Digital quality control method and system based on induction logging equipment

Abstract

The invention discloses a digital quality control method and system based on induction logging equipment, and the method comprises the steps: monitoring the state of the induction logging equipment, and obtaining the actual induction logging data when the induction logging equipment operates normally; performing standard data format processing on the actual induction logging data to realize multi-source heterogeneous data fusion of the actual induction logging data to obtain a logging standard database and a logging abnormal database; establishing a training induction logging prediction model based on the logging standard database, and performing prediction by using the induction logging prediction model to obtain prediction data of induction logging; and based on a logging standard database and a logging abnormal database, identifying whether the prediction data is abnormal data, and realizing induction logging data quality control. According to the method, the LSTM machine learning algorithm is effectively utilized to carry out digital quality control on the induction logging data.

Description

Technical Field

[0001] The present invention relates to the technical field of oil and gas exploration, and in particular to a digital quality control method and system based on induction logging equipment. Background Art

[0002] Qualified logging data is the premise and basis of logging interpretation. Logging data includes instrument calibration and calibration, logging data, and related data. The quality of the original logging data directly affects the reliability of the interpretation results and the exploration and development process of a region. Due to the characteristics of large data volume and multi-source heterogeneity, the logging processing and interpretation process faces difficulties such as multi-solution and uncertainty, and the difficulty of oil and gas identification is increasing. Therefore, it is urgent to use the advantages of artificial intelligence in big data to apply to the field of logging technology.

[0003] In order to solve the problems of multi-solution and uncertainty faced by the above logging data in the process of logging processing and interpretation, artificial intelligence methods are generally used instead of traditional methods to solve the difficulties caused by the large volume and multi-source heterogeneity of logging data. In recent years, more and more people have explored the application of artificial intelligence in the field of logging technology. Studies have found that various machine learning algorithms can break away from the assumed limitations of traditional rock physics volume models, bring a larger function space for logging data interpretation, and help logging data processing and interpretation personnel to discover knowledge from high-dimensional space in a nonlinear way.

[0004] However, there is limited research on the quality of well logging data in the field of artificial intelligence. On the one hand, it is due to the large volume and multi-source heterogeneity of well logging data. On the other hand, it is due to the complexity of the real formation environment and the uncertainty of well logging data quality. As a result, machine learning algorithms cannot effectively predict and classify data due to data uncertainty and different data structures. Therefore, there is an urgent need for a method to use artificial intelligence to control the quality of digital well logging data. Summary of the invention

[0005] The purpose of the present invention is to provide a digital quality control method and system based on induction logging equipment, which effectively utilizes the LSTM machine learning algorithm to perform digital quality control on induction logging data.

[0006] To achieve the above object, the present invention provides a digital quality control method based on induction logging equipment, comprising:

[0007] Monitor the status of the induction logging equipment and obtain actual induction logging data when the induction logging equipment operates normally;

[0008] Processing the actual induction logging data in a standard data format to achieve multi-source heterogeneous data fusion of the actual induction logging data, and obtaining a logging standard database and a logging anomaly database;

[0009] Establishing and training an induction logging prediction model based on a well logging standard database, and using the induction logging prediction model to make predictions to obtain induction logging prediction data;

[0010] Based on the well logging standard database and the well logging anomaly database, it is identified whether the predicted data is abnormal data, thereby achieving quality control of induction logging data.

[0011] Furthermore, the status of the induction logging equipment is detected to obtain actual induction logging data when the induction logging equipment operates normally, including:

[0012] The monitor is used to monitor the power supply, generator frequency, oil temperature balance, physical scale and circuit status in the induction logging equipment, and output the operating status through the corresponding status bit;

[0013] If the operating status of each status bit is normal or warning, it is determined that the induction logging equipment operates normally, and actual induction logging data monitored by the induction logging equipment is obtained;

[0014] The operating status of each status bit includes: fault, warning and normal.

