A method and apparatus for optimizing drilling parameters

Abstract

The application discloses a drilling parameter optimization method and device, relates to the technical field of oil and gas well drilling, and aims to specifically optimize the parameter values of drilling parameters under complex working conditions. The main technical scheme of the application is as follows: a well logging time series data set is received; the well logging time series data set is processed through an empirical mode decomposition algorithm to generate three groups of feature data of each drilling parameter; the three groups of feature data of each drilling parameter are input into a convolutional neural network classification model to output a downhole state classification result; an algorithm for determining a preset optimization range is used to generate an optimization range of the parameter values of each drilling parameter under the downhole state classification result; within the optimization range of the parameter values of each drilling parameter, drilling parameter value combinations between the drilling parameters are input into a mechanical drilling speed prediction model to obtain mechanical drilling speed prediction values of the drilling parameter value combinations; and a particle swarm optimization algorithm is used to maximize the mechanical drilling speed prediction values to iteratively process the drilling parameter value combinations, so that an optimal drilling parameter value combination is obtained.

Description

Technical Field

[0001] This invention relates to the field of oil and gas well drilling technology, and in particular to a method and apparatus for optimizing drilling parameters. Background Technology

[0002] In oil and gas drilling operations, drilling parameters (such as drilling pressure, rotation speed, and displacement) have a direct impact on drilling efficiency, cost, and safety. Therefore, it is necessary to optimize the values ​​of drilling parameters to improve drilling efficiency, reduce costs, and ensure operational safety.

[0003] Currently, a common method for optimizing drilling parameters is to collect historical data and manually analyze a large amount of historical well data. Using indicative data such as drill bit footage and drilling machinery speed of penetration, a range of matching engineering parameters is recommended, and the analysis results have a certain degree of generality. This also means that when facing complex operating conditions such as drill bit vibration and tubing string pressure, this method lacks specificity and is difficult to effectively optimize drilling parameter values ​​under such complex conditions. Summary of the Invention

[0004] In view of the above problems, the present invention provides a drilling parameter optimization method and apparatus, the main purpose of which is to optimize the parameter values ​​of drilling parameters in a targeted manner under complex working conditions.

[0005] To solve the above-mentioned technical problems, the present invention proposes the following solution:

[0006] In a first aspect, the present invention provides a drilling parameter optimization method, the method comprising:

[0007] Receive logging time-series dataset transmitted by logging tool, wherein the dataset contains at least a numerical sequence of drilling parameters, including drilling pressure and rotation speed, and a timestamp corresponding to each value;

[0008] The logging time series dataset is decomposed into three sets of feature data corresponding to each drilling parameter by using the empirical mode decomposition algorithm to generate three sets of feature data, including low-frequency feature data, medium-frequency feature data and high-frequency feature data.

[0009] The three sets of feature data corresponding to each drilling parameter are input into a pre-trained convolutional neural network classification model to output downhole state classification results, which include normal drilling state, abnormal drill bit vibration state, and tubing string pressure state.

[0010] The algorithm determines the parameter value optimization range corresponding to each drilling parameter in the downhole state classification result based on the preset optimization range;

[0011] Within the optimization range of the parameter values ​​corresponding to each drilling parameter, the combination of drilling parameter values ​​between each drilling parameter is input into the pre-trained mechanical drilling rate prediction model to obtain the mechanical drilling rate prediction value corresponding to each combination of drilling parameter values.

[0012] The particle swarm optimization algorithm is used to iteratively calculate the combination of drilling parameter values ​​among various drilling parameters with the goal of maximizing the predicted mechanical drilling rate, so as to obtain the optimal combination of drilling parameter values ​​among various drilling parameters.

[0013] In a second aspect, the present invention provides a drilling parameter optimization device, the device comprising:

[0014] The data receiving unit is used to receive the logging time-series dataset transmitted by the logging instrument. The dataset includes at least a numerical sequence of drilling parameters, including drilling pressure and rotation speed, and a timestamp corresponding to each value.

[0015] The data decomposition unit is used to perform multi-scale decomposition processing on the logging time series dataset received by the data receiving unit through the empirical mode decomposition algorithm, and generate three sets of feature data corresponding to each drilling parameter. The feature data includes low-frequency feature data, medium-frequency feature data and high-frequency feature data.

[0016] The data input unit is used to input the three sets of feature data corresponding to each drilling parameter obtained by the data decomposition unit into the pre-trained convolutional neural network classification model to output the downhole state classification result. The classification result includes normal drilling state, abnormal drill bit vibration state, and tubing string pressure state.

[0017] The range determination unit is used to generate the parameter value optimization range corresponding to each drilling parameter in the downhole state classification result obtained by the data input unit according to the preset optimization range determination algorithm;

[0018] The drilling speed prediction unit is used to input the drilling parameter value combination between each drilling parameter into the pre-trained mechanical drilling speed prediction model within the optimization range of the parameter values ​​corresponding to each drilling parameter obtained by the range determination unit, so as to obtain the mechanical drilling speed prediction value corresponding to each drilling parameter value combination.

[0019] The parameter optimization unit is used to perform iterative calculations on the combination of drilling parameter values ​​among various drilling parameters with the objective of maximizing the mechanical drilling rate predicted by the drilling rate prediction unit, using a particle swarm optimization algorithm, to obtain the optimal combination of drilling parameter values ​​among various drilling parameters.

[0020] To achieve the above objectives, according to a third aspect of the present invention, a storage medium is provided, the storage medium including a stored program, wherein, when the program is executed, the device where the storage medium is located is controlled to perform the drilling parameter optimization method of the first aspect described above.

