Mechanical drilling speed prediction method, related method and device

Abstract

The invention relates to a mechanical drilling speed prediction method, a related method and a related device. The method comprises the following steps: obtaining an optimized drilling rate prediction model obtained by optimizing XGboost based on a genetic algorithm; the obtained drilling data of the long horizontal section of the shale gas horizontal well to be predicted are input into the optimized drilling speed prediction model, and a drilling speed prediction result is obtained; wherein the optimized drilling rate prediction model is obtained by the following steps: acquiring initial drilling data and mechanical drilling rate of a long horizontal section of a historical shale gas horizontal well; analyzing the initial drilling data and the mechanical drilling speed by using a correlation analysis method to obtain a training drilling data set; training a pre-constructed initial drilling speed prediction model according to the training drilling data set to obtain a trained drilling speed prediction model; wherein the initial drilling speed prediction model is constructed by adopting XGboost; and optimizing and training the trained drilling speed prediction model by using a genetic algorithm until a preset condition is met, and obtaining the optimized drilling speed prediction model.

Description

Technical Field

[0001] The present invention relates to the field of exploration technology, and in particular to a mechanical drilling rate prediction method, related methods and devices. Background Art

[0002] In the process of oil and gas resource exploration and development, improving drilling efficiency and shortening drilling cycles are the goals pursued by engineering and technical personnel, and using artificial intelligence methods to predict mechanical drilling speed is an important way to improve drilling efficiency and shorten drilling cycles. Many scholars have conducted research on the problem of predicting mechanical drilling speed for horizontal sections of shale gas horizontal wells or other formation characteristics.

[0003] With the rapid development of big data technology and the rapid growth of data size, the use of machine learning methods to mine data and apply it to the drilling process has significantly improved the prediction accuracy of mechanical drilling speed compared with the method based on physical models. The mechanical drilling speed prediction model established by intelligent algorithms can predict the drilling speed results more quickly and with better quality, thereby helping to shorten the drilling cycle, reduce drilling costs, improve drilling quality and drilling efficiency, enhance the safety and controllability of drilling, and make the drilling process more efficient and safe.

[0004] Mechanical penetration rate is an important indicator for evaluating drilling efficiency. There are many factors affecting mechanical penetration rate, which can be divided into two categories: one is uncontrollable factors, which mainly characterize the formation properties and rock properties, such as rock type, rock drillability, mud content, pore volume, etc.; the other is controllable factors, which are mainly drilling fluid parameters and operating parameters during the drilling process, such as drilling fluid density, viscosity, displacement, drilling pressure, speed, torque, etc. Accurately establishing a prediction model between mechanical penetration rate and influencing factors is crucial for predicting mechanical penetration rate and optimizing engineering parameters. Summary of the invention

[0005] In order to obtain a more accurate mechanical drilling rate prediction result, an embodiment of the present invention provides a mechanical drilling rate prediction method, related methods and devices.

[0006] In a first aspect, an embodiment of the present application provides a method for predicting mechanical drilling speed, the method comprising:

[0007] Obtain the optimized drilling speed prediction model based on genetic algorithm optimization of XGboost;

[0008] Input the acquired drilling data of the long horizontal section of the shale gas horizontal well to be predicted into the optimized drilling rate prediction model to obtain a drilling rate prediction result;

[0009] The optimized drilling speed prediction model is obtained in the following way:

[0010] Obtaining initial drilling data and mechanical drilling rate of a long horizontal section of a historical shale gas horizontal well; the initial drilling data includes: drilling pressure, drill bit speed, motor speed, surface speed, displacement, natural gamma, standpipe pressure, depth measurement, well inclination, azimuth, axial vibration and lateral vibration;

[0011] Analyze the initial drilling data and the mechanical drilling rate using a correlation analysis method to obtain a training drilling data set;

[0012] Training a pre-built initial drilling speed prediction model according to the training drilling data set to obtain a trained drilling speed prediction model; wherein the initial drilling speed prediction model is constructed using XGboost;

[0013] The trained drilling speed prediction model is optimized and trained using a genetic algorithm until a preset condition is met, thereby obtaining the optimized drilling speed prediction model.

[0014] In one or some optional implementations of the embodiments of the present application, the method of optimizing and training the trained drilling rate prediction model using a genetic algorithm until a preset condition is met to obtain the optimized drilling rate prediction model includes:

[0015] Optimizing the parameters of the trained drilling speed prediction model using a genetic algorithm;

[0016] When the loss function of the trained drilling speed prediction model reaches a minimum value, the optimized drilling speed prediction model is obtained; wherein the loss function is shown in the following formula 1:

[0017]

[0018] In the formula, ROP represents the actual mechanical penetration rate, It represents the predicted value obtained by the drilling speed prediction model.

[0019] In one or some optional implementations of the embodiments of the present application, the pre-built initial drilling rate prediction model is trained according to the training drilling data set to obtain a trained drilling rate prediction model, including:

[0020] According to the data in the training drilling data set, the pre-built initial drilling speed prediction model is trained by using parameter adjustment and early stopping;

[0021] When the determination coefficient R of the drilling speed prediction model 2 When the preset threshold is reached, the trained drilling speed prediction model is obtained.

[0022] In one or some optional implementations of the embodiment of the present application, before using the correlation analysis method to analyze the initial drilling data and the mechanical drilling rate to obtain a training drilling data set, the method further includes:

[0023] The drilling data is cleaned to obtain an initial drilling data set; wherein the data cleaning includes processing missing data, detecting redundant data, and eliminating abnormal data.

[0024] In one or some optional implementations of the embodiments of the present application, the drilling data is cleaned to obtain an initial drilling data set; wherein the data cleaning includes processing missing data, detecting redundant data and removing abnormal data, specifically including:

[0025] Identifying missing data in the drilling data, filling in missing values ​​or deleting missing values;

[0026] Deleting redundant data repeatedly recorded in the drilling data;

[0027] The noise data in the drilling data is removed by a digital filtering algorithm.