[0015] Furthermore, the actual induction logging data is processed into a standard data format to realize multi-source heterogeneous data fusion of the actual induction logging data, and obtain a logging standard database and a logging anomaly database, including:

[0016] performing data processing on the actual induction logging data to obtain processed induction logging data;

[0017] The processed induction logging data is processed through a standard data format to achieve multi-source heterogeneous data fusion, and to establish a well logging standard database and well logging anomaly database with a unified data structure;

[0018] The data processing includes: raw data preprocessing, borehole correction processing, skin effect correction processing and synthetic focusing processing.

[0019] Furthermore, an induction logging prediction model is established and trained based on the logging standard database, and the induction logging prediction model is used for prediction to obtain prediction data of induction logging, including:

[0020] Establish a multi-layer simulated formation model based on the data in the well logging standard database;

[0021] Based on the data of the multi-layer simulated formation model, the training set data and the test set data are obtained by dividing them in proportion;

[0022] Based on the LSTM neural network, the induction logging prediction model is trained using the training set data to obtain the trained induction logging prediction model;

[0023] The test set data is used to predict the trained induction logging prediction model to obtain the prediction data of induction logging.

[0024] Furthermore, based on the well logging standard database and the well logging anomaly database, identifying whether the predicted data is abnormal data and implementing induction logging data quality control includes:

[0025] Using the three-times standard deviation criterion, the predicted data of the induction logging is discriminated from the data in the logging standard database to determine whether the predicted data is abnormal data;

[0026] When the predicted data is consistent with the data in the well logging standard database, the predicted data is determined to be normal data;

[0027] When the predicted data exceeds the data range in the well logging standard database, the predicted data is determined to be abnormal data, and the abnormal data is marked or updated.

[0028] Further, marking or updating the abnormal data includes:

[0029] Compare the abnormal data with the data in the well logging abnormality database:

[0030] If the abnormal data or data of the same type as the abnormal data exists in the well logging abnormality database, marking the data;

[0031] If the abnormal data or data of the same type as the abnormal data does not exist in the well logging abnormality database, the abnormal data and its data type are added to the well logging abnormality database for updating.

[0032] Based on the same inventive concept, the present invention also provides a digital quality control system based on induction logging equipment, the system comprising:

[0033] A monitoring unit, used to monitor the status of the induction logging equipment and obtain actual induction logging data when the induction logging equipment operates normally;

[0034] A fusion unit, used for processing the actual induction logging data in a standard data format to achieve multi-source heterogeneous data fusion of the actual induction logging data, and obtain a logging standard database and a logging anomaly database;

[0035] A prediction unit is used to establish and train an induction logging prediction model based on a well logging standard database, and to perform prediction using the induction logging prediction model to obtain prediction data of the induction logging;

[0036] The identification unit is used to identify whether the predicted data is abnormal data based on the well logging standard database and the well logging abnormality database, so as to realize the quality control of the induction well logging data.

[0037] Furthermore, the monitoring unit includes a monitor,

[0038] The monitor is provided with a plurality of status bits;

[0039] The monitor is used to monitor the power supply, generator frequency, oil temperature balance, physical scale and circuit status in the induction logging equipment, and output the operating status through the corresponding status bit;

[0040] The operating status of each of the status bits includes: fault, warning and normal;

[0041] If the operating status of each status bit is normal or warning, it is determined that the induction logging equipment operates normally, and actual induction logging data detected by the induction logging equipment is obtained.

[0042] Based on the same inventive concept, an embodiment of the present invention further provides an electronic device, comprising: a memory and a processor; the processor is used to read and execute a computer program stored in the memory to implement the aforementioned digital quality control method based on induction logging equipment.

[0043] Based on the same inventive concept, an embodiment of the present invention further provides a computer storage medium, wherein the computer storage medium stores computer executable instructions, and when the computer executable instructions are executed, the aforementioned digital quality control method based on induction logging equipment is implemented.

