Method and device for predicting blockage of tubular column

Abstract

The invention provides a method and device for predicting blocking of a tubular column, and the method comprises the steps: constructing a training sample based on the obtained historical well data of a to-be-logged well by combining the actual blocking condition corresponding to the historical trajectory data as a label; based on the training sample, constructing a target jamming prediction model; inputting real-time hook load data into the target blocking prediction model to obtain a corresponding blocking prediction result; based on the real-time hook load data, determining a static drill jamming coefficient and a dynamic drill jamming coefficient; and on the basis of the static drilling tool jamming coefficient, the dynamic drilling tool jamming coefficient and the jamming prevention prediction result, whether an early warning is given out or not is determined. By judging the change trend of the static drilling tool sticking coefficient and the dynamic drilling tool sticking coefficient and combining the current tripping working condition, drilling tool sticking possibly occurring in the future can be predicted and evaluated, so that drilling risks are effectively avoided, the probability and frequency of drilling engineering risks are reduced, and the well building period is shortened.

Description

Technical Field

[0001] The present invention relates to the technical field of oil drilling engineering, and particularly to a method and device for predicting pipe string sticking Background Art

[0002] During the process of running in or pulling out the pipe string during drilling, sticking accidents may occur. The occurrence of such accidents will lead to a lot of non-productive time, reduce the drilling efficiency, lengthen the drilling cycle and cause serious economic losses. Sticking is generally divided into differential sticking and mechanical sticking. The sticking caused by the wellbore quality is usually called mechanical sticking, while the sticking caused by the pressure difference between the wellbore pressure and the formation pressure is called differential sticking. Improper handling of sticking events may induce collapse sticking, drill pipe breakage or even burying of drill pipes and scrapping of drill pipes. The prediction and evaluation of sticking can effectively avoid the occurrence of drilling risks, reduce the probability and frequency of drilling engineering risks, and shorten the well construction cycle. Therefore, establishing the evaluation and early warning of sticking is of great significance for analyzing downhole conditions and making correct decisions.

[0003] Currently, the prediction method for pipe string sticking mainly relies on machine learning methods. By selecting different models to train the given data and selecting the model with high fitting degree to evaluate and predict sticking, but this method has high requirements for data, and the fitted model is not universal, and its prediction effect still has room for improvement. Summary of the Invention

[0004] The present invention provides a method and device for predicting pipe string sticking, which is used to identify the type and magnitude of sticking risks in the wellbore during the operation process and reduce non-productive events during the pipe string tripping operation in a data-driven manner.

[0005] In a first aspect, the present invention provides a method for predicting pipe string sticking, including:

[0006] Based on the historical well data of the well to be measured obtained, combined with the actual sticking situation corresponding to the historical trajectory data as a label, a training sample is constructed;

[0007] Based on the training sample, a target sticking prediction model is constructed;

[0008] The real-time hook load data is input into the target sticking prediction model to obtain the corresponding sticking prediction result;

[0009] Based on the real-time hook load data, a static sticking coefficient and a dynamic sticking coefficient are determined;

[0010] Based on the static sticking coefficient, the dynamic sticking coefficient and the sticking prediction result, it is determined whether to issue an early warning.

[0011] Optionally, based on the real-time hook load data, determining a static sticking coefficient and a dynamic sticking coefficient includes:

[0012] Extracting the hoisting and lowering hook load and the critical hook load during the hoisting and lowering of the traveling block from the real-time hook load data;

[0013] Calculating the static sticking coefficient according to the difference between the hoisting and lowering hook load and the critical hook load, compared with the difference between the hoisting and lowering hook load and the weight of the traveling block.

[0014] Optionally, based on the real-time hook load data, determining a static sticking coefficient and a dynamic sticking coefficient further includes:

[0015] Extracting the median value of the hoisting and lowering hook load of each stand in multiple stands from the real-time hook load data;

[0016] Fitting the median value of the hoisting and lowering hook load by a linear method to obtain the corresponding change slope;

[0017] Determining the dynamic sticking coefficient according to the change slope and the difference between the hoisting and lowering hook load and the weight of the traveling block.