[0028] In one or some optional implementations of the embodiments of the present application, the correlation analysis method is used to analyze the initial drilling data set and the mechanical drilling rate of the long horizontal section of the shale gas horizontal well to obtain a training drilling data set, including:

[0029] Solving multiple correlation coefficients between multiple data in the initial drilling data and the mechanical drilling rate by a correlation analysis method;

[0030] The plurality of drilling data whose correlation coefficients reach a preset threshold are used as the training drilling data set.

[0031] In one or some optional implementations of the embodiment of the present application, the method further includes: normalizing the training drilling data set based on the following formula 2:

[0032]

[0033] Among them, x′ k is the result after standardization, x k For samples that need to be standardized, is the sample mean, and n is the total number of original samples.

[0034] In a second aspect, an embodiment of the present application provides a method for training a mechanical drilling rate prediction model, the method comprising:

[0035] Obtaining initial drilling data and mechanical drilling rate of a long horizontal section of a historical shale gas horizontal well; the initial drilling data includes: drilling pressure, drill bit speed, motor speed, surface speed, displacement, natural gamma, standpipe pressure, depth measurement, well inclination, azimuth, axial vibration and lateral vibration;

[0036] Analyze the initial drilling data and the mechanical drilling rate using a correlation analysis method to obtain a training drilling data set;

[0037] Training a pre-built initial drilling speed prediction model according to the training drilling data set to obtain a trained drilling speed prediction model; wherein the initial drilling speed prediction model is constructed using XGboost;

[0038] The trained drilling speed prediction model is optimized and trained using a genetic algorithm until a preset condition is met, thereby obtaining the optimized drilling speed prediction model.

[0039] In a third aspect, an embodiment of the present application provides a mechanical drilling rate prediction device, the device comprising:

[0040] The first acquisition module is used to acquire the initial drilling data and mechanical drilling rate of the long horizontal section of the historical shale gas horizontal well; the initial drilling data includes: drilling pressure, drill bit speed, motor speed, surface speed, displacement, natural gamma, standpipe pressure, depth measurement, well inclination angle, azimuth angle, axial vibration and lateral vibration;

[0041] A first processing module is used to analyze the initial drilling data and the mechanical drilling rate using a correlation analysis method to obtain a training drilling data set;

[0042] A first training module is used to train a pre-built initial drilling speed prediction model according to the training drilling data set to obtain a trained drilling speed prediction model; wherein the initial drilling speed prediction model is constructed using XGboost;

[0043] A second training module is used to optimize and train the trained drilling speed prediction model using a genetic algorithm until a preset condition is met to obtain the optimized drilling speed prediction model;

[0044] The second acquisition module is used to obtain an optimized drilling speed prediction model obtained by optimizing XGboost based on a genetic algorithm;

[0045] The first prediction module is used to input the acquired drilling data of the long horizontal section of the shale gas horizontal well to be predicted into the optimized drilling speed prediction model to obtain the drilling speed prediction result.

[0046] In a fourth aspect, an embodiment of the present application provides a mechanical drilling speed prediction model training device, the device comprising:

[0047] The first acquisition module is used to acquire the initial drilling data and mechanical drilling rate of the long horizontal section of the historical shale gas horizontal well; the initial drilling data includes: drilling pressure, drill bit speed, motor speed, surface speed, displacement, natural gamma, standpipe pressure, depth measurement, well inclination angle, azimuth angle, axial vibration and lateral vibration;

[0048] A first processing module is used to analyze the initial drilling data and the mechanical drilling rate using a correlation analysis method to obtain a training drilling data set;

[0049] A first training module is used to train a pre-built initial drilling speed prediction model according to the training drilling data set to obtain a trained drilling speed prediction model; wherein the initial drilling speed prediction model is constructed using XGboost;

[0050] The second training module is used to optimize and train the trained drilling speed prediction model using a genetic algorithm until a preset condition is met to obtain the optimized drilling speed prediction model.

[0051] In a fifth aspect, an embodiment of the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the mechanical drilling rate prediction method as described above, and / or the mechanical drilling rate prediction model training method as described above.

[0052] In a sixth aspect, an embodiment of the present application provides a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the mechanical drilling rate prediction method as described above and / or the mechanical drilling rate prediction model training method as described above are implemented.

[0053] In a seventh aspect, an embodiment of the present application provides a computer program product comprising instructions. When the computer program product is run on a computer device, the computer device executes the mechanical drilling rate prediction method as described above, and / or the mechanical drilling rate prediction model training method as described above.

[0054] In an eighth aspect, an embodiment of the present application provides a chip, the chip including a processor and a communication interface, the communication interface and the processor are coupled, and the processor is used to run a computer program or instructions to implement the mechanical drilling rate prediction method as described above, and / or the mechanical drilling rate prediction model training method as described above.

[0055] The beneficial effects of the above technical solution provided by the embodiment of the present invention include at least:

[0056] The mechanical drilling rate prediction method provided by the embodiment of the present application obtains the drilling data of the long horizontal section of the shale gas horizontal well and inputs the optimized drilling rate prediction model to realize the mechanical drilling rate prediction for the long horizontal section of the shale gas horizontal well; XGboost is used to build the drilling rate prediction model and train it. XGboost, as a powerful integrated learning algorithm, uses parallel computing to accelerate the training process, effectively reduces the computing cost, and can build a high-performance prediction model to more accurately model the relationship between drilling data and mechanical drilling rate; the genetic algorithm is used to optimize the training of the drilling rate prediction model, effectively improve the generalization ability of the model, and make the model better applicable to new data by searching for the best parameter combination. It can effectively solve the problem of the difficulty of drilling rate prediction, and can predict the drilling rate results with better quality and faster speed, thereby assisting in shortening the drilling cycle, reducing drilling costs, and improving drilling quality and efficiency. By adopting a suitable model to obtain a more accurate mechanical drilling rate prediction result, and then using an artificial intelligence algorithm to optimize the model, a more accurate mechanical drilling rate prediction result is obtained to guide on-site actual drilling.

