Wellbore safety intelligent monitoring method based on multi-source data analysis

Abstract

This invention relates to the field of mine monitoring technology, and particularly to an intelligent monitoring method for mine wall safety based on multi-source data analysis. The method obtains the environmental interference intensity based on the amplitude and frequency characteristics of data fluctuations in the mine wall state time series; obtains a low-amplitude consistency index based on the continuous distribution and trend characteristics of low-amplitude data in the mine wall state time series; obtains a low-amplitude coordination index based on the cooperative correlation characteristics between low-amplitude data in the mine wall state time series and low-amplitude data in other mine wall state time series; and obtains the signal reliability based on the low-amplitude consistency index and the low-amplitude coordination index. This invention obtains principal component decision factors and principal components based on the environmental interference intensity, signal reliability, and the eigenvalues ​​of the mine wall state time series in the PCA algorithm; and improves the accuracy of mine wall structural state monitoring by using principal components.

Description

Technical Field

[0001] This invention relates to the field of mine monitoring technology, and in particular to an intelligent monitoring method for well wall safety based on multi-source data analysis. Background Technology

[0002] During the operation of mines, tunnels, and underground engineering projects, the well walls are subjected to complex loads such as surrounding rock stress, construction interference, and groundwater effects over a long period. With time, the well wall structure may gradually develop problems such as localized cracking, significant deformation, or decreased load-bearing capacity. If these issues are not detected and addressed in a timely manner, they can easily lead to a series of safety accidents such as well wall detachment and collapse. Therefore, continuous monitoring of the well wall structure is necessary to understand its condition and identify potential risks in advance. In engineering, displacement, strain, vibration, and acoustic emission sensors are typically deployed at different key locations on the well wall to collect data on overall deformation, local stress, dynamic response, and cracking, thus forming multi-source monitoring data reflecting the well wall structure's condition. Due to the large number of measurement dimensions and sensors deployed, multi-source monitoring data from the same time period often exhibits high-temperature characteristics, and these high-temperature characteristics are correlated and redundant. Direct condition assessment is not only highly complex but also easily affected by redundant information. Therefore, Principal Component Analysis (PCA) is typically used to reduce the dimensionality of the multi-source monitoring data. The resulting principal components are then used as well wall structure condition features and input into a statistical anomaly detection model for safety status assessment.

[0003] Traditional PCA dimensionality reduction methods use maximizing the overall variance of the data as the basis for principal component selection. However, this method has certain limitations in the current scenario. During the long-term service of the wellbore, the evolution of the wellbore structure from a stable normal state to an unstable state usually exhibits obvious stage characteristics. In the early stage of degradation, micro-cracks and minor plastic deformations may occur in local areas of the wellbore. At this time, the overall displacement of the wellbore is small, and the vibration response and strain changes are mostly manifested as low-amplitude, low-energy abnormal fluctuations. As the degradation progresses, local anomalies gradually evolve into macroscopic deformations, which will manifest as larger amplitude changes in monitoring data such as displacement, strain, or vibration. Therefore, early degradation characteristics are usually characterized by short duration, small amplitude, and low energy level, and are easily masked by high-energy components such as mechanical equipment vibration and environmental noise. External environmental interference occupies the main energy distribution of monitoring data for most of the time, thus making the overall structure of multi-source monitoring data show a high-energy background fluctuation superimposed with low-energy degradation characteristics. Traditional PCA tends to prioritize retaining high-energy feature components that contribute significantly to variance during dimensionality reduction, while low-energy feature components related to early wellbore degradation are easily weakened or even discarded. As a result, the principal components after dimensionality reduction mainly reflect the interference characteristics of the external environment and lack sufficient information representation ability for wellbore structural degradation, ultimately affecting the accuracy of wellbore structural condition monitoring.