[0044] Technical effects and advantages of the present invention: The present invention effectively utilizes the long short-term memory (LSTM) machine learning algorithm and the 3σ criterion to perform status monitoring, induction logging data prediction and abnormal data monitoring of induction logging equipment, thereby realizing digital quality control of induction logging equipment. Compared with traditional processing methods, the digital quality control method is more intelligent, convenient, accurate and fast in effective control, and can reduce labor and reduce costs.

[0045] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures pointed out in the description, claims and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0047] Figure 1 This is a flowchart of a method for digital quality control of induction logging equipment according to an embodiment of the present invention;

[0048] Figure 2 A flowchart of monitoring the status of induction logging equipment in an embodiment of the present invention;

[0049] Figure 3 This is a flow chart of multi-source heterogeneous data fusion in an embodiment of the present invention;

[0050] Figure 4 A flowchart of induction logging data prediction based on a machine learning algorithm in an embodiment of the present invention;

[0051] Figure 5 A flowchart of identifying abnormal induction logging data in an embodiment of the present invention;

[0052] FIG6( a ) is a schematic diagram of a straight line anomaly data type in a three-layer model according to an embodiment of the present invention;

[0053] FIG6( b ) is a schematic diagram of a curve disorder abnormal data type in a three-layer model according to an embodiment of the present invention;

[0054] FIG6( c ) is a schematic diagram of a frequency disorder abnormal data type in a three-layer model according to an embodiment of the present invention;

[0055] FIG. 7( a ) is a schematic diagram of a straight line anomaly data type in a five-layer model according to an embodiment of the present invention;

[0056] FIG7( b ) is a schematic diagram of a curve disorder abnormal data type in a 5-layer model according to an embodiment of the present invention;

[0057] FIG7( c ) is a schematic diagram of a frequency disorder abnormal data type in a 5-layer model according to an embodiment of the present invention;

[0058] Figure 8 This is a structural schematic diagram of a digital quality control system based on induction logging equipment according to an embodiment of the present invention;

[0059] Fig. 9 The figure is a schematic diagram of the structure of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0060] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0061] To address the deficiencies of the prior art, the present invention discloses a digital quality control method based on induction logging equipment. Figure 1 As shown, the following steps are included:

[0062] Step S1: Monitor the status of the induction logging equipment and obtain actual induction logging data when the induction logging equipment operates normally, such as Figure 2 As shown, including:

[0063] The monitor is used to monitor the power supply, generator frequency, oil temperature balance, physical scale and circuit status in the induction logging equipment, and output the operating status through the corresponding status bit;

[0064] The monitor has five status bits, and the operating status of each status bit includes three different states: fault, warning, and normal.

[0065] If the operating status of each status bit is normal or warning, it means that the induction logging equipment operates normally, and the actual induction logging data monitored by the induction logging equipment is obtained;

[0066] If the operating status of any one of the five status bits is fault, it means that the induction logging equipment is operating abnormally, and the induction logging equipment is adjusted or the monitor is replaced to re-measure.

[0067] Step S2: Process the actual induction logging data into a standard data format to achieve multi-source heterogeneous data fusion of the actual induction logging data, and obtain a logging standard database and a logging anomaly database, such as Figure 3 As shown, including:

[0068] Step S201: Processing the actual induction logging data to obtain processed induction logging data, specifically including:

[0069] The logging data processing software LEAD is used to perform raw data preprocessing, borehole correction processing, skin effect correction processing and synthetic focusing processing on the actual induction logging data in sequence: first, the parameters are set through the raw data preprocessing module in the LEAD software, such as selecting the model of the array induction logging equipment and the input data unit; then, the borehole correction module in the LEAD software is used to perform borehole correction processing on the actual induction logging data to eliminate the borehole effect; finally, the true resolution synthesis module in the LEAD software is used to simultaneously eliminate the two-dimensional environmental influence and the skin effect, and 5 different detection depth curves of 3 resolutions (i.e., 5 focused induction logging data) are obtained to realize skin effect correction processing and synthetic focusing processing.