[0018] Optionally, based on the static sticking coefficient, the dynamic sticking coefficient and the sticking prediction result, determining whether to issue an early warning includes:

[0019] Obtaining a static sticking alarm threshold value and a dynamic sticking alarm threshold value;

[0020] Successively determining whether the static sticking coefficient exceeds the static sticking alarm threshold value, and / or whether the dynamic sticking coefficient exceeds the dynamic sticking alarm threshold value, and / or whether the sticking prediction result is the existence of sticking; if so, an early warning is issued.

[0021] Optionally, based on the training samples, constructing a target sticking prediction model includes:

[0022] Inputting the training samples into an initially constructed sticking model to generate corresponding sticking categories;

[0023] Establishing a confusion matrix between the sticking categories and the actual sticking situations, and calculating the training error in the confusion matrix;

[0024] Based on the training error, adjusting the sticking model by a backpropagation algorithm to obtain optimal network parameters, and generating the target hierarchical sticking model by using the optimal network parameters.

[0025] In a second aspect, the present invention provides a prediction device for pipe string sticking, including:

[0026] A training sample construction module, configured to construct training samples based on the acquired historical well data of the well to be measured, in combination with the actual sticking situation corresponding to the historical trajectory data as labels.

[0027] A model construction module, configured to construct a target sticking prediction model based on the training samples.

[0028] A prediction module, configured to input real-time hook load data into the target sticking prediction model to obtain a corresponding sticking prediction result.

[0029] A sticking coefficient acquisition module, configured to determine a static sticking coefficient and a dynamic sticking coefficient based on the real-time hook load data.

[0030] A judgment module, configured to determine whether to issue a warning based on the static sticking coefficient, the dynamic sticking coefficient, and the sticking prediction result.

[0031] Optionally, the sticking coefficient acquisition module includes:

[0032] A first extraction sub-module, configured to extract the hoisting hook load and the critical hook load during the hoisting and lowering of the traveling block from the real-time hook load data.

[0033] A static sticking coefficient calculation sub-module, configured to calculate the static sticking coefficient according to the difference between the hoisting and lowering hook load and the critical hook load, and the difference between the hoisting and lowering hook load and the traveling block weight.

[0034] Optionally, the sticking coefficient acquisition module further includes:

[0035] A second extraction sub-module, configured to extract the median value of the hoisting and lowering hook load of each stand in multiple stands from the real-time hook load data.

[0036] A change slope determination sub-module, configured to fit the median value of the hoisting and lowering hook load by a linear method to obtain a corresponding change slope.

[0037] A dynamic sticking coefficient determination sub-module, configured to determine the dynamic sticking coefficient according to the change slope and the difference between the hoisting and lowering hook load and the traveling block weight.

[0038] In a third aspect, the present application provides an electronic device, including a processor and a memory, where the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the steps in the method provided in the first aspect above are run.

[0039] In a fourth aspect, the present application provides a storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps in the method provided in the first aspect above are run.

[0040] As can be seen from the above technical solutions, the present invention has the following advantages:

[0041] The present invention provides a method and device for predicting pipe string sticking. The method includes: constructing a training sample based on the historical well data of the well to be measured obtained, and combining the actual sticking situation corresponding to the historical trajectory data as a label; constructing a target sticking prediction model based on the training sample; inputting real-time hook load data into the target sticking prediction model to obtain a corresponding sticking prediction result; determining a static sticking coefficient and a dynamic sticking coefficient based on the real-time hook load data; and determining whether to issue a warning based on the static sticking coefficient, the dynamic sticking coefficient, and the sticking prediction result. By judging the change trends of the static sticking coefficient and the dynamic sticking coefficient and combining the current tripping operation conditions, it is possible to realize the prediction and evaluation of possible sticking in the future, thereby effectively avoiding the occurrence of drilling risks, reducing the probability and frequency of the occurrence of drilling engineering risks, and shortening the well construction period. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0043] Figure 1 It is a flow chart of the first embodiment of the method for predicting pipe string sticking of the present invention;

[0044] Figure 2 It is a flow chart of the second embodiment of the method for predicting pipe string sticking of the present invention;

[0045] Figure 3 It is a structural block diagram of the embodiment of the device for predicting pipe string sticking of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0046] The embodiments of the present invention provide a method and device for predicting pipe string sticking, which are used to identify the types and magnitudes of sticking risks in the wellbore during the operation process and reduce non-production events during the pipe string tripping operation in a data-driven manner.