[0057] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 It is a flow chart of a mechanical drilling rate prediction method provided in an embodiment of the present application;

[0059] Figure 2 It is a schematic diagram of the steps of the genetic algorithm provided in the embodiment of the present application;

[0060] Figure 3 is a schematic diagram of the relationship between the mechanical drilling speed and the well depth before and after filtering provided in an embodiment of the present application;

[0061] Figure 4 is a schematic diagram of correlation analysis between drilling data and mechanical drilling rate provided in an embodiment of the present application;

[0062] Figure 5 It is a schematic diagram of the prediction results of the drilling speed prediction model optimized by the genetic algorithm provided in the embodiment of the present application on the validation set;

[0063] Figure 6 It is a schematic diagram of the drilling speed prediction result based on the new well in the Duverney work area provided in the embodiment of the present application;

[0064] Figure 7 A schematic diagram of the steps of the mechanical drilling rate prediction model training method provided in an embodiment of the present application;

[0065] Figure 8 A schematic diagram of the structure of a mechanical drilling speed prediction device provided in an embodiment of the present application;

[0066] Fig. 9 A schematic diagram of the structure of a mechanical drilling rate prediction model training device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0067] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present application.

[0068] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, wholes, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or clusters thereof.

[0069] It should also be understood that the term “and / or” used in the specification and appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0070] As used in the specification and appended claims of this application, the term "if" can be interpreted as "when" or "uponce" or "in response to determining" or "in response to detecting", depending on the context. Similarly, the phrase "if it is determined" or "if [described condition or event] is detected" can be interpreted as meaning "uponce it is determined" or "in response to determining" or "uponce [described condition or event] is detected" or "in response to detecting [described condition or event]", depending on the context.

[0071] In addition, in the description of the present application specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0072] References to "one embodiment" or "some embodiments" etc. described in the specification of this application mean that one or more embodiments of the present application include specific features, structures or characteristics described in conjunction with the embodiment. Therefore, the statements "in one embodiment", "in some embodiments", "in some other embodiments", "in some other embodiments", etc. that appear in different places in this specification do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in other ways. The terms "including", "comprising", "having" and their variations all mean "including but not limited to", unless otherwise specifically emphasized in other ways.

[0073] It should be understood that the size of the serial numbers of the steps in the following embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0074] In order to illustrate the technical solution of the present application, a specific embodiment is provided below for illustration.

[0075] The inventors found that in the prior art, most of the research on mechanical drilling rate prediction is based on neural network prediction. The training of neural networks usually requires more iterations and computing resources, and the computing cost is relatively high. The accuracy of the obtained prediction results still has a lot of room for improvement. Secondly, in practical applications, training prediction has never been conducted for long horizontal section drilling data of shale gas horizontal wells. Therefore, there is an urgent need for a method for mechanical drilling rate prediction that meets the prediction accuracy requirements required in practical applications and overcomes the defects of the prior art. Based on this, the inventors made the present invention after further research, providing a mechanical drilling rate prediction method, related methods and devices.

[0076] Embodiment 1

[0077] See also Figure 1 , is a mechanical drilling speed prediction method provided in Example 1 of the present application, referring to Figure 1 As shown, the method includes:

[0078] S101: Obtain an optimized drilling speed prediction model based on genetic algorithm optimization of XGboost.

[0079] Among them, the optimized drilling speed prediction model is obtained by the following method:

[0080] Obtain the initial drilling data and mechanical drilling rate of the long horizontal section of the historical shale gas horizontal well; the initial drilling data include: drilling pressure, drill bit speed, motor speed, surface speed, displacement, natural gamma, standpipe pressure, depth, well inclination, azimuth, axial vibration and lateral vibration;

[0081] Using correlation analysis method to analyze the initial drilling data set and mechanical drilling speed to obtain a training drilling data set;

[0082] Training a pre-built initial drilling speed prediction model according to the training drilling data set to obtain a trained drilling speed prediction model; wherein the initial drilling speed prediction model is built using XGboost;

[0083] The trained drilling speed prediction model is optimized by using a genetic algorithm until a preset condition is met to obtain the optimized drilling speed prediction model.

[0084] In the embodiment of the present application, the above-mentioned initial drilling data and mechanical drilling rate of the long horizontal section of the historical shale gas horizontal well are obtained, and the initial drilling data and mechanical drilling rate of the long horizontal section of the shale gas horizontal well are obtained respectively, and the parameter characteristics include: drilling pressure (Weight on bit), drill bit speed (Bit RPM), motor speed (Motor RPM), surface speed (Surface RPM), displacement (Total Pump Output), natural gamma (Gamma), standpipe pressure (Standpipe pressure), depth (Measured depth), inclination (Inclination), azimuth (Azimuth), axial vibration (AxialVibration) and lateral vibration (Lateral Vibration), as well as predicted target mechanical drilling rate (Rate OfPenetration). Then the above data statistics are organized on an electronic table as initial drilling data. Due to the differences in the data types collected at different sites, the above 12 parameter features are necessary parameters, but are not limited to these parameters, and other drilling data can be added on this basis.

