Well drilling jamming early warning method, device and equipment and readable storage medium

Abstract

The invention discloses a drilling tool jamming early warning method. The method comprises the steps that a current comprehensive logging parameter set at the current moment is obtained; comparing the current comprehensive logging parameter group with a preset parameter threshold group to obtain a threshold diagnosis result; acquiring a historical comprehensive logging parameter group of each preset time period before the current comprehensive logging parameter group; calculating the change trend of each comprehensive logging parameter according to the current comprehensive logging parameter group and each historical comprehensive logging parameter group to obtain a parameter change trend diagnosis result; performing drilling jamming fault diagnosis according to the comprehensive logging parameter group by utilizing a jamming fault diagnosis model to obtain a model diagnosis result; and when it is determined that fault early warning needs to be carried out according to the threshold value diagnosis result, the parameter change trend diagnosis result and the model diagnosis result, drilling jamming early warning is carried out. According to the method and the device, intelligent and real-time judgment of the drill jamming fault is realized, and the timeliness of risk early warning is improved. The invention further discloses a device, equipment and a storage medium, which have corresponding technical effects.

Description

Technical Field

[0001] The present application relates to the technical field of deep shale gas complex formation drilling, and particularly to a drilling sticking warning method, device, equipment and computer-readable storage medium. Background Technique

[0002] In recent years, Sinopec has accelerated the exploration and development of shale gas resources. The formation pressure coefficient in the development work area is high and the drilling geological conditions are complex, resulting in frequent drilling complications and failures. Among them, the sticking failure is the most prominent, which seriously restricts the safe and efficient development of shale gas.

[0003] The existing sticking failure identification is mainly an expert knowledge system, or the engineering anomalies are analyzed by using conventional logging data. Moreover, only a small amount of measurement and calculation work is automatically completed by the computer, and most of the analysis and judgment rely on manual work. Due to reasons such as individual knowledge, experience and responsibility differences, it often leads to the inability to detect and handle drilling anomalies and complex situations in a timely manner, increasing the risk. In fact, it is not very realistic to require the operator to concentrate on observing the changes in the monitoring data and quickly judge the accident. Therefore, the existing sticking identification technology has defects such as poor comprehensive utilization ability of monitoring information, insufficient timely risk warning, and too strong subjectivity.

[0004] In summary, how to effectively solve the problems of poor comprehensive utilization ability of monitoring information, insufficient timely risk warning, and too strong subjectivity in the existing sticking failure identification method is an urgent problem to be solved by those skilled in the art at present. Summary of the Invention

[0005] The purpose of the present application is to provide a drilling sticking warning method, which realizes the intelligent and real-time judgment of sticking failures and improves the timeliness of risk warning; another purpose of the present application is to provide a drilling sticking warning device, equipment and computer-readable storage medium.

[0006] To solve the above technical problems, the present application provides the following technical solutions:

[0007] A drilling sticking warning method includes:

[0008] Obtain the current comprehensive logging parameter group of this time;

[0009] Compare the current comprehensive logging parameter group of this time with a preset parameter threshold group to obtain a threshold diagnosis result;

[0010] Obtain the historical comprehensive logging parameter groups of each preset time period before the current comprehensive logging parameter group of this time;

[0011] Calculate the change trend of each comprehensive logging parameter according to the current comprehensive logging parameter group of this time and each historical comprehensive logging parameter group to obtain a parameter change trend diagnosis result;

[0012] Use the stuck-pipe fault diagnosis model to perform stuck-pipe fault diagnosis on the current comprehensive mud logging parameter group to obtain the model diagnosis result;

[0013] When it is determined that fault warning is required based on the threshold diagnosis result, the parameter change trend diagnosis result, and the model diagnosis result, perform stuck-pipe warning for drilling.

[0014] In a specific implementation manner of the present application, comparing the current comprehensive mud logging parameter group with a preset parameter threshold group includes:

[0015] Obtain the current drilling working condition; wherein, the current drilling working condition is one of tripping out, tripping in, and drilling.

[0016] Compare the current comprehensive mud logging parameter group with the preset parameter threshold group corresponding to the current drilling working condition; wherein, the preset parameter threshold group is calculated based on the frictional resistance value and the comprehensive mud logging parameter group when a stuck-pipe fault occurs under the current drilling working condition.

[0017] In a specific implementation manner of the present application, using the stuck-pipe fault diagnosis model to perform stuck-pipe fault diagnosis on the current comprehensive mud logging parameter group includes:

[0018] Use the stuck-pipe fault diagnosis model optimized by the particle swarm optimization algorithm for the BP neural network to perform stuck-pipe fault diagnosis on the current comprehensive mud logging parameter group.

[0019] In a specific implementation manner of the present application, calculating the change trend of each comprehensive mud logging parameter based on the current comprehensive mud logging parameter group and each historical comprehensive mud logging parameter group to obtain the parameter change trend diagnosis result includes:

[0020] For each current comprehensive mud logging parameter in the current comprehensive mud logging parameter group, calculate the first moving average corresponding to the first sliding window length and the second moving average corresponding to the second sliding window length respectively according to the current comprehensive mud logging parameter and the corresponding historical comprehensive mud logging parameters in each historical comprehensive mud logging parameter group.

[0021] Calculate the moving average deviation according to the first moving average and the second moving average.

[0022] Use the sliding window linear regression function to calculate the first slope value of the regression of the current comprehensive mud logging parameter and the corresponding historical comprehensive mud logging parameters in the first sliding window length at the current moment according to the moving average deviation, and calculate the second slope value of the regression of the current comprehensive mud logging parameter and the corresponding historical comprehensive mud logging parameters in the second sliding window length at the current moment.

[0023] Obtain first weighting factors corresponding to the respective first slope values, and calculate a first average local slope corresponding to the first sliding window length according to the respective first slope values and the respective first weighting factors;

[0024] Obtain second weighting factors corresponding to the respective second slope values, and calculate a second average local slope corresponding to the second sliding window length according to the respective second slope values and the respective second weighting factors;

[0025] Calculate a difference in the current average local slope at the current moment according to the first average local slope and the second average local slope;

[0026] Obtain historical differences in average local slope corresponding to respective preset time periods before the current moment;

[0027] Determine a diagnostic result of the parameter change trend according to the difference in the current average local slope and the respective historical differences in average local slope.

[0028] In a specific implementation manner of the present application, determining the diagnostic result of the parameter change trend according to the difference in the current average local slope and the respective historical differences in average local slope includes:

[0029] Judge whether the average local slope in each subsequent time period is greater than the average local slope in the previous time period in each adjacent time period, and whether each average local slope is greater than 0, according to the difference in the current average local slope and the respective historical differences in average local slope;

[0030] If so, determine that a fault warning needs to be given.

[0031] In a specific implementation manner of the present application, before comparing the current comprehensive mud logging parameter group with a preset parameter threshold group after obtaining the current comprehensive mud logging parameter group, it further includes:

[0032] Judge whether there is a current comprehensive mud logging parameter in the current comprehensive mud logging parameter group that is an abnormal jump value;

[0033] If so, eliminate the current comprehensive mud logging parameter that is an abnormal jump value.