[0004] Therefore, how to accurately obtain principal component data characterizing the wellbore structure state during the dimensionality reduction process of PCA on multi-source monitoring data has become an urgent problem to be solved. Summary of the Invention

[0005] In view of this, embodiments of the present invention provide a wellbore safety intelligent monitoring method based on multi-source data analysis to solve the problem that existing PCA dimensionality reduction methods can easily affect the accuracy of wellbore structure status monitoring. The specific technical solution adopted is as follows:

[0006] Acquire time series of multiple wellbore statuses monitored by different types of sensors at different locations on the wellbore;

[0007] The environmental disturbance intensity is obtained based on the data fluctuation amplitude and frequency characteristics in the wellbore state time series; the low amplitude consistency index is obtained based on the continuous distribution and trend characteristics of low amplitude data in the wellbore state time series.

[0008] A low-amplitude co-correlation index is obtained based on the co-correlation characteristics between low-amplitude data in the wellbore state time series and low-amplitude data in other wellbore state time series; the signal reliability of the wellbore state time series is obtained based on the low-amplitude consistency index and the low-amplitude co-correlation index.

[0009] Principal component decision factors are obtained based on the environmental interference intensity, the signal reliability, and the eigenvalues ​​of the wellbore state time series in the PCA algorithm; principal components are obtained through the PCA algorithm based on the principal component decision factors of different wellbore state time series; and wellbore structure state is monitored based on the principal components.

[0010] Furthermore, the step of obtaining the environmental disturbance intensity based on the data fluctuation amplitude characteristics and data fluctuation frequency characteristics in the wellbore state time series includes:

[0011] In the formula, R represents the environmental disturbance intensity of the wellbore state time sequence. Indicates normalization. This represents the maximum value of the wellbore condition time series. The minimum value of the wellbore condition time series is represented by T, where T represents the number of data points in the wellbore condition time series. This represents the value of the t-th data point. Indicates the first The value of each data point. Indicates the first The value of each data point. This represents a conditional function, when... When less than the constant 0, The value is a constant 1, otherwise the value is a constant 0.

[0012] Furthermore, the step of obtaining the low-amplitude consistency index based on the continuous distribution characteristics and trend characteristics of the low-amplitude data in the wellbore condition time series includes:

[0013] In the formula, Q represents the low-amplitude consistency index, T represents the number of data points in the wellbore condition time series, and S represents the quantization function. This represents the value of the t-th data point in the wellbore state time series. Indicates the first The values ​​of the data points, where E represents the preset low amplitude threshold. This indicates that the wellbore state time series satisfies the t-th and t-th conditions. The number of data points that are simultaneously less than the preset low amplitude threshold. This represents an exponential function with the natural constant as its base, and H represents the number of data points in the wellbore state time series that are less than a preset low amplitude threshold. This represents the value of the h-th low-amplitude data point. This represents the fitted value of the h-th low-amplitude data point after fitting the low-amplitude data in the wellbore condition time series. The wellbore state time sequence indicates that the following conditions are met. The quantity.

[0014] Further, the step of obtaining the low-amplitude co-index based on the co-correlation characteristics between low-amplitude data in the wellbore state time series and low-amplitude data in other wellbore state time series includes:

[0015] In the formula, W represents the low-amplitude coordination index, and M represents the number of other wellbore state time series. The Pearson correlation coefficient represents the time series of the wellbore state and the time series of the m-th other wellbore state. Y represents the number of data points in the wellbore status time series, and Y represents the number of times when all wellbore status time series are simultaneously less than the preset low amplitude threshold at the same time.

[0016] Further, the step of obtaining the signal reliability of the wellbore state time series based on the low-amplitude consistency index and the low-amplitude coordination index includes:

[0017] The signal reliability of the wellbore state time series is obtained by calculating the average of the low-amplitude consistency index and the low-amplitude coordination index.

[0018] Furthermore, the step of obtaining the principal component decision factor based on the eigenvalues ​​of the environmental interference intensity, the signal reliability, and the wellbore state time series in the PCA algorithm includes:

[0019] In the formula, D represents the principal component decision factor of the wellbore state time series, R represents the environmental disturbance intensity of the wellbore state time series, and A represents the eigenvalue of the wellbore state time series in the PCA algorithm. This represents the maximum eigenvalue of all wellbore state time series in the PCA algorithm. denoted by , represents the minimum eigenvalue of all wellbore state time series in the PCA algorithm, and F represents the signal reliability of the wellbore state time series.

