

[0112] Step 205: Calculate the alarm evaluation index R according to the first overflow probability R1 and the second overflow probability R2.

[0113] Specifically, the first overflow probability and the second overflow probability can be weighted and calculated to obtain the alarm evaluation index, as shown in the following formula (1):

[0114] R = W1R1 + W2R2 (1)

[0115] In the formula, R1 and R2 are the first overflow probability and the second overflow probability respectively, W1 and W2 are the weights corresponding to R1 and R2. For example, W1 is defaulted to 0.3 and W2 is defaulted to 0.7; R is the alarm evaluation index, and its range is 0 to 1.

[0116] It should be noted that the above weights W1 and W2 can be adaptively adjusted according to the changes in the drilling block, and the embodiments of the present invention do not make limitations thereto.

[0117] Further, in another non - restrictive embodiment of the method of the present invention, the following steps may further be included: When the alarm evaluation index exceeds the set threshold, an overflow alarm is issued.

[0118] Further, in another non - restrictive embodiment of the method of the present invention, the following steps may further be included: Predetermine multiple alarm levels, and the alarm evaluation index ranges corresponding to different alarm levels are different; Determine the current alarm level according to the alarm evaluation index; Perform an overflow alarm according to the current alarm level.

[0119] For example, when R > 0.4, remind to observe the change of parameters; when R > 0.6, issue a minor alarm; when R > 0.8, officially issue an overflow alarm.

[0120] The well drilling overflow warning method provided by the present invention, on the one hand, uses a wavelet analysis diagnosis model based on well drilling pressure, and uses wavelet analysis of pressure signals to constrain the results to determine whether an overflow occurs; on the other hand, it uses an overflow warning model established based on historical logging data to perform real-time diagnosis based on on-site real-time data. The overflow probability calculation alarm evaluation index obtained by combining the two methods is used to accurately judge the early overflow of the well drilling. Further, an alarm is issued according to the alarm evaluation index, and an overflow signal is timely provided for subsequent overflow well control of the well, so that the problem can be processed as soon as possible, avoiding blowout accidents and ensuring the safety of well site personnel.

[0121] Taking the data of a certain well site as an example, the results show that there are a total of 8.78 million data. The average correct rate of the final model is 89.52%, the false alarm rate is 8.42%, and the F1 index is 0.69. It can be seen that using the solution of the present invention can effectively improve the accuracy of overflow warning.

[0122] Correspondingly, the present invention also provides a well drilling overflow warning device, as Figure 3 shown, which is a schematic structural diagram of the device.

[0123] The well drilling overflow warning device includes the following modules:

[0124] A data acquisition module 301 for real-time acquisition of well drilling data;

[0125] A working condition determination module 302 for determining the current working condition according to the well drilling data;

[0126] A first analysis module 303 for determining a first overflow probability according to the well drilling data and a pre-established wavelet analysis diagnosis model;

[0127] A second analysis module 304 for extracting overflow characteristic parameters from the well drilling data, inputting the overflow characteristic parameters into an overflow diagnosis model corresponding to the current working condition, and obtaining a second overflow probability according to the output of the overflow diagnosis model;

[0128] A calculation module 305 for calculating an alarm evaluation index according to the first overflow probability and the second overflow probability.

[0129] The above-mentioned wavelet analysis diagnosis model and overflow diagnosis model can be pre-established by a corresponding model construction module using the historical logging data and overflow conditions of multiple wells. The specific establishment process of the model can refer to the description in the method embodiment of the present invention above and will not be elaborated here.

[0130] It should be noted that the model construction module can be a part of the device or independent of the device, and the embodiments of the present invention do not make any limitations in this regard.

[0131] Further, in another embodiment of the device of the present invention, an alarm module (not shown in the figure) may also be included for giving an alarm according to the alarm evaluation index.

[0132] Specifically, the alarm module may give an overflow alarm when the alarm evaluation index exceeds a set threshold; or give an alarm according to a plurality of preset different alarm levels, that is, determine the current alarm level according to the alarm evaluation index; and give an overflow alarm according to the current alarm level.

[0133] For the specific implementation manners of the various modules in the device of the present invention, reference may be made to the descriptions in the foregoing method embodiments of the present invention, and details are not described herein again.

[0134] Correspondingly, the present invention also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is run by a processor, it executes Figure 1 or Figure 2 some or all of the steps of the above process.

[0135] It should be noted that the terms "including" and "having" and any variations thereof in the description and claims of the present invention and the above-mentioned drawings are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0136] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments. Moreover, the system embodiments described above are merely illustrative. The modules and units described as separate components may or may not be physically separated, that is, they may be located on one network unit, or may be distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art can understand and implement it without creative work.

[0137] In specific implementation, for each module / unit included in the various devices and products described in the above embodiments, it may be a software module / unit, a hardware module / unit, or may be partly a software module / unit and partly a hardware module / unit.

