A method and apparatus for determining the location of a casing fault.
By training a neural network model on casing fault data, the problem of being unable to determine the location of casing faults in existing technologies has been solved, enabling accurate fault location and drilling decision support.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- PETROCHINA CO LTD
- Filing Date
- 2024-12-31
- Publication Date
- 2026-06-30
AI Technical Summary
Existing casing fault detection technologies cannot accurately pinpoint the location of faults, resulting in drilling operations being unable to repair them in a timely manner.
By acquiring historical fault data of the target bushing, a neural network model is trained, including parameter matching, pre-training, format structure matching, and optimization training, until the model's prediction accuracy reaches a threshold, thus identifying the location of the bushing fault.
It enables accurate location of casing faults, provides decision-making advice for drilling and cementing, and reduces detection costs.
Smart Images
Figure CN122310131A_ABST
Abstract
Description
Technical Field
[0001] This manual belongs to the field of geological exploration technology, and in particular relates to a method and apparatus for determining the location of a casing failure. Background Technology
[0002] Casing is a device fixed to the wellbore to support it, preventing wellbore collapse and controlling fluid flow within the wellbore strata. During drilling, if the casing malfunctions (e.g., wear, leakage, or breakage), it not only affects the smooth progress of drilling operations but also creates safety hazards. Therefore, when a casing malfunction occurs, drilling personnel need to determine the nature of the malfunction during the drilling process.
[0003] However, existing technical solutions for determining casing failure can only predict whether a failure has occurred, but cannot determine the location of the failure, making it impossible for drilling personnel to repair the failure in a timely manner.
[0004] Existing technologies have not yet provided an effective solution for accurately determining the location of casing faults. Summary of the Invention
[0005] This specification provides a method and apparatus for determining the location of a casing fault. By training a neural network model with historical fault data of the casing, the resulting target neural network model can accurately determine the location of the casing fault.
[0006] Specifically, the method and apparatus for determining the location of a bushing fault are implemented as follows:
[0007] A method for determining the location of a bushing fault, comprising:
[0008] Acquire historical fault data of the target casing, wherein the historical fault data includes various types of parameter data that affect the location of casing faults;
[0009] The parameter data affecting the location of the casing fault of the various types are matched with multiple neurons in the initial neural network model according to the parameter category to obtain the matching results;
[0010] According to the matching results, the historical fault data is input into the initial neural network model to pre-train the initial neural network model and obtain the first neural network model.
[0011] According to the target optimization parameters of the first neural network model, the format structure matching of the historical fault data is performed to obtain the format structure matching historical fault data.
[0012] The parameters of the first neural network model are updated using historical fault data after format structure matching to obtain an intermediate neural network model.
[0013] The real-time fault data of the target bushing in the real-time database is obtained, and the intermediate neural network model is optimized and trained until the prediction accuracy of the intermediate neural network model reaches a preset threshold, thus obtaining the target neural network model.
[0014] The location of the fault in the target bushing under actual working conditions is identified using the target neural network model.
[0015] In one implementation, according to the matching result, the historical fault data is input into the initial neural network model to pre-train the initial neural network model to obtain a first neural network model, including:
[0016] The historical fault data is preprocessed to obtain the target fault dataset;
[0017] The target fault dataset is divided to determine the training set and test set of the initial neural network model;
[0018] Based on the matching results, the target fault data of each target in the training set are input into the matching neurons to train the initial neural network model and obtain the first neural network model.
[0019] Based on the matching results, the target fault data of each target in the test set are input into the matching neurons to test the performance of the first neural network model.
[0020] In one implementation, the parameters of the first neural network model are updated using historical fault data with matched format structure to obtain an intermediate neural network model, including:
[0021] The parameter data affecting the location of the casing fault in the historical fault data after format matching is used as the target parameter dataset;
[0022] Determine the decay coefficient, adjustment factor, change vector, and initial update coefficient of each parameter in the target parameter dataset;
[0023] Based on the decay coefficient and change vector of each parameter in the target parameter dataset, determine the state variables of each parameter data respectively;
[0024] Based on the decay coefficient, adjustment factor, initial update coefficient, change vector, and state variable of each parameter in the target parameter dataset, the first neural network model is updated to obtain an intermediate neural network model.
[0025] In one implementation, the first neural network model is updated with parameters based on the decay coefficient, adjustment factor, initial update coefficient, change vector, and state variable of each parameter in the target parameter dataset to obtain an intermediate neural network model, including:
[0026] Based on the adjustment factor, initial update coefficient, and state variable of each parameter data in the target parameter dataset, the target update coefficient of each parameter data is determined respectively;
[0027] Determine whether the target update coefficient of each parameter in the target parameter dataset is less than a preset threshold;
[0028] The parameter data whose target update coefficient is greater than a preset threshold is used as the parameter data to be optimized, and the related parameters of the parameter data whose target update coefficient is greater than the preset threshold are used as the related parameters of the parameter data to be optimized. The related parameters are the decay coefficient, adjustment factor, target update coefficient, change vector and state variable.
[0029] Based on the attenuation coefficient, adjustment factor, change vector and state variable of the parameter data to be optimized, and the parameter data to be optimized, the target update coefficient of the parameter data to be optimized is updated until the target update coefficient of the parameter data to be optimized is less than a preset threshold.
[0030] When the target update coefficients of all parameters in the target parameter dataset are less than a preset threshold, the first neural network model at this time is used as the intermediate neural network model.
[0031] In one implementation, the target update coefficient of the parameter data to be optimized is updated based on the decay coefficient, adjustment factor, change vector, and state variable of the parameter data to be optimized, until the target update coefficient of the parameter data to be optimized is less than a preset threshold, including:
[0032] The parameter data to be optimized is updated based on the target update coefficient and change vector of the parameter data to be optimized;
[0033] Determine the change vector of the updated parameter data to be optimized;
[0034] The state variables of the updated parameter data to be optimized are determined based on the state variables and decay coefficients of the parameter data to be optimized before the update, and the change vector of the parameter data to be optimized after the update.
