CN114912789A

CN114912789A - Drilling underground risk early warning method, equipment and storage medium

Info

Publication number: CN114912789A
Application number: CN202210502332.4A
Authority: CN
Inventors: 武胜男; 张苹茹; 张来斌; 樊建春; 郑文培
Original assignee: China University of Petroleum Beijing
Current assignee: China University of Petroleum Beijing
Priority date: 2022-05-10
Filing date: 2022-05-10
Publication date: 2022-08-16

Abstract

The application provides a drilling underground risk early warning method, drilling underground risk early warning equipment and a storage medium. The method comprises the following steps: in the drilling process, acquiring real-time monitoring data of a plurality of characteristic parameters corresponding to the research working conditions within a set time length; carrying out nonlinear fitting processing on the real-time monitoring data to obtain the change trend of the characteristic parameters in the future time; carrying out risk analysis on the real-time monitoring data to obtain first early warning levels of a plurality of characteristic parameters in future time; and sending out an early warning signal for researching the working condition according to the variation trend and the first early warning level. The accuracy of the downhole risk early warning of the well drilling can be improved through the method and the device.

Description

Drilling underground risk early warning method, equipment and storage medium

Technical Field

The application relates to a drilling technology, in particular to a drilling underground risk early warning method, equipment and a storage medium.

Background

In recent years, offshore drilling activities are gradually carried out towards deep sea and complex geological conditions, and the possibility of underground complex events of drilling is gradually increased in complex and unfamiliar stratums. Among them, gas cut, well leakage, overflow and other downhole complex events all affect normal drilling activities, and if corresponding measures are not taken in time, even the well bore is scrapped and even the personnel are injured and killed.

Therefore, how to accurately early warn the underground risk of the well drilling to reduce the loss to the maximum extent has great significance for realizing safe and efficient exploratory well drilling.

Disclosure of Invention

The application provides a drilling underground risk early warning method, equipment and a storage medium, which are used for accurately carrying out drilling underground risk early warning.

In one aspect, the application provides a method for downhole risk early warning of drilling, comprising:

in the drilling process, acquiring real-time monitoring data of a plurality of characteristic parameters corresponding to the research working conditions within a set time length;

carrying out nonlinear fitting processing on the real-time monitoring data to obtain the change trend of the characteristic parameters in the future time;

carrying out risk analysis on the real-time monitoring data to obtain first early warning levels of a plurality of characteristic parameters in future time;

and sending out an early warning signal for researching the working condition according to the variation trend and the first early warning level.

Optionally, the non-linear fitting processing is performed on the real-time monitoring data to obtain a variation trend of the characteristic parameter in the future time length, including:

selecting a data sample from real-time monitoring data within a set time length corresponding to one characteristic parameter in the characteristic parameters;

and carrying out nonlinear fitting processing on the data samples by adopting a dynamic fitting real-time parameter prediction model to obtain the change trend of the characteristic parameters in the future time.

Optionally, risk analysis is performed on the real-time monitoring data to obtain a first early warning level of the plurality of characteristic parameters in a future time period, including:

determining a judgment threshold value of the characteristic parameter according to real-time monitoring data of the characteristic parameter within the corresponding set duration by adopting a dynamic fusion model, wherein the judgment threshold value is used for determining the occurrence probability of the research working condition;

according to the judgment threshold, obtaining the occurrence probability of the research working condition oriented by the characteristic parameter;

performing information probability fusion on the occurrence probability of the research working conditions respectively corresponding to the plurality of characteristic parameters through a D-S evidence fusion principle to obtain an information probability fusion result;

and determining a first early warning level of the plurality of characteristic parameters in future time according to the information probability fusion result.

Optionally, the sending out an early warning signal for the research condition according to the variation trend and the first early warning level includes:

determining a second early warning level according to the variation trend;

sending out an early warning signal facing the research working condition according to the second early warning level and the first early warning level;

wherein, the rule of issuing the early warning signal according to the second early warning level is the same as the rule of issuing the early warning signal according to the first early warning level, and the rule comprises: and if the duration of any one of the second early warning grade and the first early warning grade is greater than or equal to the first time threshold, the early warning grade is increased, and an early warning signal is sent out based on the increased early warning grade.

Optionally, determining a second warning level according to the variation trend includes:

determining at least one prediction sample on the variation trend;

aiming at each prediction sample in at least one prediction sample, determining an early warning grade of the prediction value of the prediction sample;

determining an early warning grade corresponding to the characteristic parameter according to the early warning grade of the predicted value of the at least one prediction sample;

and determining a second early warning grade corresponding to the research working condition according to the early warning grades corresponding to the characteristic parameters.

denoising the real-time monitoring data by using a wavelet denoising model to obtain smooth data;

and carrying out nonlinear fitting processing on the smooth data to obtain the variation trend of the characteristic parameters in the future time length.

Optionally, denoising the real-time monitoring data by using the wavelet denoising model to obtain smooth data, including:

determining wavelet basis functions and decomposition layer numbers according to real-time monitoring data;

decomposing the real-time monitoring data by utilizing a Marait algorithm based on the wavelet basis function and the decomposition layer number to obtain a low-frequency coefficient and a high-frequency coefficient;

and performing wavelet reconstruction on the low-frequency coefficient and the quantized high-frequency coefficient to obtain smooth data.

In a second aspect, the present application provides a drilling downhole risk early warning device, comprising:

the data acquisition module is used for acquiring real-time monitoring data of a plurality of characteristic parameters within a set time corresponding to a research working condition in the drilling process;

the data processing module is used for carrying out nonlinear fitting processing on the real-time monitoring data to obtain the change trend of the characteristic parameters in the future time length; performing risk analysis on the real-time monitoring data to obtain first early warning levels of the plurality of characteristic parameters in the future time;

and the accident early warning module is used for sending an early warning signal facing the research working condition according to the change trend and the first early warning level.

In a third aspect, the present application provides a drilling downhole risk early warning device, comprising: a memory, a processor;

a memory for storing executable instructions;

a processor configured to execute executable instructions to implement the downhole drilling risk pre-warning method of the first aspect.