[0020] Furthermore, the step of obtaining principal components using the PCA algorithm based on the principal component decision factors of different wellbore state time series includes:

[0021] The principal components are sorted in descending order based on the principal component decision factors of all wellbore state time series, and the principal components are obtained using the PCA algorithm.

[0022] The beneficial effects of the embodiments of the present invention compared with the prior art are as follows:

[0023] In this invention, the environmental interference intensity is obtained based on the data fluctuation amplitude and frequency characteristics in the wellbore condition time series. This can characterize the impact of factors such as surrounding rock weight, equipment influence, and construction interference on the wellbore structural condition monitoring, thereby initially determining the possibility that the wellbore condition time series is affected by external environmental factors. A low-amplitude consistency index is obtained based on the continuous distribution and trend characteristics of low-amplitude data in the wellbore condition time series. Since early-stage wellbore structural degradation exhibits a continuous low-amplitude distribution, the low-amplitude consistency index can reflect the possibility that the wellbore condition time series represents the actual wellbore structural degradation. A low-amplitude coordination index is obtained based on the collaborative correlation characteristics between low-amplitude data in the wellbore condition time series and low-amplitude data in other wellbore condition time series. This can further determine the possibility that the wellbore condition time series represents the actual wellbore structural degradation based on the collaborative characteristics of low-amplitude data from different types of sensors. Finally, the principal component decision factor is obtained based on the environmental interference intensity, signal reliability, and the eigenvalues ​​of the wellbore condition time series in the PCA algorithm. This avoids the difficulty in directly obtaining the principal components from the eigenvalues ​​in the original PCA algorithm, which struggles to characterize the actual wellbore structural degradation state. This invention obtains principal components using the PCA algorithm based on principal component decision factors of different wellbore state time series, and then monitors the wellbore structure state based on the principal components, thereby improving the accuracy of wellbore structure monitoring. Attached Figure Description

[0024] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 This is a flowchart of a wellbore safety intelligent monitoring method based on multi-source data analysis provided in Embodiment 1 of the present invention. Detailed Implementation

[0026] Embodiments of this disclosure are described in detail below, with examples of these embodiments illustrated in the accompanying drawings. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this disclosure, and should not be construed as limiting it.

[0027] It should be noted that the terms "first," "second," etc., used in this disclosure and the accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this disclosure described herein can be implemented in orders other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this disclosure. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this disclosure.

[0028] To illustrate the technical solution of the present invention, specific embodiments are described below.

[0029] The specific scenario addressed by this invention is as follows: Traditional PCA tends to prioritize retaining high-energy feature components with large variance contributions during dimensionality reduction, while low-energy feature components related to early wellbore degradation are easily weakened or even discarded during dimensionality reduction. This results in the principal components after dimensionality reduction mainly reflecting the interference characteristics of the external environment, lacking sufficient ability to represent information on wellbore structural degradation, ultimately affecting the accuracy of wellbore structural state monitoring. Therefore, this invention improves the method of obtaining principal components in the PCA algorithm by analyzing the wellbore state time series, enabling the principal component data after dimensionality reduction to more accurately represent the wellbore structural state and improve monitoring accuracy.

[0030] This invention provides a method for intelligent monitoring of wellbore safety based on multi-source data analysis, such as... Figure 1 As shown, the method includes the following steps:

[0031] Step S1: Obtain the time sequence of multiple wellbore statuses monitored by different types of sensors at different locations on the wellbore.

