Well leakage sample category imbalance processing method, system, medium and equipment

Abstract

The invention discloses a well leakage sample category imbalance processing method and system, a medium and equipment. The well leakage sample category imbalance processing method comprises the following steps: a data preprocessing step: preprocessing extracted abnormal data; a sample balance judgment step: judging the number of the normal samples n1 and the number of the well leakage samples n2, and judging whether the number of the normal samples n1 and the number of the well leakage samples n2 are unbalanced; a step of determining a base sample required by sample expansion: screening a minority class of samples in the normal samples and the well leakage samples to form the base sample; a base sample oversampling step: performing oversampling treatment on the screened base sample to obtain a balanced sample; and a sample equalization step: combining the balanced sample with the original well leakage sample to obtain an equalized well leakage sample, and forming a well leakage monitoring sample set by the equalized well leakage sample and the original normal sample, thereby solving the class imbalance problem of well leakage monitoring by performing condition limitation on the boundary selection problem of minority class samples in the algorithm, and improving the accuracy of well leakage monitoring. Therefore, popularization and application of a machine learning algorithm in the field of well leakage monitoring are facilitated.

Description

Technical Field

[0001] The present invention relates to the field of sample imbalance processing, and in particular to a method, system, medium and equipment for processing imbalance of lost circulation sample types. Background Art

[0002] In drilling operations, timely diagnosis of complex problems is crucial to ensure operational safety and efficiency. At present, drilling site staff mainly rely on the long-term experience and expertise of experts in the drilling field to diagnose the complexity of drilling by observing and analyzing ground characterization parameters during the drilling process (such as fluctuations in the big hook load, changes in riser pressure, increases and decreases in outlet displacement, etc.). Drilling experts have formed a relatively fixed judgment process through induction and summary. However, this method is limited by the limitations of human resources. Faced with massive drilling data, it is difficult to conduct a comprehensive and in-depth analysis, and it is impossible to fully explore the complex and subtle internal connection between drilling parameters and drilling conditions. Therefore, the ability and accuracy of diagnosing complex drilling problems based solely on manual experience are obviously limited.

[0003] In recent years, with the rapid development of artificial intelligence technology and data analysis, machine learning algorithms have demonstrated powerful data processing and pattern recognition capabilities in various fields. Applying machine learning algorithms to the analysis of drilling data and establishing intelligent diagnostic methods for complex drilling problems has become a very promising solution. However, in practical applications, the characteristics of drilling data have brought new challenges to machine learning algorithms. In particular, the distribution difference between well leakage data and normal data is very large, and the number of normal data far exceeds that of well leakage data. This uneven distribution of data often leads to the tendency of machine learning algorithm models to mistakenly identify minority samples (such as well leakage samples) as majority samples (such as normal samples) when processing. This misjudgment phenomenon will greatly reduce the accuracy and reliability of the well leakage monitoring system, thereby affecting the safety and efficiency of drilling operations. Summary of the invention

[0004] The technical problem solved by the present invention is to provide a method for processing imbalanced categories of lost circulation samples for improving the accuracy of intelligent lost circulation monitoring.

[0005] The technical solution adopted by the present invention to solve the technical problem is: a method for processing imbalanced well leakage sample categories, comprising the following steps:

[0006] Data preprocessing step: preprocess the extracted abnormal data;

[0007] Sample balance judgment step: judge the number of normal sample n1 and lost circulation sample n2 to determine whether there is an imbalance between the two numbers;

[0008] The steps for determining the base samples required for sample expansion are as follows: screening the minority samples in the normal samples and the lost circulation samples to form the base samples;

[0009] Base sample oversampling step: oversampling is performed on the screened base sample to obtain a balanced sample;

[0010] Sample balancing step: The balanced sample is combined with the original well leakage sample to obtain the balanced well leakage sample. The balanced well leakage sample and the original normal sample together constitute the well leakage monitoring sample set.

[0011] Furthermore, in the data preprocessing step, the extracted abnormal data is preprocessed, specifically:

[0012] If the abnormal data is an outlier, the Sigma criterion is used to remove the outlier;

[0013] If the abnormal data is a missing value, the missing value is filled using the mean filling method of adjacent values.

