Visual analysis method for wellbore risk

Abstract

The invention relates to a shaft risk visualization analysis method. The method comprises the following steps: a data acquisition step: acquiring attribute data of a preset type of multi-source data source of a shaft; a preprocessing step: preprocessing the attribute data of various preset types; a risk prediction step: inputting a pre-constructed artificial intelligence model based on the pre-processed attribute data of various preset types, and performing risk prediction; and a real-time visualization step: according to a preset panoramic visualization algorithm, generating a corresponding visual component and a display layout relationship thereof according to the association relationship of the pre-processed attribute data of various preset types, and displaying the visual component and the display layout relationship thereof.

Description

Technical Field

[0001] The invention relates to the technical field of oil and gas production equipment, and in particular to a wellbore risk visualization analysis method. Background Art

[0002] Wellbore is an important equipment in oil and gas production. In production practice, it is necessary to analyze the status of the wellbore during production and conduct risk warning.

[0003] The study found that the wellbore analysis of the existing technology has the following defects:

[0004] (1) Analysis based on a single data source. For example, existing technologies usually only perform risk analysis based on cementing data, ignoring other data sources, resulting in incomplete analysis results.

[0005] (2) Existing analysis methods mainly rely on experience and judgment, rely heavily on expert knowledge, and lack intelligent technology;

[0006] (3) Visual analysis technology is insufficient, mainly manifested in:

[0007] ① The existing technology does not provide visual display of data, and only provides risk identification alarm based on single data;

[0008] ② In some scenarios where visualization is provided, the effectiveness of data analysis (such as the impact of missing data) is not indicative enough;

[0009] ③ The screen display of data in table format makes it difficult to show the relationship between data and fail to make full use of the display space;

[0010] ④Lack of interactive functions with users. Summary of the invention

[0011] The present invention provides a wellbore risk visualization analysis method. Compared with the existing technology, it can obtain wellbore attribute data from multiple sources, process the multi-source data, perform risk analysis calculations based on an artificial intelligence model, and perform visualization based on the processed multi-source data, and provide a human interaction function, which not only meets the needs of risk analysis, but also meets the needs of real-time visualization of data in practice.

[0012] To achieve the above object, the present invention adopts the following technical solutions:

[0013] In a first aspect, the present application provides a wellbore risk visualization analysis method, comprising:

[0014] Data acquisition step: acquiring attribute data of a multi-source data source of a preset type of the wellbore;

[0015] Preprocessing step: preprocessing various preset types of attribute data;

[0016] Risk prediction step: Based on the pre-processed attribute data of various preset types, the pre-built artificial intelligence model is input to perform risk prediction; and

[0017] Real-time visualization step: According to the preset panoramic visualization algorithm, according to the association relationship of various preset types of attribute data after preprocessing, the corresponding visualization components and their display layout relationships are generated and displayed.

[0018] In one implementation, the preset type of multi-source data source includes a sensor and a database;

[0019] The attribute data includes production data, cementing data and logging data.

[0020] In one implementation, the preprocessing includes: data cleaning and denoising, data normalization, missing data filling and inference.

[0021] In one implementation, in the real-time visualization step, a visualization component of a two-dimensional bar chart, a stacked chart, or a scatter chart corresponding to each attribute data is generated according to the result of the attribute data preprocessing; and

[0022] Use different colors, shapes or sizes to represent the numerical results of each attribute data after preprocessing.

[0023] In one implementation, a preset prompt method is used to display missing data filling and inferred data in the preprocessed attribute data.

[0024] In one implementation, the preset prompt method includes a dotted frame, a transparent display, or a gradient color.

[0025] In one implementation, the real-time visualization step further includes: determining the relative layout relationship of each visualization component according to the hierarchical relationship and / or data volume of each attribute data after preprocessing.

[0026] In one implementation, in the real-time visualization step, the method further includes:

[0027] Get the user's filtering operation on the visualization component;

[0028] A comparison or trend analysis is performed on the attribute data corresponding to at least one visualization component selected by the user's filtering operation, and the analysis result is displayed.