In a fourth aspect, the present application provides a computer-readable storage medium having stored thereon computer-executable instructions for implementing the downhole risk pre-warning method of the first aspect when executed by a processor

According to the method, the device and the storage medium for the downhole risk early warning of the drilling well, in the drilling process, real-time monitoring data of a plurality of characteristic parameters within a set time corresponding to a research working condition are obtained; carrying out nonlinear fitting processing on the real-time monitoring data to obtain the change trend of the characteristic parameters in the future time; carrying out risk analysis on the real-time monitoring data to obtain first early warning levels of a plurality of characteristic parameters in future time; and sending out an early warning signal for researching the working condition according to the variation trend and the first early warning level. The method comprises the steps of conducting early warning processing on research working conditions through nonlinear fitting processing and risk analysis, and then sending out early warning signals by combining the obtained change trend and a first early warning level so as to improve accuracy of well drilling underground risk early warning.

Drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and together with the description, serve to explain the principles of the application.

Fig. 1 is a schematic view of an application scenario of a downhole drilling risk early warning method provided in an embodiment of the present application;

FIG. 2 is a flow chart of a downhole drilling risk early warning method provided by an embodiment of the present application;

FIG. 3a is a smooth trend graph provided by an embodiment of the present application;

FIG. 3b is a graph of an ascending trend provided by an embodiment of the present application;

FIG. 3c is a graph of a trend of the present application;

fig. 4 is a flowchart for obtaining a first warning level according to an embodiment of the present application;

FIG. 5 is a graph comparing curves before and after denoising provided by the embodiment of the present application;

FIG. 6a is a first graph comparing predicted sample data and actual data provided by the embodiment of the present application;

FIG. 6b is a graph comparing predicted sample data and actual data provided in the embodiment of the present application;

FIG. 6c is a graph comparing predicted sample data and actual data provided by the embodiment of the present application;

fig. 7 is a comparison graph of overflow probability and information probability fusion results respectively predicted by three feature parameters provided in the embodiment of the present application;

FIG. 8 is a schematic structural diagram of a downhole drilling risk early warning device provided in an embodiment of the present application;

fig. 9 is a schematic structural diagram of a downhole drilling risk early warning device according to another embodiment of the present application.

With the above figures, there are shown specific embodiments of the present application, which will be described in more detail below. These drawings and written description are not intended to limit the scope of the inventive concepts in any manner, but rather to illustrate the inventive concepts to those skilled in the art by reference to specific embodiments.

Detailed Description

Reference will now be made in detail to the exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, like numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present application, as detailed in the appended claims.

Offshore drilling is often performed in deep sea or under complex geological conditions, and the well control risk of offshore platforms is high. When a risk event occurs in a well drilling well, the treatment takes a long time, and even if the treatment is not timely, well blowout can occur, so that the well bore is scrapped or even casualties occur. And some risk events such as: gas cut, lost circulation, flooding, etc. are precursors to the occurrence of a blowout. If the possibility of the risks occurring in the future can be timely detected and early warning is given, the occurrence of blowout accidents can be avoided. Meanwhile, the research for accelerating the intelligent drilling technology has important significance for realizing the updating and upgrading of the drilling technology, the integral transformation and upgrading of the oil and gas exploration development industry and the guarantee of energy safety.

Based on the problems, the application provides a method, equipment and a storage medium for downhole risk early warning of drilling, and the method can timely and accurately early warn downhole risks of drilling in the early stage of risk occurrence, so that loss is reduced to the maximum extent, and therefore the method and the equipment have great significance for realizing safe and efficient exploratory drilling.

The application can be used for offshore drilling, such as drilling in an offshore high-temperature and high-pressure environment, and the like.

Fig. 1 is a schematic view of an application scenario of the downhole drilling risk early warning method provided in the embodiment of the present application. As shown in fig. 1, during actual drilling, real-time monitoring data of characteristic parameters can be obtained by means of field measurement and the like. Wherein the monitoring device 10 may be disposed within a wellbore and the computing device 20 disposed outside of the well. In particular, the monitoring device 10 may be configured to collect real-time monitoring data of at least one characteristic parameter, and the detection device 10 includes a wellhead monitoring device, a wellbore riser monitoring device, and a downhole while-drilling monitoring device. Where the computing device 20 may be a computer, for example. The monitoring device 10 transmits the acquired real-time monitoring data to the computing device 20, and the computing device 20 performs the downhole drilling risk early warning by using the downhole drilling risk early warning method provided by the application based on the acquired real-time monitoring data.

The following describes the technical solutions of the present application and how to solve the above technical problems with specific embodiments. The following several specific embodiments may be combined with each other, and details of the same or similar concepts or processes may not be repeated in some embodiments. Embodiments of the present application will be described below with reference to the accompanying drawings.

Fig. 2 is a flowchart of a downhole drilling risk early warning method according to an embodiment of the present disclosure. Referring to fig. 2, the downhole drilling risk early warning method comprises the following steps:

and S11, acquiring real-time monitoring data of a plurality of characteristic parameters corresponding to the study working conditions within the set duration in the drilling process.

For example, studying conditions may include: gas cut, lost circulation, flooding, and the like. For example, when the downhole drilling risk early warning is performed, it is first required to determine which research condition the early warning is performed on, and the early warning result obtained thereafter is also the early warning result of the research condition. For example, if the research condition is flooding, the early warning result represents the possibility of flooding.

The characteristic parameters in the steps are also various, and the characteristic parameters required to be selected in different research working conditions can be the same or different, and only the characteristic parameters capable of representing the working condition conditions need to be selected. The following are exemplary: if the study condition is flooding, the selected characteristic parameters can be drilling fluid flow (MFOP), drilling mud pit average volume (TVA), riser pressure (SPPA) and the like.

It should be noted that, the selected characteristic parameters of the drilling well may be different in different operation environments, and the selected characteristic parameters may effectively reflect the corresponding research conditions. Depending on the characteristics of these characteristic parameters, different characteristic parameters may be acquired at different locations in the well, for example: at least one of a bottom hole, a wellbore, and a wellhead.

And S12, carrying out nonlinear fitting processing on the real-time monitoring data to obtain the change trend of the characteristic parameters in the future time length.

Optionally, when the non-linear fitting processing is performed on the real-time monitoring data, a fitting function is required, the fitting function can be selected at will, and only the fact that the fitting can be effectively performed on the real-time monitoring data is required to be guaranteed.

In addition, each characteristic parameter has corresponding real-time monitoring data, so that each characteristic parameter can obtain corresponding variation trend in future time, and the variation trends are independent.

And S13, performing risk analysis on the real-time monitoring data to obtain a first early warning level of the plurality of characteristic parameters in the future time.

In this step, risk analysis is performed on the real-time monitoring data, which may be that risk analysis is performed on the real-time monitoring data corresponding to at least one characteristic parameter, and then at least one risk analysis result is fused to obtain a first early warning level.