[0035] The target update coefficient of the updated parameter data is determined based on the target update coefficient and adjustment factor of the parameter data to be optimized before the update, and the state variables of the parameter data to be optimized after the update.
[0036] In one embodiment, the target neural network model is used to identify the location of the fault in the target bushing under actual operating conditions, including:
[0037] Obtain actual fault data of the target bushing under actual working conditions;
[0038] Extract various types of parameter data affecting the location of bushing faults from the actual fault data;
[0039] The various types of parameter data affecting the location of the casing fault, extracted from the actual fault data, are input into the target neural network model to obtain the location of the fault of the target casing under actual working conditions.
[0040] Based on the location of the failure of the target casing under actual working conditions, determine the distance between the failure location and the drilling platform.
[0041] A device for determining the location of a bushing fault, comprising:
[0042] The acquisition module is used to acquire historical fault data of the target casing, wherein the historical fault data includes various types of parameter data that affect the location of the casing fault;
[0043] The first matching module is used to match the parameter data of the various types affecting the location of the casing fault with multiple neurons in the initial neural network model according to the parameter category, and obtain the matching result;
[0044] The first training module is used to input the historical fault data into the initial neural network model according to the matching result, so as to pre-train the initial neural network model and obtain the first neural network model.
[0045] The second matching module is used to perform format structure matching on the historical fault data according to the target optimization parameters of the first neural network model to obtain the format structure matched historical fault data.
[0046] The update module is used to update the parameters of the first neural network model using historical fault data after format structure matching, so as to obtain an intermediate neural network model.
[0047] The second training module is used to acquire real-time fault data of the target bushing in the real-time database, optimize and train the intermediate neural network model until the prediction accuracy of the intermediate neural network model reaches a preset threshold, and obtain the target neural network model.
[0048] The identification module is used to identify the location of the fault in the target bushing under actual working conditions through the target neural network model.
[0049] An electronic device includes a processor and a memory for storing processor-executable instructions, wherein the processor, when executing the instructions, implements the steps of any of the methods described above.
[0050] A computer-readable storage medium having computer instructions stored thereon, which, when executed by a processor, implement the steps of any of the methods described above.
[0051] A computer program product comprising a computer program that, when executed by a processor, implements the steps of any of the methods described above.
[0052] The method and apparatus for determining the location of casing failures provided in this application involve acquiring historical failure data of the target casing, wherein the historical failure data includes various types of parameter data affecting the location of the casing failure; matching the various types of parameter data affecting the location of the casing failure with multiple neurons in an initial neural network model according to parameter categories to obtain matching results; inputting the historical failure data into the initial neural network model according to the matching results to pre-train the initial neural network model to obtain a first neural network model; performing format structure matching on the historical failure data according to the target optimization parameters of the first neural network model to obtain format structure-matched historical failure data; updating the parameters of the first neural network model using the format structure-matched historical failure data to obtain an intermediate neural network model; acquiring real-time failure data of the target casing in a real-time database and optimizing and training the intermediate neural network model until the prediction accuracy of the intermediate neural network model reaches a preset threshold to obtain a target neural network model; and identifying the location of the failure of the target casing under actual working conditions using the target neural network model. This scheme can accurately determine the location of casing failures, thereby enabling accurate judgment of drilling and cementing decisions based on the casing failure situation. Attached Figure Description
[0053] To more clearly illustrate the embodiments of this specification, the accompanying drawings used in the embodiments will be briefly introduced below. The drawings described below are only some embodiments recorded in this specification. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0054] Figure 1 This is a flowchart of one embodiment of the method for determining the location of a bushing failure provided in this application;
[0055] Figure 2 This is a schematic diagram illustrating a method for matching historical fault data with multiple neurons in one embodiment of the method for determining the location of a bushing fault provided in this application.
[0056] Figure 3 This is a flowchart illustrating a specific embodiment of the method for real-time determination of bushing faults using neural networks provided in this application.
[0057] Figure 4 This is a hardware structure block diagram of an electronic device for a method of determining the location of a bushing fault provided in this application;
[0058] Figure 5 This is a schematic diagram of the module structure of one embodiment of the bushing fault location determination device provided in this application. Detailed Implementation
[0059] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this specification, and not all embodiments. Based on the embodiments in this specification, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this specification.
[0060] To address the problem that existing technologies cannot determine the location of bushing failures during fault detection, this application provides a method for determining the location of bushing failures. Although this application provides method operation steps or device structures as shown in the following embodiments or accompanying drawings, more or fewer operation steps or module units may be included in the method or device based on conventional or non-inventive effort. In steps or structures where there is no logically necessary causal relationship, the execution order of these steps or the module structure of the device is not limited to the execution order or module structure described in the embodiments and accompanying drawings of this application. When the method or module structure is applied in actual devices or terminal products, it can be executed sequentially or in parallel according to the method or module structure shown in the embodiments or accompanying drawings (e.g., in a parallel processor or multi-threaded processing environment, or even a distributed processing environment).
[0061] Specifically, such as Figure 1 As shown, the method for determining the location of the bushing fault described above may include the following steps:
[0062] S101: Obtain historical fault data of the target casing, wherein the historical fault data includes various types of parameter data that affect the location of casing faults.
[0063] Specifically, the historical fault database of the target casing may contain various types of casing fault data. In order to improve the availability of data and simplify the training of subsequent neural network models, the parameter categories that affect the location of casing faults can be determined first. Then, the corresponding parameter data can be filtered out from the historical fault database as the historical fault data of the target casing. Furthermore, the various types of parameter data that affect the location of casing faults include formation pressure, casing depth, drilling pressure, drilling speed, drilling fluid density, drilling fluid pool volume, drilling fluid flowback rate, maximum shut-in pressure, and maximum hollowing depth when the casing is clean water, etc.