[0032] When monitoring the structural condition of a wellbore, various types of sensors are typically used, such as displacement sensors, strain sensors, vibration sensors, and acoustic emission sensors. Displacement sensors can be installed on the inner side of the wellbore or on the surface of the support structure to monitor radial convergence or relative displacement changes. Strain sensors can be installed on the surface of the wellbore support structure (such as lining or steel structure support components) or embedded inside the structural material to monitor the strain response of the wellbore structure under stress. Vibration sensors can be deployed on the surface of the wellbore structure or on the wellbore support components to collect vibration acceleration information generated by the wellbore under external disturbances or structural responses. Acoustic emission sensors can be installed on the surface of the wellbore support structure or the inner lining of the wellbore to capture transient acoustic emission signals released during the initiation and propagation of microcracks inside the material. The implementer can determine the acquisition method according to the implementation scenario. Since different types of sensors have different sampling frequencies, it is necessary to perform time alignment processing on the sequences collected by different types of sensors. Interpolation is used to make different types of data correspond one-to-one on the same time axis. The aligned data is then standardized so that monitoring data of different dimensions can be analyzed together. After preprocessing, each type of sensor corresponds to a wellbore state time series at each monitoring location. All wellbore state time series form multi-source monitoring data for monitoring the wellbore structure state.

[0033] Step S2: Obtain the environmental disturbance intensity based on the data fluctuation amplitude and frequency characteristics in the wellbore status time series; obtain the low amplitude consistency index based on the continuous distribution and trend characteristics of low amplitude data in the wellbore status time series.

[0034] During long-term service, the well wall structure is continuously subjected to the combined effects of various factors such as the weight of the surrounding rock, construction interference, equipment operation, and groundwater infiltration. These factors manifest as high-amplitude, rapidly changing signal fluctuations in the well wall's state time series. For example, when the well wall is subjected to environmental vibration or mechanical equipment interference, local areas of the well wall will undergo minute displacements or stress changes instantaneously with the external force, causing significant fluctuations in the data collected by the sensors within a short period. Simultaneously, since external disturbances are usually continuous and frequent, the signal will exhibit multiple directional changes in time series, resulting in a high fluctuation frequency. Therefore, when the well wall's state time series exhibits high-amplitude and frequent fluctuations, it reflects that external factors dominate the current period, rather than low-amplitude, slow changes caused by minor degradation of the well wall's own state. Thus, the intensity of environmental interference can be obtained based on the amplitude and frequency characteristics of the data fluctuations in the well wall's state time series. Preferably, in this embodiment of the invention, the step of obtaining the environmental interference intensity includes:

[0035]

[0036] In the formula, R represents the environmental disturbance intensity of the wellbore condition time series. Indicates normalization. This represents the maximum value of the wellbore condition time series. The minimum value representing the wellbore condition time series. This reflects the amplitude fluctuation range of the wellbore structure under the influence of high-energy external factors such as environment, construction, and equipment during the specified period. A larger difference indicates stronger high-energy background interference during that period, making early structural degradation signals such as localized crack propagation and minor structural deformation more likely to be masked, thus reducing the visibility of structural degradation information in the wellbore condition time series. T represents the number of data points in the wellbore condition time series. This represents the value of the t-th data point. Indicates the first The value of each data point. Indicates the first The value of each data point. This represents a conditional function, when... When less than the constant 0, The value is a constant 1 otherwise, a constant 0. When the product is less than 0, it means that the data magnitude fluctuated at three adjacent acquisition times. The value of this judgment function reflects the number of times the signal magnitude and direction changed within that time period. The larger the value, the more times the signal magnitude and direction changed, the stronger the high-frequency disturbances experienced by the wellbore status time series, and the more rapidly and frequently the signal exhibits amplitude fluctuation characteristics. Therefore, the greater the intensity of environmental interference, the more likely the wellbore status time series is to represent signal data interfered by external high-energy factors, and the more likely the signal data representing the degradation characteristics of the wellbore structure itself during that time period is to be covered by high-energy background data.