[0034] In a specific implementation manner of the present application, before diagnosing a drilling sticking fault according to the current comprehensive mud logging parameter group by using a drilling sticking fault diagnosis model, it further includes:

[0035] Normalize each current comprehensive mud logging parameter in the current comprehensive mud logging parameter group.

[0036] A drilling sticking warning device includes:

[0037] The current parameter group acquisition module is used to acquire the current comprehensive mud logging parameter group at the current moment;

[0038] The threshold diagnosis result acquisition module is used to compare the current comprehensive mud logging parameter group with a preset parameter threshold group to obtain a threshold diagnosis result;

[0039] The historical parameter group acquisition module is used to acquire the historical comprehensive mud logging parameter groups for each preset time period before the current comprehensive mud logging parameter group;

[0040] The change trend diagnosis result acquisition module is used to calculate the change trends of the comprehensive mud logging parameters based on the current comprehensive mud logging parameter group and each historical comprehensive mud logging parameter group to obtain a parameter change trend diagnosis result;

[0041] The model diagnosis result acquisition module is used to perform drilling sticking fault diagnosis on the current comprehensive mud logging parameter group using a sticking fault diagnosis model to obtain a model diagnosis result;

[0042] The sticking warning module is used to perform drilling sticking warning when it is determined that a fault warning is required based on the threshold diagnosis result, the parameter change trend diagnosis result, and the model diagnosis result.

[0043] A drilling sticking warning device, comprising:

[0044] A memory for storing a computer program;

[0045] A processor for implementing the steps of the drilling sticking warning method as described above when executing the computer program.

[0046] A computer-readable storage medium having a computer program stored thereon, and the computer program, when executed by a processor, implements the steps of the drilling sticking warning method as described above.

[0047] The drilling sticking warning method provided by this application acquires the current comprehensive mud logging parameter group at the current moment; compares the current comprehensive mud logging parameter group with a preset parameter threshold group to obtain a threshold diagnosis result; acquires the historical comprehensive mud logging parameter groups for each preset time period before the current comprehensive mud logging parameter group; calculates the change trends of the comprehensive mud logging parameters based on the current comprehensive mud logging parameter group and each historical comprehensive mud logging parameter group to obtain a parameter change trend diagnosis result; performs drilling sticking fault diagnosis on the current comprehensive mud logging parameter group using a sticking fault diagnosis model to obtain a model diagnosis result; and performs drilling sticking warning when it is determined that a fault warning is required based on the threshold diagnosis result, the parameter change trend diagnosis result, and the model diagnosis result.

[0048] As can be seen from the above technical solutions, after obtaining the current comprehensive mud logging parameter group, the stuck pipe fault diagnosis is carried out according to the current comprehensive mud logging parameter group by combining threshold diagnosis, abnormal diagnosis of parameter change trend, and artificial intelligence diagnosis using the stuck pipe fault diagnosis model, realizing the intelligent and real-time judgment of the stuck pipe fault, improving the comprehensive utilization ability of monitoring information, enhancing the timeliness of risk warning, avoiding the influence of subjective consciousness, and improving the accuracy of the stuck pipe fault warning.

[0049] Correspondingly, the present application also provides a stuck pipe warning device, equipment, and computer-readable storage medium corresponding to the above-mentioned stuck pipe warning method for drilling, which have the above technical effects and will not be elaborated herein. Description of the Drawings

[0050] In order to more clearly illustrate the technical solutions in the embodiments of the present application 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 following drawings are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0051] Figure 1 It is a flowchart of an implementation of the stuck pipe warning method for drilling in an embodiment of the present application;

[0052] Figure 2 It is another flowchart of an implementation of the stuck pipe warning method for drilling in an embodiment of the present application;

[0053] Figure 3 It is a real-time stuck pipe warning flowchart based on abnormal diagnosis of parameter change trend in an embodiment of the present application;

[0054] Figure 4 It is a flowchart of steps of a particle swarm optimization BP neural network algorithm in an embodiment of the present application;

[0055] Figure 5 It is a stuck pipe fault warning BP neural network structure diagram in an embodiment of the present application;

[0056] Figure 6 It is a schematic diagram of a stuck pipe fault warning BP neural network optimized based on a particle swarm algorithm in an embodiment of the present application;

[0057] Figure 7 It is a schematic diagram of the structure of a stuck pipe warning system for drilling in an embodiment of the present application;

[0058] Figure 8 It is a schematic diagram of the structure of the downhole part of a stuck pipe warning system for drilling in an embodiment of the present application;

[0059] Figure 9It is a structural block diagram of a drilling stuck - pipe warning device in an embodiment of the present application;

[0060] Figure 10 It is a structural block diagram of a drilling stuck - pipe warning equipment in an embodiment of the present application;

[0061] Figure 11 It is a specific structural schematic diagram of a drilling stuck - pipe warning equipment provided by an embodiment of the present application.

[0062] The markings in the attached drawings are as follows:

[0063] 1 - Drill bit, 2 - Downhole part of the drilling stuck - pipe warning system, 3 - Comprehensive mud logging instrument, 4 - Surface part of the drilling stuck - pipe warning system, 5 - Surface receiving instrument for downhole measurement data, 21 - Core barrel, 22 - Downhole parameter measurer, 23 - Downhole data processor, 24 - Pulse generator, 25 - Current - limiting ring, 26 - Outer barrel. Detailed implementation manners

[0064] In order to enable those skilled in the art to better understand the solution of the present application, the present application will be further described in detail below with reference to the accompanying drawings and specific implementation manners. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without creative efforts belong to the scope of protection of the present application.

[0065] See Figure 1 , Figure 1 It is an implementation flowchart of a drilling stuck - pipe warning method in an embodiment of the present application. This method may include the following steps:

[0066] S101: Obtain the current set of comprehensive mud logging parameters at the current moment.

[0067] During the drilling process, monitor the comprehensive mud logging parameters to obtain the current set of comprehensive mud logging parameters at the current moment. The set of comprehensive mud logging parameters may include hook load, standpipe pressure, torque, rate of penetration, rotary speed, weight on bit, etc.

[0068] S102: Compare the current set of comprehensive mud logging parameters with a preset set of parameter thresholds to obtain a threshold diagnosis result.

[0069] Pre-set a parameter threshold group, where the parameter threshold group contains the parameter thresholds corresponding to each comprehensive mud logging parameter in the comprehensive mud logging parameter group. After obtaining the current comprehensive mud logging parameter group at the current moment, compare the current comprehensive mud logging parameter group with the pre-set parameter threshold group to determine whether there is a comprehensive mud logging parameter exceeding the corresponding pre-set parameter threshold range, obtain a determination result, and determine a threshold diagnosis result according to the determination result. The threshold diagnosis result can include two items, one is whether there is a comprehensive mud logging parameter exceeding the corresponding pre-set parameter threshold range, and the other is the number of comprehensive mud logging parameters exceeding the corresponding pre-set parameter threshold range.