[0021] To achieve the above objectives, according to a fourth aspect of the present invention, a processor is provided for running a program, wherein the program executes the drilling parameter optimization method of the first aspect described above.

[0022] To achieve the above objectives, according to a fifth aspect of the present invention, a computer program product is provided, the computer program product comprising a computer program stored in a computer-readable storage medium, at least one processor being able to read the computer program from the computer-readable storage medium, and the at least one processor executing the computer program being able to execute the drilling parameter optimization method of the first aspect described above.

[0023] By employing the above technical solution, this invention provides a drilling parameter optimization method and apparatus. The logging instrument periodically transmits a logging time-series dataset, including numerical sequences of drilling parameters such as drilling pressure and rotational speed, along with timestamps corresponding to each value. Subsequently, an empirical mode decomposition algorithm is used to perform multi-scale decomposition processing on the logging time-series data, generating low-frequency, mid-frequency, and high-frequency feature data corresponding to each drilling parameter. Then, the three sets of feature data corresponding to each drilling parameter are input into a pre-trained convolutional neural network classification model, outputting downhole state classification results, including normal drilling state, abnormal drill bit vibration state, and tubing string pressure state. The parameter value optimization range for each drilling parameter corresponding to the downhole state classification results generated by the algorithm is determined according to a preset optimization range. Within the parameter value optimization range of each drilling parameter, the parameter value combinations between each drilling parameter are input into a pre-trained mechanical drilling rate prediction model to obtain the mechanical drilling rate prediction value corresponding to each drilling parameter value combination. Finally, a particle swarm optimization algorithm is employed to iteratively calculate the parameter value combinations among various drilling parameters with the objective of maximizing the predicted mechanical drilling rate, thereby obtaining the optimal parameter value combinations among all drilling parameters. Compared with existing technologies, this invention can make targeted adjustments to the drilling parameter value combinations based on specific downhole conditions, thus helping to maintain optimal drilling efficiency under corresponding downhole conditions.

[0024] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, and in order to make the above and other objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0025] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings:

[0026] Figure 1 A flowchart of a drilling parameter optimization method provided by an embodiment of the present invention is shown;

[0027] Figure 2 A flowchart of another drilling parameter optimization method provided by an embodiment of the present invention is shown;

[0028] Figure 3 This diagram illustrates a block diagram of a drilling parameter optimization device provided in an embodiment of the present invention.

[0029] Figure 4 A block diagram of another drilling parameter optimization device provided in an embodiment of the present invention is shown. Detailed Implementation

[0030] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0031] To address the limitations of current drilling parameter optimization methods, the inventors dedicated significant effort to creative research, ultimately proposing a novel method for drilling parameter optimization. This method completely abandons existing practices. Its implementation relies on a drilling parameter optimization system capable of communicating with the drilling site logging instrument. This system can receive real-time logging time-series datasets transmitted by the logging instrument and use this data for subsequent analysis and processing to optimize the drilling parameters.

[0032] Next, combined Figure 1 The drilling parameter optimization method proposed in this invention is described below, and its specific execution steps are as follows: Figure 1 As shown, it includes:

[0033] 101. Receive the logging time series dataset transmitted by the logging instrument.

[0034] 102. The logging time series dataset is decomposed into three sets of feature data corresponding to each drilling parameter by using the empirical mode decomposition algorithm.

[0035] In this embodiment of the invention, the logging instrument can periodically transmit logging time-series datasets at intervals of 1 to 3 seconds.

[0036] Each logging time-series dataset not only contains numerical sequences of key drilling parameters such as drilling pressure (kN) and rotational speed (RPM), but also records the corresponding data acquisition timestamps. Specifically, the drilling parameters covered in the logging time-series datasets are quite comprehensive, including but not limited to:

[0037] Well depth (m), drill bit position (m), hook height (m), hook load (kN), drilling pressure (kN), rotary table speed (RPM), torque (kN•m), riser pressure (MPa), drilling time (min / m), instantaneous drilling rate (m / hr), inlet density (g / cm³), outlet density (g / cm³), inlet flow rate (L / s), and outlet flow rate (L / s), etc.

[0038] After successfully receiving the well logging time series dataset, the system will use the Empirical Mode Decomposition (EMD) algorithm to perform multi-scale decomposition processing on the well logging time series data. Through this processing, three sets of feature data can be generated for each drilling parameter: low-frequency feature data, medium-frequency feature data, and high-frequency feature data.

[0039] 103. Input the three sets of feature data corresponding to each drilling parameter into the pre-trained convolutional neural network classification model to output the downhole state classification result.

[0040] The downhole condition classification results specifically cover three categories: normal drilling condition, abnormal drill bit vibration condition, and tubing string pressure condition.

[0041] In this embodiment, the low-frequency, medium-frequency, and high-frequency feature data corresponding to each drilling parameter can be input into the convolutional neural network classification model simultaneously. It is important to note that during the operation of the convolutional neural network model, the three sets of feature data for each drilling parameter will sequentially undergo key processing steps such as feature concatenation and feature integration. This is because only by conducting in-depth analysis and processing of the integrated features can a more accurate downhole condition classification result be obtained.

[0042] In addition to the method described above for obtaining downhole condition classification results using a convolutional neural network model, this invention also provides other implementation approaches. Specifically, statistical indicators such as the mean, standard deviation, kurtosis, and spectral entropy of each drilling parameter can be extracted based on three sets of feature data for each parameter, and these indicators can be used to comprehensively characterize the downhole condition classification results. In this process, a mapping table of correspondence between these statistical indicators and the downhole condition classification results can be pre-constructed, and then the calculated indicators are matched with the mapping table to obtain the final downhole condition classification result.