[0047] In order to make the objectives, features, and advantages of the present invention more obvious and understandable, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the embodiments described below are only some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0048] Example 1. Please refer to Figure 1 , Figure 1 which is a flowchart of the steps of Example 1 of a prediction method for pipe string sticking in the present invention, including:

[0049] S101. Based on the historical well data of the well to be measured obtained, and combining the actual sticking situation corresponding to the historical trajectory data as a label, construct a training sample;

[0050] S102. Based on the training sample, construct a target sticking prediction model;

[0051] S103. Input the real-time hook load data into the target sticking prediction model to obtain the corresponding sticking prediction result;

[0052] S104. Based on the real-time hook load data, determine the static sticking coefficient and the dynamic sticking coefficient;

[0053] S105. Based on the static sticking coefficient, the dynamic sticking coefficient and the sticking prediction result, determine whether to issue a warning.

[0054] A prediction method for pipe string sticking provided by an embodiment of the present invention includes: based on the historical well data of the well to be measured obtained, and combining the actual sticking situation corresponding to the historical trajectory data as a label, construct a training sample; based on the training sample, construct a target sticking prediction model; input the real-time hook load data into the target sticking prediction model to obtain the corresponding sticking prediction result; based on the real-time hook load data, determine the static sticking coefficient and the dynamic sticking coefficient; based on the static sticking coefficient, the dynamic sticking coefficient and the sticking prediction result, determine whether to issue a warning. By judging the change trends of the static sticking coefficient and the dynamic sticking coefficient, and combining the current tripping operation conditions, it is possible to realize the prediction and evaluation of possible sticking in the future, thereby effectively avoiding the occurrence of drilling risks, reducing the probability and frequency of the occurrence of drilling engineering risks, and shortening the well construction period.

[0055] Example 2. Please refer to Figure 2 , Figure 2 which is a flowchart of the steps of Example 2 of a prediction method for pipe string sticking in the present invention, including:

[0056] S201. Based on the historical well data of the well to be measured obtained, and combining the actual sticking situation corresponding to the historical trajectory data as a label, construct a training sample;

[0057] In the embodiment of the present invention, historical well data of the well to be measured under the condition of no sticking and historical well data under the condition of sticking (including time series, hook load, bit depth, hook weight, tripping load, etc.) are searched, and according to the corresponding actual sticking situation, after normalizing the historical well data, training samples are constructed.

[0058] S202, input the training samples into the initially constructed sticking model to generate corresponding sticking categories;

[0059] S203, establish a confusion matrix between the sticking categories and the actual sticking situation, and calculate the training error in the confusion matrix;

[0060] S204, based on the training error, adjust the sticking model through the backpropagation algorithm to obtain optimal network parameters, and use the optimal network parameters to generate the target hierarchical sticking model;

[0061] In the embodiment of the present invention, by establishing a confusion matrix between the expected result (sticking category) and the actual result (actual sticking situation), calculate the percentages of false positives and false negatives in the confusion matrix. Use the percentages of false positives and false negatives as the training error, and continuously modify the parameters of the sticking model according to the training error until the percentages of FP and FN are less than 1%.

[0062] A confusion matrix is a tabular form used in machine learning and statistics to evaluate the performance of a classification model. It compares the predicted results with the true results and classifies them into four different categories: true positives, true negatives, false positives, and false negatives.

[0063] Among them, true positives represent the number of positive examples (positive samples) correctly classified as positive examples by the model, true negatives represent the number of negative examples (negative samples) correctly classified as negative examples by the model, false positives represent the number of negative examples misclassified as positive examples by the model, and false negatives represent the number of positive examples misclassified as negative examples by the model.

[0064] The confusion matrix provides detailed information about the performance of the model and can be used to calculate a series of metrics, such as accuracy, recall, precision, and F1 score, etc., to evaluate the classification ability of the model on different categories.