[0064] Specifically, the aforementioned historical fault data can be data in a single mode obtained from the same drilling equipment, data in different modes obtained from the same drilling equipment, data in a single mode obtained from different drilling equipment, and data in different modes obtained from different drilling equipment. For data in different modes from the same drilling equipment, as well as data in a single mode and data in different modes from different drilling equipment, the data can be processed uniformly first and then combined.
[0065] S102: The parameter data of the various types affecting the location of the casing fault are matched with multiple neurons in the initial neural network model according to the parameter category to obtain the matching result.
[0066] Specifically, neurons can be set up according to the parameter categories of the parameter data affecting the location of the casing fault, and each neuron can only receive parameter data of the parameter category it matches. For example, ... Figure 2 As shown, the parameter data affecting the location of casing failure are [A1, A2, A3, A4, A5, A6, A7, A8], and the corresponding parameter categories are [drilling pressure, drilling speed, drilling pressure, formation pressure, drilling speed, maximum shut-in pressure, drilling pressure, drilling speed]. The parameter categories of historical failure data are drilling pressure, drilling speed, formation pressure, and maximum shut-in pressure, which are four types. Accordingly, four neurons are set according to the parameter categories. The four neurons are matched with the parameter categories of historical failure data, so that there is a one-to-one correspondence between the four parameter categories and the four neurons. Finally, the matching results between the four parameter categories and the four neurons are obtained.
[0067] S103: According to the matching result, the historical fault data is input into the initial neural network model to pre-train the initial neural network model and obtain the first neural network model.
[0068] Specifically, pre-training the aforementioned initial neural network model requires preprocessing historical fault data. Furthermore, the historical fault data can be audited first. Data auditing can include completeness checks and accuracy checks. The completeness check mainly checks whether the combined historical fault data is missing or complete. The accuracy check mainly checks whether the combined historical fault data can truly reflect the fault status of the target casing, whether it conforms to the actual situation of drilling operations, and whether the filtered data is correct, logical, and whether there are contradictions between the various data. Then, based on the data audit results, missing data in the historical fault data is supplemented and flawed data is removed.
[0069] Furthermore, in order to pre-train the initial neural network model, historical fault data can be divided into a training set and a test set. Based on the matching results, the initial neural network model is trained using the training set to obtain the first neural network model. Then, the first neural network model is tested using the test set. Specifically, the initial neural network model can be pre-trained according to the following steps:
[0070] The historical fault data is preprocessed to obtain the target fault dataset;
[0071] The target fault dataset is divided to determine the training set and test set of the initial neural network model;
[0072] Based on the matching results, the target fault data of each target in the training set are input into the matching neurons to train the initial neural network model and obtain the first neural network model.
[0073] Based on the matching results, the target fault data of each target in the test set are input into the matching neurons to test the performance of the first neural network model.
[0074] S104: According to the target optimization parameters of the first neural network model, perform format structure matching on the historical fault data to obtain format structure matched historical fault data.
[0075] After obtaining the first neural network model, it can be further determined whether the target optimization parameters of the first neural network model meet the preset requirements. If the target optimization parameters do not meet the preset requirements, considering that the historical fault data contains all parameter data that affect the location of the casing fault at the time of the target fault, using all historical fault data to update the target optimization parameters of the first neural network model would significantly affect the update rate of the first neural network model. Therefore, based on the degree of influence of parameter data of each parameter category in the historical fault data on the target optimization parameters of the first neural network model, one or more parameter categories with the greatest influence can be selected from the historical fault data to optimize and train the first neural network model, thereby reducing the amount of data involved in model training and improving the training efficiency of the model. For example, assuming that the target optimization parameter A of the first neural network model does not meet the preset requirements, and the parameter data of parameter categories B and C have the greatest influence on the target optimization parameter A, then only parameter data of parameter categories B and C can be selected from the historical fault data to optimize and train the first neural network model.
[0076] Furthermore, to ensure the consistency of the selected historical fault data, the selected historical fault data can be further formatted and matched to unify the data format of the selected historical fault data, thereby making the training results of the first neural network model more accurate.
[0077] S105: Update the parameters of the first neural network model using historical fault data after format structure matching to obtain an intermediate neural network model.
[0078] To ensure that the parameters of the first neural network model meet the preset requirements of the target optimization parameters and thus improve the prediction accuracy of the first neural network model, the parameters of the first neural network model can be updated based on historical fault data after format structure matching. Specifically, the parameter update of the first neural network model may include the following steps:
[0079] S1: Use the parameter data that affects the location of the casing fault in the historical fault data after format matching as the target parameter dataset.
[0080] S2: Determine the decay coefficient, adjustment factor, change vector, and initial update coefficient of each parameter in the target parameter dataset.
[0081] Specifically, for each parameter in the target parameter dataset, the decay coefficient, adjustment factor, change vector and initial update coefficient can be set in the first neural network model. The first neural network model can adaptively adjust the model parameters for different parameter data.
[0082] S3: Determine the state variables of each parameter data based on the attenuation coefficient and change vector of each parameter data in the target parameter dataset.
[0083] S4: Based on the decay coefficient, adjustment factor, initial update coefficient, change vector and state variable of each parameter in the target parameter dataset, update the parameters of the first neural network model to obtain the intermediate neural network model.
[0084] Specifically, the first neural network model updates its parameters by successively adjusting the initial update coefficients, change vectors, and intermediate state data of each parameter in the target parameter dataset, so that the model parameters can adapt to parameter data of different parameter categories. Furthermore, the parameter update of the first neural network model may include the following steps:
[0085] S1: Determine the target update coefficient of each parameter data according to the adjustment factor, initial update coefficient and state variable of each parameter data in the target parameter dataset.
[0086] S2: Determine whether the target update coefficient of each parameter data in the target parameter dataset is less than a preset threshold;
[0087] S3: The parameter data with a target update coefficient greater than a preset threshold is taken as the parameter data to be optimized, and the relevant parameters of the parameter data with a target update coefficient greater than the preset threshold are taken as the relevant parameters of the parameter data to be optimized. The relevant parameters are the decay coefficient, adjustment factor, target update coefficient, change vector and state variable.