[0043] 104. Determine the parameter value optimization range for each drilling parameter corresponding to the downhole state classification results generated by the algorithm based on the preset optimization range.

[0044] In this embodiment, the step of generating the parameter value optimization range for each drilling parameter can be achieved through various methods, as follows:

[0045] The first method involves pre-constructing an optimization rule system corresponding to each downhole condition. Then, based on these optimization rules and the numerical sequences of each drilling parameter, preliminary optimization ranges for each drilling parameter are generated. On this basis, multi-dimensional constraints are applied to the preliminary optimization ranges of each parameter. After comprehensive consideration and screening, the final optimization range for each drilling parameter value is determined.

[0046] The second method involves pre-setting the basic optimization range for each drilling parameter, and then adaptively adjusting the parameter values ​​based on the actual downhole conditions. For example, when strong drill bit vibration is detected, the upper limit of the drilling pressure and rotation speed can be appropriately lowered from the original optimization range to reduce equipment wear and avoid the risk of failure caused by excessive vibration.

[0047] Furthermore, when downhole conditions are relatively stable, the optimization range of various drilling parameters can be appropriately widened, such as pursuing maximum drilling speed to improve drilling efficiency. Simultaneously, corresponding dynamic adjustment strategies can be formulated based on different combinations of operating conditions. For example, reducing rotational speed or drilling pressure when vibration is significant, and adjusting the tubing load range when the risk of pressure buildup increases, can ensure stable and efficient drilling operations.

[0048] 105. Within the optimization range of the parameter values ​​corresponding to each drilling parameter, input the combination of drilling parameter values ​​between each drilling parameter into the pre-trained mechanical drilling rate prediction model to obtain the mechanical drilling rate prediction value corresponding to each combination of drilling parameter values.

[0049] 106. Using the particle swarm optimization algorithm, with the goal of maximizing the predicted mechanical drilling rate, iterative calculations are performed on the combination of drilling parameter values ​​among various drilling parameters to obtain the optimal combination of drilling parameter values ​​among various drilling parameters.

[0050] In this invention, a mapping relationship between various drilling parameters and mechanical drilling rate can be established in advance based on a large amount of historical logging data, and a Long Short-Term Memory (LSTM) network model can be trained based on this. This model learns the influence of drilling parameters on mechanical drilling rate under different operating conditions, thereby obtaining a well-trained mechanical drilling rate prediction model.

[0051] Subsequently, in step 105, multiple combinations of drilling parameter values ​​can be randomly generated within the optimization range corresponding to each drilling parameter. These combinations of drilling parameter values ​​are then input into a pre-trained mechanical drilling rate prediction model to obtain the predicted mechanical drilling rate value for each combination of parameter values.

[0052] Next, a particle swarm optimization algorithm is employed to iteratively calculate the combinations of drilling parameters with the goal of maximizing the predicted mechanical drilling rate. By continuously adjusting and optimizing these combinations, the optimal combination of drilling parameters that maximizes the mechanical drilling rate is ultimately found. This method not only considers the complex influencing factors under different operating conditions but also ensures optimal drilling efficiency in actual operation through optimization algorithms.

[0053] Based on the above Figure 1 As can be seen from the implementation method, the drilling parameter optimization method provided by this invention involves the logging instrument periodically transmitting a logging time-series dataset, which includes numerical sequences of drilling parameters such as drilling pressure and rotational speed, along with the timestamps corresponding to each value. Subsequently, the empirical mode decomposition algorithm is used to perform multi-scale decomposition processing on the logging time-series data, generating low-frequency, mid-frequency, and high-frequency feature data corresponding to each drilling parameter. Then, the three sets of feature data corresponding to each drilling parameter are input into a pre-trained convolutional neural network classification model, outputting downhole state classification results, including normal drilling state, abnormal drill bit vibration state, and tubing string pressure state. The parameter value optimization range for each drilling parameter corresponding to the downhole state classification results generated by the algorithm is determined according to a preset optimization range. Within the parameter value optimization range of each drilling parameter, the parameter value combinations between each drilling parameter are input into a pre-trained mechanical drilling rate prediction model to obtain the mechanical drilling rate prediction value corresponding to each drilling parameter value combination. Finally, a particle swarm optimization algorithm is employed to iteratively calculate the parameter value combinations among various drilling parameters with the objective of maximizing the predicted mechanical drilling rate, thereby obtaining the optimal parameter value combinations among all drilling parameters. Compared with existing technologies, this invention can make targeted adjustments to the drilling parameter value combinations based on specific downhole conditions, thus helping to maintain optimal drilling efficiency under corresponding downhole conditions.

[0054] Furthermore, as a response to Figure 1 Further refinement and extension of the illustrated embodiment, this invention also provides another drilling parameter optimization method, such as... Figure 2 As shown, the specific steps are as follows:

[0055] 201. Receive the logging time series dataset transmitted by the logging instrument.

[0056] The implementation method of step 201 is the same as that of step 101, and can achieve the same technical effect and solve the same technical problem, so it will not be repeated here.

[0057] 202. The logging time series dataset is decomposed into three sets of feature data corresponding to each drilling parameter by using the empirical mode decomposition algorithm.

[0058] In this embodiment, the following steps can be performed on each drilling parameter in the logging time series dataset to obtain three sets of feature data corresponding to each drilling parameter. The steps are as follows:

[0059] 1. Identify local maxima and local minima in a numerical sequence;

[0060] 2. Use spline interpolation to obtain the upper envelope formed by each local maxima and the lower envelope formed by each local minima;

[0061] 3. Calculate the average value between the values ​​at the same position between the upper and lower envelopes to obtain the average value sequence (e.g., the average value between the first point on the upper envelope and the first point on the lower envelope).

