Well drilling abnormal state judgment method, device, computer equipment and system

Abstract

The invention provides a well drilling abnormal state judgment method and device, computer equipment and a system. The method comprises the steps of obtaining to-be-detected while-drilling data; performing anomaly detection on the to-be-detected while-drilling data based on a pre-trained auto-encoder model to obtain a first detection result; performing anomaly detection on the to-be-detected while-drilling data based on a pre-trained isolated forest model to obtain a second detection result; performing anomaly detection on the to-be-detected while-drilling data based on a pre-trained single-classification support vector machine model to obtain a third detection result; and obtaining a drilling state abnormity judgment result based on the first detection result, the second detection result and the third detection result. The method provided by the invention is high in automation degree and good in stability, and can improve the recognition efficiency and accuracy of the abnormal state in the drilling process, thereby improving the reliability of the intelligent drilling system.

Description

Technical Field

[0001] The present application relates to the technical field of oil and gas exploration, and in particular to a method, device, computer equipment and system for judging abnormal drilling status. Background Art

[0002] Automated intelligent drilling is an important development direction of drilling engineering, and the automatic optimization and control of drilling parameters during drilling is a prerequisite for realizing intelligent drilling. Drilling conditions include drilling, tripping, circulation, reaming, back reaming and circulation, etc. Drilling conditions can be divided into normal and abnormal conditions, among which abnormal conditions include well leakage, overflow, obstruction, drill bit sticking, drill bit mud packing, broken drill tools, etc. Timely detection of abnormal conditions in drilling conditions can effectively avoid serious accidents, help improve drilling efficiency and ensure drilling safety.

[0003] The drilling status can be reflected by a variety of information, such as drilling pressure, rotation speed, displacement, pump pressure, torque, mechanical drilling speed, hook load, inlet density, outlet density, inlet temperature, outlet temperature, inlet conductivity, outlet conductivity, etc. The information dimension is high and complex. Traditionally, the drilling site mainly relies on expert experience or simple rules based on expert experience to process the real-time drilling and logging curves and give judgments on the downhole drilling status, with low automation, poor accuracy and reliability. Summary of the invention

[0004] In order to solve the problem that it is difficult to judge the abnormal state of drilling, the present application provides a method, device, computer equipment and system for judging the abnormal state of drilling, which can improve the efficiency and accuracy of identifying the abnormal state during the drilling process.

[0005] On the one hand, a method for determining an abnormal drilling state is provided, the method comprising:

[0006] Obtain the drilling data to be tested;

[0007] Performing anomaly detection on the drilling while drilling data to be tested based on a pre-trained autoencoder model to obtain a first detection result;

[0008] Performing anomaly detection on the drilling data to be tested based on a pre-trained isolation forest model to obtain a second detection result;

[0009] Performing anomaly detection on the drilling while drilling data to be tested based on a pre-trained single-classification support vector machine model to obtain a third detection result;

[0010] A drilling status abnormality judgment result is obtained based on the first detection result, the second detection result and the third detection result.

[0011] In some embodiments,

[0012] The method further comprises:

[0013] Obtain historical drilling data under normal working conditions as training samples;

[0014] Training the autoencoder model based on the training samples;

[0015] The autoencoder model includes a restoration error calculation formula;

[0016] The restoration error calculation formula includes:

[0017]

[0018] Among them, MSE restore is the restoration error of the training sample, R is the total number of rows of the training sample, row is the row number of the training sample, C is the total number of columns of the training sample, col is the column number of the training sample, and x row，col is the actual value of the training sample at a certain position, For x row，col The corresponding predicted value.

[0019] In some embodiments,

[0020] The first detection result includes a first abnormality coefficient. The pre-trained autoencoder model performs abnormality detection on the drilling data to be tested to obtain the first detection result, including:

[0021] Performing anomaly detection on the drilling while drilling data to be tested based on the autoencoder model to obtain a first anomaly coefficient;

[0022] The first abnormal coefficient calculation formula includes:

[0023]

[0024] Among them, CV is the first abnormal coefficient, MSE is the mean square error, restore To restore the error.

[0025] In some embodiments, the method further comprises:

[0026] Obtain historical drilling data under normal working conditions as training samples;

[0027] Training an isolation forest model based on the training samples, wherein the isolation forest model includes a plurality of isolated trees;

[0028] The isolation forest model includes an anomaly score calculation formula:

[0029]

[0030] Where x is the input LWD data, n is the dimension of the input LWD data, h(x) is the average height of the isolated tree, c(n) is the average value of h(x) given n, and E(h(x)) is the average value of h(x) of the isolated tree set.