[0088] S4: Based on the decay coefficient, adjustment factor, change vector, and state variable of the parameter data to be optimized, and the parameter data to be optimized, update the target update coefficient of the parameter data to be optimized until the target update coefficient of the parameter data to be optimized is less than a preset threshold.
[0089] S5: When the target update coefficients of all parameters in the target parameter dataset are less than the preset threshold, the first neural network model at this time is used as the intermediate neural network model.
[0090] Furthermore, to ensure that the target update coefficient of the parameter to be optimized is less than a preset threshold, the parameter data to be optimized, the change vector of the parameter data to be optimized, and the intermediate state data can be iteratively updated to update the target update coefficient of the parameter to be optimized until the target update coefficient of the parameter data to be optimized is less than the preset threshold. Specifically, updating the target update coefficient of the parameter data to be optimized based on the decay coefficient, adjustment factor, change vector, state variable, and the parameter data to be optimized can include:
[0091] S1: Update the parameter data to be optimized according to the target update coefficient and change vector of the parameter data to be optimized.
[0092] S2: Determine the change vector of the updated parameter data to be optimized.
[0093] S3: Determine the state variables of the updated parameter data to be optimized based on the state variables and decay coefficients of the parameter data to be optimized before the update, and the change vector of the parameter data to be optimized after the update.
[0094] S4: Determine the target update coefficient of the updated parameter data based on the target update coefficient and adjustment factor of the parameter data to be optimized before the update, and the state variables of the parameter data to be optimized after the update.
[0095] S106: Obtain real-time fault data of the target bushing in the real-time database, optimize and train the intermediate neural network model until the prediction accuracy of the intermediate neural network model reaches a preset threshold, and obtain the target neural network model.
[0096] Existing neural network models are generally trained based on historical data. However, during formation drilling operations, there are numerous parameter categories that affect the location of casing faults, and these parameter categories may involve multiple parameter types, such as drilling fluid parameters, formation parameters, and process parameters. Since the purpose of each drilling operation may be different, the parameter categories and parameter types involved in the obtained parameter data affecting the location of casing faults may differ for different drilling operations in the same formation. Therefore, it cannot be guaranteed that the historical fault data of the target casing can cover the real-time fault data in the real-time fault database. If the neural network model is trained only based on historical fault data, the accuracy of the obtained neural network model is not high enough.
[0097] To further improve the prediction accuracy of the neural network model, this application further optimizes and trains the above-mentioned intermediate neural network model based on data from the real-time fault database of the target casing, and tests the optimized intermediate neural network model using real-time fault data until the prediction accuracy of the intermediate neural network model reaches a preset threshold or the number of training iterations reaches a preset number. The intermediate neural network model at this point is then used as the target neural network model.
[0098] S107: The location of the fault in the target bushing under actual working conditions is identified through the target neural network model.
[0099] Specifically, in actual working conditions, real-time fault data of the target casing can be obtained first. Then, various types of parameter data affecting the location of casing faults can be extracted from the real-time fault data and input into the aforementioned target neural network model to obtain the location of the fault in the target casing under actual working conditions. Based on the location of the fault, the distance between the location of the fault and the drilling platform can be determined.
[0100] The above approach allows for the training of a target neural network model by sequentially matching the data types of various parameters affecting the location of casing faults with the data types of multiple neurons, matching the format and structure of historical fault data with the target optimization parameters of the neural network model, and training the neural network model with the actual fault data in the real-time fault database of the target casing. This results in a target neural network model, which can accurately determine the location of casing faults and provide decision-making suggestions for drilling and cementing.
[0101] The above method will be described below with reference to a specific embodiment. However, it is worth noting that this specific embodiment is only for better illustration of this application and does not constitute an improper limitation of this application.
[0102] In this example, considering the problem that existing technologies cannot determine the location of casing faults when detecting them, a method for real-time casing fault determination using neural networks is proposed. This method combines historical fault data from a historical fault database with real-time fault data from a real-time fault database to train a neural network model, thereby obtaining a target neural network model. By inputting casing fault data from actual working conditions into the target neural network model, the location of the casing fault can be accurately determined, providing decision-making suggestions for drilling and cementing, and reducing the cost of casing fault detection.
[0103] Specifically, such as Figure 3 As shown, the above-mentioned method for real-time determination of bushing faults using neural networks may include the following steps:
[0104] S301: Obtain historical fault data affecting the location of faults affecting the target bushing, wherein the historical fault data contains various types of parameter data.
[0105] Specifically, the aforementioned historical fault data consists of data from different drilling rigs in different patterns. The fault data from different patterns are combined, and further, the combined fault data can be clustered. For example, the similarity between each fault data is determined, and then the fault data is clustered based on the similarity between the fault data.
[0106] Specifically, the parameter types in the historical fault data include formation pressure, casing depth, drilling pressure, drilling speed, drilling fluid density, total pool volume, flowback rate, maximum shut-in pressure, and maximum scour depth when using clean water.
[0107] S302: Perform data preprocessing on historical fault data.
[0108] Specifically, historical fault data can be filtered to obtain data that is more conducive to data mining. Then, the filtered data can be combined and checked to fill in missing data and remove flawed data.
[0109] The aforementioned checks on the screened data can include completeness checks and accuracy checks. In addition, the completeness check mainly checks whether the screened data is missing or complete. The accuracy check mainly checks whether the screened data can truly reflect the fault status of the target casing, whether it conforms to the actual situation of drilling operations, and whether the screened data is correct, logical, and whether there are any contradictions between the various data.
[0110] Specifically, by preprocessing the fault-related data to be trained, the input data for data mining can be enhanced by filling in missing items, removing redundant items, and unifying the input data. This addresses issues such as inconsistent input data definitions and outdated input data, ensuring the completeness and correctness of the input data and improving the accuracy of subsequent data mining results.
[0111] Specifically, batch management can be set for the data type of historical fault data to unify historical fault data and make subsequent data mining simpler.