[0062] 4. Subtract the average value sequence point by point from the numerical sequence corresponding to the drilling parameters to obtain a new numerical sequence;

[0063] 5. Extract three sets of feature data corresponding to drilling parameters from the new numerical sequence according to the preset feature extraction conditions. This step also includes the following sub-steps:

[0064] (1) Check whether the new numerical sequence meets the feature extraction conditions. The feature extraction conditions are the difference between the number of maximum and minimum points in the numerical sequence (the maximum and minimum points here are the maximum and minimum points in the new numerical sequence) and the number of zero-crossing points by a preset number;

[0065] (2) If satisfied, the new numerical sequence is determined as high-frequency feature data;

[0066] (3) Subtract the high-frequency feature data from the numerical sequence corresponding to the drilling parameters to obtain the remaining numerical sequence, and repeat (1)-(3) of steps 1 to 5 above for the remaining numerical sequence until the medium-frequency feature data and low-frequency feature data are obtained.

[0067] 203. Input the three sets of feature data corresponding to each drilling parameter into the pre-trained convolutional neural network classification model to output the downhole state classification result.

[0068] In this step, the three sets of feature data corresponding to each drilling parameter are normalized to obtain standardized feature data for each parameter. Then, the standardized feature data of each parameter are concatenated into a multi-dimensional feature vector in a preset order. This multi-dimensional feature vector is then input into a pre-trained convolutional neural network classification model, where features are extracted and dimensionality is gradually reduced through convolutional layers. Finally, in the last fully connected layer of the convolutional neural network classification model, the extracted features are integrated to determine the corresponding classification, and the downhole state classification result is output.

[0069] It should be noted that there can be multiple fully connected layers. Within these layers, linear transformations and non-linear activation functions can be applied to the features, thereby integrating different feature information. Finally, the output layer (such as a Softmax layer) calculates the probability distribution of each category, and the category with the highest probability is selected as the downhole state classification result.

[0070] 204. Obtain the set of preset optimization rules corresponding to the downhole state classification results.

[0071] 205. Based on the triangular membership function, combined with the numerical sequence of each drilling parameter and the corresponding set of preset optimization rules, generate the parameter value optimization range for each drilling parameter.

[0072] In steps 204-205, a set of preset optimization rules corresponding to the downhole state classification results can be obtained. Then, based on the triangular membership function, combined with the numerical sequence of each drilling parameter and the corresponding set of preset optimization rules, a preliminary optimization range for each drilling parameter is generated. Finally, multi-dimensional constraints are applied to the preliminary optimization range of each parameter to obtain the parameter value optimization range for each drilling parameter.

[0073] Among them, the downhole state classification results are: normal drilling corresponding to the first preset optimization rule set, abnormal drill bit vibration state corresponding to the second preset optimization rule set, and tubing string pressure state corresponding to the third preset optimization rule set.

[0074] Among them, the preset optimization rule set is a parameter adjustment strategy predefined according to different downhole conditions (normal drilling, abnormal drill bit vibration, and tubing string pressure). Its core is to map the downhole condition classification results into the optimization direction and range of drilling parameters through mathematical rules and expert experience.

[0075] Each rule set contains the following information: Target parameters: Drilling parameters that need to be optimized (such as drilling pressure, rotation speed, pump pressure, etc.). Adjustment direction: Whether the parameter should be increased, decreased, or kept stable.

[0076] Membership function type: Fuzzy logic functions (such as triangular membership functions) used to generate parameter optimization intervals.

[0077] Experience weighting: the priority of adjusting different parameters (e.g., prioritizing the reduction of drilling pressure when the drill bit vibrates).

[0078] For example, the first preset set of optimization rules corresponding to a normal drilling state may include:

[0079] Rule objective: Maximize the rate of drilling within a safe range;

[0080] Drilling pressure: Allowed to be adjusted within ±10% of the historical average;

[0081] Rotation speed: Match the optimal rotation speed range based on lithology;

[0082] Pump pressure: Maintain stability to reduce fluctuations.

[0083] For example, the second set of preset optimization rules for abnormal drill bit vibration may include:

[0084] Rule objective: To suppress vibration and prevent drill bit damage;

[0085] Drilling pressure: Forced to be reduced to 70%~90% of the historical average;

[0086] Rotational speed: Avoid the resonant frequency range;

[0087] Pump pressure: Increase by 10%~20% to enhance well bottom cleaning.

[0088] For example, the third set of preset optimization rules corresponding to the tubing support pressure may include:

[0089] Rule objective: Reduce friction and restore free movement of the tubing;

[0090] Drilling pressure: Reduced to 50%~70% of historical average;

[0091] Rotation speed: Reduce to the minimum safe value;

[0092] Pump pressure: Increase by 15%~25% to lubricate the well wall.

[0093] Next, let's take the drilling pressure parameter as an example to illustrate the above scheme:

[0094] Downhole condition classification result: Abnormal drill bit vibration (triggered the second preset optimization rule set).

[0095] Target parameter: WOB (Weight on Drill);

[0096] Real-time data (i.e., the value with the closest timestamp to the current time in the latest received logging time series data): WOB = 18 tons (historical average 20 tons);

[0097] Call the second preset set of optimization rules, which may include: Drilling pressure (WOB) adjustment rules: rule objective: reduce drilling pressure to reduce vibration energy.

[0098] Membership function: Triangular function, vertex at 80% of the historical mean (16 tons), left boundary 14 tons, right boundary 18 tons.