Or, performing risk analysis on the real-time monitoring data, which may be performing risk analysis on the real-time monitoring data corresponding to all the characteristic parameters, to obtain a first early warning level.

In some embodiments, the first warning level may include a primary warning, a secondary warning, and a tertiary warning, and the severity of the research condition represented by the three increases in sequence.

It should be noted that step S12 and step S13 are independent from each other, so there is no precedence order, and both steps may be performed simultaneously.

And S14, sending out an early warning signal for the research working condition according to the variation trend and the first early warning level.

Illustratively, the type of the warning signal may be various, and may be at least one of a graphic, a text, a sound, and the like. And when the operator receives the early warning signal, taking corresponding measures according to the meaning represented by the early warning signal. Corresponding to the first early warning level, when the first early warning level is primary early warning and secondary early warning, the sent early warning signal can be a graphic prompt; when the first early warning level is three-level early warning, the risk is high, an alarm is needed, and the sent early warning signal can be an alarm sound.

According to the embodiment of the application, in the drilling process, real-time monitoring data of a plurality of characteristic parameters corresponding to the research working conditions within the set duration are obtained; carrying out nonlinear fitting processing on the real-time monitoring data to obtain the change trend of the characteristic parameters in the future time; carrying out risk analysis on the real-time monitoring data to obtain first early warning levels of a plurality of characteristic parameters in future time; and sending out an early warning signal for researching the working condition according to the variation trend and the first early warning level. The method comprises the steps of conducting early warning processing on research working conditions through nonlinear fitting processing and risk analysis, and then sending out early warning signals by combining the obtained change trend and a first early warning level so as to improve accuracy of well drilling underground risk early warning.

On the basis of the foregoing embodiments, in some embodiments, the performing nonlinear fitting processing on the real-time monitoring data to obtain a variation trend of the characteristic parameter in a future time period includes: denoising the real-time monitoring data by using a wavelet denoising model to obtain smooth data; and then carrying out nonlinear fitting processing on the smooth data to obtain the variation trend of the characteristic parameters in the future time length.

The reason is that the local environment of the drilling work is complex, the acquired real-time monitoring data may be data formed by combining target data and noise, the purpose of noise reduction is to suppress the noise to recover the target data, and in general, the noise is represented by a high-frequency signal, and the target data is represented by a low-frequency signal or a steady signal. The denoising method utilizes the variable scale characteristic in wavelet transformation to achieve the suppression of noise. Selecting a proper threshold, wherein the wavelet coefficient larger than the threshold is considered to be generated by the signal of the target data and is reserved; the wavelet coefficient smaller than the threshold is considered as noise generation and is set to be zero, so that the purpose of denoising is achieved.

Corresponding to the characteristics of the target data, in order to reduce unnecessary noise reduction processing, after the real-time monitoring data is obtained, generally, feature judgment is performed on the real-time monitoring data first to judge whether the real-time monitoring data is a stationary signal, and if the real-time monitoring data is a stationary signal, it indicates that the real-time monitoring data can be directly processed and analyzed without denoising. Otherwise, the real-time monitoring data can be processed and analyzed only by denoising.

Correspondingly, if the real-time monitoring data need to be denoised, risk analysis is carried out on the real-time monitoring data according to the steps to obtain a first early warning level of a plurality of characteristic parameters in the future time length, and the method comprises the following steps of: denoising the real-time monitoring data by using a wavelet denoising model to obtain smooth data; and then carrying out risk analysis on the smooth data to obtain a first early warning level of the plurality of characteristic parameters in the future time length.

Therefore, when the real-time monitoring data contains noise, in order to make the real-time monitoring data easier to analyze and the analysis result more accurate, generally, the wavelet denoising model may be used to perform denoising processing on the real-time monitoring data to obtain smooth data. Correspondingly, the processing and analyzing objects in the following steps S12 and S13 are also changed from the real-time monitoring data to the de-noised smooth data, and finally the variation trend and the first early warning level are obtained.

Specifically, the denoising processing of the real-time monitoring data by using the wavelet denoising model to obtain the smooth data includes the following steps:

firstly, determining wavelet basis functions and decomposition layer numbers according to real-time monitoring data.

In practical applications, it is generally desirable that the selected wavelet basis functions satisfy the following conditions: orthogonality, high vanishing moment, tautness, symmetry, or antisymmetry. However, the wavelet basis functions have characteristics when processing signals, and no wavelet basis function can achieve the optimal denoising effect on all types of signals. Therefore, when in use, a proper wavelet basis function is selected according to the characteristics of the signal.

The decomposition layer number represents the decomposition times of the high-frequency signal and the low-frequency signal, and the larger the decomposition layer number is, the more obvious the different characteristics of the noise and the target data are, and the more beneficial the separation of the two. On the other hand, the larger the number of decomposition layers is, the more easily information is lost in the finally obtained target data, and the final denoising effect is influenced to a certain extent. Therefore, an appropriate number of decomposition layers is selected for application.

And then, decomposing the real-time monitoring data by utilizing a Marait algorithm based on the wavelet basis function and the decomposition layer number to obtain a low-frequency coefficient and a high-frequency coefficient.

The specific process of decomposing the real-time monitoring data is as follows: firstly, the real-time monitoring data is transformed by using a wavelet basis function, wherein the frequency information of the real-time monitoring data can be obtained by scaling the scale of the wavelet basis function. And then decomposing the transformed real-time monitoring data, wherein if the 0 level is in the direction of the highest resolution, the decomposition of the first layer can obtain a high-frequency coefficient of a layer-1 and a low-frequency coefficient of the layer-1, the decomposition of the second layer can decompose the low-frequency coefficient of the layer-1 into a high-frequency coefficient of a layer-2 and a low-frequency coefficient of a layer-2, and so on, if the number of decomposition layers is 5, the final decomposition result is a low-frequency coefficient of the layer-5 and a high-frequency coefficient of the layer-1 to the layer-5.

Specifically, if the original real-time monitoring data is f (x), the result obtained after decomposition is as follows:

in the above formula, J is the number of decomposition layers, and the low frequency part is

The high frequency part is D _j f。

In addition, in the process of denoising, the selection of a threshold is also involved, the threshold is a value for distinguishing high frequency from low frequency, and a soft threshold or a hard threshold can be selected. The specific selection condition is based on real-time monitoring data.

And finally, performing wavelet reconstruction on the low-frequency coefficient and the quantized high-frequency coefficient to obtain smooth data.