[0070] S103: Obtain the historical comprehensive mud logging parameter groups for each pre-set time period before the current comprehensive mud logging parameter group.

[0071] The change trend of the fault symptoms corresponding to the stuck-pipe fault needs to be determined by analyzing the measurement data over a certain period of time, rather than a single value at the current time point. The abnormal change of the parameter is reflected in the curve shape as an increase or decrease in the curve slope. When a stuck-pipe occurs, it is prominently manifested as floating changes in the hook load, torque, and standpipe pressure. To determine the change trend of the comprehensive mud logging parameters, it is necessary to analyze the comprehensive mud logging parameter groups for multiple time periods. After obtaining the current comprehensive mud logging parameter group at the current moment, obtain the historical comprehensive mud logging parameter groups for each pre-set time period before the current comprehensive mud logging parameter group.

[0072] S104: Calculate the change trend of each comprehensive mud logging parameter based on the current comprehensive mud logging parameter group and each historical comprehensive mud logging parameter group to obtain a parameter change trend diagnosis result.

[0073] After obtaining the current comprehensive mud logging parameter group and each historical comprehensive mud logging parameter group, calculate the change trend of each comprehensive mud logging parameter based on the current comprehensive mud logging parameter group and each historical comprehensive mud logging parameter group to obtain a parameter change trend diagnosis result.

[0074] S105: Use the stuck-pipe fault diagnosis model to perform stuck-pipe fault diagnosis on the current comprehensive mud logging parameter group to obtain a model diagnosis result.

[0075] Pre-train a stuck-pipe fault diagnosis model for performing stuck-pipe diagnosis based on comprehensive mud logging parameters. After obtaining the current comprehensive mud logging parameter group at the current moment, use the stuck-pipe fault diagnosis model to perform stuck-pipe fault diagnosis on the current comprehensive mud logging parameter group to obtain a model diagnosis result.

[0076] In a specific implementation manner of the present application, before step S105, the method may further include the following steps:

[0077] Normalize each current comprehensive mud logging parameter in the current comprehensive mud logging parameter group.

[0078] After obtaining the current comprehensive mud logging parameter group and each historical comprehensive mud logging parameter group, before using the stuck pipe fault diagnosis model to diagnose the stuck pipe fault according to the current comprehensive mud logging parameter group, normalize each current comprehensive mud logging parameter in the current comprehensive mud logging parameter group. The calculation formula is as follows:

[0079]

[0080] Where x min is the minimum value of the same comprehensive mud logging parameter within the preset time interval, and x max is the maximum value of the same comprehensive mud logging parameter within the preset time interval. x k is the normalized parameter value.

[0081] By normalizing the comprehensive mud logging parameters, the problem of increased error in the network prediction results caused by the large difference in the order of magnitude of the input and output data is avoided.

[0082] S106: When it is determined that a fault warning is required based on the threshold diagnosis result, the parameter change trend diagnosis result, and the model diagnosis result, issue a stuck pipe warning for drilling.

[0083] During the drilling process, if the drilling tools in the well cannot be lifted or lowered, and in severe cases, cannot even be rotated, this complex situation is called stuck pipe.

[0084] After obtaining the threshold diagnosis result, the parameter change trend diagnosis result, and the model diagnosis result respectively, combine the above three diagnosis results to determine whether a fault warning is required. Generally, it is set that as long as it is determined that a fault warning is required according to one of the three diagnosis results, it is confirmed that there is a stuck pipe risk and a fault warning is required. When it is determined that a fault warning is required based on the threshold diagnosis result, the parameter change trend diagnosis result, and the model diagnosis result, issue a stuck pipe warning for drilling. It realizes the intelligent and real-time judgment of the stuck pipe fault, improves the comprehensive utilization ability of monitoring information, improves the timeliness of risk warning, avoids the influence of subjective consciousness, improves the accuracy of the stuck pipe fault warning, and reduces the fault loss.

[0085] As can be seen from the above technical solution, after obtaining the current comprehensive mud logging parameter group at the current moment, by combining threshold diagnosis, abnormal diagnosis of parameter change trends, and artificial intelligence diagnosis using the stuck pipe fault diagnosis model to diagnose the stuck pipe fault according to the current comprehensive mud logging parameter group, it realizes the intelligent and real-time judgment of the stuck pipe fault, improves the comprehensive utilization ability of monitoring information, improves the timeliness of risk warning, avoids the influence of subjective consciousness, and improves the accuracy of the stuck pipe fault warning.

[0086] It should be noted that based on the above embodiments, the embodiments of the present application also provide corresponding improvement solutions. In the subsequent embodiments, for the same steps or corresponding steps involved in the above embodiments, reference can be made to each other, and the corresponding beneficial effects can also be referred to each other, and will not be repeated one by one in the following improved embodiments.

[0087] See Figure 2 , Figure 2 which is another implementation flowchart of the drilling stuck - pipe warning method in the embodiments of the present application. The method may include the following steps:

[0088] S201: Obtain the current set of comprehensive mud logging parameters.

[0089] To achieve real - time dynamic warning of stuck - pipe faults, it is necessary to establish a connection interface between the comprehensive mud logging parameters and the stuck - pipe fault warning model. WITS (Wellsite Information Transfer Standard) is a communication format used to transfer various on - site data from one computer system to another. The comprehensive mud logging parameters need to go through the following five steps when being transmitted from the site to the stuck - pipe warning software for data analysis:

[0090] ① The mud logging company obtains on - site real - time data through sensors on the tool;

[0091] ② The collected data is transmitted to the ground through WITS for storage or other purposes;

[0092] ③ Through the WITS data transmission software, the data is transmitted to the software in real - time in the form of WITS files or in the form of TCP / IP protocol;

[0093] ④ The real - time transmitted data is saved to the local database for other purposes;

[0094] ⑤ The stuck - pipe fault software directly reads the data in the local database for dynamic warning of stuck - pipe faults.

[0095] Using C#, SQL database, and MATLAB to compile the stuck - pipe dynamic warning software, and pre - compiling the data transmission interface between the comprehensive mud logging parameters and the stuck - pipe fault warning software, realizing the real - time data transmission between the comprehensive mud logging parameters and the stuck - pipe fault warning software, and achieving the intelligent and real - time judgment of stuck - pipe faults.

[0096] According to the data to be transmitted, find the corresponding parameters in the WITS record: hook load, standpipe pressure, torque, rate of penetration, rotary speed, and weight on bit. As shown in Table 1, Table 1 is the table of mud logging parameters transmitted based on the WITS standard.

[0097] Table 1

[0098] Chinese / English English Abbreviation Bit Depth DBTM Drilling Time TIME Hookload HKLA Weight on bit WOBA Rotary speed RPMA Rotary torque TQA Standpipe pressure SPPA Rate of penetration ROP

[0099] S202: Determine whether there is a current comprehensive logging parameter with an abnormal jump value in the current comprehensive logging parameter group. If yes, execute step S203; if not, execute step S204.