[0070] After data processing, we obtained 14 original induction logging data, 14 induction logging data after wellbore correction, 5 induction logging data after synthetic focusing, and 8 induction logging data including well diameter, mud, temperature, quality control, natural potential, and gamma ray, for a total of 41 processed induction logging data.

[0071] Step S202: Processing the processed induction logging data in a standard data format to achieve multi-source heterogeneous data fusion, and establishing a well logging standard database and a well logging anomaly database with a unified data structure; specifically including:

[0072] Through the measurement of multiple wells, the processed induction logging data measured in different regions are processed in a unified and standardized standard data format to perform multi-source heterogeneous data fusion, and a standard logging standard database and logging anomaly database with a unified data structure are established.

[0073] Among them, multi-source heterogeneous data fusion is to establish a database with a unified data structure after data processing based on different data sources.

[0074] Step S3: Establishing and training an induction logging prediction model based on a well logging standard database, and using the induction logging prediction model to perform predictions to obtain prediction data for induction logging; Figure 4 As shown, including:

[0075] A multi-layer simulated formation model is established based on the data in the logging standard database; the data based on the multi-layer simulated formation model are divided in a ratio of 8:2 to obtain training set data and test set data; based on the LSTM neural network, the induction logging prediction model is trained using the training set data to obtain the trained induction logging prediction model; the trained induction logging prediction model is predicted using the test set data to obtain the induction logging prediction data.

[0076] Among them, in this embodiment, a 3-layer and 5-layer simulated formation model is established, and the established multi-layer model data is predicted using the long short-term memory algorithm (LSTM algorithm). In this embodiment, only the data of the 3-layer and 5-layer simulated formation models are predicted.

[0077] In the process of training the induction logging prediction model, the training time of the induction logging prediction model is realized by adjusting the number of iterations according to different sample numbers. In the induction logging prediction model training of this embodiment, the number of iterations is set to 500 times. When the number of iterations reaches 500 times, the neural network training has become stable and the prediction model training is completed.

[0078] The present invention uses a long short-term memory algorithm (LSTM algorithm) to predict the established multi-layer model data, mainly to predict array data. According to the correlation between array data, an array with the best correlation with other array data is found to predict the remaining array data.

[0079] Step S4: Based on the logging standard database and the logging anomaly database, identify whether the predicted data is abnormal data to achieve quality control of induction logging data; Figure 5 As shown, specifically including:

[0080] Using the three-times standard deviation criterion (3σ criterion) to discriminate the predicted data of the induction logging from the data in the logging standard database, so as to determine whether the predicted data is abnormal data;

[0081] When the predicted data is consistent with the data in the logging standard database, the predicted data is judged to be normal data, which also indicates that the data in the logging standard database (i.e. the actual induction logging data measured) is accurate;

[0082] When the predicted data exceeds the data range in the logging standard database, the predicted data is determined to be abnormal data, which also indicates that the data in the logging standard database (ie, the actual induction logging data measured) is abnormal, and the abnormal data is marked or updated.

[0083] Among them, marking or updating abnormal data includes:

[0084] Compare the anomaly data with the data in the well logging anomaly database:

[0085] If there is abnormal data or data of the same type as the abnormal data in the well logging anomaly database, the data is marked;

[0086] If the abnormal data or data of the same type as the abnormal data does not exist in the well logging anomaly database, the abnormal data and its data type are added to the well logging anomaly database for updating.

[0087] Therefore, the present invention uses the 3σ criterion to identify abnormal data types and implement quality control of induction logging data.

[0088] Based on the anomaly database, this embodiment only lists three common anomaly data types including: straight line anomaly, curve disorder anomaly, and frequency disorder anomaly. Figure 6(a) is the straight line anomaly type of the 3-layer model, Figure 6(b) is the curve disorder anomaly type of the 3-layer model, Figure 6(c) is the frequency disorder anomaly type of the 3-layer model, Figure 7(a) is the straight line anomaly type of the 5-layer model, Figure 7(b) is the curve disorder anomaly type of the 5-layer model, and Figure 7(c) is the frequency disorder anomaly type of the 5-layer model.