[0029] In one implementation, the risk prediction step includes:

[0030] Adopting correlation analysis, a preset number of key attribute data are screened from various preset types of attribute data after preprocessing;

[0031] The features of a preset number of key attribute data are extracted, fused, and input into the artificial intelligence model to predict risk type and probability.

[0032] In one implementation, the artificial intelligence model includes Transformer and XGBOOST.

[0033] The technical solution of the present invention realizes real-time risk analysis, performs visual display, and provides human interaction functions through multi-source data processing and feature fusion technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 is a schematic diagram of a wellbore risk visualization analysis method in one embodiment of the present invention;

[0035] Figure 2 is a schematic diagram of a risk prediction step in one embodiment of the present invention;

[0036] Figure 3 is a schematic diagram of a real-time visualization step in an embodiment of the present invention;

[0037] Figure 4 This is an example diagram of a panoramic visualization of a wellbore in one embodiment of the present invention. DETAILED DESCRIPTION

[0038] In order to make the purpose, technical solution and advantages of the embodiment of the present invention clearer, the technical solution of the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings of the embodiment of the present invention. Obviously, the described embodiment is a part of the embodiment of the present invention, not all of the embodiments. Based on the described embodiment of the present invention, all other embodiments obtained by ordinary technicians in this field belong to the scope of protection of the present invention.

[0039] In view of the defects and problems of existing technologies, such as Figure 1 , the present application provides a wellbore risk visualization analysis method, comprising:

[0040] (1) Data acquisition step: acquiring attribute data of a multi-source data source of a preset type of wellbore;

[0041] Specifically, the preset types of multi-source data sources include sensors and databases.

[0042] Various attribute data of wellbore data, including production data, cementing data and logging data.

[0043] With the advancement of sensing technology, wellbore data involves massive and complex physical, chemical, and mechanical property data, including wellbore size, steel grade, cement type, cementing quality, etc.

[0044] (2) Preprocessing step: preprocessing various preset types of attribute data;

[0045] Specifically, data preprocessing is a critical step in the process of wellbore risk visualization analysis. In order to ensure the accuracy, completeness and availability of data, we proposed a series of preprocessing steps to provide high-quality basic data for subsequent visualization and analysis.

[0046] ①Data cleaning and noise removal:

[0047] Wellbore data obtained from external data sources (such as databases, sensors, or other devices) often contain noise, outliers, or erroneous data. For example, the data collected by the sensor may be disturbed by the environment, resulting in abnormal values. First, these outliers are removed through data cleaning algorithms, and statistical analysis methods (such as median filtering or sliding window method) are used to remove noise. For data points with large errors, they can be eliminated through anomaly detection algorithms (such as Z-Score or IQR method).

[0048] ②Data standardization and normalization:

[0049] Attributes from different data sources may use different units or scales. For example, the unit of borehole size is meters, while the steel grade data may use other standards. In order to make these data uniform, a standardized method is used to adjust the scale of each attribute and convert the attributes of different dimensions into a unified standard range (such as the [0,1] interval) for subsequent visualization analysis and modeling.

[0050] ③ Missing data filling and inference:

[0051] Wellbore data often have missing values, especially when collected by sensors, some data may be lost due to failures, occlusions, etc. When processing missing data, the present invention uses the following techniques:

[0052] Interpolation method: Use the adjacent valid data for linear interpolation or polynomial interpolation to fill the missing segments. This method is suitable for situations where there is less missing data and can ensure a smooth transition of data.

[0053] Data inference: Use machine learning models (such as KNN, regression models, etc.) to predict missing data points. For example, infer missing wellbore dimensions or steel grades through the relationship between existing data points.

[0054] Interval estimation: For wellbore data with obvious hierarchical structure, interval estimation is used to fill in the missing data through context information and data intervals to ensure that the filled data is reasonable and contextually consistent.