[0085] In the embodiment of the present application, before using the correlation analysis method to analyze the initial drilling data and the mechanical drilling speed to obtain the training drilling data set, it also includes: cleaning the drilling data to obtain the initial drilling data set; wherein, data cleaning includes processing missing data, detecting redundant data and removing abnormal data, specifically including: identifying missing data in the drilling data, filling missing values ​​or deleting missing values; deleting redundant data recorded repeatedly in the drilling data; removing noise data in the drilling data by a digital filtering algorithm. Among them, identifying missing data in the drilling data, filling missing values ​​or deleting missing values ​​is a step of processing missing values, and the method includes deleting missing values ​​or filling missing values. The method of filling missing values ​​includes interpolation, random sampling filling, professional knowledge filling, etc.; removing redundant data recorded repeatedly in the drilling data is a step of detecting redundant data, in order to delete redundant data recorded repeatedly due to equipment failure or abnormality; removing noise data in the drilling data by a digital filtering algorithm is a step of removing abnormal data, and removing noise data by a digital filtering algorithm extracts useful data information. Among them, data with large mutations between two adjacent points in the time domain recorded by the drilling equipment is unnecessary noise data.

[0086] In the embodiment of the present application, the initial drilling data set and the mechanical drilling rate of the long horizontal section of the shale gas horizontal well are analyzed by the correlation analysis method to obtain a training drilling data set, which specifically includes: solving multiple data in the initial drilling data and the mechanical drilling rate by the correlation analysis method to obtain multiple correlation coefficients; taking multiple drilling data whose correlation coefficient reaches a preset threshold as the training drilling data set. The data in the initial drilling data that show a strong correlation with the long horizontal section of the shale gas horizontal well are solved by the correlation analysis method, and the data whose absolute value of the correlation coefficient is between 0.4 and 1.0 is judged to be strongly correlated data, the data whose absolute value of the correlation coefficient is between 0.1 and 0.4 is weakly correlated data, and the data whose absolute value of the correlation coefficient is between 0 and 0.1 is irrelevant data. The correlation coefficient here can be selected as the Pearson correlation coefficient, and the correlation coefficient in specific use is not limited here. It is thus concluded that which of the above 12 drilling data is strongly correlated with the prediction of mechanical drilling rate: riser pressure, drilling pressure, drill bit speed, axial vibration, lateral vibration, displacement, and ground speed, so the training drilling data will be composed of these strongly correlated drilling data.

[0087] In the correlation analysis process, by establishing the correlation coefficient image between different drilling parameters and mechanical penetration rate, the important drilling parameters related to the long horizontal section of shale gas horizontal wells were screened out, and then the training drilling data set and mechanical penetration rate were standardized based on the following formula 2:

[0088]

[0089] Among them, x′ k is the result after standardization, x k For samples that need to be standardized, is the average value of the sample, and n is the total number of original samples. Standardization can scale data of different magnitudes to between -1 and 1, and satisfy the normal distribution state with a mean of 0 and a variance of 1. This facilitates the subsequent training of the drilling speed prediction model.

[0090] In an embodiment of the present application, an initial drilling speed prediction model is trained according to the training drilling data set to obtain a trained drilling speed prediction model. It is necessary to pre-build an initial drilling speed prediction model, which is built using XGboost. XGboost uses a decision tree as a basic learner and integrates by iteratively building multiple decision trees, each of which attempts to correct the prediction error of the previous tree. First, it is necessary to select the hyperparameters of the XGboost model, such as the learning rate, the number of trees, the depth of the tree, the regularization term, etc. These hyperparameters will be important goals in the optimization training.

[0091] The construction of the XGboost model is carried out synchronously with the training, and the steps include: initializing the model, the initial XGboost model contains only one tree, and initializing the model parameters (such as weights, biases, etc.); forward propagation to calculate the loss, input each drilling data sample into the XGboost model, obtain the output predicted drilling speed, and calculate the loss value between the predicted drilling speed and the actual drilling speed according to the loss function; back propagation, according to the loss value, back propagation updates the gradient of each node; constructing a new decision tree, using the negative gradient of each node as the target variable, and constructing a new decision tree to improve the prediction results of the current XGboost model; feature splitting, using the feature splitting strategy to determine the node splitting method inside the new decision tree, find the best features and splitting points to maximize the splitting gain of the decision tree; updating the tree structure, adding the new decision tree to the XGboost model, and determining the node splitting position according to the feature splitting strategy, and updating the decision tree structure; repeating the steps of iterative forward propagation to calculate the loss, back propagation and updating the tree structure until the preset condition is reached to stop the iterative training. The preset condition can be set to reach a predetermined total number of decision trees or the difference between the losses of two adjacent iterations does not exceed a preset threshold.

[0092] The efficient and flexible distributed gradient boosting XGboost algorithm is used for iterative training. The initial drilling speed prediction model is trained by parameter adjustment and early stopping to avoid overfitting problems. Multiple decision trees are used to build an integrated model of XGboost to improve prediction accuracy, and the regularization strength of the regularization term in the cost function is adjusted to control the complexity of the model. The model formula for mechanical drilling speed prediction can be expressed as shown in Formula 3:

[0093]

[0094] Using the coefficient of determination R 2 As an evaluation index, this evaluation index is used to indicate the fitting effect of the initial model for predicting mechanical drilling speed. 2 As an evaluation index to characterize the fitting effect, R 2 The larger it is, the closer it is to 1, and the better the drilling speed prediction model fits the actual mechanical drilling speed; on the contrary, R 2 The smaller it is, that is, the closer it is to 0, the worse the fitting effect of the drilling speed prediction model on the actual mechanical drilling speed.

[0095] In the embodiment of the present application, the drilling speed prediction model is optimized and trained using a genetic algorithm through XGboost training until the preset conditions are reached to obtain an optimized drilling speed prediction model. The model parameters of the drilling speed prediction model after training are optimized using a genetic algorithm, thereby optimizing the drilling speed prediction model. The model parameters here mainly refer to hyperparameters, i.e., learning rate, number of trees, depth of trees, regularization term, minimum child weight, etc. By simulating the process of natural selection and genetics mechanism in Darwin's theory of evolution, the type determination, data input, random combination, and random change of drilling parameters are carried out. The loss function formula is used to solve the fitting effect of the drilling speed prediction model, and finally an optimized drilling speed prediction model is obtained, which can more accurately predict the drilling speed.