[0065] S205, input the real-time hook load data into the target sticking prediction model to obtain the corresponding sticking prediction result;

[0066] S206, extract the tripping hook load and critical hook load during the hook tripping from the real-time hook load data;

[0067] S207. Calculate the static sticking coefficient based on the difference between the tripping hook load and the critical hook load, compared with the difference between the tripping hook load and the weight of the traveling block.

[0068] In the embodiment of the present invention, the hook load (tripping hook load) at the maximum speed value and the minimum hook load (critical hook load) during tripping of the traveling block are obtained, and the static sticking coefficient is calculated based on the difference between the tripping hook load and the critical hook load compared with the difference between the tripping hook load and the weight of the traveling block.

[0069] S208. Extract the median value of the tripping hook load for each stand in multiple stands from the real-time hook load data.

[0070] S209. Fit the median value of the tripping hook load by a linear method to obtain the corresponding change slope.

[0071] S210. Determine the dynamic sticking coefficient based on the change slope and the difference between the tripping hook load and the weight of the traveling block.

[0072] In the embodiment of the present invention, the slope of the change in the median value of the tripping hook load for each stand in multiple stands over a period of time is fitted by a linear method. The dynamic sticking coefficient is recorded by multiplying the slope by the depth change and dividing by the difference between the tripping hook load and the weight of the traveling block, and this coefficient is used to evaluate the loss of dynamic hook load.

[0073] S211. Obtain the static sticking alarm threshold value and the dynamic sticking alarm threshold value.

[0074] S212. Determine in sequence whether the static sticking coefficient exceeds the static sticking alarm threshold value, and / or whether the dynamic sticking coefficient exceeds the dynamic sticking alarm threshold value, and / or whether the sticking prediction result is the existence of sticking; if so, issue a warning.

[0075] In specific implementation, the real-time hook load data on site is used as an input parameter and input into the generated target sticking prediction model to obtain the corresponding sticking prediction result. By judging the static sticking coefficient and the preset static sticking coefficient alarm threshold value, as well as the dynamic sticking coefficient and the preset dynamic sticking coefficient alarm threshold value, the real-time sticking risk is judged, and then it is determined whether to issue a warning.

[0076] A method for predicting pipe string sticking in an embodiment of the present invention includes: constructing a training sample based on the historical well data of a well to be measured obtained, and combining the actual sticking situation corresponding to the historical trajectory data as a label; constructing a target sticking prediction model based on the training sample; inputting real-time hook load data into the target sticking prediction model to obtain a corresponding sticking prediction result; determining a static sticking coefficient and a dynamic sticking coefficient based on the real-time hook load data; and determining whether to issue a warning based on the static sticking coefficient, the dynamic sticking coefficient, and the sticking prediction result. By judging whether the static sticking coefficient exceeds the static sticking alarm threshold value, whether the dynamic sticking coefficient exceeds the dynamic sticking alarm threshold value, and whether the sticking prediction result indicates sticking, the prediction and evaluation of possible sticking in the future can be realized, so as to effectively avoid the occurrence of drilling risks, reduce the probability and frequency of the occurrence of drilling engineering risks, and shorten the well construction period.

[0077] Embodiment 3, please refer to Figure 3 , Figure 3 which is a structural block diagram of an embodiment of a device for predicting pipe string sticking of the present invention, and includes:

[0078] A training sample construction module 301, configured to construct a training sample based on the historical well data of a well to be measured obtained, and combining the actual sticking situation corresponding to the historical trajectory data as a label;

[0079] A model construction module 302, configured to construct a target sticking prediction model based on the training sample;

[0080] A prediction module 303, configured to input real-time hook load data into the target sticking prediction model to obtain a corresponding sticking prediction result;

[0081] A sticking coefficient acquisition module 304, configured to determine a static sticking coefficient and a dynamic sticking coefficient based on the real-time hook load data;

[0082] A judgment module 305, configured to determine whether to issue a warning based on the static sticking coefficient, the dynamic sticking coefficient, and the sticking prediction result.