[0099] Optimization range generation for drill pressure (WOB): Real-time data is processed using a triangular membership function (current WOB = 18 tons, historical average 20 tons): Left boundary: 14 tons (70% of historical average), vertex: 16 tons (80% of historical average), right boundary: 18 tons (90% of historical average). Preliminary optimization range: 14 tons ≤ WOB ≤ 18 tons, optimal value is close to 16 tons.

[0100] Based on the initial optimization range, the following actual constraints are added: Equipment limitation: the maximum drilling pressure of the drilling rig is 25 tons; Safety margin: a 10% safety margin must be retained when adjusting the drilling pressure. Therefore, the final parameter optimization range is: Drilling pressure (WOB): 14 tons ≤ WOB ≤ 16.2 tons (the original initial range was 14-18 tons, and the right boundary was corrected to 16.2 tons after adding the safety margin).

[0101] 206. Within the optimization range of the parameter values ​​corresponding to each drilling parameter, input the combination of drilling parameter values ​​between each drilling parameter into the pre-trained mechanical drilling rate prediction model to obtain the mechanical drilling rate prediction value corresponding to each combination of drilling parameter values.

[0102] The implementation method of step 206 is the same as that of step 105, and can achieve the same technical effect and solve the same technical problem, so it will not be repeated here.

[0103] 207. Using the particle swarm optimization algorithm, with the goal of maximizing the predicted mechanical drilling rate, iterative calculations are performed on the combination of drilling parameter values ​​among various drilling parameters to obtain the optimal combination of drilling parameter values ​​among various drilling parameters.

[0104] In this step, the particle swarm can be initialized first. Each particle in the particle swarm corresponds to a set of drilling parameter values ​​at each position. The initial position of each particle is randomly generated by each drilling parameter within the corresponding parameter value optimization range.

[0105] Subsequently, the predicted mechanical drilling rate corresponding to each combination of drilling parameter values ​​can be used as the fitness value of each particle at the corresponding position. Then, the initial position of each particle is recorded as the particle's individual historical best position, and the position with the highest fitness value is selected from all the individual historical best positions of all particles to determine the global historical best position.

[0106] Furthermore, the movement speed of each particle can be adjusted according to the velocity update formula of the particle swarm optimization algorithm. Based on the adjusted movement speed, the position of each particle is updated. It is important to note in this step that the updated particle positions represent drilling parameter values ​​that are within the optimization range of each drilling parameter.

[0107] Then, the fitness value corresponding to the drilling parameter value combination represented by each particle at the updated position is calculated, and the individual historical best position of each particle and the global historical best position of the particle swarm are updated according to the fitness value corresponding to the particle at the updated position. Finally, the above steps are repeated until the preset maximum number of iterations is reached, and the drilling parameter value combination corresponding to the global historical best position recorded by the particle swarm during the iteration process is determined as the optimal parameter value combination.

[0108] The speed update formula is as follows:

[0109] (Formula 1)

[0110] In the formula, vnew is the adjusted moving speed, w is the inertia weight, c1 is the individual learning factor, c2 is the global learning factor, rand() is a random number in the range [0,1], vold is the particle's current speed, pbest is the individual optimal position, gbest is the global historical optimal position, and xold is the combination of drilling parameter values ​​represented by the particle at its current position.

[0111] Furthermore, as a response to the above Figure 1 In addition to the method shown, this embodiment of the invention also provides a drilling parameter optimization device for optimizing the above-mentioned drilling parameters. Figure 1 The method shown is implemented accordingly. This device embodiment corresponds to the foregoing method embodiment. For ease of reading, this device embodiment will not repeat the details of the foregoing method embodiment, but it should be clear that the device in this embodiment can implement all the contents of the foregoing method embodiment. Figure 3 As shown, the device includes:

[0112] The data receiving unit 301 is used to receive the logging time series dataset transmitted by the logging instrument. The dataset includes numerical sequences of drilling parameters such as drilling pressure and rotation speed, as well as timestamps corresponding to each value.

[0113] The data decomposition unit 302 is used to perform multi-scale decomposition processing on the logging time series dataset received by the data receiving unit 301 using the empirical mode decomposition algorithm to generate three sets of feature data corresponding to each drilling parameter. The feature data includes low-frequency feature data, medium-frequency feature data and high-frequency feature data.

[0114] The data input unit 303 is used to input the three sets of feature data corresponding to each drilling parameter obtained by the data decomposition unit 302 into the pre-trained convolutional neural network classification model to output the downhole state classification result, which includes normal drilling state, abnormal drill bit vibration state and tubing string pressure state.

[0115] The range determination unit 304 is used to generate the parameter value optimization range corresponding to each drilling parameter obtained by the downhole state classification result from the data input unit 303 according to the preset optimization range determination algorithm;

[0116] The drilling speed prediction unit 305 is used to input the drilling parameter value combination between each drilling parameter into the pre-trained mechanical drilling speed prediction model within the parameter value optimization range corresponding to each drilling parameter obtained by the range determination unit 304, so as to obtain the mechanical drilling speed prediction value corresponding to each drilling parameter value combination.

[0117] The parameter optimization unit 306 is used to perform iterative calculations on the combination of drilling parameter values ​​among various drilling parameters with the objective of maximizing the mechanical drilling rate predicted by the drilling rate prediction unit 305, using a particle swarm optimization algorithm, to obtain the optimal combination of drilling parameter values ​​among various drilling parameters.