[0031] In some embodiments, the method further comprises:

[0032] Obtain historical drilling data under normal working conditions as training samples;

[0033] Training the single-classification support vector machine model based on the training samples;

[0034] In the single-classification support vector machine model, normal historical drilling data are positive examples, and abnormal historical drilling data are negative examples.

[0035] In some embodiments, the abnormal drilling status determination result obtained based on the first detection result, the second detection result, and the third detection result includes:

[0036] If the number of abnormalities in the first detection result, the second detection result and the third detection result is greater than a preset abnormal threshold, the drilling state is determined to be abnormal.

[0037] On the other hand, a drilling abnormal state judgment device is provided, the device comprising:

[0038] A data acquisition module, used to acquire the drilling data to be tested;

[0039] An autoencoder detection module is used to perform anomaly detection on the drilling data to be tested based on a pre-trained autoencoder model to obtain a first detection result;

[0040] An isolation forest detection module is used to perform anomaly detection on the drilling data to be tested based on a pre-trained isolation forest model to obtain a second detection result;

[0041] A support vector machine detection module, used for performing anomaly detection on the drilling data to be tested based on a pre-trained single-classification support vector machine model to obtain a third detection result;

[0042] The abnormality judgment module is used to obtain a drilling status abnormality judgment result based on the first detection result, the second detection result and the third detection result.

[0043] In some embodiments, the apparatus further comprises an autoencoder model training module for:

[0044] Obtain historical drilling data under normal working conditions as training samples;

[0045] Training the autoencoder model based on the training samples;

[0046] The autoencoder model includes a restoration error calculation formula;

[0047] The restoration error calculation formula includes:

[0048]

[0049] Among them, MSE restore is the restoration error of the training sample, R is the total number of rows of the training sample, row is the row number of the training sample, C is the total number of columns of the training sample, col is the column number of the training sample, and x row，col is the actual value of the training sample at a certain position, For x row，col The corresponding predicted value.

[0050] In some embodiments, the first detection result includes a first abnormality coefficient, and the autoencoder detection module is specifically used to perform abnormality detection on the drilling-while-drilling data to be tested based on the autoencoder model to obtain the first abnormality coefficient;

[0051] The first abnormal coefficient calculation formula includes:

[0052]

[0053] Among them, CV is the first abnormal coefficient, MSE is the mean square error, restore To restore the error.

[0054] In some embodiments, the apparatus further comprises an isolation forest model training module, configured to:

[0055] Obtain historical drilling data under normal working conditions as training samples;

[0056] Training an isolation forest model based on the training samples, wherein the isolation forest model includes a plurality of isolated trees;

[0057] The isolation forest model includes an anomaly score calculation formula:

[0058]

[0059] Where x is the input LWD data, n is the dimension of the input LWD data, h(x) is the average height of the isolated tree, c(n) is the average value of h(x) given n, and E(h(x)) is the average value of h(x) of the isolated tree set.

[0060] In some embodiments, the apparatus further comprises a support vector machine model training module, configured to:

[0061] Obtain historical drilling data under normal working conditions as training samples;

[0062] Training the single-classification support vector machine model based on the training samples;

[0063] In the single-classification support vector machine model, normal historical drilling data are positive examples, and abnormal historical drilling data are negative examples.

[0064] In some embodiments, the abnormality determination module is specifically used to:

[0065] The abnormal drilling status judgment result obtained based on the first detection result, the second detection result and the third detection result includes:

[0066] If the number of abnormalities in the first detection result, the second detection result and the third detection result is greater than a preset abnormal threshold, the drilling state is determined to be abnormal.

[0067] On the other hand, a computer device is provided, which includes a processor and a memory, wherein the memory stores at least one instruction, at least one program, code set or instruction set, and the processor can load and execute at least one instruction, at least one program, code set or instruction set to implement the method for determining abnormal drilling status provided in the above-mentioned application embodiment.

[0068] On the other hand, a drilling abnormal state judgment system is provided, comprising the computer device as described above.

[0069] On the other hand, a computer-readable storage medium is provided, in which at least one instruction, at least one program, code set or instruction set is stored. The processor can load and execute at least one instruction, at least one program, code set or instruction set to implement the method for determining abnormal drilling status provided in the above-mentioned embodiment of the present application.