In the decomposed expression, the low-frequency coefficient portion is a portion of the target data and should be stored, and the high-frequency coefficient portion is a noise portion and should be removed. Therefore, the high frequency coefficient needs to be quantized, and then the low frequency coefficient and the quantized high frequency coefficient need to be wavelet reconstructed to form smooth data after noise reduction.

In some embodiments, the specific implementation manner of performing nonlinear fitting processing on the real-time monitoring data to obtain the variation trend of the characteristic parameter in the future duration is as follows: firstly, aiming at one characteristic parameter in a plurality of characteristic parameters, selecting a data sample from real-time monitoring data corresponding to the characteristic parameter, and carrying out nonlinear fitting processing on the data sample by adopting a dynamic fitting real-time parameter prediction model to obtain the change trend of the characteristic parameter in the future time.

Specifically, a section of monitoring data is selected from the database of real-time monitoring data, and then a data sample is selected from the section of monitoring data, if the sample is the data sampleThe sampling period is s seconds, that is, one data sample is selected every s seconds, if i data samples are selected, monitoring data with the time length of i × s seconds is needed, that is, the size of a time window is i × s seconds, and the set of the selected data samples is [ x s ] ₁ 、x ₂ 、x ₃ 、…、x _i ]. And then performing curve fitting on the part of the data samples, namely finding out a curve function with higher correlation with the data samples. By using the curve function, the prediction values of the prediction samples can be obtained, and if the prediction values of the w prediction samples are to be obtained, the set of the prediction values of the obtained prediction samples is [ x ] _i+1 、x _i+2 、x _i+3 、...、x _i+w ]And obtaining the variation trend of the characteristic parameters in the future time length according to the set of the predicted values of the prediction samples.

It should be noted that the above monitoring data is dynamically stored data, i.e. one data sample enters the database every s seconds, and is slide-advanced in the fitting model, i.e. one old sample exits every new sample is advanced, and the size of the time window is not changed all the time, so as to realize the dynamic prediction function of the model.

In addition, the fitting function in this embodiment is a fourier approximation function, and different fourier approximation function levels can be selected according to different data samples, and the basic type of the fourier function is shown as follows:

in the above formula, n is a constant, T is time, T ₀ Is a period, ω ₀ Is the frequency.

It can be understood that, in the above-mentioned specific step of performing the non-linear fitting process on the real-time monitoring data to obtain the variation trend of the characteristic parameters in the future time period, each characteristic parameter needs to be performed in the step to obtain the respective variation trend of all the characteristic parameters. The multiple characteristic parameters are mutually independent when the step is carried out, so that the step can be simultaneously carried out by the multiple characteristic parameters, the time is shortened, and the efficiency is improved.

As shown in fig. 3a, 3b and 3c, the above-mentioned variation trend includes a steady, ascending and descending. The curve is stable and normal, when the change trend is ascending or descending, the curve can be further judged to obtain a second early warning grade, so that the severity of the research working condition represented by the change trend can be displayed more visually, and relative measures can be taken more quickly according to the second early warning grade.

In some embodiments, the second warning level also includes a first warning level, a second warning level and a third warning level, and the severity of the research conditions represented by the first warning level, the second warning level and the third warning level increases in sequence. In addition, different prompts can be provided according to different early warning levels. When the second early warning level is the first-level early warning and the second-level early warning, the sent prompt can be a graphic prompt; when the second early warning level is three-level early warning, the risk is high, an alarm is needed, and the sent prompt can be an alarm sound. The second warning level may be further determined according to the following steps.

In the process of determining the second early warning level according to the variation trend, the method comprises the following steps: and determining at least one prediction sample on the variation trend, wherein the number of the prediction samples is not determined. And aiming at the determined one or more prediction samples, determining the early warning level of the prediction value of the prediction sample. And determining the early warning grade corresponding to the characteristic parameter according to the early warning grade of the predicted value. And determining a second early warning grade corresponding to the research working condition according to the early warning grades corresponding to the characteristic parameters.

Illustratively, if the predicted value of the determined prediction sample is x _i+1 When determining whether or not an early warning is required, and what kind of early warning is required, first, it is determined that x is _i+1 -x _i Relation to threshold values of predicted values of characteristic parameters, x in the above equation _i Refers to the selected data sample, and the data sample is x _i+1 Adjacent to each other. The threshold value of the predicted value is a fixed value corresponding to the characteristic parameter, and each characteristic parameter has the threshold value of the predicted value corresponding to the characteristic parameter. In some embodiments, the threshold for the predicted value comprises: threshold minimum limit X _u Threshold value intermediate limit value X _c And a threshold maximum limit value X _l When x is _i+1 -x _i Exceeding a threshold minimum limit X _u In time, early warning is triggered, but the exceeding amount determines the early warning level. When X is present _u ≤x _i+1 -x _i ＜X _c When X is detected, the first-stage warning is performed _c ≤x _i+1 -x _i ＜X _l Then, for the second-level warning, when X _l ≤x _i+1 -x _i And then, three-level early warning is carried out.

Optionally, the slope of the connection line of the prediction sample may also be calculated, and the early warning level is determined by comparing the slope with the slope of the threshold, where the slope of the threshold refers to the slope of the connection line of the threshold and the data sample at the previous time, and the slope of the prediction sample may be determined by the following formula:

in the above formula, k is the slope of the prediction sample; x is the number of _i+1 Is a prediction sample; x is the number of _i Selecting a data sample; and the two samples are adjacent samples. t is t _i+1 Is predicted to be x _i+1 The time of day; t is t _i Is predicted to be x _i The time of day; when two samples are adjacent samples, t _i+1 -t _i The value of (a) is the time interval over which the data sample is taken.

Accordingly, the slope of the threshold includes: a threshold minimum limit slope, a threshold intermediate limit slope, and a threshold maximum limit slope. Can be determined by the following formula:

in the above formula, K _u Is the threshold minimum limit slope; k _c Is the threshold median limit slope; k is _l Is the threshold maximum limit slope; x _u Is a threshold minimum limit; x _c Is a threshold intermediate limit; x _l Is a threshold maximum limit; t is t _i+1 -t _i The value of (a) is the time interval over which the data sample is taken. When K is _u ≤k＜K _c When it is, for the first-level warning, when K _c ≤k＜K _l When it is, it is a two-stage early warning, when K _l And when k is less than or equal to k, performing three-stage early warning.