[0118] Furthermore, as a response to the above Figure 2 In addition to the method shown, this embodiment of the invention also provides another drilling parameter optimization device for optimizing the above-mentioned drilling parameters. Figure 2 The method shown is implemented accordingly. This device embodiment corresponds to the foregoing method embodiment. For ease of reading, this device embodiment will not repeat the details of the foregoing method embodiment, but it should be clear that the device in this embodiment can implement all the contents of the foregoing method embodiment. Figure 4 As shown, the device includes:

[0119] The data receiving unit 301 is used to receive the logging time series dataset transmitted by the logging instrument. The dataset includes numerical sequences of drilling parameters such as drilling pressure and rotation speed, as well as timestamps corresponding to each value.

[0120] The data decomposition unit 302 is used to perform multi-scale decomposition processing on the logging time series dataset received by the data receiving unit 301 using the empirical mode decomposition algorithm to generate three sets of feature data corresponding to each drilling parameter. The feature data includes low-frequency feature data, medium-frequency feature data and high-frequency feature data.

[0121] The data input unit 303 is used to input the three sets of feature data corresponding to each drilling parameter obtained by the data decomposition unit 302 into the pre-trained convolutional neural network classification model to output the downhole state classification result, which includes normal drilling state, abnormal drill bit vibration state and tubing string pressure state.

[0122] The range determination unit 304 is used to generate the parameter value optimization range corresponding to each drilling parameter obtained by the downhole state classification result from the data input unit 303 according to the preset optimization range determination algorithm;

[0123] The drilling speed prediction unit 305 is used to input the drilling parameter value combination between each drilling parameter into the pre-trained mechanical drilling speed prediction model within the parameter value optimization range corresponding to each drilling parameter obtained by the range determination unit 304, so as to obtain the mechanical drilling speed prediction value corresponding to each drilling parameter value combination.

[0124] The parameter optimization unit 306 is used to perform iterative calculations on the combination of drilling parameter values ​​among various drilling parameters with the objective of maximizing the mechanical drilling rate predicted by the drilling rate prediction unit 305, using a particle swarm optimization algorithm, to obtain the optimal combination of drilling parameter values ​​among various drilling parameters.

[0125] In one optional implementation, the data decomposition unit 302 includes:

[0126] The pole determination module 3021 is used to identify local maxima and local minima in the numerical sequence of each drilling parameter in the logging time series dataset.

[0127] The envelope determination module 3022 is used to obtain the upper envelope composed of local maxima and the lower envelope composed of local minima obtained by the pole determination module 3021 using spline interpolation.

[0128] The mean determination module 3023 is used to calculate the average value between the values ​​at the same position between the upper and lower envelopes obtained by the envelope determination module 3022, and obtain the average value sequence.

[0129] The new sequence determination module 3024 is used to subtract the average value sequence obtained by the mean determination module 3023 from the numerical sequence corresponding to the drilling parameters point by point to obtain a new numerical sequence.

[0130] The data extraction module 3025 is used to extract three sets of feature data corresponding to drilling parameters from the new numerical sequence determined by the new sequence determination module 3024 according to preset feature extraction conditions.

[0131] In one optional implementation, the data extraction module 3025 is specifically used for:

[0132] Check whether the new numerical sequence meets the feature extraction condition, which is that the number of maximum and minimum points in the numerical sequence differs from the number of zero-crossing points by a preset number.

[0133] If the conditions are met, the new numerical sequence is identified as high-frequency feature data;

[0134] Subtract the high-frequency feature data from the numerical sequence corresponding to the drilling parameters to obtain the remaining numerical sequence, and repeat the above steps for the remaining numerical sequence until the mid-frequency feature data and low-frequency feature data are obtained.

[0135] In one optional implementation, the data input unit 303 is specifically used for:

[0136] The three sets of feature data corresponding to each drilling parameter are normalized to obtain the standardized feature data corresponding to each drilling parameter.

[0137] The standardized feature data of each drilling parameter are concatenated into a multi-dimensional feature vector according to a preset order;

[0138] The multidimensional feature vector is input into a pre-trained convolutional neural network classification model, where features are extracted and dimensionality is gradually reduced through convolutional layers.

[0139] In the fully connected layer of the convolutional neural network classification model, the extracted features are integrated to determine the corresponding classification and output the downhole state classification result.

[0140] In one optional implementation, the range determination unit 304 is specifically used for:

[0141] Obtain the preset optimization rule set corresponding to the downhole state classification result, wherein the downhole state classification result of normal drilling corresponds to the first preset optimization rule set, the downhole state classification result of abnormal drill bit vibration corresponds to the second preset optimization rule set, and the downhole state classification result of tubing support corresponds to the third preset optimization rule set.

[0142] Based on the triangular membership function, combined with the numerical sequence of each drilling parameter and the corresponding set of preset optimization rules, a preliminary optimization range for each drilling parameter is generated;

[0143] By applying multi-dimensional constraints to the initial optimization range of each parameter, the optimization range of each drilling parameter value is obtained.

[0144] In one optional implementation, the parameter optimization unit 306 is specifically used for:

[0145] Initialize the particle swarm, where each particle in the swarm corresponds to a set of drilling parameter values ​​at each position. The initial position of each particle is randomly generated by each drilling parameter within the corresponding parameter value optimization range.

[0146] The predicted mechanical drilling rate corresponding to each combination of drilling parameter values ​​is used as the fitness value of each particle at the corresponding location.

[0147] The initial position of each particle is recorded as the individual historical best position of the particle, and the position with the highest fitness value is selected from all the individual historical best positions of the particles and determined as the global historical best position.

[0148] The movement speed of each particle is adjusted according to the velocity update formula of the particle swarm optimization algorithm.