[0070] On the other hand, a computer program product or a computer program is provided, the computer program product or the computer program includes computer program instructions, and the computer program instructions are stored in a computer-readable storage medium. The processor reads the computer instructions from the computer-readable storage medium and executes the computer instructions, so that the computer device executes the drilling abnormal state judgment method described in any of the above embodiments.

[0071] The beneficial effects brought by the technical solution provided by the present application include at least: the embodiment of the present invention provides a method, device, computer equipment and system for judging abnormal drilling status, the method comprising obtaining the drilling data to be tested; performing abnormality detection on the drilling data to be tested based on a pre-trained autoencoder model to obtain a first detection result; performing abnormality detection on the drilling data to be tested based on a pre-trained isolation forest model to obtain a second detection result; performing abnormality detection on the drilling data to be tested based on a pre-trained single-classification support vector machine model to obtain a third detection result; and obtaining a drilling status abnormality judgment result based on the first detection result, the second detection result and the third detection result. The method provided by the embodiment of the present invention has a high degree of automation and good stability, and can improve the recognition efficiency and accuracy of abnormal status during drilling, thereby improving the reliability of the intelligent drilling system. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0073] Figure 1 A schematic diagram of an implementation flow of a method for determining an abnormal drilling state provided by an exemplary embodiment of the present application is shown;

[0074] Figure 2 A schematic diagram of a drilling process abnormality judgment model based on an autoencoder in a drilling abnormal state judgment method provided by an exemplary embodiment of the present application is shown;

[0075] Figure 3 A schematic diagram of a drilling process abnormality judgment model based on an isolation forest in a drilling abnormality state judgment method provided by an exemplary embodiment of the present application is shown;

[0076] Figure 4 A schematic diagram of a drilling process abnormality judgment model based on a single-classification support vector machine in a drilling abnormality state judgment method provided by an exemplary embodiment of the present application is shown;

[0077] Figure 5 A schematic diagram of an abnormal working condition judgment process in a method for judging an abnormal drilling state provided by an exemplary embodiment of the present application is shown;

[0078] Figure 6 A schematic diagram showing an abnormality judgment result obtained by a method for judging abnormal drilling state provided by an exemplary embodiment of the present application is shown;

[0079] Figure 7Another implementation flow chart of a method for determining abnormal drilling status provided by an exemplary embodiment of the present application is shown;

[0080] Figure 8 A structural diagram of a drilling abnormal state judgment device provided by an exemplary embodiment of the present application is shown;

[0081] Fig. 9 A schematic diagram showing the structure of a computer device corresponding to a method for determining an abnormal drilling state provided by an exemplary embodiment of the present application is shown;

[0082] Fig.10 A schematic diagram of the structure of a drilling abnormal state judgment system provided by an exemplary embodiment of the present application is shown. DETAILED DESCRIPTION

[0083] In order to make the objectives, technical solutions and advantages of the present application more clear, the implementation methods of the present application will be further described in detail below in conjunction with the accompanying drawings.

[0084] The abnormal drilling state judgment method provided in the present application can improve the efficiency and accuracy of identifying abnormal states during drilling, thereby improving the reliability of the intelligent drilling system.

[0085] Embodiment 1

[0086] Figure 1 A schematic diagram of the implementation flow of a method for determining abnormal drilling status provided by an embodiment of the present invention is shown.

[0087] See also Figure 1 The method for determining abnormal drilling status provided by an embodiment of the present invention may include steps 101 to 105.

[0088] Step 101: Obtain the drilling data to be measured.

[0089] The downhole data may include depth, drilling pressure, rotation speed, displacement, pump pressure, torque, mechanical drilling speed, hook load, inlet density, outlet density, inlet temperature, outlet temperature, inlet conductivity, outlet conductivity, etc.

[0090] In some embodiments, step 101 includes storing and normalizing the extracted drilling process data. Optionally, the data storage format is table storage, and the normalization operation is performed column by column.

[0091] Table 1 shows an example of a while-drilling data set provided by an embodiment of the present invention.