[0014] Furthermore, in the sample balance judgment step, the number of normal samples n1 and lost circulation samples n2 is judged to determine whether there is an imbalance problem between the two numbers. Specifically, the ratio r of the number of normal samples n1 and lost circulation samples n2 is calculated. When r>3, it is determined that the sample set has a serious class imbalance problem, and the normal sample n1 is divided into the majority class sample, and the lost circulation sample n2 is divided into the minority class sample.

[0015] Furthermore, in the base sample determination step required for sample expansion, the minority class samples in the normal sample n1 and the lost circulation sample n2 are screened to form base samples, specifically, the k nearest neighbor samples of each minority class sample are calculated, and the number of heterogeneous samples in the k nearest neighbor samples is k',

[0016] when A small number of neighbor samples are heterogeneous samples, which are difficult for the model to identify. These samples are used as base samples.

[0017] When k = k', all neighbor samples are heterogeneous samples and are not processed as noise;

[0018] When all neighbor samples are of the same type, they belong to the category that is easy for the model to identify and no processing is performed.

[0019] Furthermore, in the base sample oversampling step, an oversampling process is performed on the screened base sample to obtain a balanced sample, specifically:

[0020] If there are 4 similar samples among the k nearest neighbors of the basic sample, the basic sample is judged as a safe sample and no processing is performed;

[0021] If all the k nearest neighbors of a basic sample are heterogeneous samples, the basic sample is identified as noise;

[0022] If there are two similar samples among the k nearest neighbors of the basic sample, the basic sample is identified as a boundary sample, and the synthetic sample is calculated between the boundary sample and the similar samples among the k nearest neighbor samples.

[0023] Furthermore, the synthetic sample x is obtained by calculating the boundary sample and the similar samples in the k nearest neighbor samples. new , the specific calculation method is:

[0024] x new =x i +γ(x'-x i ), where γ is a random number between 0 and 1, and x i is a boundary sample, and x' is a similar sample among the k nearest neighbor samples.

[0025] The present invention also discloses a system for processing imbalanced well leakage sample categories, including the above-mentioned method for processing imbalanced well leakage sample categories, including a data preprocessing module, a sample balance judgment module, a base sample determination module required for sample expansion, a base sample oversampling module and a sample balancing module:

[0026] The data preprocessing module is used to preprocess the extracted abnormal data;

[0027] The sample balance judgment module is used to judge the number of normal samples n1 and lost circulation samples n2 to determine whether there is an imbalance between the two numbers;

[0028] The base sample determination module required for sample expansion is used to screen minority samples in normal samples and lost circulation samples to form base samples;

[0029] The base sample oversampling module is used to perform oversampling processing on the screened base samples to obtain balanced samples;

[0030] The sample balancing module is used to combine the balanced sample with the original well leakage sample to obtain a balanced well leakage sample, and the balanced well leakage sample and the original normal sample together constitute a well leakage monitoring sample set.

[0031] Furthermore, in the data preprocessing module, the extracted abnormal data is preprocessed, specifically:

[0032] If the abnormal data is an outlier, the Sigma criterion is used to remove the outlier;

[0033] If the abnormal data is a missing value, the missing value is filled using the mean filling method of adjacent values.

[0034] The present invention also discloses a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the method for processing imbalanced well leakage sample categories described above are implemented.

[0035] The present invention also discloses a computer device, comprising a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other via the communication bus; wherein:

[0036] The memory is used to store computer programs;

[0037] The processor is used to execute the steps of the above-mentioned method for processing imbalance of lost circulation sample categories by running the program stored in the memory.

[0038] The beneficial effect of the present invention is that the method solves the class imbalance problem of well leakage monitoring by conditionally restricting the boundary selection problem of minority class samples existing in the algorithm, thereby being more conducive to the promotion and application of machine learning algorithms in the field of well leakage monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 Schematic diagram of a flow chart of a method for processing imbalanced types of lost circulation samples according to an embodiment of the present invention.