[0149] Based on the adjusted movement speed, the position of each particle is updated, where the updated particle position represents a combination of drilling parameter values ​​that are within the optimization range of each drilling parameter value.

[0150] Calculate the fitness value corresponding to the combination of drilling parameter values ​​represented by each particle at the updated position, and update the individual historical best position of each particle and the global historical best position in the particle swarm based on the fitness value corresponding to the particle at the updated position.

[0151] Repeat the above steps until the preset maximum number of iterations is reached, and determine the combination of drilling parameter values ​​corresponding to the global historical best position recorded by the particle swarm during the iteration process as the optimal combination of parameter values.

[0152] In one optional implementation, the speed update formula is:

[0153]

[0154] In the formula, vnew is the adjusted moving speed, w is the inertia weight, c1 is the individual learning factor, c2 is the global learning factor, rand() is a random number in the range [0,1], vold is the particle's current speed, pbest is the individual optimal position, gbest is the global historical optimal position, and xold is the combination of drilling parameter values ​​represented by the particle at its current position.

[0155] Furthermore, embodiments of the present invention also provide a storage medium for storing a computer program, wherein the computer program, when running, controls the device where the storage medium is located to execute the above-described... Figure 1 and Figure 2 The drilling parameter optimization method described in [the document].

[0156] Furthermore, embodiments of the present invention also provide a processor for running a program, wherein the program executes the above-described... Figure 1 and Figure 2 The drilling parameter optimization method described in [the document].

[0157] Furthermore, embodiments of the present invention also provide a computer program product, the computer program product including a computer program stored in a computer-readable storage medium, at least one processor can read the computer program from the computer-readable storage medium, and the at least one processor can perform the above-described actions when executing the computer program. Figure 1 and Figure 2 The drilling parameter optimization method described in [the document].

[0158] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0159] It is understood that the relevant features in the above methods and apparatus can be referenced interchangeably. Furthermore, the terms "first," "second," etc., in the above embodiments are used to distinguish between embodiments and do not represent the superiority or inferiority of any particular embodiment.

[0160] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0161] The algorithms and displays provided herein are not inherently related to any particular computer, virtual system, or other device. Various general-purpose systems can also be used in conjunction with the teachings herein. The required structure for constructing such systems is apparent from the above description. Furthermore, this invention is not directed to any particular programming language. It should be understood that the contents of the invention described herein can be implemented using various programming languages, and the above description of specific languages ​​is for the purpose of disclosing the best mode of implementation of the invention.

[0162] In addition, the memory may include non-permanent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0163] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0164] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0165] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0166] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0167] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0168] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, like read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0169] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0170] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0171] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0172] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A method for optimizing drilling parameters, characterized in that, The method includes: Receive logging time-series dataset transmitted by logging tool, wherein the dataset contains at least a numerical sequence of drilling parameters, including drilling pressure and rotation speed, and a timestamp corresponding to each value; The logging time series dataset is decomposed into three sets of feature data corresponding to each drilling parameter by using the empirical mode decomposition algorithm to generate three sets of feature data, including low-frequency feature data, medium-frequency feature data and high-frequency feature data. The three sets of feature data corresponding to each drilling parameter are input into a pre-trained convolutional neural network classification model to output downhole state classification results, which include normal drilling state, abnormal drill bit vibration state, and tubing string pressure state. The algorithm determines the parameter value optimization range corresponding to each drilling parameter in the downhole state classification result based on the preset optimization range; Within the optimization range of the parameter values ​​corresponding to each drilling parameter, the combination of drilling parameter values ​​between each drilling parameter is input into the pre-trained mechanical drilling rate prediction model to obtain the mechanical drilling rate prediction value corresponding to each combination of drilling parameter values. The particle swarm optimization algorithm is used to iteratively calculate the combination of drilling parameter values ​​among various drilling parameters with the goal of maximizing the predicted mechanical drilling rate, so as to obtain the optimal combination of drilling parameter values ​​among various drilling parameters. The well logging time series dataset is decomposed into three sets of feature data corresponding to each drilling parameter by using the empirical mode decomposition algorithm at multiple scales, including: For each drilling parameter in the logging time series dataset, local maxima and local minima are identified in the numerical sequence. The upper envelope formed by each local maxima and the lower envelope formed by each local minima are obtained by using spline interpolation. Calculate the average value among the values ​​at the same position between the upper and lower envelopes to obtain the average value sequence; The average value sequence is subtracted point by point from the numerical sequence corresponding to the drilling parameters to obtain a new numerical sequence. Check whether the new numerical sequence meets the feature extraction condition, which is that the number of maximum and minimum points in the numerical sequence differs from the number of zero-crossing points by a preset number. If the conditions are met, the new numerical sequence is identified as high-frequency feature data; Subtract the high-frequency feature data from the numerical sequence corresponding to the drilling parameters to obtain the remaining numerical sequence, and repeat the above steps for the remaining numerical sequence until the mid-frequency feature data and low-frequency feature data are obtained.

2. The method according to claim 1, characterized in that, The three sets of feature data corresponding to each drilling parameter are input into a pre-trained convolutional neural network classification model, which outputs downhole state classification results, including: The three sets of feature data corresponding to each drilling parameter are normalized to obtain the standardized feature data corresponding to each drilling parameter. The standardized feature data of each drilling parameter are concatenated into a multi-dimensional feature vector according to a preset order; The multidimensional feature vector is input into a pre-trained convolutional neural network classification model, where features are extracted and dimensionality is gradually reduced through convolutional layers. In the fully connected layer of the convolutional neural network classification model, the extracted features are integrated to determine the corresponding classification and output the downhole state classification result.