[0100] The burrs or mutations of the engineering parameter curve may be abnormal jitter values, or they may be sudden changes in downhole working conditions. For abnormal jitter values, burrs or mutations should be removed, and for sudden changes in downhole working conditions, they should be retained. After obtaining the current comprehensive logging parameter group at the current moment, it is determined whether there is a current comprehensive logging parameter with an abnormal jitter value in the current comprehensive logging parameter group. If yes, step S203 is executed, and if not, step S204 is executed.

[0101] The method for identifying abnormal jump values ​​is mainly the rational method. Assuming that the engineering parameter response at a certain moment is an unknown rational number A, the calculation formula is as follows:

[0102] A=[[min(X i ),max(X i )],p(x)];(1≤i≤N);

[0103] Among them, X i is the logging value at a certain depth, and p(x) is the credibility distribution function of the logging value.

[0104] Contrast outliers and true mutations: If X i If it is an outlier, it is isolated and there are few data with similar values ​​in its neighborhood. If it is a true mutation, then i There are many data with similar values ​​in the neighborhood of X. i The more data similar to it in the neighborhood, the greater the value of X. i The credibility of X is high. i The credibility of is small. The credibility function is:

[0105]

[0106] Among them, n i Represents X i Neighborhood|X i -X j |≤λ;(i-δ<j<i+δ) contains X j , λ is the allowable error value, and j is the number of data points within the credibility range.

[0107] We can use the expectation of an undetermined rational number A Instead, abnormal jump values ​​can be eliminated.

[0108] S203: Eliminate the current comprehensive mud logging parameters that are abnormal jump values.

[0109] When it is determined that there are current comprehensive mud logging parameters that are abnormal jump values in the current comprehensive mud logging parameter group, eliminate the current comprehensive mud logging parameters that are abnormal jump values, so as to avoid misjudgment of the drilling fault diagnosis results caused by abnormal jump values.

[0110] S204: Obtain the current drilling working condition.

[0111] Among them, the current drilling working condition is one of tripping out, tripping in, and drilling.

[0112] After obtaining the current comprehensive mud logging parameter group at the current moment, obtain the current drilling working condition, so as to determine which one of tripping out, tripping in, and drilling it belongs to.

[0113] S205: Compare the current comprehensive mud logging parameter group with the preset parameter threshold group corresponding to the current drilling working condition to obtain a threshold diagnosis result.

[0114] Among them, the preset parameter threshold group is calculated based on the friction value and the comprehensive mud logging parameter group when a sticking accident occurs under the current drilling working condition.

[0115] The preset parameter threshold groups corresponding to different working conditions are different. Collect the relevant data of the drilled wells with sticking accidents in advance, calculate the friction value when sticking occurs according to the wellbore structure, drill string assembly, drilling fluid performance, and drilling parameters, and calculate the preset parameter threshold group based on the friction value and the comprehensive mud logging parameter group when a sticking accident occurs under the current drilling working condition. After obtaining the current drilling working condition, compare the current comprehensive mud logging parameter group with the preset parameter threshold group corresponding to the current drilling working condition to obtain a threshold diagnosis result.

[0116] S206: Obtain the historical comprehensive mud logging parameter groups of each preset time period before the current comprehensive mud logging parameter group.

[0117] S207: For each current comprehensive mud logging parameter in the current comprehensive mud logging parameter group, calculate the first moving average corresponding to the first sliding window length and the second moving average corresponding to the second sliding window length respectively according to the current comprehensive mud logging parameter and the corresponding historical comprehensive mud logging parameters in each historical comprehensive mud logging parameter group.

[0118] After obtaining the historical comprehensive logging parameter groups for each preset time period before obtaining the current comprehensive logging parameter group, for each current comprehensive logging parameter in the current comprehensive logging parameter group, according to the current comprehensive logging parameter and the corresponding historical comprehensive logging parameters in each historical comprehensive logging parameter group, calculate the first moving average corresponding to the first sliding window length and the second moving average corresponding to the second sliding window length respectively. Taking the abnormal diagnosis of the change trend of the bit torque monitored in real time as an example, the first moving average of the bit torque monitoring value corresponding to the first sliding window length at time t is calculated as M α , and the second moving average of the bit torque monitoring value corresponding to the second sliding window length is M β , where α and β are the sliding window lengths.

[0119] S208: Calculate the moving average deviation according to the first moving average and the second moving average.

[0120] After calculating the first moving average corresponding to the first sliding window length and the second moving average corresponding to the second sliding window length respectively, calculate the moving average deviation according to the first moving average and the second moving average. The calculation formula is as follows:

[0121] ΔM = M α - M β (α < β).

[0122] S209: Use the sliding window linear regression function to calculate the first slope value of the regression of the current comprehensive logging parameter and the corresponding historical comprehensive logging parameters in the first sliding window length according to the moving average deviation, and calculate the second slope value of the regression of the current comprehensive logging parameter and the corresponding historical comprehensive logging parameters in the second sliding window length.

[0123] After calculating the moving average deviation, use the sliding window linear regression function to calculate the first slope value of the regression of the current comprehensive logging parameter and the corresponding historical comprehensive logging parameters in the first sliding window length according to the moving average deviation, and calculate the second slope value of the regression of the current comprehensive logging parameter and the corresponding historical comprehensive logging parameters in the second sliding window length. The first slope value and the second slope value can be uniformly represented by K i,t , only the number of the second slope values is greater than the number of the first slope values.

[0124] S210: Obtain the first weighting factors corresponding to the first slope values respectively, and calculate the first average local slope corresponding to the first sliding window length according to the first slope values and the first weighting factors.

[0125] After calculating the first slope value and the second slope value, obtain the first weighting factors w corresponding to the first slope values respectively i, calculate the first average local slope corresponding to the first sliding window length according to each first slope value and each first weighting factor. The calculation formula is:

[0126]

[0127]

[0128] Among them, λ1 is the position to control the transition, λ2 is the sharpness to control the transition, λ1 = 0.5 - α, λ2 = 0.1. When i = t - α, the weighting factor is close to 0; when t - α < i ≤ t, the weighting factor is close to 1.

[0129] S211: Obtain the second weighting factor corresponding to each second slope value, and calculate the second average local slope corresponding to the second sliding window length according to each second slope value and each second weighting factor.

[0130] After calculating the first slope value and the second slope value, obtain the second weighting factor corresponding to each second slope value, and calculate the second average local slope corresponding to the second sliding window length according to each second slope value and each second weighting factor. The calculation formula is:

[0131]

[0132] S212: Calculate the difference of the current average local slope at the current moment according to the first average local slope and the second average local slope.

[0133] After calculating the first average local slope and the second average local slope, calculate the difference of the current average local slope at the current moment according to the first average local slope and the second average local slope. The calculation formula is as follows:

[0134]

[0135] S213: Obtain the historical average local slope differences corresponding to each preset time period before the current moment.