[0092] Table 1

[0093]

[0094]

[0095] Referring to Table 1, in a specific example, the judgment criteria for the drilling process in the drilling process data include:

[0096] (WOB i ≥1)&(Q i ≥1)&(H i ≥H i-1 )

[0097] Among them, WOB i is the drilling pressure of the ith row, in kN; Q i is the displacement of the ith row, in L / s; H i is the well depth of the ith row, in units of; H i-1 is the well depth of the i-1th row, in meters; i is the sequence number of the sample in the data table.

[0098] The proposed while-drilling process data is stored in a table and normalized column by column. The normalized calculation formula includes:

[0099]

[0100] Among them, x j ′ is the normalized value of the jth row in a column in the drilling data table: x j is the value of the jth row in a column in the drilling data table before normalization; x min is the minimum value of a column in the drilling data table, x max is the maximum value of a column in the drilling data table; j is the sequence number of a sample in the drilling data table.

[0101] In the process of intelligent drilling, timely identification of abnormal conditions can achieve rapid response, switch the drilling control mode from automatic to manual, avoid complex situations handled by intelligent algorithms, and ensure the reliability of the system and drilling safety. The method provided by the embodiment of the present invention provides automated anomaly detection for high-dimensional, nonlinear, and complex response mechanism systems based on unsupervised machine learning technology, improves the efficiency and automation of abnormal state detection, and can improve the reliability of the intelligent drilling system.

[0102] Step 102: performing anomaly detection on the while-drilling data to be tested based on a pre-trained autoencoder model to obtain a first detection result;

[0103] In some embodiments, the method further comprises training the autoencoder model.

[0104] In some embodiments, the method further comprises:

[0105] Obtain historical drilling data under normal working conditions as training samples;

[0106] The autoencoder model is trained based on the training samples.

[0107] Autoencoder is an unsupervised learning method with a flexible structure and good adaptability to data sets of different sizes and dimensions. It can be trained end-to-end through a special neural network structure, and can learn the inherent laws of high-dimensional data through the network structure. It can be used for data cooling and outlier detection.

[0108] Optionally, the training step includes: constructing an encoder and a decoder respectively, wherein an input parameter of the encoder and an output parameter of the decoder are normalized logging data.

[0109] Figure 2 A schematic diagram of a drilling process abnormality judgment model based on an autoencoder in the method provided by an embodiment of the present invention is shown.

[0110] See also Figure 2 In a specific example, the input parameters of the encoder and the output parameters of the decoder are normalized drilling data, including depth, drilling pressure, rotation speed, displacement, pump pressure, torque, mechanical drilling speed, hook load, inlet density, outlet density, inlet temperature, outlet temperature, inlet conductivity, and outlet conductivity.

[0111] Furthermore, the model training process includes inputting the normalized drilling process data set into the autoencoder training model. After the model training is completed, the data used for training is input into the anomaly detection model and the output results are recorded, and the restoration error and anomaly coefficient in the original sample set are calculated.

[0112] In a specific example, the original sample restoration error calculation formula includes:

[0113]

[0114] Among them, MSE restore is the restoration error of the training sample, R is the total number of rows of the training sample, row is the row number of the training sample, C is the total number of columns of the training sample, col is the column number of the training sample, and x row，col is the actual value of the training sample at a certain position, For x row，col The corresponding predicted value.

[0115] In some embodiments, the first detection result includes a first abnormality coefficient, and step 102 includes:

[0116] Based on the autoencoder model, anomaly detection is performed on the drilling data to be tested to obtain a first anomaly coefficient

[0117] The calculation formula of data anomaly coefficient includes:

[0118]

[0119] Among them, CV is the abnormality of a certain line or section of drilling data. If CV≤1, the data is in the normal range, and if CV>1, the data is in the abnormal range. The larger the CV value, the greater the abnormality of the data.

[0120] Step 103: Perform anomaly detection on the drilling data to be tested based on the pre-trained isolation forest model to obtain a second detection result.

[0121] In some embodiments, the method further comprises: training the isolation forest model.

[0122] Obtain historical drilling data under normal working conditions as training samples;

[0123] An isolation forest model is trained based on the training samples, where the isolation forest model includes a plurality of isolated trees.

[0124] Figure 3 A schematic diagram of a drilling process abnormality judgment model based on isolation forest in the method provided by an embodiment of the present invention is shown.

[0125] See also Figure 3 The Isolation Forest Algorithm is an unsupervised method that can identify outliers in data by isolating sample points. Usually, the number of outliers in a sample is small and they are far away from the main sample, so outliers can be isolated earlier in the Isolation Forest partition. In practical applications, the Isolation Forest Algorithm is suitable for anomaly detection of continuous data, so it can be used as the basis of anomaly detection models.