[0037] Furthermore, during the long-term service of the wellbore, the evolution of the wellbore structure from a stable state to an unstable state typically undergoes a slow early stage. During this stage, due to the combined effects of various complex factors such as the weight of the surrounding rock, construction work, and the operation of mechanical equipment, local stress in the rock mass may gradually concentrate, leading to the initiation of micro-cracks that slowly propagate along weak surfaces, or micro-plastic deformation of the material in local areas, resulting in a slow convergence of the overall structure. These physical changes are characterized by small amplitudes and slow changes, exhibiting low-amplitude, slow-changing, and continuously occurring micro-responses in the wellbore state time series corresponding to displacement, strain, vibration, or acoustic emission. Compared to random fluctuations caused by external factors, these responses exhibit significant low-amplitude and temporal continuity characteristics, meaning the wellbore state time series continuously displays low-amplitude characteristics over a long period, thus providing reference signal data for early wellbore structural deterioration. Therefore, a low-amplitude consistency index can be obtained based on the continuous distribution and trend characteristics of low-amplitude data in the wellbore state time series. Preferably, in this embodiment of the invention, the step of obtaining the low-amplitude consistency index includes:

[0038]

[0039] In the formula, Q represents the low-amplitude consistency index, T represents the number of data points in the wellbore condition time series, and S represents the quantization function. This represents the value of the t-th data point in the wellbore state time series. Indicates the first The value of each data point; E represents the preset low amplitude threshold. In this embodiment of the invention, the preset low amplitude threshold is the first quartile obtained by calculating the interquartile range through long-term historical data. The implementer can determine it according to the implementation scenario. This indicates that the wellbore state time series satisfies the t-th and t-th conditions. The number of data points that are simultaneously less than the preset low amplitude threshold. The magnitude of this value can characterize the degree of continuous distribution of low amplitude data in this period. The larger the value, the more likely there are two adjacent moments that are simultaneously less than the preset low amplitude threshold in this period. The stronger the continuity of the distribution of low amplitude data, the more likely the wellbore state time series is to characterize the early slow deterioration of the wellbore structure. This represents an exponential function with the natural constant as its base, and H represents the number of data points in the wellbore state time series that are less than a preset low amplitude threshold. This represents the value of the h-th low-amplitude data point. This represents the fitted value of the h-th low-amplitude data point after fitting the low-amplitude data in the wellbore condition time series. The smaller the value, the closer the low-amplitude data points are to their corresponding fitted values, the more stable the data distribution of low-amplitude data, and the more obvious the stable distribution characteristics along the established data trend direction. This indicates that the wellbore state time series satisfies The quantity. This means that the magnitude and direction of the closely spaced low-amplitude data have changed. The smaller the number of values ​​less than 0, the more continuous the trend of low-amplitude data changes within that period, and the less significant the change in direction. During the long-term operation of the wellbore structure, under the continuous influence of the external environment, the deformation process of the wellbore structure usually has a clear direction of force. Its direction of change usually maintains a high degree of consistency within a local timeframe. For example, when the wellbore gradually undergoes slight convergence under the pressure of the surrounding rock, the displacement data will show a continuous slight increase or decrease trend within a continuous time period; similarly, during the gradual expansion of local microcracks or the slow plastic deformation of the structural material, monitoring data such as strain or acoustic emission will also gradually accumulate along the same direction of change within a certain time range. Therefore, when a wellbore state time series exhibits both low-amplitude continuous changes and a generally consistent trend in direction of change within that period, it is more likely that such changes originate from a slow cumulative deformation process of the wellbore structure under continuous stress. The larger the low-amplitude consistency index, the more likely the wellbore state time series reflects a slow, cumulative process of micro-deformation or crack evolution in the wellbore structure.

[0040] Step S3: Obtain the low-amplitude co-correlation index based on the co-correlation characteristics between low-amplitude data in the wellbore state time series and low-amplitude data in other wellbore state time series; obtain the signal reliability of the wellbore state time series based on the low-amplitude consistency index and the low-amplitude co-correlation index.