Furthermore, the included angle of the prediction sample can be determined, and the early warning grade is judged by comparing the included angle with the threshold value. The included angle of the prediction sample refers to an included angle of a connecting line of the prediction sample and the data sample at the previous moment, and the threshold included angle refers to an included angle of a connecting line of the threshold and the data sample at the previous moment. Specifically, since the included angle and the slope may be converted to each other, the angle of the connection line of the prediction sample may be determined by the following formula:

θ＝arctank

in the above equation, θ is the angle of the prediction sample, and k is the slope of the prediction sample.

Accordingly, the threshold included angle may also be determined by the following equation:

θ _u ＝arctan(K _u )；θ _c ＝arctan(K _c )；θ _l ＝arctan(K _l )

in the above formula, K _u Is the threshold minimum limit slope; k _c Is the threshold median limit slope; k is _l Is the threshold maximum limit slope; theta _u Is the minimum threshold included angle; theta _c Is the included angle of the middle limit value of the threshold value; theta _l Is the angle of the maximum limit angle of the threshold. When theta is _u ≤θ＜θ _c When the temperature of the water is higher than the set temperature,for the first-level warning, when theta _c ≤θ＜θ _l When it is, for two-stage warning, when theta _l When theta is less than or equal to theta, three-level early warning is carried out.

When the early warning grade of the predicted value of the predicted sample is determined through the predicted value of the predicted sample, the slope of the predicted sample or the included angle of the predicted sample, the early warning grade of the corresponding characteristic parameter is determined, the relation between the early warning grade and the early warning grade is regulated by an operator, and the early warning grade are in one-to-one correspondence in some embodiments, namely when the early warning grade of the predicted value of the predicted sample is first-level early warning, the early warning grade of the corresponding characteristic parameter is also first-level early warning; and when the early warning level of the predicted value of the prediction sample is the third-level early warning, the early warning level of the corresponding characteristic parameter is also the third-level early warning.

Because each research working condition corresponds to a plurality of characteristic parameters, after the early warning grade of each characteristic parameter is determined, the plurality of characteristic parameters can obtain a plurality of early warning grades. At this time, the second warning level of the research working condition can be determined in various ways. One mode is that a plurality of early warning levels of the plurality of characteristic parameters are all the second early warning levels, that is, the second early warning levels are a plurality. In another mode, a plurality of early warning levels of the characteristic parameters are compared, and the maximum early warning level is set as the second early warning level.

In some embodiments, as shown in fig. 4, performing risk analysis on the real-time monitoring data to obtain a first warning level of the plurality of characteristic parameters in a future time period includes:

s131, determining a judgment threshold value of the characteristic parameter according to the real-time monitoring data of the characteristic parameter within the corresponding set time length by adopting a dynamic fusion model, wherein the judgment threshold value is used for determining the occurrence probability of the research working condition.

To determine the judgment threshold of the characteristic parameter, first, a threshold of a predicted value of the characteristic parameter is determined, which includes: threshold minimum limit X _u And a threshold maximum limit value X _l . The decision threshold is then calculated according to:

X _RU ＝μ-X _u

X _RL ＝μ+X _l

in the above formula, X _u Is a threshold minimum limit; x _l Is a threshold maximum limit; x (t) _i ) Is a time of t _i Real-time sample values of the time; x _RU A lower limit value of the dynamic threshold; x _RL An upper limit value of the dynamic threshold; mu is a dynamic propulsion mean value of the characteristic parameter; δ is a decision threshold.

And S132, obtaining the occurrence probability of the characteristic parameters facing the research working condition according to the judgment threshold.

And solving the occurrence probability of the characteristic parameters facing the research working condition according to the corresponding relation between the judgment threshold and the occurrence probability of the research working condition. In some implementations, the decision threshold δ is divided into fuzzy partitions to achieve the probability m of occurrence of the study condition _x The division rule is shown in the following table:

0＜δ≤0.2	0.2＜δ≤0.4	0.4＜δ≤0.5	δ>0.5
				0＜m _x ≤0.4	0.4＜m _x ≤0.8	0.8＜m _x ≤1	1

in the above table, δ and m for each column _x Corresponding to each other, and when delta is greater than 0 and not more than 0.5, m _x Is twice the value of delta.

In the above table, m _x To investigate the probability of occurrence of a condition, i.e. to represent the likelihood that the condition under investigation will occur in the future, m is therefore _x The value range of (1) is between 0 and 1, and the closer the value is to 1, the higher the possibility of the corresponding research working condition is. Each research working condition corresponds to a plurality of characteristic parameters, and each characteristic parameter has a judgment on the occurrence probability of the research working condition. Therefore, step S132 needs to be performed for each characteristic parameter corresponding to the research condition to calculate the occurrence probability of the research condition corresponding to each characteristic parameter. In addition, since there are a plurality of occurrence probabilities of the study behavior corresponding to a plurality of characteristic parameters, m is used _x To represent the occurrence probability of the study condition, m _x The corner marks may be replaced by any number or letter to represent different probabilities of occurrence of the investigation regime resulting from different characteristic parameters, e.g. m ₁ Probability of occurrence, m, of a condition of investigation for a characteristic parameter of drilling fluid flow ₂ The probability of occurrence of the investigated operating conditions for this characteristic parameter of the riser pressure.

And S133, performing information probability fusion on the occurrence probabilities of the research working conditions respectively corresponding to the plurality of characteristic parameters according to a D-S evidence fusion principle to obtain an information probability fusion result.

Since the calculated occurrence probabilities of the research conditions corresponding to the plurality of characteristic parameters are not necessarily the same, but it cannot be determined which characteristic parameters or which characteristic parameters obtain more accurate occurrence probabilities of the research conditions, it is necessary to perform information probability fusion on the different occurrence probabilities of the research conditions.

The specific information probability fusion process is as follows:

firstly, a decision distance of the occurrence probability of the research working condition corresponding to any two characteristic parameters is calculated, namely, the difference of the occurrence probability of the research working condition obtained by representing the two characteristic parameters is calculated. The calculation method is as follows:

d _ij ＝2|(m _i -0.5)(m _i -m _j )|

in the above formula, the occurrence probability of the research conditions corresponding to the two characteristic parameters is m _i And m _j ，d _ij Represents the result m _i Can refuse m _j The distance of (c). The number of the characteristic parameters is multiple, and d is calculated pairwise _ij A decision matrix D can be obtained, which is shown as:

then, a similarity matrix R of the decision matrix D is calculated according to the following formula, and the calculation manner of each element in R is as follows:

r _ij ＝1-d _ij (0＜r _ij ＜1)

element r in matrix D _ij Can represent the consistency of the occurrence probability of the research working condition obtained by the two characteristic parameters. r is _ij The larger the value, the better the degree of uniformity, r _ij The smaller the value, the worse the degree of consistency.