[0083] In an optional embodiment, the sticking coefficient acquisition module 304 includes:

[0084] A first extraction sub-module, configured to extract the hoisting and lowering hook load and the critical hook load during the hoisting and lowering of the traveling block from the real-time hook load data;

[0085] A static sticking coefficient calculation sub-module, configured to calculate the static sticking coefficient according to the difference between the hoisting and lowering hook load and the critical hook load, and the difference between the hoisting and lowering hook load and the weight of the traveling block.

[0086] In an alternative embodiment, the sticking coefficient acquisition module 304 further includes:

[0087] A second extraction sub-module, configured to extract the median value of the hook load during tripping for each stand in multiple stands from the real-time hook load data;

[0088] A change slope determination sub-module, configured to fit the median value of the hook load during tripping by a linear method to obtain a corresponding change slope;

[0089] A dynamic sticking coefficient determination sub-module, configured to determine the dynamic sticking coefficient according to the change slope and the difference between the hook load during tripping and the weight of the traveling block.

[0090] In an alternative embodiment, the judgment module 305 includes:

[0091] A threshold value acquisition sub-module, configured to acquire a static sticking alarm threshold value and a dynamic sticking alarm threshold value;

[0092] A judgment sub-module, configured to sequentially judge whether the static sticking coefficient exceeds the static sticking alarm threshold value, and / or whether the dynamic sticking coefficient exceeds the dynamic sticking alarm threshold value, and / or whether the sticking prediction result is that sticking exists; if so, a warning is issued.

[0093] In an alternative embodiment, the model construction module 302 includes:

[0094] An input sub-module, configured to input the training sample into an initially constructed sticking model to generate a corresponding sticking category;

[0095] An error determination sub-module, configured to establish a confusion matrix between the sticking category and the actual sticking situation, and calculate the training error in the confusion matrix;

[0096] A target model construction sub-module, configured to adjust the sticking model by a backpropagation algorithm based on the training error to obtain optimal network parameters, and generate the target hierarchical sticking model by using the optimal network parameters.

[0097] Embodiment 4. The embodiment of the present invention further provides an electronic device, including a memory and a processor. A computer program is stored in the memory. When the computer program is executed by the processor, the processor executes the steps of a prediction method for pipe string sticking in any one of the embodiments, including:

[0098] Based on the historical well data of the well to be measured obtained, a training sample is constructed by combining the actual sticking situation corresponding to the historical trajectory data as a label;

[0099] Based on the training sample, a target sticking prediction model is constructed.

[0100] Input the real-time hook load data into the target sticking prediction model to obtain the corresponding sticking prediction result;

[0101] Based on the real-time hook load data, determine the static sticking coefficient and the dynamic sticking coefficient;

[0102] Based on the static sticking coefficient, the dynamic sticking coefficient and the sticking prediction result, determine whether to issue a warning.

[0103] In an alternative embodiment, based on the real-time hook load data, determining the static sticking coefficient and the dynamic sticking coefficient includes:

[0104] Extract the hoisting and lowering hook load and the critical hook load during the hoisting and lowering of the traveling block from the real-time hook load data;

[0105] According to the difference between the hoisting and lowering hook load and the critical hook load, and the difference between the hoisting and lowering hook load and the traveling block weight, calculate the static sticking coefficient.

[0106] In an alternative embodiment, based on the real-time hook load data, determining the static sticking coefficient and the dynamic sticking coefficient further includes:

[0107] Extract the median value of the hoisting and lowering hook load of each stand in multiple stands from the real-time hook load data;

[0108] By a linear method, fit the median value of the hoisting and lowering hook load to obtain the corresponding change slope;

[0109] According to the change slope and the difference between the hoisting and lowering hook load and the traveling block weight, determine the dynamic sticking coefficient.

[0110] In an alternative embodiment, based on the static sticking coefficient, the dynamic sticking coefficient and the sticking prediction result, determining whether to issue a warning includes:

[0111] Obtain the static sticking alarm threshold value and the dynamic sticking alarm threshold value;

[0112] Successively determine whether the static sticking coefficient exceeds the static sticking alarm threshold value, and / or whether the dynamic sticking coefficient exceeds the dynamic sticking alarm threshold value, and / or whether the sticking prediction result is that sticking exists; if so, issue a warning.