[0126] In a specific example, the steps of constructing an isolation forest include first training a plurality of isolated trees to form an isolation forest, inputting a normalized logging data set into each isolated tree, calculating an average height of the isolated tree, and then calculating an abnormal score value of each logging data.

[0127] In a specific example, the anomaly score calculation formula for isolation forest includes:

[0128]

[0129] Where x is the input while-drilling data, n is the dimension of the input while-drilling data, h(x) is the average height of the isolated tree, c(n) is the average value of h(x) given n, and E(h(x)) is the average value of h(x) for the set of isolated trees. The larger the s value, the greater the abnormality of the data.

[0130] Step 104: performing anomaly detection on the drilling while drilling data to be tested based on the pre-trained single-classification support vector machine model to obtain a third detection result.

[0131] In some embodiments, the method further comprises a support vector machine model training step.

[0132] Obtain historical drilling data under normal working conditions as training samples;

[0133] Training the single-classification support vector machine model based on the training samples;

[0134] In the single-classification support vector machine model, normal historical drilling data are positive examples, and abnormal historical drilling data are negative examples.

[0135] Figure 4 A schematic diagram of a drilling process abnormality judgment model based on a single-classification support vector machine in the method provided by an embodiment of the present invention is shown.

[0136] See also Figure 4 The single-class support vector machine algorithm is an unsupervised single-classification method, which is to find a hyperplane or hypersphere in the dimensional space of the data set to divide the positive examples in the sample set. The normal logging data is a positive example, represented by a value of 1, and the abnormal logging data is a negative example, represented by a value of -1.

[0137] In some embodiments, after constructing a single-classification support vector machine model, the normalized logging data is input into the model, and the model can automatically learn to generate a hyperplane or hypersphere to distinguish positive examples from negative examples, and add abnormal labels to abnormal logging data samples.

[0138] Step 105: Obtaining a drilling status abnormality judgment result based on the first detection result, the second detection result, and the third detection result.

[0139] In some embodiments, step 105 includes:

[0140] If the number of abnormalities in the first detection result, the second detection result and the third detection result is greater than a preset abnormal threshold, the drilling state is determined to be abnormal.

[0141] In some embodiments, the while-drilling data are extracted and normalized, and the processed data are respectively input into anomaly detection models, anomaly scores are calculated, and abnormal states are determined.

[0142] The downhole data may include depth, drilling pressure, rotation speed, displacement, pump pressure, torque, mechanical rotation speed, hook load, inlet density, outlet density, inlet temperature, outlet temperature, inlet conductivity, and outlet conductivity.

[0143] Optionally, after calculating the outlier value using the autoencoder and isolation forest model, an abnormal state judgment is given according to the set threshold. On the other hand, the single-classification support vector machine model directly gives the abnormal judgment.

[0144] Optionally, a voting method is used to combine the abnormal judgment results of the three models, and the final judgment result is determined according to the number of votes for abnormal judgment. Each method is given one vote, and when the number of votes for abnormal judgment is greater than 2, the path data is marked as abnormal.

[0145] Figure 5 A schematic diagram of the abnormal operating condition judgment process in the method provided by an embodiment of the present invention is shown.

[0146] See also Figure 5 , the normalized logging data set is input into the autoencoder, isolation forest model and single-classification support vector machine respectively. The autoencoder data outlier CV is used for anomaly judgment, the isolation forest output outlier S is used for anomaly judgment, and the single-classification support vector machine is used for anomaly judgment. Finally, a comprehensive anomaly judgment result is obtained based on the three anomaly judgment results.

[0147] Figure 6 A schematic diagram of an abnormality judgment result obtained by the method provided in an embodiment of the present invention is shown.

[0148] See also Figure 6 , the embodiment of the present invention can obtain the abnormality evaluation result at the current time based on various logging data.

[0149] The embodiment of the present invention provides a drilling process anomaly detection method based on the integration of autoencoder, isolation forest and single-classification support vector machine. It can establish and train anomaly recognition models based on drilling information and normal operating condition data of drilled wells, and make a comprehensive judgment on the abnormal state of unknown state data. It can effectively improve the efficiency, accuracy and automation of anomaly detection in the drilling process, and can serve as an auxiliary unit for automatic control of drilling parameters in an intelligent drilling system, effectively improving the reliability of the intelligent drilling system.