And then calculating to obtain a final information probability fusion result through the following formula:

in the above formula, Su (m) _i ) The support vector corresponding to the characteristic parameter in the sample is obtained; alpha is alpha _i The reliability vector of the characteristic parameters in the sample is obtained;

a function value is assigned for the average trust of the feature parameter. n is the number of the selected and calculated characteristic parameters; and m is an information probability fusion result.

Optionally, in some embodiments, first early warning levels corresponding to different working conditions are researched, and an information probability fusion result obtained at this time is recorded as m (a) _N ) Wherein A is _N Can be A ₁ 、A ₂ 、A ₃ Thus representing different study conditions, m (A) _N ) Can be calculated by the following formula:

corresponding to, in the formula

And

the same way of calculation.

Wherein the reliability vector α _i The larger the value of (A), the larger the weight of the result corresponding to the value of (A) is, the higher the credibility of the characteristic parameter prediction result is, the larger the influence on the information probability fusion result is; and otherwise, the smaller the weight representing the corresponding result is, the lower the reliability of the characteristic parameter prediction result is, and the smaller the influence on the information probability fusion result is.

S134, determining a first early warning level of the plurality of characteristic parameters in the future time length according to the information probability fusion result.

The information probability fusion result is a decimal and can also be expressed as a percentage, and the corresponding relation between the first early warning level and the information probability fusion result can be divided by self.

In some embodiments, when m < 50%, there is low risk and no pre-warning is given; when m is more than or equal to 50% and less than 75%, the early warning is a first-stage early warning, when m is more than or equal to 75% and less than 90%, the early warning is a second-stage early warning, and when m is more than or equal to 90% and less than 100%, the early warning is a third-stage early warning. m (A) _N ) Is divided in the same manner as m.

And when the change trend and the first early warning level are determined, a corresponding early warning signal is sent out. The form of the warning signal may be varied. The curve of the change trend can be directly displayed, the grade of the second early warning grade can be displayed according to the second early warning grade determined by the change trend, and corresponding light or alarm sound can be output according to the corresponding grade. The first early warning level can also directly display the level or output corresponding light or alarm sound according to the corresponding level.

When the first early warning level and the second early warning level are output in the same way, the output modes can be the same or different. As long as they do not affect each other.

In addition, when the first early warning level or the second early warning level carries out early warning for the first time, the early warning is carried out for the first level. If the early warning duration exceeds a first threshold value, the early warning level is increased by one level, and if the early warning level is maximum, the early warning level is kept unchanged. If the early warning duration is smaller than the second threshold, the early warning level is decreased by one level, and if the early warning level is minimum, the early warning level is kept unchanged.

To facilitate an understanding of all of the steps described above, examples are given herein. In the example, a certain drilling part database is selected as a data source, the research working condition is overflow, and the characteristic parameters are relative drilling fluid outlet flow (MFOP), average drilling mud pit volume (TVA) and riser pressure (SPPA).

Because the obtained real-time monitoring data contains noise, the three characteristic parameters are respectively subjected to denoising processing. For example, a data curve before and after denoising of the real-time monitoring data of the average volume of the drilling mud pit is shown in fig. 5, wherein when denoising is performed on the real-time monitoring data of the average volume of the drilling mud pit, a db7 wavelet basis function is selected for performing 5-layer decomposition; selecting a db7 wavelet basis function for 4-layer decomposition when denoising real-time monitoring data of the riser pressure; and when denoising is carried out on the real-time monitoring data relative to the outlet flow of the drilling fluid, selecting a db7 wavelet basis function to carry out 5-layer decomposition. And soft thresholds are selected for all three.

And obtaining smooth data after denoising, and respectively carrying out nonlinear fitting processing on the smooth data of the three. Firstly, selecting data samples, taking the average volume of a mud pit of a drilling machine as an example, selecting 20 data samples, and selecting one data sample every 5 seconds.

The fitting function selects a Fourier approximation function according to the data sample, and determines the grade of the Fourier approximation function according to respective smooth data of the fitting function, the fitting function and the data sample. The Fourier approximation function relative to the outlet flow of the drilling fluid is selected to be 5 grades, the Fourier approximation function of the average volume of the drilling mud pool is selected to be 4 grades, and the Fourier approximation function of the pressure of the vertical pipe is selected to be 6 grades. The fourier approximation function of the average volume of the drilling mud pit is as follows:

f(t)＝a ₀ +a ₁ ·cos(ω _t )+b ₁ ·sin(ωt)+a ₂ ·cos(2ωt)+b ₂ ·sin(2ωt)+a ₃ ·cos(3ωt)+b ₃ ·sin(3ωt)+a ₄ ·cos(4ωt)+b ₄ ·sin(4ωt)

and after the predicted value of the predicted sample is predicted according to the formula, taking the average volume of the drilling mud pool as an example, and judging the second early warning level. Specifically, the slope of a prediction sample is greater than or equal to 0.1, and primary early warning is performed when the slope of the prediction sample is less than 0.15; the slope of the prediction sample is greater than or equal to 0.15, and secondary early warning is performed when the slope of the prediction sample is less than 0.2; and when the slope of the prediction sample is greater than or equal to 0.2, the three-stage early warning is realized. Namely, the angle of the included angle of the prediction sample is greater than or equal to 5.71 degrees, and the first-stage early warning is performed when the angle of the included angle of the prediction sample is less than 8.53 degrees; the angle of the included angle of the prediction sample is greater than or equal to 8.53 degrees, and secondary early warning is performed when the angle of the included angle of the prediction sample is less than 11.3 degrees; and predicting the three-level early warning when the included angle of the sample is greater than or equal to 11.3 degrees. After the predicted value is obtained, the second early warning grade can be obtained by comparing the predicted value with the numerical value, and no specific calculation is carried out here.

Fig. 6a, 6b and 6c are graphs comparing curves of predicted sample data and actual data, and it can be found that the prediction is more accurate due to higher goodness of fit within a certain time.