[0113] In an alternative embodiment, based on the training samples, constructing the target sticking prediction model includes:

[0114] Input the training samples into the initially constructed sticking model to generate the corresponding sticking categories;

[0115] Establish a confusion matrix for the stuck pipe category and the actual stuck pipe situation, and calculate the training error in the confusion matrix;

[0116] Based on the training error, adjust the stuck pipe model through the backpropagation algorithm to obtain the optimal network parameters, and use the optimal network parameters to generate the target hierarchical stuck pipe model.

[0117] Embodiment 5. The embodiment of the present invention also provides a computer storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of a prediction method for pipe string sticking in any embodiment are implemented, including:

[0118] Based on the historical well data of the well to be measured obtained, combined with the actual stuck pipe situation corresponding to the historical trajectory data as a label, construct a training sample;

[0119] Based on the training sample, construct a target stuck pipe prediction model;

[0120] Input the real-time hook load data into the target stuck pipe prediction model to obtain the corresponding stuck pipe prediction result;

[0121] Based on the real-time hook load data, determine the static sticking coefficient and the dynamic sticking coefficient;

[0122] Based on the static sticking coefficient, the dynamic sticking coefficient and the stuck pipe prediction result, determine whether to issue a warning.

[0123] In an alternative embodiment, based on the real-time hook load data, determining the static sticking coefficient and the dynamic sticking coefficient includes:

[0124] Extract the hoisting hook load and the critical hook load during the hoisting and lowering of the traveling block from the real-time hook load data;

[0125] According to the difference between the hoisting and lowering hook load and the critical hook load, and the difference between the hoisting and lowering hook load and the weight of the traveling block, calculate the static sticking coefficient.

[0126] In an alternative embodiment, based on the real-time hook load data, determining the static sticking coefficient and the dynamic sticking coefficient further includes:

[0127] Extract the median value of the hoisting and lowering hook load of each stand in multiple stands from the real-time hook load data;

[0128] By a linear method, fit the median value of the hoisting and lowering hook load to obtain the corresponding change slope;

[0129] According to the change slope and the difference between the hoisting and lowering hook load and the weight of the traveling block, determine the dynamic sticking coefficient.

[0130] In an alternative embodiment, based on the static sticking coefficient, the dynamic sticking coefficient, and the sticking prediction result, determining whether to issue a warning includes:

[0131] Obtaining a static sticking alarm threshold value and a dynamic sticking alarm threshold value;

[0132] Sequentially determining whether the static sticking coefficient exceeds the static sticking alarm threshold value, and / or whether the dynamic sticking coefficient exceeds the dynamic sticking alarm threshold value, and / or whether the sticking prediction result indicates sticking; if so, issuing a warning.

[0133] In an alternative embodiment, constructing a target sticking prediction model based on the training samples includes:

[0134] Inputting the training samples into an initially constructed sticking model to generate corresponding sticking categories;

[0135] Establishing a confusion matrix between the sticking categories and the actual sticking situations, and calculating the training error in the confusion matrix;

[0136] Based on the training error, adjusting the sticking model through the backpropagation algorithm to obtain optimal network parameters, and using the optimal network parameters to generate the target hierarchical sticking model.

[0137] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0138] In several embodiments provided in the present application, it should be understood that the methods, devices, electronic devices, and storage media disclosed in the present invention can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division, and there may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the couplings or direct couplings or communication connections shown or discussed with each other can be through some interfaces, and the indirect couplings or communication connections of the devices or units can be in electrical, mechanical, or other forms.

[0139] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0140] In addition, in each embodiment of the present invention, each functional unit may be integrated into a processing unit, may exist separately physically for each unit, 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 software functional unit.

[0141] If the above-mentioned integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it may be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, may be embodied in the form of a software product. This computer software product is stored in a readable storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in each embodiment of the present invention. The foregoing readable storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical discs that can store program codes.

[0142] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of each embodiment of the present invention.