[0136] After calculating the difference of the current average local slope, obtain the historical average local slope differences corresponding to each preset time period before the current moment.

[0137] For example, select the time periods t1, t2, t3, t4, t5, t6 before the t moment (the length of the time period is set to Δt, and the value is assigned according to the actual situation) to analyze the change trend of the monitoring value.

[0138] S214: Determine the diagnostic result of the parameter change trend according to the difference of the current average local slope and each historical average local slope difference.

[0139] After obtaining the historical average local slope differences, determine the diagnosis result of the parameter change trend based on the current average local slope difference and the historical average local slope differences.

[0140] In a specific implementation manner of the present application, step S214 may include the following steps:

[0141] Step 1: Based on the current average local slope difference and the historical average local slope differences, determine whether the average local slope of the latter time period in each adjacent time period is greater than that of the previous time period, and whether each average local slope is greater than 0. If so, execute Step 2; if not, determine that no fault warning is required.

[0142] Step 2: Determine that a fault warning is required.

[0143] For ease of description, the above two steps can be combined for explanation.

[0144] Based on the current average local slope difference and the historical average local slope differences, determine whether the average local slope of the latter time period in each adjacent time period is greater than that of the previous time period, and whether each average local slope is greater than 0. If so, it indicates that the average local slope has been increasing, indicating an abnormal increasing trend in the monitored value, indicating that the probability of sticking gradually increases, and determine that a fault warning is required; if not, determine that no fault warning is required.

[0145] For example, select 0.5, 1, 1.5, 2, 2.5, 3 min (the time interval can be adjusted) before time t respectively for parameter change trend analysis. If the slope has been increasing and the average local slope within 0.5 min before time t is greater than 0, it indicates an abnormal increasing trend in the engineering parameter. For example, during the process of pulling out the drill string, if the average local slope of the hook load measurement parameter has been increasing within 6 time periods of 0.5, 1, 1.5, 2, 2.5, 3 min before time t, and the average local slope within 0.5 min before time t is greater than 0, a sticking warning is issued.

[0146] See Figure 3 , Figure 3 which is a real-time warning flowchart for sticking based on abnormal diagnosis of parameter change trend in an embodiment of the present application. The specific steps of the sticking automatic monitoring algorithm are as follows:

[0147] (1) First, before using the sticking monitoring algorithm, eliminate abnormal points from the real-time data, identify the working conditions of the drilling state through real-time drilling parameters, and determine whether the operation belongs to any transient activity. The above preprocessing process can establish a reasonable starting point for risk warning and avoid false alarms caused by transient operations or abnormal points.

[0148] (2) Calculate the key characterization parameters for stuck pipe based on the data of the current and previous time steps: hook load, torque, deviation of the moving average of standpipe pressure, and local trend characteristics at 0.5, 1, 1.5, 2, 2.5, and 3 minutes before time t. If the average moving deviation indicates a local increasing trend and the absolute value of the average local slope of the data in the 6 time periods has been increasing, a stuck pipe warning is issued according to the stuck pipe fault judgment criteria under different working conditions.

[0149] S215: Use the stuck pipe fault diagnosis model based on the particle swarm optimization BP neural network to perform stuck pipe fault diagnosis for the current comprehensive mud logging parameter group, and obtain the model diagnosis result.

[0150] Pre-train the stuck pipe fault diagnosis model based on the particle swarm optimization BP neural network. After obtaining the current comprehensive mud logging parameter group, use the stuck pipe fault diagnosis model based on the particle swarm optimization BP neural network to perform stuck pipe fault diagnosis for the current comprehensive mud logging parameter group, and obtain the model diagnosis result.

[0151] See Figure 4 , Figure 4 See, which is the flow chart of a particle swarm optimization BP neural network algorithm step in the embodiment of the present application. The stuck pipe fault warning model based on the particle swarm optimization BP neural network needs to design three major parts, including: the design of the BP neural network, the design of the particle swarm algorithm, and the preprocessing of the model input parameters.

[0152] See Figure 5 , Figure 5 See, which is the structure diagram of a stuck pipe fault warning BP neural network in the embodiment of the present application.

[0153] The design of the BP neural network is as follows:

[0154] ① Input layer design

[0155] According to expert knowledge judgment, select the characterization parameters with strong correlation with stuck pipe faults, and set 6 input neurons, namely: hook load, standpipe pressure, torque, rotary speed, rotary table speed, and weight on bit.

[0156] ② Output layer design

[0157] The main problem solved by the present invention is the identification of downhole stuck pipe faults. Only two judgments need to be made, whether a stuck pipe fault occurs, and at the same time, if so, what is the probability. Therefore, there are 2 output neurons. The expected output vector for a stuck pipe fault is q1 = (1, 0), and the expected output vector for no stuck pipe fault is q2 = (0, 1).

[0158] ③ Hidden layer design

[0159] The number of hidden layers is calculated by a classical formula, and the value of the hidden layer ranges from 1 to 10. In this application, through research, the number of hidden layers is set to 10, and the obtained result is the most consistent with the actual situation.

[0160] Finally, a BP neural network structure is established, as Figure 5 shown.

[0161] See Figure 6 , Figure 6 which is a schematic diagram of a BP neural network optimized by a particle swarm algorithm for stuck-pipe fault warning in an embodiment of this application.

[0162] The design of the particle swarm algorithm is as follows:

[0163] ① Particle encoding

[0164] The neural network structure is 6-10-2. There are a total of 6×10 + 10×2 = 80 weights and 10 + 2 = 12 thresholds. Then the number of particle swarm optimization parameters is 92, and the length of the particle is 92.

[0165] The present invention uses vector encoding, and particle i is encoded as:

[0166]

[0167] ② Fitness function

[0168] The root mean square error of BP network training is used as the fitness function.

[0169] Using the established stuck-pipe fault warning and identification model, combined with the on-site practice of shale gas drilling, the comprehensive logging parameters of the wells where stuck-pipe faults have occurred are selected as training samples to train the established model and optimize the model parameters. The control elements in the BP neural network of the stuck-pipe fault warning model and the particle swarm algorithm are set. As shown in Table 2, Table 2 is the basic control element table of the particle swarm optimization BP neural network algorithm.

[0170] Table 2

[0171]

[0172] Model training simulation:

[0173] Combined with the logging and well history data of 6 drilled wells with stuck pipe failures in the shale gas work area, 12 groups of working condition data were selected as samples, with 6 groups of data corresponding to stuck pipe failures and 6 groups of data corresponding to non-stuck pipe working conditions; the comprehensive logging parameters monitored within a certain time domain before and after the occurrence of stuck pipe failures were selected and normalized according to the maximum-minimum method. The first 10 groups of samples were selected as training samples to train the constructed stuck pipe failure model, and the other two groups of samples were used as verification samples to verify the accuracy of the constructed model. As shown in Table 3, Table 3 is the network training simulation sample table.