[0089] Based on the same inventive concept, the embodiment of the present invention also provides a digital quality control system based on induction logging equipment, such as Figure 8 As shown, the system comprises:

[0090] A monitoring unit, used to monitor the status of the induction logging equipment and obtain actual induction logging data when the induction logging equipment operates normally;

[0091] A fusion unit, used for processing the actual induction logging data in a standard data format to realize multi-source heterogeneous data fusion of the actual induction logging data, and obtaining a logging standard database and a logging anomaly database;

[0092] A prediction unit is used to establish and train an induction logging prediction model based on a well logging standard database, and to perform prediction using the induction logging prediction model to obtain prediction data of the induction logging;

[0093] The identification unit is used to identify whether the predicted data is abnormal data based on the well logging standard database and the well logging abnormality database, so as to realize the quality control of the induction well logging data.

[0094] In some specific embodiments, the monitoring unit includes a monitor,

[0095] The monitor is provided with a plurality of status bits;

[0096] The monitor is used to monitor the power supply, generator frequency, oil temperature balance, physical scale and circuit status in the induction logging equipment, and output the operating status through the corresponding status bit;

[0097] The operating status of each status bit includes: fault, warning and normal;

[0098] Wherein, if the operation status of each status bit is normal or warning, it is determined that the induction logging equipment operates normally, and the actual induction logging data detected by the induction logging equipment is obtained;

[0099] If any one of the status bits is a fault, it is determined that the induction logging equipment is operating abnormally, and the induction logging equipment is adjusted or the monitor is replaced to re-measure.

[0100] Regarding the system in the above embodiment, the specific manner in which each unit module performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.

[0101] Based on the same inventive concept, an embodiment of the present invention further provides an electronic device, whose structure is as follows: Fig. 9 As shown, it includes: a memory and a processor, wherein the processor is used to read and execute the computer program stored in the memory to implement the aforementioned digital quality control method based on induction logging equipment.

[0102] Based on the same inventive concept, an embodiment of the present invention further provides a computer storage medium, wherein the computer storage medium stores computer executable instructions, and when the computer executable instructions are executed, the aforementioned digital quality control method based on induction logging equipment is implemented.

[0103] The present invention effectively utilizes the long short-term memory (LSTM) machine learning algorithm and the 3σ criterion to perform status monitoring, induction logging data prediction and abnormal data monitoring of induction logging equipment, thereby realizing digital quality control of induction logging equipment. Compared with traditional processing methods, the digital quality control method is more intelligent, convenient, accurate and fast in effective control, and can reduce labor and reduce costs.

[0104] Finally, it should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments or to make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A digital quality control method based on induction logging equipment, characterized in that: include: Monitor the status of the induction logging equipment and obtain actual induction logging data when the induction logging equipment operates normally; Processing the actual induction logging data in a standard data format to achieve multi-source heterogeneous data fusion of the actual induction logging data, and obtaining a logging standard database and a logging anomaly database; Establishing and training an induction logging prediction model based on a well logging standard database, and using the induction logging prediction model to make predictions to obtain induction logging prediction data; Based on the well logging standard database and the well logging anomaly database, it is identified whether the predicted data is abnormal data, thereby achieving quality control of induction logging data.

2. A digital quality control method based on induction logging equipment according to claim 1, characterized in that: Detect the status of the induction logging equipment and obtain the actual induction logging data when the induction logging equipment is operating normally, including: The monitor is used to monitor the power supply, generator frequency, oil temperature balance, physical scale and circuit status in the induction logging equipment, and output the operating status through the corresponding status bit; If the operating status of each status bit is normal or warning, it is determined that the induction logging equipment operates normally, and actual induction logging data monitored by the induction logging equipment is obtained; The operating status of each status bit includes: fault, warning and normal.