[0040] Figure 2 It is a module framework diagram of a method for processing imbalanced types of lost circulation samples according to an embodiment of the present invention.

[0041] Figure 3 Schematic diagram of the sample balancing algorithm according to an embodiment of the present invention. DETAILED DESCRIPTION

[0042] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0043] like Figure 1 As shown, the embodiment of the present application discloses a method for processing imbalance of lost circulation sample categories, comprising the following steps:

[0044] Data preprocessing step: preprocess the extracted abnormal data;

[0045] Sample balance judgment step: judge the number of normal sample n1 and lost circulation sample n2 to determine whether there is an imbalance between the two numbers;

[0046] The steps for determining the base samples required for sample expansion are as follows: screening the minority samples in the normal samples and the lost circulation samples to form the base samples;

[0047] Base sample oversampling step: oversampling is performed on the screened base sample to obtain a balanced sample;

[0048] Sample balancing step: The balanced sample is combined with the original well leakage sample to obtain the balanced well leakage sample. The balanced well leakage sample and the original normal sample together constitute the well leakage monitoring sample set.

[0049] This method solves the class imbalance problem of well leakage monitoring by imposing conditional restrictions on the boundary selection problem of minority class samples in the algorithm, which is more conducive to the promotion and application of machine learning algorithms in the field of well leakage monitoring.

[0050] Specifically, the original drilling data may contain "dirty data" caused by sensor failure or improper operation of on-site personnel, such as outliers, missing values, etc. Since effective feature information cannot be extracted due to incomplete or abnormal information, this type of data needs to be preprocessed.

[0051] In this embodiment, in the data preprocessing step, the extracted abnormal data is preprocessed, specifically:

[0052] If the abnormal data is an outlier, the Sigma criterion is used to remove the outlier;

[0053] If the abnormal data is a missing value, the missing value is filled using the mean filling method of adjacent values.

[0054] It should be explained that the above-mentioned sigma criterion and the mean filling method of adjacent values ​​are both existing outlier processing methods, and no improvements are made here.

[0055] In this method, the acquired data is first preprocessed to remove dirty data therein, thereby ensuring the validity of the acquired data and thus ensuring the accuracy of subsequent data processing.

[0056] In this embodiment, in the sample balance judgment step, the number of normal samples n1 and well leakage samples n2 is judged to determine whether there is an imbalance problem in the number of the two. Specifically, the ratio r of the number of normal samples n1 and well leakage samples n2 is calculated. When r>3, it is determined that the sample set has a serious class imbalance problem, and the normal sample n1 is divided into the majority class sample, and the well leakage sample n2 is divided into the minority class sample.

[0057] The setting of this step can be used to judge the balance tilt of the samples. When there is a class imbalance problem, the subsequent class balance problem is carried out.

[0058] In this embodiment, in the base sample determination step required for sample expansion, the minority class samples in the normal sample n1 and the lost circulation sample n2 are screened to form base samples, specifically: the k nearest neighbor samples of each minority class sample are calculated, and the number of heterogeneous samples in the k nearest neighbor samples is k',

[0059] when A small number of neighbor samples are heterogeneous samples, which are difficult for the model to identify. These samples are used as base samples.

[0060] When k = k', all neighbor samples are heterogeneous samples and are not processed as noise;

[0061] When all neighbor samples are of the same type, they belong to the category that is easy for the model to identify and no processing is performed.

[0062] Specifically, since the traditional oversampling method uses all minority class samples as base samples, although it can solve the problem of data imbalance to a certain extent, when minority class samples are simply copied to increase their number, the model may overfit these repeated samples. If there are noise or outliers in the minority class samples, copying these samples may amplify these noise or outliers into the data set. Therefore, the above method can solve this problem by screening minority class samples before oversampling and selecting samples that are beneficial to model training as base samples.