[0150] Embodiment 2

[0151] Figure 7 A schematic diagram of another implementation flow of the method for determining abnormal drilling status provided by an embodiment of the present invention is shown.

[0152] See also Figure 7 In a specific example, the implementation process of the drilling abnormal state judgment method provided by the embodiment of the present invention is as follows.

[0153] First, the dataset is constructed and preprocessed, and the anomaly detection model is established for the dataset.

[0154] On the one hand, anomaly detection of the drilling process is carried out based on the anomaly detection model, and on the other hand, an automatic judgment system for abnormal conditions in the drilling process is constructed.

[0155] In summary, the embodiment of the present invention proposes a drilling process anomaly detection method based on autoencoder, isolation forest, and single-classification support vector machine, which can train the autoencoder based on the logging information of the normal working condition of the drilled section to establish an anomaly detection model, input the logging data to be detected into the anomaly detection model, calculate the restoration error between the model output result and the actual data, calculate the anomaly coefficient based on the restoration error, and judge the abnormal state and level in combination with the threshold. The method has a high degree of automation and good stability, and can effectively improve the efficiency and accuracy of abnormal state identification during drilling, and improve the reliability of the drilling system.

[0156] Embodiment 3

[0157] Figure 8 A schematic diagram of the structure of a drilling abnormal state judgment device provided by an embodiment of the present invention is shown.

[0158] See also Figure 8 The drilling abnormal state judgment device provided by the embodiment of the present invention may include:

[0159] The data acquisition module 201 is used to acquire the drilling data to be measured;

[0160] An autoencoder detection module 202 is used to perform anomaly detection on the drilling data to be tested based on a pre-trained autoencoder model to obtain a first detection result;

[0161] An isolation forest detection module 203 is used to perform anomaly detection on the drilling data to be tested based on a pre-trained isolation forest model to obtain a second detection result;

[0162] A support vector machine detection module 204 is used to perform anomaly detection on the drilling data to be tested based on a pre-trained single-classification support vector machine model to obtain a third detection result;

[0163] The abnormality judgment module 205 is used to obtain a drilling status abnormality judgment result based on the first detection result, the second detection result and the third detection result.

[0164] In some embodiments, the apparatus further comprises an autoencoder model training module for:

[0165] Obtain historical drilling data under normal working conditions as training samples;

[0166] Training the autoencoder model based on the training samples;

[0167] The autoencoder model includes a restoration error calculation formula;

[0168] The restoration error calculation formula includes:

[0169]

[0170] Among them, MSE restore is the restoration error of the training sample, R is the total number of rows of the training sample, row is the row number of the training sample, C is the total number of columns of the training sample, col is the column number of the training sample, and x row，col is the actual value of the training sample at a certain position, For x row，col The corresponding predicted value.

[0171] In some embodiments, the first detection result includes a first abnormal coefficient, and the autoencoder detection module 202 is specifically used to:

[0172] Performing anomaly detection on the drilling while drilling data to be tested based on the autoencoder model to obtain a first anomaly coefficient;

[0173] The first abnormal coefficient calculation formula includes:

[0174]

[0175] Among them, CV is the first abnormal coefficient, MSE is the mean square error, restore To restore the error.

[0176] In some embodiments, the apparatus further comprises an isolation forest model training module, configured to:

[0177] Obtain historical drilling data under normal working conditions as training samples;

[0178] Training an isolation forest model based on the training samples, wherein the isolation forest model includes a plurality of isolated trees;

[0179] The isolation forest model includes an anomaly score calculation formula:

[0180]

[0181] Where x is the input LWD data, n is the dimension of the input LWD data, h(x) is the average height of the isolated tree, c(n) is the average value of h(x) given n, and E(h(x)) is the average value of h(x) of the isolated tree set.

[0182] In some embodiments, the apparatus further comprises a support vector machine model training module, configured to:

[0183] Obtain historical drilling data under normal working conditions as training samples;

[0184] Training the single-classification support vector machine model based on the training samples;

[0185] In the single-classification support vector machine model, normal historical drilling data are positive examples, and abnormal historical drilling data are negative examples.