And calculating to obtain the overflow probability of the three parameters according to the threshold value and the smooth data of the predicted value of the three parameters and the corresponding relation between the judgment threshold value and the overflow probability. And then obtaining an information probability fusion result according to a D-S evidence fusion principle. And judging a first early warning grade corresponding to the information probability fusion result. The data obtained are shown in the following table:

serial number	Time	MFOP	TVA	SPPA	Fusion probability	Early warning result
							1	13:23:35	0.7474	0.0102	0.5660	0.7129	First-level warning
2	13:23:40	0.7708	0.0504	0.4851	0.6459	First-level warning
							3	13:23:45	0.7932	0.0922	0.3622	0.5547	First-level warning
4	13:23:50	0.8148	0.1352	0.2152	0.4625	Is normal and normal
							5	13:23:55	0.8360	0.1796	0.0845	0.4186	Is normal
6	13:24:00	0.8570	0.2251	0.0194	0.4363	Is normal
							7	13:24:05	0.8779	0.2712	0.1104	0.4801	Is normal
8	13:24:10	0.8992	0.3174	0.1777	0.5313	First-level warning
							9	13:24:15	0.9214	0.3635	0.2115	0.5727	First-level warning
10	13:24:20	0.9445	0.4091	0.2131	0.8007	First-level warning
							11	13:24:25	0.9682	0.4543	0.1782	0.6171	First-level warning
12	13:24:30	0.9921	0.4991	0.1181	0.6383	First-level warning

In addition, fig. 7 also shows the comparison of the overflow probability predicted by each of the three characteristic parameters and the information probability fusion result.

And the first early warning grade result and the second early warning grade result are output in the form of alarm sound to serve as early warning signals.

The following are embodiments of the apparatus of the present application that may be used to perform embodiments of the method of the present application. For details which are not disclosed in the embodiments of the apparatus of the present application, reference is made to the embodiments of the method of the present application.

Fig. 8 is a schematic structural diagram of a downhole drilling risk early warning device according to an embodiment of the present application. As shown in fig. 8, the downhole drilling risk early warning apparatus 800 includes: a data acquisition module 801, a data processing module 802 and an accident early warning module 803. Wherein:

the data acquisition module 801 is configured to acquire real-time monitoring data of a plurality of characteristic parameters within a set time corresponding to a research condition in a drilling process.

The data processing module 802 is configured to perform nonlinear fitting processing on the real-time monitoring data to obtain a variation trend of the characteristic parameters in the future time length, perform risk analysis on the real-time monitoring data to obtain first early warning levels of the plurality of characteristic parameters in the future time length;

and the accident early warning module 803 is used for sending an early warning signal for researching the working condition according to the variation trend and the first early warning level.

Optionally, the data processing module 802 performs nonlinear fitting processing on the real-time monitoring data, and a specific process of obtaining a variation trend of the characteristic parameter in the future time length is as follows: aiming at one characteristic parameter in the characteristic parameters, selecting a data sample from real-time monitoring data within a set time length corresponding to the characteristic parameter; and carrying out nonlinear fitting processing on the data samples by adopting a dynamic fitting real-time parameter prediction model to obtain the change trend of the characteristic parameters in the future time.

Optionally, the specific process of the data processing module 802 performing risk analysis on the real-time monitoring data to obtain the first early warning level of the plurality of characteristic parameters in the future time length is as follows:

performing information probability fusion on the occurrence probabilities of the research working conditions respectively corresponding to the plurality of characteristic parameters through a D-S evidence fusion principle to obtain an information probability fusion result;

and determining a first early warning level of the plurality of characteristic parameters in the future time length according to the information probability fusion result.

Further, when the accident warning module 803 sends out a warning signal for researching a working condition according to the variation trend and the first warning level, the specific process is as follows:

determining a second early warning level according to the variation trend;

In some embodiments, when the data processing module 802 determines the second warning level according to the variation trend, the specific process is as follows: determining at least one prediction sample on the variation trend; aiming at each prediction sample in at least one prediction sample, determining an early warning grade of the prediction value of the prediction sample; determining an early warning grade corresponding to the characteristic parameter according to the early warning grade of the predicted value of the at least one prediction sample; and determining a second early warning grade corresponding to the research working condition according to the early warning grades corresponding to the characteristic parameters.

In some embodiments, the specific process of the data processing module 802 performing non-linear fitting processing on the real-time monitoring data to obtain the variation trend of the characteristic parameter in the future time length is as follows: denoising the real-time monitoring data by using a wavelet denoising model to obtain smooth data; and carrying out nonlinear fitting processing on the smooth data to obtain the variation trend of the characteristic parameters in the future time length.

In some embodiments, the data processing module 802 performs denoising processing on the real-time monitoring data by using the wavelet denoising model, and the specific process of obtaining the smooth data is as follows: determining wavelet basis functions and decomposition layer numbers according to real-time monitoring data; decomposing the real-time monitoring data by utilizing a Marait algorithm based on the wavelet basis function and the decomposition layer number to obtain a low-frequency coefficient and a high-frequency coefficient; and performing wavelet reconstruction on the low-frequency coefficient and the quantized high-frequency coefficient to obtain smooth data.

It should be noted that the downhole drilling risk early warning device provided by the application can be used for executing the downhole drilling risk early warning method embodiment, the implementation principle and the technical effect are similar, and details are not repeated here.

It should be noted that the division of the modules of the above apparatus is only a logical division, and the actual implementation may be wholly or partially integrated into one physical entity, or may be physically separated. And these modules can be realized in the form of software called by processing element; or may be implemented entirely in hardware; and part of the modules can be realized in the form of calling software by the processing element, and part of the modules can be realized in the form of hardware. For example, the processing module may be a processing element separately set up, or may be implemented by being integrated in a chip of the apparatus, or may be stored in a memory of the apparatus in the form of program code, and a function of the processing module may be called and executed by a processing element of the apparatus. Other modules are implemented similarly. In addition, all or part of the modules can be integrated together or can be independently realized. The processing element here may be an integrated circuit with signal processing capabilities. In implementation, each step of the above method or each module above may be implemented by an integrated logic circuit of hardware in a processor element or an instruction in the form of software.

For example, the above modules may be one or more integrated circuits configured to implement the above methods, such as: one or more Application Specific Integrated Circuits (ASICs), or one or more microprocessors (DSPs), or one or more Field Programmable Gate Arrays (FPGAs), among others. For another example, when some of the above modules are implemented in the form of a processing element scheduler code, the processing element may be a general-purpose processor, such as a Central Processing Unit (CPU) or other processor capable of calling program code. As another example, these modules may be integrated together, implemented in the form of a system-on-a-chip (SOC).