Claims

1. A prediction method for pipe sticking, characterized in that Including: Based on the historical well data of the well to be measured obtained, combined with the actual sticking situation corresponding to the historical trajectory data as a label, construct a training sample; Based on the training sample, construct a target sticking prediction model; Input the real-time hook load data into the target sticking prediction model to obtain the corresponding sticking prediction result; Based on the real-time hook load data, determine the static sticking coefficient and the dynamic sticking coefficient; Based on the static sticking coefficient, the dynamic sticking coefficient and the sticking prediction result, determine whether to issue a warning.

2. The prediction method for pipe string sticking according to claim 1, characterized in that Based on the real-time hook load data, determining the static sticking coefficient and the dynamic sticking coefficient includes: Extract the hoisting and lowering hook load and the critical hook load during the hoisting and lowering of the traveling block from the real-time hook load data; According to the difference between the hoisting and lowering hook load and the critical hook load, and the difference between the hoisting and lowering hook load and the traveling block weight, calculate the static sticking coefficient.

3. The prediction method for pipe string sticking according to claim 2, wherein Based on the real-time hook load data, determining the static sticking coefficient and the dynamic sticking coefficient further includes: Extract the median value of the hoisting and lowering hook load of each stand in multiple stands from the real-time hook load data; By a linear method, fit the median value of the hoisting and lowering hook load to obtain the corresponding change slope; According to the change slope and the difference between the hoisting and lowering hook load and the traveling block weight, determine the dynamic sticking coefficient.

4. The prediction method for pipe string sticking according to claim 1, characterized in that, Based on the static sticking coefficient, the dynamic sticking coefficient and the sticking prediction result, determining whether to issue a warning includes: Obtain the static sticking alarm threshold value and the dynamic sticking alarm threshold value; Successively judge whether the static sticking coefficient exceeds the static sticking alarm threshold value, and / or whether the dynamic sticking coefficient exceeds the dynamic sticking alarm threshold value, and / or whether the sticking prediction result is that sticking exists; if so, issue a warning.

5. The prediction method for pipe string sticking according to claim 1, characterized in that, Based on the training sample, constructing a target sticking prediction model includes: Input the training sample into the initially constructed sticking model to generate the corresponding sticking category; Establish a confusion matrix between the sticking category and the actual sticking situation, and calculate the training error in the confusion matrix; Based on the training error, adjust the sticking model by the backpropagation algorithm to obtain the optimal network parameters, and use the optimal network parameters to generate the target hierarchical sticking model.

6. A prediction device for pipe string sticking, characterized in that, Including: A training sample construction module for constructing a training sample based on the historical well data of the well to be measured obtained, combined with the actual sticking situation corresponding to the historical trajectory data as a label; A model construction module for constructing a target sticking prediction model based on the training sample; A prediction module for inputting the real-time hook load data into the target sticking prediction model to obtain the corresponding sticking prediction result; A sticking coefficient acquisition module for determining the static sticking coefficient and the dynamic sticking coefficient based on the real-time hook load data; A judgment module for determining whether to issue a warning based on the static sticking coefficient, the dynamic sticking coefficient and the sticking prediction result.

7. The prediction device for pipe string sticking according to claim 6, wherein The sticking coefficient acquisition module includes: A first extraction sub-module for extracting the hoisting and lowering hook load and the critical hook load during the hoisting and lowering of the traveling block from the real-time hook load data; A static sticking coefficient calculation sub-module for calculating the static sticking coefficient according to the difference between the tripping hook load and the critical hook load and the difference between the tripping hook load and the weight of the traveling block.

8. The predicting device for pipe string sticking according to claim 7, characterized in that, The sticking coefficient acquisition module further includes: A second extraction sub-module for extracting the median value of the tripping hook load of each stand in multiple stands from the real-time hook load data; A change slope determination sub-module for fitting the median value of the tripping hook load by a linear method to obtain the corresponding change slope; A dynamic sticking coefficient determination sub-module for determining the dynamic sticking coefficient according to the change slope and the difference between the tripping hook load and the weight of the traveling block.

9. An electronic device, characterized in that, It includes a processor and a memory, and the memory stores computer-readable instructions. When the computer-readable instructions are executed by the processor, the method according to any one of claims 1-5 is run.

10. A storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, the method according to any one of claims 1-5 is run.

Images