[0055] In order to prevent users from misunderstanding the filled data, the present invention uses transparent areas, dotted lines or gradient colors to mark missing areas, so that users can clearly distinguish between valid data and filled data, ensuring that the impact of missing data is reasonably reflected.

[0056] (3) Risk prediction step: Based on the pre-processed attribute data of various preset types, the pre-built artificial intelligence model is input to perform risk prediction;

[0057] Specifically, they include:

[0058] ①Use correlation analysis to filter out a preset number of key attribute data from various preset types of attribute data after preprocessing;

[0059] For example, the key attribute data can be analyzed by using the correlation coefficient method. The correlation coefficient method can be the Pearson coefficient method, and the calculation formula is:

[0060]

[0061] Where X i and Y i are the observed values ​​of variables X and Y, respectively, where and are the average values ​​of variables X and Y, respectively. The Pearson correlation coefficient ranges from -1 to 1. The larger the absolute value of the coefficient, the stronger the linear relationship between the variables.

[0062] ② Extract the features of a preset number of key attribute data, fuse them, and input them into the artificial intelligence model to predict risk type and probability.

[0063] by Figure 2 For example, the attribute data includes production data, cementing data and logging data.

[0064] Risk classification is performed based on personal data and fused based on time features.

[0065] Extract the key features of cementing quality in different well sections, including the number of well sections with good cementing quality, the longest good well section length, the average good well section length, etc.

[0066] For the logging gas measurement data, the well sections with excessively high hydrocarbon gas, carbon dioxide and sulfide gas measurement values ​​are screened out, and the relative position information of the well sections with poor cementing quality is counted. Based on the spatial information, the well sections that are susceptible to invasion by passing oil and gas layers are obtained, thereby locating the positions with higher risks in the wellbore.

[0067] The Transformer model is used to process the time series data of the fused production data. The self-attention mechanism of the Transformer model can capture the long-distance dependencies in the time series data. The XGBOOST gradient boosting decision tree model is used to conduct risk assessment on the features after the cementing data and logging data are fused to identify key risk factors. Combining the prediction results of the Transformer and XGBOOST models, the spatiotemporal weighted method is used to comprehensively evaluate the overall risk level of the wellbore, taking into account the risk contribution under different time and space conditions.

[0068] (4) Real-time visualization step: According to the preset panoramic visualization algorithm, the corresponding visualization components and their display layout relationships are generated and displayed according to the association relationship of various preset types of attribute data after preprocessing.

[0069] like Figure 3 and Figure 4 , illustrating the real-time visualization steps.

[0070] Specifically, this step is visualized in the following way:

[0071] ①Multiple attributes are displayed side by side:

[0072] In order to effectively display multiple attributes of the wellbore structure, the present invention uses a variety of methods such as two-dimensional bar charts, stacked charts, and scatter plots to display data. Each data attribute is encoded by different colors, shapes, and sizes, which helps users intuitively identify the relationship between multiple attributes in the same chart. For example:

[0073] Wellbore size: Indicated by the length of the bar graph.

[0074] Steel grade type: Distinguished by the depth of color and thickness of the bars.

[0075] Cement Type: Different types are distinguished by different marking shapes.

[0076] This multi-dimensional display allows users to quickly understand the different layers, attributes and relationships of the wellbore, thereby improving the efficiency and accuracy of data analysis.

[0077] ② Missing data identification:

[0078] In order to ensure that missing data can be clearly identified, the present invention uses dotted boxes, transparent areas and gradient colors to highlight missing areas. For example:

[0079] Dotted line mark: For parts with missing data, use a dotted line box to mark them to remind users that the data is incomplete.

[0080] Gradient color: Use gradient colors in areas with missing data to indicate the degree of missing data. The darker the gradient color, the greater the impact of the missing data.

[0081] Transparent Areas: Use transparency effects for areas with missing data to further avoid misleading users by filling in data.