In the above embodiments, the implementation may be wholly or partially realized by software, hardware, firmware, or any combination thereof. When implemented in software, may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. The procedures or functions according to the embodiments of the present application are all or partially generated when the computer program instructions are loaded and executed on a computer. The computer may be a general purpose computer, a special purpose computer, a network of computers, or other programmable device. The computer instructions may be stored in a computer readable storage medium or transmitted from one computer readable storage medium to another, for example, the computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wire (e.g., coaxial cable, fiber optic, Digital Subscriber Line (DSL)) or wirelessly (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device, such as a server, a data center, etc., that incorporates one or more of the available media. The usable medium may be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., Solid State Disk (SSD)), among others.

Fig. 9 is a schematic structural diagram of a downhole drilling risk early warning device according to another embodiment of the present application. Referring to fig. 9, the downhole risk early warning device 900 includes a collector 901, a memory 902, a processor 903, and an alarm 904. The collector 901 is used for collecting data; the memory 902 is used to store executable instructions; the processor 903 is configured to execute executable instructions to implement the downhole risk forewarning method of drilling described above; the alarm 904 is used to alarm the outside.

The embodiment of the application also provides a computer-readable storage medium, wherein computer-executable instructions are stored in the computer-readable storage medium, and when the computer-executable instructions are executed by a processor, the method for early warning the risk in the well drilling downhole is realized.

The computer-readable storage medium may be implemented by any type of volatile or non-volatile memory device or combination thereof, such as Static Random Access Memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic or optical disk. Computer-readable storage media can be any available media that can be accessed by a general purpose or special purpose computer.

An exemplary readable storage medium is coupled to the processor such the processor can read information from, and write information to, the readable storage medium. Of course, the readable storage medium may also be an integral part of the processor. The processor and the readable storage medium may reside in an Application Specific Integrated Circuits (ASIC). Of course, the processor and the readable storage medium may also reside as discrete components in the detection apparatus for sensing holes.

The embodiment of the application also provides a computer program product, which comprises a computer program, and the computer program is executed by a processor to realize the drilling downhole risk early warning method.

Those of ordinary skill in the art will understand that: all or a portion of the steps of implementing the above-described method embodiments may be performed by hardware associated with program instructions. The program may be stored in a computer-readable storage medium. When executed, the program performs steps comprising the method embodiments described above; and the aforementioned storage medium includes: various media that can store program codes, such as ROM, RAM, magnetic or optical disks.

Finally, it should be noted that: the above embodiments are only used for illustrating the technical solutions of the present disclosure, and not for limiting the same; while the present disclosure has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art will understand that: the technical solutions described in the foregoing embodiments may still be modified, or some or all of the technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present disclosure.

Claims

1. A downhole drilling risk early warning method is characterized by comprising the following steps:

carrying out nonlinear fitting processing on the real-time monitoring data to obtain the change trend of the characteristic parameters in the future time length;

performing risk analysis on the real-time monitoring data to obtain first early warning levels of the plurality of characteristic parameters in future time;

and sending out an early warning signal facing the research working condition according to the variation trend and the first early warning level.

2. The downhole drilling risk early warning method according to claim 1, wherein the non-linear fitting processing is performed on the real-time monitoring data to obtain a change trend of the characteristic parameter in a future time period, and the method comprises the following steps:

aiming at one characteristic parameter in the characteristic parameters, selecting a data sample from the real-time monitoring data corresponding to the characteristic parameter within the set duration;

and carrying out nonlinear fitting processing on the data sample by adopting a dynamic fitting real-time parameter prediction model to obtain the change trend of the characteristic parameters in the future time.

3. The downhole drilling risk pre-warning method of claim 1, wherein the risk analyzing the real-time monitoring data to obtain a first pre-warning level of the plurality of characteristic parameters in a future time period comprises:

determining a judgment threshold value of the characteristic parameter according to real-time monitoring data of the characteristic parameter corresponding to the set duration by adopting a dynamic fusion model, wherein the judgment threshold value is used for determining the occurrence probability of the research working condition;

obtaining the occurrence probability of the research working condition faced by the characteristic parameter according to the judgment threshold;

performing information probability fusion on the occurrence probabilities of the research working conditions respectively corresponding to the plurality of characteristic parameters according to a D-S evidence fusion principle to obtain an information probability fusion result;

4. A downhole drilling risk pre-warning method according to claim 1, wherein said emitting a pre-warning signal for the study condition based on the trend and the first pre-warning level comprises:

determining a second early warning level according to the change trend;

wherein, the rule for issuing the early warning signal according to the second early warning level is the same as the rule for issuing the early warning signal according to the first early warning level, and the rule comprises: and if the duration of any one of the second early warning grade and the first early warning grade is greater than or equal to a first time threshold, improving the early warning grade, and sending out an early warning signal based on the improved early warning grade.

5. A downhole drilling risk pre-warning method according to claim 4, wherein determining a second pre-warning level according to the trend of change comprises:

determining at least one prediction sample on the trend of change;

aiming at each prediction sample in the at least one prediction sample, determining an early warning grade of the prediction value of the prediction sample;

determining an early warning grade corresponding to the characteristic parameter according to the early warning grade of the predicted value of at least one prediction sample;

6. The downhole drilling risk early warning method according to any one of claims 1 to 5, wherein the non-linear fitting processing of the real-time monitoring data to obtain the variation trend of the characteristic parameter in a future time period comprises:

and carrying out nonlinear fitting processing on the smooth data to obtain the change trend of the characteristic parameters in the future time length.

7. The downhole drilling risk pre-warning method according to claim 6, wherein the denoising processing of the real-time monitoring data by using the wavelet denoising model to obtain the smoothed data comprises:

determining a wavelet basis function and the decomposition layer number according to the real-time monitoring data;

and performing wavelet reconstruction on the low-frequency coefficient and the high-frequency coefficient after quantization processing to obtain smooth data.

8. A drilling downhole risk early warning device, comprising:

the data processing module is used for carrying out nonlinear fitting processing on the real-time monitoring data to obtain the change trend of the characteristic parameters in the future time length, carrying out risk analysis on the real-time monitoring data to obtain first early warning levels of the characteristic parameters in the future time length;

9. A drilling downhole risk early warning apparatus, comprising: a memory, a processor;

the memory to store executable instructions;

the processor configured to execute the executable instructions to implement the downhole drilling risk pre-warning method of any one of claims 1 to 7.

10. A computer readable storage medium having stored therein computer executable instructions for implementing the downhole drilling risk warning method of any one of claims 1 to 7 when executed by a processor.