[0096] The above operation process of optimizing and training the drilling speed prediction model through genetic algorithm is as follows: Figure 2 As shown, it specifically includes: initializing the hyperparameters of the drilling speed prediction model after training; encoding the original parameters by genetic algorithm; determining the fitness function, which is used to evaluate the prediction effect of the model; selection, using the selection operation to select excellent individuals according to the value of the fitness function as the parents of the next generation population; crossover, performing crossover operations on the selected individuals to generate new individuals. The crossover operation can combine the hyperparameter combinations of different individuals to generate new hyperparameter configurations; mutation, performing mutation operations on the newly generated individuals to introduce randomness to avoid falling into the local optimal solution. The mutation operation can fine-tune the value of the hyperparameters, and add the crossed and mutated individuals to the population to form a new population; repeating the steps of selection, crossover, mutation and updating the population until the loss function reaches the preset target. The corresponding optimized hyperparameters and optimized drilling speed prediction model are obtained. Among them, the loss function is shown in Formula 1:

[0097]

[0098] In the formula, ROP represents the actual mechanical penetration rate, It represents the predicted value obtained by the drilling speed prediction model.

[0099] The embodiment of the present application uses a genetic algorithm to optimize the XGboost model to predict the mechanical drilling rate, and the prediction result has a high accuracy. The XGboost algorithm can effectively avoid overfitting, support input feature weights, and use genetic algorithms to simulate the biological evolution process to determine, input, combine, and change drilling parameters, and finally obtain a high-precision mechanical drilling rate prediction result.

[0100] S102: Input the acquired drilling data of the long horizontal section of the shale gas horizontal well to be predicted into the optimized drilling rate prediction model to obtain a drilling rate prediction result.

[0101] In an embodiment of the present application, the drilling data of the long horizontal section of the shale gas horizontal well to be predicted will be obtained in the above step S102. According to the results of analyzing the drilling data and the mechanical penetration rate using the correlation analysis method in step S101, the parameters that are strongly correlated with the mechanical penetration rate include riser pressure, drilling pressure, drill bit speed, axial vibration, lateral vibration, displacement, and ground speed, a total of 7 drilling parameters. The above 7 data in the drilling data to be predicted are input into the optimized drilling rate prediction model, and the drilling rate prediction output is optimized to obtain the drilling rate prediction result.

[0102] In order to explain the embodiment of the present invention more clearly, in a specific example, see Figure 3-Figure 6 As shown, the mechanical drilling speed prediction method of the embodiment of the present invention is described in detail as follows:

[0103] The first step is to obtain the mud logging data and well logging data during the on-site drilling of the long horizontal section of the shale gas horizontal well in the Duverney work area, as shown in Table 1.

[0104] Table 1 Initial drilling data and mechanical drilling rate example table of Duverney work area

[0105]

[0106] The second step is to clean the initial drilling data and mechanical drilling speed. Data cleaning includes processing missing data, detecting redundant data, and eliminating abnormal data. The specific operations are as follows:

[0107] First, deal with missing data: identify missing data sets in the Duverney work area. The near-drilling parameters collected in the Duverney work area are indeed relatively small, and the number of samples occupied by the data is far less than the total number of samples, so the missing data can be deleted; then, detect redundant data: delete redundant data with repeated records caused by equipment failure or anomalies in the Duverney work area; finally, eliminate abnormal data: in actual drilling work in the Duverney work area, instrument recording errors will cause abnormal noise in some input data. By using the SG (Savitzky-Golay) algorithm for least squares convolution smoothing low-pass filtering, the outliers of the data can be effectively reduced. The filtered data also maintains the trend of the original data, and more refined data information is extracted.

[0108] The relationship between the filtered mechanical drilling speed and the measured depth is shown as follows Figure 3 As shown, the light color is the mechanical drilling speed before filtering, and the dark color is the mechanical drilling speed after filtering.

[0109] In the third step, the correlation analysis method was used to solve the data in the initial drilling data of the Duverney work area that showed a strong correlation with the long horizontal section of the shale gas horizontal well.

[0110] The correlation analysis method is used to analyze the drilling data set and the mechanical drilling rate of the long horizontal section of the shale gas horizontal well, specifically including: solving the correlation between multiple data in the drilling data and the mechanical drilling rate through the correlation analysis method, the correlation coefficient used is the Pearson correlation coefficient, and judging that the absolute value of the Pearson correlation coefficient is between 0.4 and 1.0 for strong correlation data, the absolute value of the Pearson correlation coefficient is between 0.1 and 0.4 for weak correlation data, and the absolute value of the Pearson correlation coefficient is between 0 and 0.1 for irrelevant data. Plot the correlation analysis as shown in Figure 4 As shown in the figure, the following data are strongly correlated with the prediction of mechanical drilling speed of the long horizontal section of shale gas horizontal wells in the Duverney area: standpipe pressure, drilling pressure, drill bit speed, axial vibration, lateral vibration, displacement, and surface speed. Therefore, these strongly correlated drilling data will form the training drilling data. The training drilling data and mechanical drilling speed are then standardized, the data is scaled to between -1 and 1, and the data mean is 0 and the variance is 1, so that the subsequent training process can be carried out smoothly.