[0082] Through these methods, missing data can be presented clearly and intuitively, ensuring that users do not misunderstand the completeness and accuracy of the data.

[0083] ③ Intelligent layout: The present invention adopts an intelligent layout algorithm, which can automatically adjust the layout of the chart according to the properties of different data (such as hierarchical relationship, data volume, etc.), avoiding the crowding of the chart and the confusion of information. For example:

[0084] Areas with high data density: The chart automatically adjusts the proportion of the area to increase the display space of this part.

[0085] Parts with closely related data: The chart adjusts the position of the graphics so that attributes with strong correlation are displayed close together, making it easier for users to observe the relationship between them from the same perspective.

[0086] This layout algorithm not only improves the clarity of data presentation, but also optimizes the space utilization of the chart.

[0087] At the same time, this step also implements interactive functions, which are as follows:

[0088] The interactive function is a highlight of the present invention, which allows users to perform flexible operations during the data visualization process, thereby improving the depth of data exploration and analysis.

[0089] ①Data screening:

[0090] Users can filter and display different data dimensions or attributes through the dynamic selection interface. For example, users can choose to display data for a specific time period, attribute data for a specific wellbore, or data for a specific steel grade. It supports real-time filtering, area zooming, data comparison and other functions, so that users can analyze data more accurately.

[0091] ②Dynamic update:

[0092] When users modify the filter conditions or perspectives, the chart will be updated in real time to ensure the real-time and dynamic nature of data visualization. This means that after each user interaction, the system will automatically adjust the displayed content according to the new filter conditions, so that users can see the latest analysis results.

[0093] ③Interactive annotations:

[0094] Each chart element (such as data points, bars, lines, etc.) is equipped with an interactive annotation function. When the user clicks on an element, the system will pop up the detailed data and analysis information of that part. For example, when clicking on a certain steel grade data, the user can view the specific physical properties, application range, manufacturer and other information of the steel grade, further improving the user's analysis efficiency.

[0095] In summary, the technical solution of this application has the following effects:

[0096] On the one hand, it can take into account multi-source data and conduct comprehensive and accurate risk prediction based on artificial intelligence models to solve problems such as single data source, poor generalization ability, and excessive reliance on expert experience.

[0097] On the other hand, regarding visual display, it has the following highlights:

[0098] (1) Intuitive identification of missing data:

[0099] The present invention introduces transparency gradient, dotted line marking, gradient color and other methods to accurately and intuitively display the missing data area, avoiding the situation where zero value filling or blank areas in traditional methods mislead users. Through color gradient, dotted line marking or transparent area, users can quickly identify which data is missing and take reasonable supplementary measures or further analysis based on the marking. This method not only improves the accuracy of data visualization, but also helps users understand the integrity and credibility of the data, thereby reducing misunderstandings caused by improper handling of missing data.

[0100] (2) Multiple attributes are displayed side by side:

[0101] One of the innovations of this invention is to display multiple wellbore attributes (such as steel grade, wellbore size, cement type, etc.) in the same chart, and each attribute is encoded by different colors, shapes, sizes, etc., so that users can get multi-dimensional information at a glance when viewing the chart. This method of displaying multiple attributes in parallel can save users from frequently switching between different charts and data views, significantly improving the efficiency of data analysis, and is particularly suitable for processing complex data structures. For example, in a chart showing the wellbore structure, multiple attributes such as steel grade, cement type and wellbore size can be presented in the same view, which is convenient for users to conduct overall analysis and comparison.

[0102] (3) Intelligent layout optimization:

[0103] The present invention uses an intelligent layout algorithm based on data features to dynamically adjust the layout of the chart according to the amount of data, hierarchical relationships, and different characteristics of attributes, thereby optimizing the space utilization of the chart. The algorithm can automatically calculate the optimal position of different data blocks, avoiding the situation in traditional charts where the chart space is crowded and the information is not clearly presented due to excessive data or complex attribute hierarchies. Intelligent layout optimization ensures the reasonable distribution of each data point, attribute, and identifier in the chart, maximizes the visualization effect, ensures that the information can be presented to the user clearly and completely, and avoids the drawbacks of data occlusion or chart content stacking due to compact space in traditional methods.