3. The method according to claim 1, characterized in that, The algorithm determines the parameter value optimization range for each drilling parameter corresponding to the downhole state classification result based on the preset optimization range, including: Obtain the preset optimization rule set corresponding to the downhole state classification result, wherein the downhole state classification result of normal drilling corresponds to the first preset optimization rule set, the downhole state classification result of abnormal drill bit vibration corresponds to the second preset optimization rule set, and the downhole state classification result of tubing support corresponds to the third preset optimization rule set. Based on the triangular membership function, combined with the numerical sequence of each drilling parameter and the corresponding set of preset optimization rules, a preliminary optimization range for each drilling parameter is generated; By applying multi-dimensional constraints to the initial optimization range of each parameter, the optimization range of each drilling parameter value is obtained.

4. The method according to claim 1, characterized in that, The particle swarm optimization algorithm is used to iteratively calculate the combination of drilling parameter values ​​among various drilling parameters with the objective of maximizing the predicted mechanical drilling rate, to obtain the optimal combination of drilling parameter values ​​among various drilling parameters, including: Initialize the particle swarm, where each particle in the swarm corresponds to a set of drilling parameter values ​​at each position. The initial position of each particle is randomly generated by each drilling parameter within the corresponding parameter value optimization range. The predicted mechanical drilling rate corresponding to each combination of drilling parameter values ​​is used as the fitness value of each particle at the corresponding location. The initial position of each particle is recorded as the individual historical best position of the particle, and the position with the highest fitness value is selected from all the individual historical best positions of the particles and determined as the global historical best position. The movement speed of each particle is adjusted according to the velocity update formula of the particle swarm optimization algorithm. Based on the adjusted movement speed, the position of each particle is updated, where the updated particle position represents a combination of drilling parameter values ​​that are within the optimization range of each drilling parameter value. Calculate the fitness value corresponding to the combination of drilling parameter values ​​represented by each particle at the updated position, and update the individual historical best position of each particle and the global historical best position in the particle swarm based on the fitness value corresponding to the particle at the updated position. Repeat the above steps until the preset maximum number of iterations is reached, and determine the combination of drilling parameter values ​​corresponding to the global historical best position recorded by the particle swarm during the iteration process as the optimal combination of parameter values.

5. The method according to claim 4, characterized in that, The speed update formula is: In the formula, vnew is the adjusted moving speed, w is the inertia weight, c1 is the individual learning factor, c2 is the global learning factor, rand() is a random number in the range [0,1], vold is the particle's current speed, pbest is the individual optimal position, gbest is the global historical optimal position, and xold is the combination of drilling parameter values ​​represented by the particle at its current position.

6. A drilling parameter optimization device, characterized in that, The device includes: The data receiving unit is used to receive the logging time-series dataset transmitted by the logging instrument. The dataset includes at least a numerical sequence of drilling parameters, including drilling pressure and rotation speed, and a timestamp corresponding to each value. The data decomposition unit is used to perform multi-scale decomposition processing on the logging time series dataset received by the data receiving unit through the empirical mode decomposition algorithm, and generate three sets of feature data corresponding to each drilling parameter. The feature data includes low-frequency feature data, medium-frequency feature data and high-frequency feature data. The data decomposition unit is further configured to: identify local maxima and local minima in the numerical sequence of each drilling parameter in the logging time-series dataset; obtain the upper envelope formed by each local maxima and the lower envelope formed by each local minima using spline interpolation; calculate the average value between the values ​​at the same position between the upper and lower envelopes to obtain an average value sequence; subtract the average value sequence point by point from the numerical sequence corresponding to the drilling parameters to obtain a new numerical sequence; check whether the new numerical sequence meets the feature extraction condition, wherein the feature extraction condition is that the number of maxima and minima in the numerical sequence differs from the number of zero-crossing points by a preset number; if it meets the condition, the new numerical sequence is identified as high-frequency feature data; subtract the high-frequency feature data from the numerical sequence corresponding to the drilling parameters to obtain the remaining numerical sequence, and repeat the above steps for the remaining numerical sequence until medium-frequency feature data and low-frequency feature data are obtained; The data input unit is used to input the three sets of feature data corresponding to each drilling parameter obtained by the data decomposition unit into the pre-trained convolutional neural network classification model to output the downhole state classification result. The classification result includes normal drilling state, abnormal drill bit vibration state, and tubing string pressure state. The range determination unit is used to generate the parameter value optimization range corresponding to each drilling parameter in the downhole state classification result obtained by the data input unit according to the preset optimization range determination algorithm; The drilling speed prediction unit is used to input the drilling parameter value combination between each drilling parameter into the pre-trained mechanical drilling speed prediction model within the optimization range of the parameter values ​​corresponding to each drilling parameter obtained by the range determination unit, so as to obtain the mechanical drilling speed prediction value corresponding to each drilling parameter value combination. The parameter optimization unit is used to perform iterative calculations on the combination of drilling parameter values ​​among various drilling parameters with the objective of maximizing the mechanical drilling rate predicted by the drilling rate prediction unit, using a particle swarm optimization algorithm, to obtain the optimal combination of drilling parameter values ​​among various drilling parameters.

7. A storage medium, characterized in that, The storage medium includes a stored program, wherein, when the program is executed, it controls the device containing the storage medium to perform the drilling parameter optimization method as described in any one of claims 1 to 5.

8. A processor, characterized in that, The processor is used to run a program, wherein the program executes the drilling parameter optimization method as described in any one of claims 1 to 5.

9. A computer program product, characterized in that, It includes a computer program that is executed by a processor as the drilling parameter optimization method as described in any one of claims 1 to 5.

Images