[0041] Because the initiation of local fractures in the wellbore simultaneously alters the stress distribution of the surrounding rock mass, minute plastic deformation leads to synchronous micro-amplitude changes in displacement, strain, and vibration signals. This means that different types of sensors at nearby monitoring points will exhibit low-amplitude micro-variation characteristics within the same time period. The time series of different wellbore conditions are characterized by low-amplitude, continuous, and synchronous behavior. This characteristic not only eliminates the possibility of sporadic noise from a single sensor but also indicates that the signal is more likely to characterize the slow, cumulative deterioration of the wellbore structure itself. Therefore, a low-amplitude synergy index can be obtained based on the synergistic correlation between low-amplitude data in the wellbore condition time series and low-amplitude data in other wellbore condition time series. Preferably, in this embodiment of the invention, the step of obtaining the low-amplitude synergy index includes:

[0042]

[0043] In the formula, W represents the low-amplitude coordination index, and M represents the number of other wellbore state time series. The Pearson correlation coefficient represents the correlation between the wellbore state time series and the m-th other wellbore state time series. It should be noted that the Pearson correlation coefficient is existing technology. The larger the absolute value of the Pearson correlation coefficient between two series, the stronger the correlation and the stronger the synergy between the two series. The number of data points representing the wellbore condition time series is denoted by Y, which represents the number of times when all wellbore condition time series simultaneously fall below a preset low amplitude threshold. A larger Y indicates a greater number of times when all wellbore condition time series simultaneously meet the condition of falling below the preset low amplitude threshold. This signifies a stronger synchronization of low amplitude changes between the wellbore condition time series and other wellbore condition time series, and a more pronounced early degradation characteristic. Therefore, a larger low amplitude coordination index indicates stronger coordination between the wellbore condition time series and other wellbore condition time series during that period, and a more synchronous occurrence of low amplitudes; the early degradation signal represented by the wellbore condition time series is more reliable.

[0044] Furthermore, by obtaining the low-amplitude coordination index and low-amplitude consistency index corresponding to different wellbore state time series, the signal reliability of the wellbore state time series can be obtained based on the low-amplitude consistency index and low-amplitude coordination index. Preferably, in this embodiment of the invention, the step of obtaining the signal reliability includes: calculating the average value of the low-amplitude consistency index and the low-amplitude coordination index to obtain the signal reliability of the wellbore state time series. The higher the signal reliability, the better the wellbore state time series can reflect the characteristics of the slow cumulative degradation of the wellbore structure itself, rather than occasional noise or interference from the external environment, and therefore it needs to be given more attention in the subsequent dimensionality reduction process.

[0045] Step S4: Obtain principal component decision factors based on the eigenvalues ​​of environmental interference intensity, signal reliability, and wellbore state time series in the PCA algorithm; obtain principal components using the PCA algorithm based on the principal component decision factors of different wellbore state time series; and monitor the wellbore structure state based on the principal components.

[0046] Because the signal amplitudes corresponding to early degradation phenomena such as the initiation of micro-cracks, subtle plastic deformation, or slow convergence in the wellbore structure are low and easily affected by external factors, being covered by high-energy background signals, the dimensionality reduction results are difficult to represent the true state of the wellbore structure. Therefore, the principal component decision factor can be obtained based on the intensity of environmental interference, the reliability of the signal, and the eigenvalues ​​of the wellbore state time series in the PCA algorithm; preferably, in this embodiment of the invention, the step of obtaining the principal component decision factor includes:

[0047]