[0138] For example, for each device or product applied to or integrated into a chip, each module / unit included therein may be implemented in the form of hardware such as a circuit. Alternatively, at least some of the modules / units may be implemented in the form of a software program that runs on a processor integrated inside the chip, and the remaining (if any) part of the modules / units may be implemented in the form of hardware such as a circuit. For each device or product applied to or integrated into a chip module, each module / unit included therein may be implemented in the form of hardware such as a circuit. Different modules / units may be located in the same component (such as a chip, a circuit module, etc.) or different components of the chip module. Alternatively, at least some of the modules / units may be implemented in the form of a software program that runs on a processor integrated inside the chip module, and the remaining (if any) part of the modules / units may be implemented in the form of hardware such as a circuit. For each device or product applied to or integrated into a terminal, each module / unit included therein may be implemented in the form of hardware such as a circuit. Different modules / units may be located in the same component (such as a chip, a circuit module, etc.) or different components inside the terminal. Alternatively, at least some of the modules / units may be implemented in the form of a software program that runs on a processor integrated inside the terminal, and the remaining (if any) part of the modules / units may be implemented in the form of hardware such as a circuit.

[0139] In addition, in each embodiment of the present application, each functional unit may be integrated into a processing unit, or each unit may be physically arranged separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware, or in the form of a combination of hardware and a software functional unit.

[0140] The above-mentioned integrated unit implemented in the form of a software functional unit may be stored in a computer-readable storage medium. The above-mentioned software functional unit stored in a storage medium includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute some steps of the methods described in each embodiment of the present application.

[0141] The above has introduced the embodiments of the present invention in detail. In this article, specific implementation manners are used to elaborate on the present invention. The descriptions of the above embodiments are only used to help understand the methods and systems of the present invention. They are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention. The content of this specification should not be construed as a limitation to the present invention. Therefore, any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for early warning of drilling overflow, characterized in that, The method includes: Pre - establishing a wavelet analysis diagnosis model and a kick diagnosis model corresponding to different drilling conditions respectively by using the historical logging data of multiple wells; Collecting drilling data in real - time; Determining the current drilling condition according to the drilling data; Determining the first kick probability according to the drilling data and the wavelet analysis diagnosis model; Extracting kick characteristic parameters from the drilling data, inputting the kick characteristic parameters into the kick diagnosis model corresponding to the current drilling condition, and obtaining the second kick probability according to the output of the kick diagnosis model; Calculating an alarm evaluation index according to the first kick probability and the second kick probability.

2. The drilling overflow warning method according to claim 1, wherein The historical logging data includes: rotary table speed, casing pressure, standpipe pressure, bit position, bit speed, bit torque, well depth.

3. The drilling overflow warning method according to claim 1, characterized in that The pre - training of the kick diagnosis model corresponding to different drilling conditions by using the historical logging data of multiple wells includes: Collecting the historical logging data of multiple wells; Pre - processing the historical logging data; Using a drilling condition identification model to determine the drilling condition corresponding to each historical logging data, and generating a training data set corresponding to each drilling condition; Establishing a wavelet analysis diagnosis model based on drilling pressure; Training to obtain a kick diagnosis model corresponding to each drilling condition by using the training data set.

4. The drilling overflow warning method according to claim 3, wherein The pre - processing of the historical logging data includes any one or more of the following: removing outliers, smoothing filtering, quantitative correlation analysis, non - parametric correlation test, data dimensionality reduction processing, dimensionless processing.

5. The drilling overflow warning method according to claim 3, characterized in that, The establishment of the wavelet analysis diagnosis model based on drilling pressure includes: Obtaining a wellbore pressure wave response signal; Decomposing the pressure wave response signal by using wavelet packet to obtain each decomposition curve; Statistically analyzing the range of the mean and variance of the decomposition curve in the kick and non - kick states; Calculating the proportion of kick data in different ranges; Determining a wavelet analysis diagnosis model based on drilling pressure according to the calculation results.

6. The drilling overflow warning method according to claim 1, characterized in that The calculation of the alarm evaluation index according to the first kick probability and the second kick probability includes: Performing weighted calculation on the first kick probability and the second kick probability to obtain an alarm evaluation index.

7. The method for warning of drilling overflow according to any one of claims 1 to 6, characterized in that, The method further includes: When the alarm evaluation index exceeds a set threshold, giving a kick alarm.

8. The drilling overflow warning method according to any one of claims 1 to 6, characterized in that, The method further includes: Pre - setting multiple alarm levels, and the alarm evaluation index ranges corresponding to different alarm levels are different; Determining the current alarm level according to the alarm evaluation index; Giving a kick alarm according to the current alarm level.

9. A drilling overflow warning device, characterized in that, The device includes: A data acquisition module for collecting drilling data in real - time; A working condition determination module for determining the current working condition according to the drilling data; A first analysis module for determining the first kick probability according to the drilling data and a pre - established wavelet analysis diagnosis model; A second analysis module for extracting kick characteristic parameters from the drilling data, inputting the kick characteristic parameters into the kick diagnosis model corresponding to the current working condition, and obtaining the second kick probability according to the output of the kick diagnosis model; A calculation module for calculating an alarm evaluation index according to the first kick probability and the second kick probability.

10. The drilling overflow warning device according to claim 9, characterized in that, The device further includes: An alarm module for giving an alarm according to the alarm evaluation index.

Images