[0104] (4) Interactive Data Exploration:

[0105] A significant innovation of the present invention is that it provides a powerful interactive data exploration function, where users can interact with data through charts, dynamically filter different data dimensions, compare different wellbore attributes, and even perform trend analysis. Users can customize to view specific data intervals, attribute combinations, and even click on any part of the chart to obtain detailed data annotations. This interactive function not only improves the user experience, but also supports more flexible and in-depth data analysis. Through interactive filtering, users can quickly discover potential data trends and relationships, provide support for decision-making, and greatly enhance the practical application value of visualization tools.

[0106] (5) Dynamic updates and real-time feedback:

[0107] Unlike traditional static charts, the present invention supports dynamic updating and real-time feedback of charts. When users modify the filter conditions or view settings, the chart will be refreshed immediately to reflect the latest data display results. This function significantly improves the smoothness and response speed of the interaction between the chart and the user. Dynamic updates ensure that users can view data changes in real time and obtain timely feedback when conducting multi-dimensional analysis, thereby helping users make immediate decisions.

[0108] In the several embodiments provided by the present invention, it should be understood that the disclosed method can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the above units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0109] The above are only preferred embodiments of the present invention and are 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 scope of protection of the present invention.

Claims

1. A wellbore risk visualization analysis method, characterized in that: include: Data acquisition step: acquiring attribute data of a multi-source data source of a preset type of the wellbore; Preprocessing step: preprocessing various preset types of attribute data; Risk prediction step: Based on the pre-processed attribute data of various preset types, the pre-built artificial intelligence model is input to perform risk prediction; as well as Real-time visualization step: According to the preset panoramic visualization algorithm, according to the association relationship of various preset types of attribute data after preprocessing, the corresponding visualization components and their display layout relationships are generated and displayed.

2. The wellbore risk visualization analysis method according to claim 1, characterized in that: The preset type of multi-source data sources includes sensors and databases; The attribute data includes production data, cementing data and logging data.

3. The wellbore risk visualization analysis method according to claim 1, characterized in that: The preprocessing includes: data cleaning and denoising, data normalization, missing data filling and inference.

4. The wellbore risk visualization analysis method according to claim 3, characterized in that: In the real-time visualization step, a visualization component of a two-dimensional bar chart, a stacked chart or a scatter chart corresponding to each attribute data is generated according to the result of the attribute data preprocessing; as well as Use different colors, shapes or sizes to represent the numerical results of each attribute data after preprocessing.

5. The wellbore risk visualization analysis method according to claim 4, characterized in that: The preset prompt method is used to display the missing data filling and inferred data in the preprocessed attribute data.

6. The wellbore risk visualization analysis method according to claim 5, characterized in that: The preset prompting method includes a dotted frame, a transparent display or a gradient color.

7. The wellbore risk visualization analysis method according to claim 4, characterized in that: The real-time visualization step also includes: determining the relative layout relationship of each visualization component according to the hierarchical relationship and / or data volume of each attribute data after preprocessing.

8. The wellbore risk visualization analysis method according to claim 7, characterized in that: In the real-time visualization step, the method further comprises: Get the user's filtering operation on the visualization component; A comparison or trend analysis is performed on the attribute data corresponding to at least one visualization component selected by the user's filtering operation, and the analysis result is displayed.

9. The wellbore risk visualization analysis method according to claim 1, characterized in that: The risk prediction step comprises: Adopt correlation analysis to filter out the key points of preset quantity from the pre-processed attribute data of various preset types. property data; The features of a preset number of key attribute data are extracted, fused, and input into the artificial intelligence model to predict risk type and probability.

10. The wellbore risk visualization analysis method according to claim 1, characterized in that: The artificial intelligence model includes Transformer and XGBOOST.

Images