[0112] S303: Based on the parameter type of the historical fault data, determine the matching results between the historical fault data and multiple neurons of the neural network model.
[0113] Specifically, assuming that the historical fault data has nine parameter types, namely formation pressure, casing depth, drilling pressure, drilling speed, drilling fluid density, total pool volume, flowback rate, maximum shut-in pressure, and maximum scour depth in clear water, nine neurons are set in the input layer of the neural network model, and these nine neurons are matched with the nine parameter types. Based on the matching results between the historical fault data and multiple neurons of the neural network model, the nine neurons are used to input parameter data of the matching parameter types. Furthermore, the output layer of the above neural network model has one neuron to output the predicted result of the casing fault location.
[0114] S304: Based on the matching results between historical fault data and multiple neurons of the neural network model, input the historical fault data into the neural network model for supervised training to obtain the initial neural network model.
[0115] The parameter data from historical fault data is divided into two sets: 70% for the training set and 30% for the test set. Based on the matching results, the parameter data from the training set is input into the corresponding neurons of the neural network model to train the model, resulting in an initial neural network model. Finally, the parameter data from the test set is input into the neurons corresponding to the initial neural network model to test it.
[0116] Furthermore, to avoid mutual interference between the input parameters of the neural network model, as well as between the input parameters of the initial neural network model, the training set and the test set can be normalized.
[0117] S305: Initialize the initial neural network model and align the parameter data in the historical fault data.
[0118] Specifically, the initial neural network model can be initialized using a double-loop recursive algorithm, and then the weight matrix of the initial neural network model can be generated. Further, the initial values of the weight matrix can be determined, and the thresholds of the hidden layer and the output layer can be set to 0. Then, the thresholds of the hidden layer and the output layer and the weight values of the weight matrix can be randomly generated using the Nguyen-Widrow algorithm, and all weight parameters from the hidden layer to the output layer can be set to 1. Finally, the weight parameters from the hidden layer to the output layer can be randomly generated using Python's random module.
[0119] Specifically, in order to ensure the consistency of historical fault data, further data alignment can be performed on the historical fault data to unify the data format of the historical fault data, thereby making the training results of the initial neural network model more accurate.
[0120] S306: Input the aligned historical fault data into the initial neural network model for incremental supervised training to obtain a candidate neural network model.
[0121] Specifically, the aligned historical fault data can be used for initialization, and the learning rate, decay coefficient, and stabilization factor of each parameter data in the aligned historical fault data can be set. Then, the gradient and cumulative squared gradient corresponding to each parameter data can be calculated. Based on the learning rate, decay coefficient, stabilization factor, initial gradient, and initial cumulative squared gradient of each parameter data, the candidate neural network model updates each parameter data and the learning rate of each parameter data in each iteration of training until the learning rate of each parameter data is less than a preset threshold. The candidate neural network model at this time is then used as the target neural network model.
[0122] Furthermore, taking a single parameter data F as an example, assuming the candidate neural network model iterates 300 times, and the preset threshold for the learning rate of parameter data F is e. -8 The process of updating the parameter data F and the learning rate of the parameter data F in each iteration of the candidate neural network model can include the following steps:
[0123] S1: Update the parameter data F based on the learning rate and gradient.
[0124] S2: Determine the gradient of the updated parameter data F.
[0125] S3: Based on the cumulative squared gradient and decay coefficient of the parameter data F before the update, and the gradient of the parameter data F after the update, determine the cumulative squared gradient of the updated parameter data F.
[0126] S4: Determine the learning rate of the updated parameter data F based on the learning rate and stabilization factor of the parameter data F before the update, and the cumulative squared gradient of the parameter data F after the update.
[0127] S5: Repeat steps S1-S4 until the learning rate of parameter data F is less than e. -8 Or, the number of iterations of the candidate neural network model reaches 300.
[0128] S307: Use candidate neural network models to train batches of real-time bushing fault data in the real-time database to obtain the target neural network model.
[0129] S308: The location of the fault in the bushing under actual working conditions is determined by the target neural network model.
[0130] In the example above, by combining historical fault data from the historical fault database and real-time fault data from the real-time fault database to train the neural network model, a target neural network model is obtained. By inputting casing fault data from actual working conditions into the target neural network model, the location of casing faults can be accurately determined, providing decision-making suggestions for drilling and cementing, and reducing the cost of casing fault detection.
[0131] The methods and embodiments provided in the above-described embodiments of this application can be executed in a mobile terminal, computer terminal, or similar computing device. Taking its operation on an electronic device as an example... Figure 4 This is a hardware structure block diagram of an electronic device for a method of determining the location of a bushing fault provided in this application. (See diagram for example.) Figure 4 As shown, the electronic device 10 may include one or more (only one is shown in the figure) processors 02 (processors 02 may include, but are not limited to, microprocessors MCUs or programmable logic devices FPGAs, etc.), a memory 04 for storing data, and a transmission module 06 for communication functions. Those skilled in the art will understand that... Figure 4 The structure shown is for illustrative purposes only and does not limit the structure of the electronic device described above. For example, electronic device 10 may also include... Figure 4 The more or fewer components shown, or having the same Figure 4 The different configurations shown.
[0132] The memory 04 can be used to store software programs and modules of application software, such as the program instructions / modules corresponding to the method for determining the location of a bushing fault in this embodiment. The processor 02 executes various functional applications and data processing by running the software programs and modules stored in the memory 04, thereby implementing the aforementioned method for determining the location of a bushing fault in the application. The memory 04 may include high-speed random access memory and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 04 may further include memory remotely located relative to the processor 02, and these remote memories can be connected to the electronic device 10 via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.
[0133] The transmission module 06 is used to receive or send data via a network. Specific examples of the network described above may include a wireless network provided by the communication provider of the electronic device 10. In one example, the transmission module 06 includes a Network Interface Controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the transmission module 06 may be a Radio Frequency (RF) module, used for wireless communication with the Internet.