[0063] In this embodiment, in the base sample oversampling step, an oversampling process is performed on the screened base sample to obtain a balanced sample, specifically:

[0064] If there are 4 similar samples among the k nearest neighbors of the basic sample, the basic sample is judged as a safe sample and no processing is performed;

[0065] If all the k nearest neighbors of a basic sample are heterogeneous samples, the basic sample is identified as noise;

[0066] If there are two similar samples among the k nearest neighbors of the basic sample, the basic sample is identified as a boundary sample, and the synthetic sample is calculated between the boundary sample and the similar samples among the k nearest neighbor samples;

[0067] The synthetic sample x is obtained by calculating the boundary sample and the similar samples in the k nearest neighbor samples. new , the specific calculation method is: new =x i +γ(x'-x i ), where γ is a random number between 0 and 1, and x i is a boundary sample, and x' is a similar sample among the k nearest neighbor samples.

[0068] Specifically, Figure 3As shown in the figure, there are 4 similar samples among the k nearest neighbors of x1. Therefore, according to the basic sample judgment rule, x1 is judged as a safe sample and no processing is performed; all the k nearest neighbors of x2 are heterogeneous samples, so x2 is judged as noise; there are 2 similar samples among the k nearest neighbors of x3, and x3 is judged as a boundary sample, that is, x new =x3+γ(x'-x3).

[0069] Finally, x1 and x3 are balanced samples. Then the balanced samples are combined with the original lost circulation samples to obtain balanced lost circulation samples. The balanced lost circulation samples and the original normal samples together constitute the lost circulation monitoring sample set.

[0070] In order to prove the superiority of the method of the present invention, the SMOTE algorithm is introduced for comparison with the present algorithm. After the equalization processing algorithm proposed by the present invention, the accuracy of the well leakage monitoring model is 94.45%, which is 4.83% higher than the model before the equalization processing and 1.05% higher than the accuracy of the model trained after the SMOTE algorithm processing.

[0071] The following is a specific embodiment of this method:

[0072] Input: unbalanced sample set D; number of minority class samples n1; number of majority class samples n2; nearest neighbor sample k; mapping relationship session between data and trend features;

[0073] Output: a balanced sample set D' containing n2*2 samples;

[0074] forpacket inD do

[0075] x'={x|session[x]=session[packet]};

[0076] Calculate each sample x of n1 i k nearest neighbor samples, the number of majority class samples k';

[0077] if k = k'then

[0078] xi is a noise sample;

[0079] End;

[0080] else if then;

[0081] x i It is a safe sample;

[0082] End;

[0083] Else;

[0084] x iis a boundary sample;

[0085] End;

[0086] i = n;

[0087] whilei>0do;

[0088] x new =x i +γ(x'-x i );

[0089] Linsert(x new );

[0090] session[L]=session[packet];

[0091] i=i-1;

[0092] End;

[0093] End;

[0094] Return D';

[0095] The present invention also discloses a system for processing imbalanced well leakage sample categories, including the above-mentioned method for processing imbalanced well leakage sample categories, including a data preprocessing module, a sample balance judgment module, a base sample determination module required for sample expansion, a base sample oversampling module and a sample balancing module:

[0096] The data preprocessing module is used to preprocess the extracted abnormal data;

[0097] The sample balance judgment module is used to judge the number of normal samples n1 and lost circulation samples n2 to determine whether there is an imbalance between the two numbers;

[0098] The base sample determination module required for sample expansion is used to screen minority samples in normal samples and lost circulation samples to form base samples;

[0099] The base sample oversampling module is used to perform oversampling processing on the screened base samples to obtain balanced samples;

[0100] The sample balancing module is used to combine the balanced sample with the original well leakage sample to obtain a balanced well leakage sample, and the balanced well leakage sample and the original normal sample together constitute a well leakage monitoring sample set.

[0101] Specifically, in the data preprocessing module, the extracted abnormal data is preprocessed, specifically:

[0102] If the abnormal data is an outlier, the Sigma criterion is used to remove the outlier;

[0103] If the abnormal data is a missing value, the missing value is filled using the mean filling method of adjacent values.