[0186] In some embodiments, the abnormality determination module 205 is specifically used to:

[0187] The abnormal drilling status judgment result obtained based on the first detection result, the second detection result and the third detection result includes:

[0188] If the number of abnormalities in the first detection result, the second detection result and the third detection result is greater than a preset abnormal threshold, the drilling state is determined to be abnormal.

[0189] In summary, the device provided by the embodiment of the present invention has a high degree of automation and good stability, and can improve the efficiency and accuracy of identifying abnormal conditions during drilling, thereby improving the reliability of the intelligent drilling system.

[0190] Embodiment 4:

[0191] Fig. 9 A schematic diagram of the structure of a computer device provided by an exemplary embodiment of the present application is shown, and the computer device includes:

[0192] The processor 301 includes one or more processing cores. The processor 301 executes various functional applications and data processing by running software programs and modules.

[0193] The receiver 302 and the transmitter 303 can be implemented as a communication component, and the communication component can be a communication chip. Optionally, the communication component can be implemented to include a signal transmission function. That is, the transmitter 303 can be used to transmit a control signal to the image acquisition device and the scanning device, and the receiver 302 can be used to receive the corresponding feedback instruction.

[0194] The memory 304 is connected to the processor 301 via a bus 305 .

[0195] The memory 304 may be used to store at least one instruction, and the processor 301 may be used to execute the at least one instruction to implement steps 101 to 102 in the above control method embodiment.

[0196] Those skilled in the art will understand that Fig. 9 It is only an example of a computer device and does not constitute a limitation of the computer device. The computer device may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, the computer device may also include a network access device, etc.

[0197] The processor 301 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0198] The memory 304 may be an internal storage unit of the computer device, such as a hard disk or memory of the computer device. The memory 304 may also be an external storage device of the computer device, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the computer device. Further, the memory 304 may also include both an internal storage unit of the computer device and an external storage device. The memory 304 is used to store the computer program and other programs and data required by the terminal device. The memory 304 may also be used to temporarily store data that has been output or is to be output.

[0199] Embodiment 5

[0200] An embodiment of the present application also provides a drilling abnormal state judgment system, including the computer device as described above.

[0201] Fig.10 A schematic diagram of the system structure provided by an embodiment of the present invention is shown.

[0202] See also Fig.10 In a specific example, the automatic judgment system of abnormal state of drilling process may include a data acquisition unit, a data storage unit, an online data preprocessing unit, an offline data preprocessing unit, an abnormality detection model building unit, an abnormality detection unit and a result visualization unit.

[0203] The data acquisition unit is used to collect the drilling data required for working condition judgment through well site sensors, including depth, drilling pressure, rotation speed, displacement, pump pressure, torque, mechanical drilling speed, hook load, inlet density, outlet density, inlet temperature, outlet temperature, inlet conductivity, and outlet conductivity.

[0204] The online data processing unit user obtains the real-time data obtained by the data acquisition unit, and performs drilling condition identification and data normalization processing.

[0205] The offline data preprocessing unit is used to obtain a large amount of historical data in the data storage unit for working condition identification and data normalization processing.

[0206] The integrated anomaly detection model building unit is used to train the anomaly detection model based on the extracted normal operating condition data and record the restoration error of the training data.

[0207] The anomaly detection unit is used to input the data from the online data preprocessing unit into the integrated anomaly detection model, calculate the anomaly coefficient of the data, make anomaly judgments, and give a comprehensive judgment result.

[0208] The result visualization unit is used to visualize the process curve and real-time data, mark the abnormal data section of a well section, and calculate and judge the abnormal values ​​of the latest set of real-time data.

[0209] Embodiment 6

[0210] An embodiment of the present application also provides a computer-readable storage medium, which stores at least one instruction, at least one program, code set or instruction set, so as to be loaded and executed by a processor to implement the above method.

[0211] Optionally, the computer readable storage medium may include: a read only memory (ROM), a random access memory (RAM), a solid state drive (SSD), or an optical disk, etc. Among them, the random access memory may include a resistance random access memory (ReRAM) and a dynamic random access memory (DRAM).

[0212] Embodiment 7

[0213] The present application also provides a computer program product or a computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes any of the methods described in the above embodiments.

[0214] The serial numbers of the above-mentioned embodiments of the present application are for description only and do not represent the pros and cons of the implementation.