[0134] At the software level, the aforementioned device for determining the location of the bushing fault can be as follows: Figure 5 As shown, it includes:
[0135] The acquisition module 501 is used to acquire historical fault data of the target casing, wherein the historical fault data includes various types of parameter data that affect the location of the casing fault;
[0136] The first matching module 502 is used to match the parameter data of the various types affecting the location of the casing fault with multiple neurons in the initial neural network model according to the parameter category, and obtain the matching result;
[0137] The first training module 503 is used to input the historical fault data into the initial neural network model according to the matching result, so as to pre-train the initial neural network model and obtain the first neural network model.
[0138] The second matching module 504 is used to perform format structure matching on the historical fault data according to the target optimization parameters of the first neural network model to obtain the format structure matched historical fault data.
[0139] The update module 505 is used to update the parameters of the first neural network model using historical fault data after format structure matching, so as to obtain an intermediate neural network model.
[0140] The second training module 506 is used to acquire real-time fault data of the target bushing in the real-time database, optimize and train the intermediate neural network model until the prediction accuracy of the intermediate neural network model reaches a preset threshold, and obtain the target neural network model.
[0141] The identification module 507 is used to identify the location of the fault in the target bushing under actual working conditions through the target neural network model.
[0142] In one implementation, the first training module 503 inputs the historical fault data into the initial neural network model according to the matching result to pre-train the initial neural network model and obtain the first neural network model, which may include the following steps:
[0143] The historical fault data is preprocessed to obtain the target fault dataset;
[0144] The target fault dataset is divided to determine the training set and test set of the initial neural network model;
[0145] Based on the matching results, the target fault data of each target in the training set are input into the matching neurons to train the initial neural network model and obtain the first neural network model.
[0146] Based on the matching results, the target fault data of each target in the test set are input into the matching neurons to test the performance of the first neural network model.
[0147] In one implementation, the update module 505 updates the parameters of the first neural network model using historical fault data after format structure matching to obtain an intermediate neural network model, which may include the following steps:
[0148] The parameter data affecting the location of the casing fault in the historical fault data after format matching is used as the target parameter dataset;
[0149] Determine the decay coefficient, adjustment factor, change vector, and initial update coefficient of each parameter in the target parameter dataset;
[0150] Based on the decay coefficient and change vector of each parameter in the target parameter dataset, determine the state variables of each parameter data respectively;
[0151] Based on the decay coefficient, adjustment factor, initial update coefficient, change vector, and state variable of each parameter in the target parameter dataset, the first neural network model is updated to obtain an intermediate neural network model.
[0152] In one implementation, updating the parameters of the first neural network model based on the decay coefficient, adjustment factor, initial update coefficient, change vector, and state variable of each parameter in the target parameter dataset to obtain an intermediate neural network model may include the following steps:
[0153] Based on the adjustment factor, initial update coefficient, and state variable of each parameter data in the target parameter dataset, the target update coefficient of each parameter data is determined respectively;
[0154] Determine whether the target update coefficient of each parameter in the target parameter dataset is less than a preset threshold;
[0155] The parameter data whose target update coefficient is greater than a preset threshold is used as the parameter data to be optimized, and the related parameters of the parameter data whose target update coefficient is greater than the preset threshold are used as the related parameters of the parameter data to be optimized. The related parameters are the decay coefficient, adjustment factor, target update coefficient, change vector and state variable.
[0156] Based on the attenuation coefficient, adjustment factor, change vector and state variable of the parameter data to be optimized, and the parameter data to be optimized, the target update coefficient of the parameter data to be optimized is updated until the target update coefficient of the parameter data to be optimized is less than a preset threshold.
[0157] When the target update coefficients of all parameters in the target parameter dataset are less than a preset threshold, the first neural network model at this time is used as the intermediate neural network model.
[0158] In one implementation, updating the target update coefficient of the parameter data to be optimized based on the decay coefficient, adjustment factor, change vector, and state variable of the parameter data to be optimized, until the target update coefficient of the parameter data to be optimized is less than a preset threshold, may include the following steps:
[0159] The parameter data to be optimized is updated based on the target update coefficient and change vector of the parameter data to be optimized;
[0160] Determine the change vector of the updated parameter data to be optimized;
[0161] The state variables of the updated parameter data to be optimized are determined based on the state variables and decay coefficients of the parameter data to be optimized before the update, and the change vector of the parameter data to be optimized after the update.
[0162] The target update coefficient of the updated parameter data is determined based on the target update coefficient and adjustment factor of the parameter data to be optimized before the update, and the state variables of the parameter data to be optimized after the update.
[0163] In one embodiment, the identification module 507 identifies the location of the fault in the target bushing under actual working conditions through the target neural network model, which may include the following steps:
[0164] Obtain actual fault data of the target bushing under actual working conditions;
[0165] Extract various types of parameter data affecting the location of bushing faults from the actual fault data;
[0166] The various types of parameter data affecting the location of the casing fault, extracted from the actual fault data, are input into the target neural network model to obtain the location of the fault of the target casing under actual working conditions.
[0167] Based on the location of the failure of the target casing under actual working conditions, determine the distance between the failure location and the drilling platform.
[0168] The embodiments of this application also provide a specific implementation of an electronic device capable of implementing all steps in the bushing fault location determination method in the above embodiments. The electronic device specifically includes: a processor, a memory, a communication interface, and a bus; wherein the processor, memory, and communication interface communicate with each other via the bus; the processor is used to call a computer program in the memory, and when the processor executes the computer program, it implements all steps in the bushing fault location determination method in the above embodiments. For example, when the processor executes the computer program, it implements the following steps:
[0169] Step 1: Obtain historical fault data of the target casing, wherein the historical fault data includes various types of parameter data that affect the location of casing faults;
[0170] Step 2: Match the parameter data of the various types affecting the location of the casing fault with multiple neurons in the initial neural network model according to the parameter category to obtain the matching results;
[0171] Step 3: According to the matching results, input the historical fault data into the initial neural network model to pre-train the initial neural network model and obtain the first neural network model;
[0172] Step 4: According to the target optimization parameters of the first neural network model, perform format structure matching on the historical fault data to obtain the format structure matched historical fault data;
[0173] Step 5: Update the parameters of the first neural network model using historical fault data after format structure matching to obtain the intermediate neural network model;
[0174] Step 6: Obtain real-time fault data of the target bushing in the real-time database, optimize and train the intermediate neural network model until the prediction accuracy of the intermediate neural network model reaches a preset threshold, and obtain the target neural network model.