[0174] Table 3

[0175]

[0176] Result analysis: The stuck pipe failure diagnosis results obtained through simulation are consistent with the actual downhole conditions, indicating that the model established through training can be applied to the stuck pipe failure early warning and identification in this block and this interval. As shown in Table 4, Table 4 is the stuck pipe failure neural network training simulation result table.

[0177] Table 4

[0178]

[0179]

[0180] S216: When it is determined that a fault early warning is required based on the threshold diagnosis result, the parameter change trend diagnosis result, and the model diagnosis result, a stuck pipe warning for drilling is carried out.

[0181] Stuck pipe can be further divided into adhesion, collapse, sand bridge, hole shrinkage, keyseat, mud packing, and falling object stuck pipe according to different causes and symptoms. First, it is necessary to combine the knowledge of drilling experts to judge and sort out and summarize the qualitative identification methods for stuck pipe failures, which is also an important basic work for stuck pipe prediction. Different types of stuck pipe correspond to different judgment methods. Combining on-site drilling practice and judging according to expert knowledge, the stuck pipe judgment methods are summarized as shown in Table 5. Table 5 is the judgment method table corresponding to different types of stuck pipe.

[0182] Table 5

[0183]

[0184]

[0185] Through the above expert knowledge judgment of stuck pipe failures, the comprehensive logging parameters that can characterize stuck pipe failures are: hook load, standpipe pressure, torque, rotary speed, rotary table speed, and drilling pressure.

[0186] Sensitivity of stuck pipe failure: Hook load > Torque > Rotary table speed > Drilling pressure > Rotary speed > Standpipe pressure.

[0187] The over-threshold diagnosis is applicable to sudden sticking accidents. Exceeding the threshold indicates an abnormal downhole state. The abnormal diagnosis of parameter variation trend is applicable to the occurrence of downhole fault signs, which can give early warning information in the early stage of the fault, and avoid the deterioration of the fault to the greatest extent. The artificial intelligence diagnosis requires a large number of drilled wells as training samples. With the increase of training samples, the warning results will be more and more accurate. Therefore, according to the characteristics of different work areas and different well sections, the three-layer mechanism will play different effects. Based on the three-layer diagnosis mechanism for sticking warning, the results are more accurate and efficient.

[0188] See Figure 7 and Figure 8 , Figure 7 is a schematic structural diagram of a drilling sticking warning system in an embodiment of the present application. Figure 8 is a schematic structural diagram of the downhole part of a drilling sticking warning system in an embodiment of the present application. The drilling sticking warning system includes two major parts: "downhole + surface". The downhole part 2 of the drilling sticking warning system includes: a core barrel 21, a downhole parameter measurer 22, a downhole data processor 23, a pulse generator 24, a choke ring 25, and an outer barrel 26, which can realize the measurement, preprocessing of parameters such as the weight on bit, torque, and equivalent circulating density of drilling fluid near the position of the downhole bit 1, and the transmission while drilling to the surface software. The surface part is mainly a set of monitoring and diagnosis software system, including two major modules: data reception and management, and risk diagnosis, which realizes the reception of downhole measurement data while drilling and comprehensive logging data, manages the data in real time, and realizes real-time warning of downhole risks based on the abnormal trend diagnosis method. This system can effectively predict and diagnose downhole risks (kick, lost circulation, hole collapse, sticking) during deep and ultra-deep drilling, ensuring drilling safety.

[0189] The downhole parameter measurer 22 can measure the actual torque, weight on bit, temperature and pressure near the position of the downhole drill bit 1 in real time. The downhole data processor 23 preprocesses the acquired downhole parameters, mines the data features of the high-frequency data, and transmits the processed data to the pulse emitter. The pulse generator 24 encodes the processed data to generate a pulse signal, which is transmitted to the downhole measurement data ground receiving instrument 5 through the pulse emitter, and the downhole measurement signal is transmitted to the ground software system through this instrument. The ground working software system of the ground part 4 of the drilling sticking warning system can receive the data measured by the downhole parameter while-drilling monitoring tool in real time, and at the same time receive the parameters in the surface comprehensive logging instrument 3 into the software, realizing the reception of the downhole while-drilling measurement data and the comprehensive logging data, and performing real-time management on the data. The abnormal trend diagnosis method is used to judge the borehole cleaning degree in real time and dynamically. Taking the sticking risk warning as an example, the downhole risk (kick, lost circulation, wellbore collapse, sticking) evaluation method realizes the diagnosis of the sticking risk through the abnormal diagnosis of the change trend of the torque of the drill bit 1 in real time. It solves the monitoring and warning of downhole complexity and failures during the drilling process, and avoids serious and vicious drilling accidents caused by untimely risk handling.

[0190] The data monitoring module of the monitoring system software realizes the standardization of data through the downhole data processor 23 and the comprehensive logging instrument 3 in the WITSO standard data format, and at the same time establishes a relevant database, and the received data can be stored in the database in real time. It realizes the function of real-time monitoring of the weight on bit, torque, and equivalent circulating density of the drilling fluid near the position of the downhole drill bit 1, and the monitoring results can be displayed in the form of curves.

[0191] Through the real-time monitoring of the weight on bit, torque, and equivalent circulating density of the drilling fluid near the position of the downhole drill bit 1 by the downhole parameter measurer 22, compared with the prior art, the obtained weight on bit and torque are more accurate, which are the actual downhole values, thus providing more accurate data for the drilling risk warning. At the same time, by combining the torque, weight on bit, pump pressure, mechanical drilling rate and cuttings returned to the surface of the surface comprehensive logging parameters, the abnormal trend diagnosis method is used to judge the downhole complexity and failures, which is more accurate and intelligent, eliminates the influence of subjective judgment by people, and the obtained results are more objective and accurate. And relevant warning software is compiled, which can predict, diagnose and evaluate the downhole risks (kick, lost circulation, collapse, sticking) in real time, dynamically, intelligently, with timeliness and reliability, helps to ensure the smooth progress of on-site construction, and avoids serious and vicious drilling accidents caused by untimely risk handling.

[0192] Corresponding to the above method embodiment, the present application also provides a drilling sticking warning device, and the drilling sticking warning device described below can be mutually referred to with the drilling sticking warning method described above.

[0193] SeeFigure 9 , Figure 9 is a structural block diagram of a drilling sticking warning device in an embodiment of the present application. The device may include:

[0194] The current parameter group acquisition module 91 is configured to acquire the current comprehensive mud logging parameter group at the current moment;

[0195] The threshold diagnosis result acquisition module 92 is configured to compare the current comprehensive mud logging parameter group with a preset parameter threshold group to obtain a threshold diagnosis result;

[0196] The historical parameter group acquisition module 93 is configured to acquire the historical comprehensive mud logging parameter groups of each preset time period before the current comprehensive mud logging parameter group;

[0197] The change trend diagnosis result acquisition module 94 is configured to calculate the change trend of each comprehensive mud logging parameter according to the current comprehensive mud logging parameter group and each historical comprehensive mud logging parameter group to obtain a parameter change trend diagnosis result;

[0198] The model diagnosis result acquisition module 95 is configured to perform drilling sticking fault diagnosis on the current comprehensive mud logging parameter group by using a sticking fault diagnosis model to obtain a model diagnosis result;

[0199] The sticking warning module 96 is configured to perform drilling sticking warning when it is determined that a fault warning is required according to the threshold diagnosis result, the parameter change trend diagnosis result, and the model diagnosis result.