[0111] The fourth step is to iteratively train the drilling rate prediction model built by XGboost based on the obtained training drilling data set of the Duverney work area. The steps include: initializing the model, forward propagation to calculate the loss, back propagation, feature splitting, updating the tree structure, repeating the steps of iterative forward propagation to calculate the loss, back propagation and updating the tree structure until the loss function reaches the minimum value, stop training, and obtain the drilling rate prediction model after training. Use the determination coefficient R 2 As an evaluation index, this evaluation index is used to indicate the fitting effect of the initial model for predicting mechanical drilling speed. 2 As an evaluation index to characterize the fitting effect, R 2 The larger it is, the closer it is to 1, and the better the drilling speed prediction model fits the actual mechanical drilling speed; on the contrary, R 2 The smaller it is, that is, the closer it is to 0, the worse the drilling speed prediction model fits the actual mechanical drilling speed. The prediction effect of the drilling speed prediction model after training on the validation set of the Duverney work area R 2 The prediction effect R of the surrounding wells (test set) is 0.93. 2 It is 0.83.

[0112] In the fifth step, the genetic algorithm is used to optimize the model parameters of the trained drilling speed prediction model, that is, the hyperparameters of XGboost, so as to optimize the drilling speed prediction model and obtain a high-precision optimized drilling speed prediction model. The hyperparameters of the trained drilling speed prediction model constructed by XGboost before optimization are shown in Table 2.

[0113] Table 2 Hyperparameters of the training drilling speed prediction model constructed by XGboost

[0114] Model Hyperparameters Value Cardinality tree (tree num) 8 Learning rate 0.2 Max depth of decision tree 130 Minimum child weight 8

[0115] The operation process of optimizing the drilling prediction model through genetic algorithm specifically includes: defining a fitness function, which will evaluate the fitness of each individual. An individual represents an XGboost hyperparameter combination, that is, the fitness function will evaluate the prediction effect of the drilling speed prediction model corresponding to the hyperparameter combination. Import the GA algorithm library, and define the population size, that is, how many individuals are included, generate the initial population, and define the parameters of the genetic algorithm, such as the crossover rate and mutation rate. Define the evolution function of the genetic algorithm, which includes operations such as selection, crossover and mutation, and use the fitness function to calculate the fitness of each individual, and select the optimal individual based on the fitness to generate the next generation population. Repeat the above selection, crossover and mutation operations. In the process of using the genetic algorithm for evolutionary iteration, use the loss function to evaluate its performance until the loss function is reduced to the minimum, stop training, and save the model. The hyperparameters of the genetic algorithm are shown in Table 3.

[0116] Table 3 Hyper parameters of genetic algorithm

[0117] Model Hyperparameters Value Iterations 100 Initial population 30 Mutation probability 0.1 Crossover probability 0.8

[0118] When the loss function finds the minimum value, the fitting effect is optimal. After the drilling speed prediction model is optimized by genetic algorithm, the prediction effect of the optimized drilling speed prediction model on the validation set of Duverney work area is R 2 is 0.96, and the effect is Figure 5 As shown in the figure, it shows that compared with the training drilling speed prediction model before optimization, the accuracy of the optimized mechanical drilling speed prediction model has been significantly improved.

[0119] The sixth step is to apply the mechanical drilling rate prediction model to new wells in the long horizontal section of shale gas horizontal wells in the Duverney work area that are not included in the training data set, that is, test data for prediction. The logging data of the 4000m to 4500m section of the well is selected, including 7 data such as drilling pressure, drill bit speed, surface speed, displacement, standpipe pressure, lateral vibration, and axial vibration. The test data is input into the optimized drilling rate prediction model to obtain the prediction effect as follows Figure 6 As shown in the figure, the determination coefficient R between the predicted mechanical drilling speed and the actual mechanical drilling speed is 2 The determination coefficient R of the unoptimized drilling speed prediction model for the adjacent well (test set) is 0.89. 2 The prediction effect of the optimized model on the adjacent wells is improved by about 0.06, which shows that the model has obtained better fitting accuracy, and the accuracy and generalization ability of the model have been improved. The original data of some well sections and the prediction results of mechanical drilling speed are shown in Table 4.

[0120] Table 4 Original data of some well sections and prediction results of mechanical drilling speed

[0121]

[0122] The prediction results of mechanical drilling speed are output as a drilling speed optimization recommendation table template containing important drilling parameters for the long horizontal section of shale gas horizontal wells in the Duverney work area, as shown in Table 5. It is easy for technicians in this field to understand and execute, and provides drilling parameter optimization suggestions for on-site drillers, thereby improving drilling efficiency.

[0123] Table 5 Drilling speed optimization suggestion table template

[0124]

[0125] Embodiment 2

[0126] Based on the same inventive concept, the present application embodiment also provides a mechanical drilling speed prediction model training method, referring to Figure 7 As shown, the method includes:

[0127] S201: Acquire initial drilling data and mechanical drilling rate of a long horizontal section of a historical shale gas horizontal well; the initial drilling data include: drilling pressure, drill bit speed, motor speed, surface speed, displacement, natural gamma, standpipe pressure, depth measurement, well inclination angle, azimuth angle, axial vibration and lateral vibration;

[0128] S202: Analyze the initial drilling data and the mechanical drilling rate using a correlation analysis method to obtain a training drilling data set;

[0129] S203: training a pre-built initial drilling speed prediction model according to the training drilling data set to obtain a trained drilling speed prediction model; wherein the initial drilling speed prediction model is constructed using XGboost;

[0130] S204: Optimize and train the trained drilling speed prediction model using a genetic algorithm until a preset condition is met to obtain the optimized drilling speed prediction model.