[0175] Step 7: Using the target neural network model, identify the location of the fault in the target bushing under actual working conditions.
[0176] Embodiments of this application also provide a computer-readable storage medium capable of implementing all steps of the bushing fault location determination method in the above embodiments. The computer-readable storage medium stores a computer program that, when executed by a processor, implements all steps of the bushing fault location determination method in the above embodiments. For example, when the processor executes the computer program, it implements the following steps:
[0177] Step 1: Obtain historical fault data of the target casing, wherein the historical fault data includes various types of parameter data that affect the location of casing faults;
[0178] Step 2: Match the parameter data of the various types affecting the location of the casing fault with multiple neurons in the initial neural network model according to the parameter category to obtain the matching results;
[0179] Step 3: According to the matching results, input the historical fault data into the initial neural network model to pre-train the initial neural network model and obtain the first neural network model;
[0180] Step 4: According to the target optimization parameters of the first neural network model, perform format structure matching on the historical fault data to obtain the format structure matched historical fault data;
[0181] Step 5: Update the parameters of the first neural network model using historical fault data after format structure matching to obtain the intermediate neural network model;
[0182] Step 6: Obtain real-time fault data of the target bushing in the real-time database, optimize and train the intermediate neural network model until the prediction accuracy of the intermediate neural network model reaches a preset threshold, and obtain the target neural network model.
[0183] Step 7: Using the target neural network model, identify the location of the fault in the target bushing under actual working conditions.
[0184] As described above, this application embodiment acquires historical fault data of the target casing, which includes various types of parameter data affecting the location of casing faults. These parameters are then matched with multiple neurons in an initial neural network model according to their categories to obtain matching results. The historical fault data is input into the initial neural network model for pre-training, resulting in a first neural network model. The historical fault data is then subjected to format structure matching according to the target optimization parameters of the first neural network model, resulting in format structure-matched historical fault data. The first neural network model is updated with the format structure-matched historical fault data to obtain an intermediate neural network model. Real-time fault data of the target casing is acquired from a real-time database, and the intermediate neural network model is optimized and trained until its prediction accuracy reaches a preset threshold, resulting in a target neural network model. The target neural network model is then used to identify the location of the target casing fault under actual operating conditions. This scheme can accurately determine the location of casing faults, thereby enabling accurate judgment of drilling and cementing decisions based on the casing fault situation.
[0185] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to interchangeably. Each embodiment focuses on its differences from other embodiments. In particular, hardware + program embodiments are relatively simple in description because they are fundamentally similar to method embodiments; relevant parts can be referred to the descriptions in the method embodiments.
[0186] The foregoing has described specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in a different order than that shown in the embodiments and may still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require the specific or sequential order shown to achieve the desired result. In some embodiments, multitasking and parallel processing are possible or may be advantageous.
[0187] While this application provides the method operation steps as described in the embodiments or flowcharts, more or fewer operation steps may be included based on conventional or non-inventive labor. The order of steps listed in the embodiments is merely one possible execution order among many and does not represent the only execution order. In actual device or client product execution, the methods shown in the embodiments or drawings can be executed sequentially or in parallel (e.g., in a parallel processor or multi-threaded processing environment).
[0188] While this specification provides method operation steps as described in the embodiments or flowcharts, more or fewer operation steps may be included based on conventional or non-inventive means. The order of steps listed in the embodiments is merely one possible execution order among many and does not represent the only execution order. In actual device or end product execution, the methods shown in the embodiments or drawings may be executed sequentially or in parallel (e.g., in a parallel processor or multi-threaded processing environment, or even a distributed data processing environment). The terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, product, 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 a process, method, product, or apparatus. Without further limitations, the presence of other identical or equivalent elements in the process, method, product, or apparatus that includes said elements is not excluded.
[0189] For ease of description, the above devices are described in terms of function, divided into various modules. Of course, in implementing the embodiments of this specification, the functions of each module can be implemented in one or more software and / or hardware components, or a module that performs the same function can be implemented by a combination of multiple sub-modules or sub-units. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division; in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces, or indirect coupling or communication connection between devices or units, and may be electrical, mechanical, or other forms.
[0190] Those skilled in the art will also know that, besides implementing the controller using purely computer-readable program code, the same functions can be achieved by logically programming the method steps, making the controller function as logic gates, switches, application-specific integrated circuits (ASICs), programmable logic controllers (PLCs), and embedded microcontrollers. Therefore, such a controller can be considered a hardware component, and the devices within it used to implement various functions can also be considered structures within that hardware component. Alternatively, the devices used to implement various functions can be considered as both software modules implementing the method and structures within a hardware component.
[0191] 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.
[0192] Those skilled in the art will understand that the embodiments of this specification can be provided as methods, systems, or computer program products. Therefore, the embodiments of this specification can take the form of entirely hardware embodiments, entirely software embodiments, or embodiments combining software and hardware aspects. Furthermore, the embodiments of this specification can take the form of computer program products implemented 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.
[0193] The embodiments described in this specification can be described in the general context of computer-executable instructions, such as program modules, that are executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform a specific task or implement a specific abstract data type. The embodiments of this specification can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.
[0194] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, system embodiments are basically similar to method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments. In the description of this specification, the terms "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of the embodiments in this specification. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described can be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification and the features of different embodiments or examples.
[0195] The above description is merely an embodiment of the present specification and is not intended to limit the embodiments of the present specification. For those skilled in the art, various modifications and variations can be made to the embodiments of the present specification. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principle of the embodiments of the present specification should be included within the scope of the claims of the embodiments of the present specification.