[0104] Specifically, in the sample balance judgment module, the number of normal samples n1 and well leakage samples n2 is judged to determine whether there is an imbalance problem between the two numbers. Specifically, the ratio r of the number of normal samples n1 and well leakage samples n2 is calculated. When r>3, it is determined that the sample set has a serious class imbalance problem, and the normal sample n1 is divided into the majority class sample, and the well leakage sample n2 is divided into the minority class sample.

[0105] Specifically, in the base sample determination module required for sample expansion, the minority class samples in the normal sample n1 and the lost circulation sample n2 are screened to form base samples, specifically: the k nearest neighbor samples of each minority class sample are calculated, and the number of heterogeneous samples in the k nearest neighbor samples is k',

[0106] when A small number of neighbor samples are heterogeneous samples, which are difficult for the model to identify. These samples are used as base samples.

[0107] When k = k', all neighbor samples are heterogeneous samples and are not processed as noise;

[0108] When all neighbor samples are of the same type, they belong to the category that is easy for the model to identify and no processing is performed.

[0109] Specifically, in the base sample oversampling module, oversampling is performed on the screened base samples to obtain balanced samples, specifically:

[0110] If there are 4 similar samples among the k nearest neighbors of the basic sample, the basic sample is judged as a safe sample and no processing is performed;

[0111] If all the k nearest neighbors of a basic sample are heterogeneous samples, the basic sample is identified as noise;

[0112] If there are two similar samples among the k nearest neighbors of the basic sample, the basic sample is identified as a boundary sample, and the synthetic sample is calculated between the boundary sample and the similar samples among the k nearest neighbor samples.

[0113] The synthetic sample x is obtained by calculating the boundary sample and the similar samples in the k nearest neighbor samples. new , the specific calculation method is: new =x i +γ(x'-x i ), where γ is a random number between 0 and 1, and x i is a boundary sample, and x' is a similar sample among the k nearest neighbor samples.

[0114] This method solves the class imbalance problem of well leakage monitoring by imposing conditional restrictions on the boundary selection problem of minority class samples in the algorithm, which is more conducive to the promotion and application of machine learning algorithms in the field of well leakage monitoring.

[0115] The present invention also discloses a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the method for processing imbalanced well leakage sample categories described above are implemented.

[0116] In addition, the computer-readable storage medium of this embodiment may adopt any combination of one or more computer-readable storage media, wherein the computer-readable storage medium includes electrical, optical, electromagnetic, infrared or semiconductor systems, devices or components, or any combination thereof.

[0117] The present invention also discloses a computer device, comprising a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other via the communication bus; wherein:

[0118] The memory is used to store computer programs;

[0119] The processor is used to execute the steps of the above-mentioned method for processing imbalance of lost circulation sample categories by running the program stored in the memory.

[0120] As an embodiment of the present invention, the communication bus mentioned in the above terminal can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. The communication bus can be divided into an address bus, a data bus, a control bus, etc.

[0121] As an implementation mode of the present invention, the communication interface is used for communication between the above-mentioned terminal and other devices.

[0122] As an embodiment of the present invention, the memory may include a random access memory (RAM) or a non-volatile memory, such as at least one disk memory. Optionally, the memory may also be at least one storage device located away from the aforementioned processor.

[0123] As an embodiment of the present invention, the above-mentioned processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components.

[0124] Different from the existing technology, this method solves the class imbalance problem of well leakage monitoring by imposing conditional restrictions on the boundary selection problem of minority class samples in the algorithm, which is more conducive to the promotion and application of machine learning algorithms in the field of well leakage monitoring.

[0125] The specific embodiments described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the present invention. 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 method for processing imbalanced well leakage sample categories, characterized in that: The steps include: Data preprocessing step: preprocess the extracted abnormal data; Sample balance judgment step: judge the number of normal sample n1 and lost circulation sample n2 to determine whether there is an imbalance between the two numbers; The steps for determining the base samples required for sample expansion are as follows: screening the minority samples in the normal samples and the lost circulation samples to form the base samples; Base sample oversampling step: oversampling is performed on the screened base sample to obtain a balanced sample; Sample balancing step: The balanced sample is combined with the original well leakage sample to obtain the balanced well leakage sample. The balanced well leakage sample and the original normal sample together constitute the well leakage monitoring sample set.