[0131] Embodiment 3

[0132] Based on the same inventive concept, the present application also provides a mechanical drilling speed prediction device, referring to Figure 8 As shown, the device comprises:

[0133] The first acquisition module 101 is used to acquire the initial drilling data and mechanical drilling rate of the long horizontal section of the historical shale gas horizontal well; the initial drilling data includes: drilling pressure, drill bit speed, motor speed, surface speed, displacement, natural gamma, standpipe pressure, depth measurement, well inclination angle, azimuth angle, axial vibration and lateral vibration;

[0134] A first processing module 102 is used to analyze the initial drilling data and the mechanical drilling rate using a correlation analysis method to obtain a training drilling data set;

[0135] The first training module 103 is used to train a pre-built initial drilling speed prediction model according to the training drilling data set to obtain a trained drilling speed prediction model; wherein the initial drilling speed prediction model is constructed using XGboost;

[0136] The second training module 104 is used to optimize and train the trained drilling speed prediction model using a genetic algorithm until a preset condition is met to obtain the optimized drilling speed prediction model;

[0137] The first prediction module 105 is used to input the drilling data of the long horizontal section of the shale gas horizontal well to be predicted into the optimized drilling speed prediction model to obtain the drilling speed prediction result.

[0138] Embodiment 4

[0139] Based on the same inventive concept, the present application also provides a mechanical drilling speed prediction model training device, referring to Fig. 9 As shown, the device comprises:

[0140] The first acquisition module 201 is used to acquire the initial drilling data and mechanical drilling rate of the long horizontal section of the historical shale gas horizontal well; the initial drilling data includes: drilling pressure, drill bit speed, motor speed, surface speed, displacement, natural gamma, standpipe pressure, depth measurement, well inclination angle, azimuth angle, axial vibration and lateral vibration;

[0141] A first processing module 202 is used to analyze the initial drilling data and the mechanical drilling rate using a correlation analysis method to obtain a training drilling data set;

[0142] The first training module 203 is used to train a pre-built initial drilling speed prediction model according to the training drilling data set to obtain a trained drilling speed prediction model; wherein the initial drilling speed prediction model is constructed using XGboost;

[0143] The second training module 204 is used to optimize and train the trained drilling speed prediction model using a genetic algorithm until a preset condition is met to obtain the optimized drilling speed prediction model.

[0144] Embodiment 5

[0145] Based on the same inventive concept, an embodiment of the present application also provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, it implements the mechanical drilling rate prediction method described in the above embodiment one, and / or the mechanical drilling rate prediction model training method described in the above embodiment two.

[0146] Embodiment 6

[0147] Based on the same inventive concept, an embodiment of the present application also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the mechanical drilling rate prediction method described in the above-mentioned embodiment 1 and / or the mechanical drilling rate prediction model training method described in the above-mentioned embodiment 2 are implemented.

[0148] Embodiment 7

[0149] Based on the same inventive concept, an embodiment of the present application also provides a computer program product comprising instructions. When the computer program product is run on a computer device, the computer device executes the mechanical drilling rate prediction method described in the above embodiment 1, and / or the mechanical drilling rate prediction model training method described in the above embodiment 2.

[0150] Embodiment 8

[0151] Based on the same inventive concept, an embodiment of the present application also provides a chip, the chip includes a processor and a communication interface, the communication interface and the processor are coupled, and the processor is used to run a computer program or instruction to implement the mechanical drilling rate prediction method described in the above embodiment 1, and / or the mechanical drilling rate prediction model training method described in the above embodiment 2.

[0152] The technicians in the relevant field 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 is divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated into a processing unit, or each unit can exist physically separately, or two or more units can be integrated into 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 device can refer to the corresponding process in the above-mentioned method embodiment, which will not be repeated here. If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such understanding, the present application implements all or part of the processes in the above-mentioned embodiment method, which can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned method embodiment when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may at least include: any entity or device capable of carrying computer program code, recording medium, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium. For example, a USB flash drive, a mobile hard disk, a magnetic disk or an optical disk. In some jurisdictions, according to legislation and patent practice, computer-readable media cannot be electric carrier signals and telecommunication signals.

[0153] The present application implements all or part of the processes in the above-mentioned embodiment method, and may also be completed through a computer program product. When the computer program product runs on a computer device, the steps in the above-mentioned method embodiment can be implemented when the computer device executes it.

[0154] 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.

[0155] Those of ordinary skill 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 this application.

[0156] In the embodiments provided in the present application, it should be understood that the disclosed devices / computer equipment and methods can be implemented in other ways. For example, the device / computer equipment embodiments described above are only schematic. For example, the division of modules or units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0157] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0158] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These 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 application, and should all be included in the protection scope of the present application.

Claims

1. A method for predicting mechanical drilling speed, characterized in that: include: Obtain the optimized drilling speed prediction model based on genetic algorithm optimization of XGboost; Input the acquired drilling data of the long horizontal section of the shale gas horizontal well to be predicted into the optimized drilling rate prediction model to obtain a drilling rate prediction result; The optimized drilling speed prediction model is obtained in the following way: Obtaining initial drilling data and mechanical drilling rate of a long horizontal section of a historical shale gas horizontal well; the initial drilling data includes: drilling pressure, drill bit speed, motor speed, surface speed, displacement, natural gamma, standpipe pressure, depth measurement, well inclination, azimuth, axial vibration and lateral vibration; Analyze the initial drilling data and the mechanical drilling rate using a correlation analysis method to obtain a training drilling data set; Training a pre-built initial drilling speed prediction model according to the training drilling data set to obtain a trained drilling speed prediction model; wherein the initial drilling speed prediction model is constructed using XGboost; The trained drilling speed prediction model is optimized and trained using a genetic algorithm until a preset condition is met, thereby obtaining the optimized drilling speed prediction model.

2. The method according to claim 1, characterized in that The method of using a genetic algorithm to optimize and train the trained drilling speed prediction model until a preset condition is met to obtain the optimized drilling speed prediction model comprises: Optimizing the parameters of the trained drilling speed prediction model using a genetic algorithm; When the loss function of the trained drilling speed prediction model reaches a minimum value, the optimized drilling speed prediction model is obtained; wherein the loss function is shown in the following formula 1: In the formula, ROP represents the actual mechanical penetration rate, It represents the predicted value obtained by the drilling speed prediction model.