Claims
1. A method for determining the location of a bushing fault, characterized in that, include: Acquire historical fault data of the target casing, wherein the historical fault data includes various types of parameter data that affect the location of casing faults; The parameter data affecting the location of the casing fault of the various types are matched with multiple neurons in the initial neural network model according to the parameter category to obtain the matching results; According to the matching results, the historical fault data is input into the initial neural network model to pre-train the initial neural network model and obtain the first neural network model. According to the target optimization parameters of the first neural network model, the format structure matching of the historical fault data is performed to obtain the format structure matching historical fault data. The parameters of the first neural network model are updated using historical fault data after format structure matching to obtain an intermediate neural network model. The real-time fault data of the target bushing in the real-time database is obtained, and the intermediate neural network model is optimized and trained until the prediction accuracy of the intermediate neural network model reaches a preset threshold, thus obtaining the target neural network model. The location of the fault in the target bushing under actual working conditions is identified using the target neural network model.
2. The method according to claim 1, characterized in that, According to the matching results, the historical fault data is input into the initial neural network model to pre-train the initial neural network model, resulting in a first neural network model, including: The historical fault data is preprocessed to obtain the target fault dataset; The target fault dataset is divided to determine the training set and test set of the initial neural network model; Based on the matching results, the target fault data of each target in the training set are input into the matching neurons to train the initial neural network model and obtain the first neural network model. Based on the matching results, the target fault data of each target in the test set are input into the matching neurons to test the performance of the first neural network model.
3. The method according to claim 1, characterized in that, The parameters of the first neural network model are updated using historical fault data after format structure matching to obtain an intermediate neural network model, including: The parameter data affecting the location of the casing fault in the historical fault data after format matching is used as the target parameter dataset; Determine the decay coefficient, adjustment factor, change vector, and initial update coefficient of each parameter in the target parameter dataset; Based on the decay coefficient and change vector of each parameter in the target parameter dataset, determine the state variables of each parameter data respectively; Based on the decay coefficient, adjustment factor, initial update coefficient, change vector, and state variable of each parameter in the target parameter dataset, the first neural network model is updated to obtain an intermediate neural network model.
4. The method according to claim 3, characterized in that, Based on the decay coefficient, adjustment factor, initial update coefficient, change vector, and state variable of each parameter in the target parameter dataset, the first neural network model is updated to obtain an intermediate neural network model, including: Based on the adjustment factor, initial update coefficient, and state variable of each parameter data in the target parameter dataset, the target update coefficient of each parameter data is determined respectively; Determine whether the target update coefficient of each parameter in the target parameter dataset is less than a preset threshold; The parameter data with a target update coefficient greater than a preset threshold are used as the parameter data to be optimized, and the related parameters of the parameter data with a target update coefficient greater than the preset threshold are used as the related parameters of the parameter data to be optimized. The related parameters are the decay coefficient, adjustment factor, target update coefficient, change vector and state variable. Based on the decay coefficient, adjustment factor, change vector and state variable of the parameter data to be optimized, and the parameter data to be optimized, the target update coefficient of the parameter data to be optimized is updated until the target update coefficient of the parameter data to be optimized is less than a preset threshold. If the target update coefficients of all parameters in the target parameter dataset are less than a preset threshold, the first neural network model at this time is used as the intermediate neural network model.
5. The method according to claim 4, characterized in that, Based on the decay coefficient, adjustment factor, change vector, and state variable of the parameter data to be optimized, and the parameter data to be optimized, the target update coefficient of the parameter data to be optimized is updated until the target update coefficient of the parameter data to be optimized is less than a preset threshold, including: The parameter data to be optimized is updated based on the target update coefficient and change vector of the parameter data to be optimized; Determine the change vector of the updated parameter data to be optimized; The state variables of the updated parameter data to be optimized are determined based on the state variables and decay coefficients of the parameter data to be optimized before the update, and the change vector of the parameter data to be optimized after the update. The target update coefficient of the updated parameter data is determined based on the target update coefficient and adjustment factor of the parameter data to be optimized before the update, and the state variables of the parameter data to be optimized after the update.
6. The method according to claim 1, characterized in that, The target neural network model is used to identify the location of the fault in the target bushing under actual working conditions, including: Obtain actual fault data of the target bushing under actual working conditions; Extract various types of parameter data affecting the location of bushing faults from the actual fault data; The various types of parameter data affecting the location of the casing fault extracted from the actual fault data are input into the target neural network model to obtain the location of the fault of the target casing under actual working conditions. Based on the location of the failure of the target casing under actual working conditions, determine the distance between the failure location and the drilling platform.
7. A device for determining the location of a bushing fault, characterized in that, include: The acquisition module is used to acquire historical fault data of the target casing, wherein the historical fault data includes various types of parameter data that affect the location of the casing fault; The first matching module is used to match the parameter data of the various types affecting the location of the casing fault with multiple neurons in the initial neural network model according to the parameter category, and obtain the matching result; The first training module is used to input the historical fault data into the initial neural network model according to the matching result, so as to pre-train the initial neural network model and obtain the first neural network model. The second matching module is used to perform format structure matching on the historical fault data according to the target optimization parameters of the first neural network model to obtain the format structure matched historical fault data. The update module is used to update the parameters of the first neural network model using historical fault data after format structure matching, so as to obtain an intermediate neural network model. The second training module is used to acquire real-time fault data of the target bushing in the real-time database, optimize and train the intermediate neural network model until the prediction accuracy of the intermediate neural network model reaches a preset threshold, and obtain the target neural network model. The identification module is used to identify the location of the fault in the target bushing under actual working conditions through the target neural network model.
8. An electronic device comprising a processor and a memory for storing processor-executable instructions, characterized in that, When the processor executes the instructions, it implements the steps of the method according to any one of claims 1 to 6.
9. A computer-readable storage medium storing computer instructions thereon, characterized in that, When the instructions are executed by the processor, they implement the steps of the method according to any one of claims 1 to 6.
10. A computer program product comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.