2. The method for processing imbalanced types of lost circulation samples according to claim 1, characterized in that: In the data preprocessing step, the extracted abnormal data is preprocessed, specifically: If the abnormal data is an outlier, the Sigma criterion is used to remove the outlier; If the abnormal data is a missing value, the missing value is filled using the mean filling method of adjacent values.

3. The method for processing imbalanced types of lost circulation samples according to claim 1, characterized in that: In the sample balance judgment step, the number of normal samples n1 and lost circulation samples n2 is judged to determine whether there is an imbalance problem between the two numbers. Specifically, the ratio r of the number of normal samples n1 and lost circulation samples n2 is calculated. When r>3, it is determined that the sample set has a serious class imbalance problem, and the normal sample n1 is divided into the majority class sample, and the lost circulation sample n2 is divided into the minority class sample.

4. The method for processing imbalanced types of lost circulation samples according to claim 1, characterized in that: In the base sample determination step required for sample expansion, the minority class samples in the normal sample n1 and the lost circulation sample n2 are screened to form base samples, specifically: the k nearest neighbor samples of each minority class sample are calculated, and the number of heterogeneous samples in the k nearest neighbor samples is k', when A small number of neighbor samples are heterogeneous samples, which are difficult for the model to identify. These samples are used as base samples. When k = k', all neighbor samples are heterogeneous samples and are not processed as noise; When all neighbor samples are of the same type, they belong to the category that is easy for the model to identify and no processing is performed.

5. The method for processing imbalanced types of lost circulation samples according to claim 1, characterized in that: In the base sample oversampling step, an oversampling process is performed on the screened base sample to obtain a balanced sample, specifically: If there are 4 similar samples among the k nearest neighbors of the basic sample, the basic sample is judged as a safe sample and no processing is performed; If all the k nearest neighbors of a basic sample are heterogeneous samples, the basic sample is identified as noise; If there are two similar samples among the k nearest neighbors of the basic sample, the basic sample is identified as a boundary sample, and the synthetic sample is calculated between the boundary sample and the similar samples among the k nearest neighbor samples.

6. The method for processing imbalanced types of lost circulation samples according to claim 5, characterized in that: The synthetic sample x is obtained by calculating the boundary sample and the similar samples in the k nearest neighbor samples. new , the specific calculation method is: x new =x i +γ(x'-x i ), where γ is a random number between 0 and 1, and x i is a boundary sample, and x' is a similar sample among the k nearest neighbor samples.

7. A system for processing imbalanced types of lost circulation samples, comprising the method for processing imbalanced types of lost circulation samples according to any one of claims 1 to 6, characterized in that: It includes data preprocessing module, sample balance judgment module, base sample determination module required for sample expansion, base sample oversampling module and sample balancing module: The data preprocessing module is used to preprocess the extracted abnormal data; The sample balance judgment module is used to judge the number of normal samples n1 and lost circulation samples n2 to determine whether there is an imbalance between the two numbers; The base sample determination module required for sample expansion is used to screen minority samples in normal samples and lost circulation samples to form base samples; The base sample oversampling module is used to perform oversampling processing on the screened base samples to obtain balanced samples; The sample balancing module is used to combine the balanced sample with the original well leakage sample to obtain a balanced well leakage sample, and the balanced well leakage sample and the original normal sample together constitute a well leakage monitoring sample set.

8. The system for processing imbalanced types of lost circulation samples according to claim 7, characterized in that: In the data preprocessing module, the extracted abnormal data is preprocessed, specifically: If the abnormal data is an outlier, the Sigma criterion is used to remove the outlier; If the abnormal data is a missing value, the missing value is filled using the mean filling method of adjacent values.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method for processing imbalanced types of lost circulation samples according to any one of claims 1 to 6 are implemented.

10. A computer device, characterized in that: It includes a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other through the communication bus; wherein: The memory is used to store computer programs; The processor is used to execute the steps of the method for processing imbalance of lost circulation sample categories according to any one of claims 1 to 6 by running the program stored in the memory.

Images