3. A digital quality control method based on induction logging equipment according to claim 1 or 2, characterized in that: The actual induction logging data is processed in a standard data format to realize multi-source heterogeneous data fusion of the actual induction logging data, and obtain a logging standard database and a logging anomaly database, including: performing data processing on the actual induction logging data to obtain processed induction logging data; The processed induction logging data is processed through a standard data format to achieve multi-source heterogeneous data fusion, and to establish a well logging standard database and well logging anomaly database with a unified data structure; The data processing includes: raw data preprocessing, borehole correction processing, skin effect correction processing and synthetic focusing processing.

4. The digital quality control method based on induction logging equipment according to claim 1 is characterized in that: Based on the logging standard database, an induction logging prediction model is established and trained, and the induction logging prediction model is used for prediction to obtain the prediction data of induction logging, including: Establish a multi-layer simulated formation model based on the data in the well logging standard database; Based on the data of the multi-layer simulated formation model, the training set data and the test set data are obtained by dividing them in proportion; Based on the LSTM neural network, the induction logging prediction model is trained using the training set data to obtain the trained induction logging prediction model; The test set data is used to predict the trained induction logging prediction model to obtain the prediction data of induction logging.

5. A digital quality control method based on induction logging equipment according to claim 1 or 4, characterized in that: Based on the well logging standard database and the well logging anomaly database, identifying whether the predicted data is abnormal data, and realizing the quality control of induction logging data, including: Using the three-times standard deviation criterion, the predicted data of the induction logging is discriminated from the data in the logging standard database to determine whether the predicted data is abnormal data; When the predicted data is consistent with the data in the well logging standard database, the predicted data is determined to be normal data; When the predicted data exceeds the data range in the well logging standard database, the predicted data is determined to be abnormal data, and the abnormal data is marked or updated.

6. A digital quality control method based on induction logging equipment according to claim 5, characterized in that: The abnormal data is marked or updated, including: Compare the abnormal data with the data in the well logging abnormality database: If the abnormal data or data of the same type as the abnormal data exists in the well logging abnormality database, marking the data; If the abnormal data or data of the same type as the abnormal data does not exist in the well logging abnormality database, the abnormal data and its data type are added to the well logging abnormality database for updating.

7. A digital quality control system based on induction logging equipment, characterized in that: The system comprises: A monitoring unit, used to monitor the status of the induction logging equipment and obtain actual induction logging data when the induction logging equipment operates normally; A fusion unit, used for processing the actual induction logging data in a standard data format to achieve multi-source heterogeneous data fusion of the actual induction logging data, and obtain a logging standard database and a logging anomaly database; A prediction unit is used to establish and train an induction logging prediction model based on a well logging standard database, and to perform prediction using the induction logging prediction model to obtain prediction data of the induction logging; The identification unit is used to identify whether the predicted data is abnormal data based on the well logging standard database and the well logging abnormality database, so as to realize the quality control of the induction well logging data.

8. A digital quality control system based on induction logging equipment according to claim 7, characterized in that: The monitoring unit comprises a monitor, The monitor is provided with a plurality of status bits; The monitor is used to monitor the power supply, generator frequency, oil temperature balance, physical scale and circuit status in the induction logging equipment, and output the operating status through the corresponding status bit; The operating status of each of the status bits includes: fault, warning and normal; If the operating status of each status bit is normal or warning, it is determined that the induction logging equipment operates normally, and actual induction logging data detected by the induction logging equipment is obtained.

9. An electronic device, characterized in that: include: Memory, processor; The processor is used to read and execute the computer program stored in the memory to implement the digital quality control method based on induction logging equipment as described in any one of claims 1-6.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, and when the computer-executable instructions are executed, the digital quality control method based on induction logging equipment described in any one of claims 1 to 6 is implemented.

Images