3. The method according to claim 1, characterized in that The pre-built initial drilling speed prediction model is trained according to the training drilling data set to obtain a trained drilling speed prediction model, including: According to the data in the training drilling data set, the pre-built initial drilling speed prediction model is trained by using parameter adjustment and early stopping; When the determination coefficient R of the drilling speed prediction model 2 When the preset threshold is reached, the trained drilling speed prediction model is obtained.

4. The method according to claim 1, characterized in that Before using the correlation analysis method to analyze the initial drilling data and the mechanical drilling rate to obtain a training drilling data set, the method further includes: The drilling data is cleaned to obtain an initial drilling data set; wherein the data cleaning includes processing missing data, detecting redundant data, and eliminating abnormal data.

5. The method according to claim 4, characterized in that The data cleaning of the drilling data is performed to obtain an initial drilling data set; wherein the data cleaning includes processing missing data, detecting redundant data and removing abnormal data, specifically including: Identifying missing data in the drilling data, filling in missing values ​​or deleting missing values; Deleting redundant data repeatedly recorded in the drilling data; The noise data in the drilling data is removed by a digital filtering algorithm.

6. The method according to claim 1, characterized in that The method of using the correlation analysis method to analyze the initial drilling data set and the mechanical drilling rate of the long horizontal section of the shale gas horizontal well to obtain a training drilling data set includes: Solving multiple correlation coefficients between multiple data in the initial drilling data and the mechanical drilling rate by a correlation analysis method; The plurality of drilling data whose correlation coefficients reach a preset threshold are used as the training drilling data set.

7. The method according to claim 6, characterized in that The method further comprises: standardizing the training drilling data set based on the following formula 2: Among them, x′ k is the result after standardization, x k is the sample that needs to be standardized, x is the mean of the sample, and n is the total number of original samples.

8. A mechanical drilling rate prediction model training method, characterized in that: include: Obtain initial drilling data and mechanical penetration rate for long horizontal sections of historical shale gas horizontal wells; The initial drilling data include: drilling pressure, drill bit speed, motor speed, surface speed, displacement, natural gamma, standpipe pressure, depth, well inclination, azimuth, axial vibration and lateral vibration; Analyze the initial drilling data and the mechanical drilling rate using a correlation analysis method to obtain a training drilling data set; Training a pre-built initial drilling speed prediction model according to the training drilling data set to obtain a trained drilling speed prediction model; wherein the initial drilling speed prediction model is constructed using XGboost; The trained drilling speed prediction model is optimized and trained using a genetic algorithm until a preset condition is met, thereby obtaining the optimized drilling speed prediction model.

9. A mechanical drilling rate prediction device, characterized in that: include: The first acquisition module is used to obtain the initial drilling data and mechanical drilling rate of the long horizontal section of the historical shale gas horizontal well; The initial drilling data include: drilling pressure, drill bit speed, motor speed, surface speed, displacement, natural gamma, standpipe pressure, depth, well inclination, azimuth, axial vibration and lateral vibration; A first processing module is used to analyze the initial drilling data and the mechanical drilling rate using a correlation analysis method to obtain a training drilling data set; A first training module is used to train a pre-built initial drilling speed prediction model according to the training drilling data set to obtain a trained drilling speed prediction model; wherein the initial drilling speed prediction model is constructed using XGboost; A second training module is used to optimize and train the trained drilling speed prediction model using a genetic algorithm until a preset condition is met to obtain the optimized drilling speed prediction model; The second acquisition module is used to obtain an optimized drilling speed prediction model obtained by optimizing XGboost based on a genetic algorithm; The first prediction module is used to input the acquired drilling data of the long horizontal section of the shale gas horizontal well to be predicted into the optimized drilling speed prediction model to obtain the drilling speed prediction result.

10. A mechanical drilling rate prediction model training device, characterized in that: include: The first acquisition module is used to obtain the initial drilling data and mechanical drilling rate of the long horizontal section of the historical shale gas horizontal well; The initial drilling data include: drilling pressure, drill bit speed, motor speed, surface speed, displacement, natural gamma, standpipe pressure, depth, well inclination, azimuth, axial vibration and lateral vibration; A first processing module is used to analyze the initial drilling data and the mechanical drilling rate using a correlation analysis method to obtain a training drilling data set; A first training module is used to train a pre-built initial drilling speed prediction model according to the training drilling data set to obtain a trained drilling speed prediction model; wherein the initial drilling speed prediction model is constructed using XGboost; The second training module is used to optimize and train the trained drilling speed prediction model using a genetic algorithm until a preset condition is met to obtain the optimized drilling speed prediction model.

11. A computer-readable storage medium, wherein instructions are stored in the computer-readable storage medium. When the instructions are executed on a terminal, the terminal executes the mechanical drilling rate prediction method according to any one of claims 1 to 7, and / or the mechanical drilling rate prediction model training method according to claim 8.

12. A computer device, characterized in that: The method comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the method for predicting the mechanical drilling rate according to any one of claims 1 to 7 and / or the method for training a mechanical drilling rate prediction model according to claim 8 is implemented.

13. A computer program product comprising instructions, which, when executed on a computer device, enables the computer device to execute the mechanical drilling rate prediction method according to any one of claims 1 to 7, and / or the mechanical drilling rate prediction model training method according to claim 8.

14. A chip, comprising a processor and a communication interface, wherein the communication interface and the processor are coupled, and the processor is used to run a computer program or instruction to implement the mechanical drilling rate prediction method according to any one of claims 1 to 7, and / or the mechanical drilling rate prediction model training method according to claim 8.

Images