[0215] Those skilled in the art can understand that all or part of the steps of implementing the above-mentioned embodiments can be completed by hardware, or by instructing the relevant hardware to complete the steps through a program. The program can be stored in a computer-readable storage medium, and the storage medium mentioned above can be a read-only memory, a disk or an optical disk, etc. Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0216] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0217] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0218] The computer program includes computer program code, which may be in source code form, object code form, executable file or some intermediate form, etc. The computer readable medium may include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium, etc. It should be noted that the content contained in the computer readable medium may be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer readable media do not include electric carrier signals and telecommunication signals.

[0219] The embodiments described above are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. A method for determining abnormal drilling status, characterized in that: The method comprises: Obtain the drilling data to be tested; Performing anomaly detection on the drilling while drilling data to be tested based on a pre-trained autoencoder model to obtain a first detection result; Performing anomaly detection on the drilling data to be tested based on a pre-trained isolation forest model to obtain a second detection result; Performing anomaly detection on the drilling while drilling data to be tested based on a pre-trained single-classification support vector machine model to obtain a third detection result; A drilling status abnormality judgment result is obtained based on the first detection result, the second detection result and the third detection result.

2. The method according to claim 1, characterized in that The method further comprises: Obtain historical drilling data under normal working conditions as training samples; Training the autoencoder model based on the training samples; The autoencoder model includes a restoration error calculation formula; The restoration error calculation formula includes: Among them, MSE restore is the restoration error of the training sample, R is the total number of rows of the training sample, row is the row number of the training sample, C is the total number of columns of the training sample, col is the column number of the training sample, and x row,col is the actual value of the training sample at a certain position, For x row,col The corresponding predicted value.

3. The method according to claim 2, characterized in that The first detection result includes a first abnormality coefficient. The pre-trained autoencoder model performs abnormality detection on the drilling data to be tested to obtain the first detection result, including: Performing anomaly detection on the drilling while drilling data to be tested based on the autoencoder model to obtain a first anomaly coefficient; The first abnormal coefficient calculation formula includes: Among them, CV is the first abnormal coefficient, MSE is the mean square error, restore To restore the error.

4. The method according to claim 1, characterized in that The method further comprises: Obtain historical drilling data under normal working conditions as training samples; Training an isolation forest model based on the training samples, wherein the isolation forest model includes a plurality of isolated trees; The isolation forest model includes an anomaly score calculation formula: Where x is the input LWD data, n is the dimension of the input LWD data, h(x) is the average height of the isolated tree, c(n) is the average value of h(x) given n, and E(h(x)) is the average value of h(x) of the isolated tree set.

5. The method according to claim 1, characterized in that The method further comprises: Obtain historical drilling data under normal working conditions as training samples; Training the single-classification support vector machine model based on the training samples; In the single-classification support vector machine model, normal historical drilling data are positive examples, and abnormal historical drilling data are negative examples.

6. The method according to any one of claims 1 to 5, characterized in that: The abnormal drilling status judgment result obtained based on the first detection result, the second detection result and the third detection result includes: If the number of abnormalities in the first detection result, the second detection result and the third detection result is greater than a preset abnormal threshold, the drilling state is determined to be abnormal.

7. A drilling abnormal state judgment device, characterized in that: The device comprises: A data acquisition module, used to acquire the drilling data to be tested; An autoencoder detection module is used to perform anomaly detection on the drilling data to be tested based on a pre-trained autoencoder model to obtain a first detection result; An isolation forest detection module is used to perform anomaly detection on the drilling data to be tested based on a pre-trained isolation forest model to obtain a second detection result; A support vector machine detection module, used for performing anomaly detection on the drilling data to be tested based on a pre-trained single-classification support vector machine model to obtain a third detection result; The abnormality judgment module is used to obtain a drilling status abnormality judgment result based on the first detection result, the second detection result and the third detection result.

8. A computer device, characterized in that: The computer device includes a processor and a memory, wherein the memory stores at least one instruction, at least one program, a code set or an instruction set, and the at least one instruction, at least one program, a code set or an instruction set is loaded and executed by the processor to implement the method for determining an abnormal drilling state as described in any one of claims 1 to 6.

9. A drilling abnormal state judgment system, characterized in that: Comprising a computer device as claimed in claim 8.

10. A computer-readable storage medium, characterized in that: The readable memory medium stores at least one instruction, at least one program, code set or instruction set, and the at least one instruction, at least one program, code set or instruction set is loaded and executed by the processor to implement the method for determining abnormal drilling status as described in any one of claims 1 to 6.

Images