[0200] It can be seen from the above technical solution that after obtaining the current comprehensive mud logging parameter group at the current moment, by combining threshold diagnosis, abnormal diagnosis of parameter change trends, and artificial intelligence diagnosis using a sticking fault diagnosis model to perform sticking fault diagnosis on the current comprehensive mud logging parameter group, the intelligent and real-time judgment of sticking faults is realized, the comprehensive utilization ability of monitoring information is improved, the timeliness of risk warning is improved, the influence of subjective consciousness is avoided, and the accuracy of sticking fault warning is improved.

[0201] In a specific embodiment of the present application, the threshold diagnosis result acquisition module 92 includes:

[0202] The drilling working condition acquisition sub-module is configured to acquire the current drilling working condition; wherein, the current drilling working condition is one of tripping out, tripping in, and drilling;

[0203] The parameter comparison sub-module is configured to compare the current comprehensive mud logging parameter group with a preset parameter threshold group corresponding to the current drilling working condition; wherein, the preset parameter threshold group is calculated according to the friction value and the comprehensive mud logging parameter group when a sticking fault occurs under the current drilling working condition.

[0204] In a specific implementation manner of the present application, the model diagnosis result obtaining module 95 is specifically a module that uses a stuck-pipe fault diagnosis model optimized by a particle swarm algorithm for a BP neural network to perform stuck-pipe fault diagnosis based on the current comprehensive logging parameter group.

[0205] In a specific implementation manner of the present application, the change trend diagnosis result obtaining module 94 includes:

[0206] A moving average calculation sub-module, configured to calculate, for each current comprehensive logging parameter in the current comprehensive logging parameter group, a first moving average corresponding to a first sliding window length and a second moving average corresponding to a second sliding window length respectively according to the current comprehensive logging parameter and the corresponding historical comprehensive logging parameters in each historical comprehensive logging parameter group;

[0207] A moving average deviation calculation sub-module, configured to calculate a moving average deviation according to the first moving average and the second moving average;

[0208] A slope value calculation sub-module, configured to use a sliding window linear regression function to calculate a first slope value of the regression of the current comprehensive logging parameter and the corresponding historical comprehensive logging parameters in the first sliding window length at the current moment according to the moving average deviation, and calculate a second slope value of the regression of the current comprehensive logging parameter and the corresponding historical comprehensive logging parameters in the second sliding window length at the current moment;

[0209] A first average local slope calculation sub-module, configured to obtain first weighting factors corresponding to the respective first slope values, and calculate a first average local slope corresponding to the first sliding window length according to the respective first slope values and the respective first weighting factors;

[0210] A second average local slope calculation sub-module, configured to obtain second weighting factors corresponding to the respective second slope values, and calculate a second average local slope corresponding to the second sliding window length according to the respective second slope values and the respective second weighting factors;

[0211] A current average local slope difference calculation sub-module, configured to calculate a current average local slope difference at the current moment according to the first average local slope and the second average local slope;

[0212] A historical average local slope difference obtaining sub-module, configured to obtain historical average local slope differences corresponding to respective preset time periods before the current moment;

[0213] A trend diagnosis result determining sub-module, configured to determine a parameter change trend diagnosis result according to the current average local slope difference and the respective historical average local slope differences.

[0214] In a specific implementation manner of the present application, the trend diagnosis result determining sub-module includes:

[0215] A judgment unit, configured to determine whether the average local slope of each subsequent time period in each adjacent time period is greater than that of the previous time period, and whether each average local slope is greater than 0, based on the current average local slope difference and each historical average local slope difference;

[0216] A stuck-pipe fault determination unit, configured to determine that a fault warning needs to be issued when it is determined that the average local slope of each subsequent time period in each adjacent time period is greater than that of the previous time period, and each average local slope is greater than 0.

[0217] In a specific embodiment of the present application, the device may further include:

[0218] A judgment module, configured to determine whether there is a current comprehensive mud logging parameter in the current comprehensive mud logging parameter group that is an abnormal jump value before comparing the current comprehensive mud logging parameter group with a preset parameter threshold group after obtaining the current comprehensive mud logging parameter group;

[0219] A parameter elimination module, configured to eliminate the current comprehensive mud logging parameter that is an abnormal jump value when it is determined that there is a current comprehensive mud logging parameter in the current comprehensive mud logging parameter group that is an abnormal jump value.

[0220] In a specific embodiment of the present application, the device may further include:

[0221] A normalization module, configured to perform normalization processing on each current comprehensive mud logging parameter in the current comprehensive mud logging parameter group before diagnosing a stuck-pipe fault in drilling using a stuck-pipe fault diagnosis model.

[0222] Corresponding to the above method embodiment, refer to Figure 10 , Figure 10 which is a schematic diagram of the stuck-pipe warning device provided by the present application. The device may include:

[0223] A memory 332, configured to store a computer program;

[0224] A processor 322, configured to implement the steps of the stuck-pipe warning method in the above method embodiment when executing the computer program.

[0225] Specifically, please refer to Figure 11 , Figure 11FIG. 0 is a schematic structural diagram of a drilling sticking warning device provided in this embodiment. The drilling sticking warning device may vary greatly due to different configurations or performances, and may include a processor (central processing units, CPU) 322 (for example, one or more processors) and a memory 332. The memory 332 stores one or more computer programs 342 or data 344. Among them, the memory 332 may be short-term storage or persistent storage. The program stored in the memory 332 may include one or more modules (not shown in the figure), and each module may include a series of instruction operations on the data processing device. Further, the processor 322 may be set to communicate with the memory 332 and execute a series of instruction operations in the memory 332 on the drilling sticking warning device 301.

[0226] The drilling sticking warning device 301 may further include one or more power supplies 326, one or more wired or wireless network interfaces 350, one or more input / output interfaces 358, and / or one or more operating systems 341.

[0227] The steps in the drilling sticking warning method described above may be implemented by the structure of the drilling sticking warning device.

[0228] Corresponding to the above method embodiment, the present application further provides a computer-readable storage medium. A computer program is stored on the computer-readable storage medium. When the computer program is executed by a processor, the following steps may be implemented:

[0229] Obtain the current comprehensive mud logging parameter group at the current moment; compare the current comprehensive mud logging parameter group with a preset parameter threshold group to obtain a threshold diagnosis result; obtain the historical comprehensive mud logging parameter groups in each preset time period before the current comprehensive mud logging parameter group; calculate the change trend of each comprehensive mud logging parameter according to the current comprehensive mud logging parameter group and each historical comprehensive mud logging parameter group to obtain a parameter change trend diagnosis result; use a sticking fault diagnosis model to perform sticking fault diagnosis on the current comprehensive mud logging parameter group to obtain a model diagnosis result; when it is determined that a fault warning needs to be performed according to the threshold diagnosis result, the parameter change trend diagnosis result, and the model diagnosis result, perform drilling sticking warning.

[0230] The computer-readable storage medium may include: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memories (ROM), random access memories (RAM), magnetic disks, or optical discs.

[0231] For the introduction of the computer-readable storage medium provided in the present application, please refer to the above method embodiment, and the present application will not elaborate here.

[0232] In this specification, the various embodiments are described in a progressive manner. Each embodiment focuses on the differences from other embodiments, and the same or similar parts among the embodiments can be referred to each other. For the devices, equipment, and computer-readable storage media disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple, and the relevant parts can be referred to the description in the method section.

[0233] Specific examples are used in this article to elaborate on the principles and implementation manners of the present application. The descriptions of the above embodiments are only used to help understand the technical solution and its core idea of the present application. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present application, several improvements and modifications can be made to the present application, and these improvements and modifications also fall within the protection scope of the claims of the present application.

Claims

1. A method for warning of stuck pipe during drilling, characterized in that, Including: Obtaining the current comprehensive logging parameter group at the current moment; Comparing the current comprehensive logging parameter group with a preset parameter threshold group to obtain a threshold diagnosis result; Obtaining historical comprehensive logging parameter groups for each preset time period before the current comprehensive logging parameter group; Calculating the change trends of the comprehensive logging parameters according to the current comprehensive logging parameter group and each of the historical comprehensive logging parameter groups to obtain a parameter change trend diagnosis result; Using a stuck-pipe fault diagnosis model to perform stuck-pipe fault diagnosis on the current comprehensive logging parameter group to obtain a model diagnosis result; When it is determined that a fault warning needs to be issued according to the threshold diagnosis result, the parameter change trend diagnosis result, and the model diagnosis result, a stuck-pipe warning for drilling is issued.

2. The drilling stuck warning method according to claim 1, characterized in that, Comparing the current comprehensive logging parameter group with a preset parameter threshold group includes: Obtaining the current drilling working condition; where the current drilling working condition is one of tripping out, tripping in, and drilling; Comparing the current comprehensive logging parameter group with the preset parameter threshold group corresponding to the current drilling working condition; where the preset parameter threshold group is calculated according to the friction resistance value and the comprehensive logging parameter group when a stuck-pipe fault occurs under the current drilling working condition.

3. The drilling sticking warning method according to claim 1, wherein Using a stuck-pipe fault diagnosis model to perform stuck-pipe fault diagnosis on the current comprehensive logging parameter group includes: Using a stuck-pipe fault diagnosis model optimized by a particle swarm algorithm for a BP neural network to perform stuck-pipe fault diagnosis on the current comprehensive logging parameter group.

4. The drilling sticking warning method according to any one of claims 1 to 3, characterized in that, Calculating the change trends of the comprehensive logging parameters according to the current comprehensive logging parameter group and each of the historical comprehensive logging parameter groups to obtain a parameter change trend diagnosis result, including: For each current comprehensive logging parameter in the current comprehensive logging parameter group, calculating a first moving average corresponding to a first sliding window length and a second moving average corresponding to a second sliding window length according to the current comprehensive logging parameter and the corresponding historical comprehensive logging parameters in each of the historical comprehensive logging parameter groups; Calculating a moving average deviation according to the first moving average and the second moving average; Using a sliding window linear regression function to calculate a first slope value of the regression of the current comprehensive logging parameter and the corresponding historical comprehensive logging parameters in the first sliding window length at the current moment according to the moving average deviation, and calculating a second slope value of the regression of the current comprehensive logging parameter and the corresponding historical comprehensive logging parameters in the second sliding window length at the current moment; Obtaining first weighting factors corresponding to the first slope values respectively, and calculating a first average local slope corresponding to the first sliding window length according to the first slope values and the first weighting factors; Obtaining second weighting factors corresponding to the second slope values respectively, and calculating a second average local slope corresponding to the second sliding window length according to the second slope values and the second weighting factors; Calculating a difference in the current average local slope at the current moment according to the first average local slope and the second average local slope; Obtaining historical average local slope differences corresponding to each preset time period before the current moment; Determine the diagnostic result of the parameter change trend according to the current average local slope difference and each of the historical average local slope differences.

5. The drilling stuck pipe warning method according to claim 4, wherein Determining the diagnostic result of the parameter change trend according to the current average local slope difference and each of the historical average local slope differences includes: According to the current average local slope difference and each of the historical average local slope differences, determine whether the average local slope of the latter time period in each adjacent time period is greater than that of the previous time period, and each average local slope is greater than 0; If so, determine that a fault warning needs to be carried out.

6. The drilling stuck pipe warning method according to claim 1, wherein Before comparing the current comprehensive mud logging parameter group with the preset parameter threshold group after obtaining the current comprehensive mud logging parameter group, it further includes: Determine whether there is a current comprehensive mud logging parameter in the current comprehensive mud logging parameter group that is an abnormal jump value; If so, eliminate the current comprehensive mud logging parameter that is an abnormal jump value.

7. The warning method for drilling sticking according to claim 1, wherein Before using the stuck pipe fault diagnosis model to perform stuck pipe fault diagnosis on the current comprehensive mud logging parameter group, it further includes: Normalize each current comprehensive mud logging parameter in the current comprehensive mud logging parameter group.

8. A drilling sticking warning device, characterized in that, It includes: A current parameter group acquisition module, configured to acquire the current comprehensive mud logging parameter group at the current moment; A threshold diagnostic result acquisition module, configured to compare the current comprehensive mud logging parameter group with the preset parameter threshold group to obtain a threshold diagnostic result; A historical parameter group acquisition module, configured to acquire the historical comprehensive mud logging parameter groups of each preset time period before the current comprehensive mud logging parameter group; A change trend diagnostic result acquisition module, configured to calculate the change trend of each comprehensive mud logging parameter according to the current comprehensive mud logging parameter group and each of the historical comprehensive mud logging parameter groups to obtain a parameter change trend diagnostic result; A model diagnostic result acquisition module, configured to use the stuck pipe fault diagnosis model to perform stuck pipe fault diagnosis on the current comprehensive mud logging parameter group to obtain a model diagnostic result; A stuck pipe warning module, configured to perform stuck pipe warning on the drilling when it is determined according to the threshold diagnostic result, the parameter change trend diagnostic result and the model diagnostic result that a fault warning needs to be carried out.

9. A drilling stuck pipe warning device, characterized in that, It includes: A memory, configured to store a computer program; A processor, configured to implement the steps of the stuck pipe warning method according to any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium, characterized in that, A computer program is stored on the computer-readable storage medium, and when the computer program is executed by the processor, the steps of the stuck pipe warning method according to any one of claims 1 to 7 are implemented.

Images