[0048] In the formula, D represents the principal component decision factor of the wellbore state time series, and R represents the environmental disturbance intensity of the wellbore state time series. A larger R indicates that the wellbore state time series is more likely to characterize external environmental disturbance factors, and even if the variance of the wellbore state time series is large, it should not be prioritized in the dimensionality reduction process. A represents the eigenvalue of the wellbore state time series in the PCA algorithm. This represents the maximum eigenvalue of all wellbore state time series in the PCA algorithm. denoted by , represents the minimum eigenvalue of all wellbore state time series in the PCA algorithm; F represents the signal reliability of the wellbore state time series. The higher the signal reliability, the more it is necessary to provide its contribution in the acquisition of principal components. This refers to the normalization of eigenvalues. It's important to note that the acquisition of these eigenvalues ​​is based on existing technology. A larger eigenvalue results in a higher priority ranking during principal component calculation, meaning the corresponding eigenvector is considered more preferentially. Furthermore, if the environmental disturbance intensity of the wellbore condition time series is greater, the corresponding eigenvalue needs to be reduced to avoid the principal components representing external interference rather than the wellbore structure's own degradation characteristics. Therefore, by calculating the principal component decision factor, the method for acquiring eigenvalues ​​in the original PCA can be improved, allowing the eigenvalue ranking to be based on the specific scenario, thus ensuring that the acquired principal components more accurately represent the wellbore structure's own degradation characteristics. A larger principal component decision factor results in a greater contribution during principal component calculation.

[0049] Furthermore, principal components can be obtained using the PCA algorithm based on the principal component decision factors of different wellbore state time series. The principal component decision factors of all wellbore state time series are then sorted from largest to smallest and obtained using the PCA algorithm. It should be noted that the PCA algorithm is existing technology, and its specific steps will not be elaborated further. By improving the eigenvalues, the accuracy of principal component acquisition can be enhanced, making the principal components more representative of the true early degradation characteristics of the wellbore structure. Finally, wellbore structure condition monitoring is performed based on the principal components. The principal components can be input as wellbore structure condition features into a statistical anomaly detection model for safety status discrimination. Principal components can fully preserve the early degradation characteristics of low-amplitude, subtle changes, while effectively suppressing the interference of high-amplitude noise such as construction disturbances, equipment operation, and environmental vibrations on the analysis results. This ensures sensitive capture of subtle degradation signals of the wellbore structure and improves monitoring accuracy. Implementers can determine the principal component-based monitoring method according to the implementation scenario; no limitations are imposed here.

[0050] In summary, this invention provides an intelligent wellbore safety monitoring method based on multi-source data analysis. It obtains the environmental interference intensity based on the data fluctuation amplitude and frequency characteristics in the wellbore state time series; obtains a low-amplitude consistency index based on the continuous distribution and trend characteristics of low-amplitude data in the wellbore state time series; obtains a low-amplitude coordination index based on the cooperative correlation characteristics between low-amplitude data in the wellbore state time series and low-amplitude data in other wellbore state time series; and obtains the signal reliability based on the low-amplitude consistency index and the low-amplitude coordination index. This invention obtains principal component decision factors and principal components based on the environmental interference intensity, signal reliability, and the eigenvalues ​​of the wellbore state time series in the PCA algorithm; and improves the accuracy of wellbore structural state monitoring by using principal components.

[0051] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A wellbore safety intelligent monitoring method based on multi-source data analysis, characterized in that, The intelligent wellbore safety monitoring method based on multi-source data analysis includes: Acquire time series of multiple wellbore statuses monitored by different types of sensors at different locations on the wellbore; The environmental disturbance intensity is obtained based on the data fluctuation amplitude and frequency characteristics in the wellbore state time series; the low amplitude consistency index is obtained based on the continuous distribution and trend characteristics of low amplitude data in the wellbore state time series. A low-amplitude co-correlation index is obtained based on the co-correlation characteristics between low-amplitude data in the wellbore state time series and low-amplitude data in other wellbore state time series; the signal reliability of the wellbore state time series is obtained based on the low-amplitude consistency index and the low-amplitude co-correlation index. Principal component decision factors are obtained based on the environmental interference intensity, the signal reliability, and the eigenvalues ​​of the wellbore state time series in the PCA algorithm; principal components are obtained through the PCA algorithm based on the principal component decision factors of different wellbore state time series; and wellbore structure state is monitored based on the principal components.

2. The method of claim 1, wherein, The step of obtaining the environmental disturbance intensity based on the data fluctuation amplitude characteristics and data fluctuation frequency characteristics in the wellbore state time series includes: In the formula, R represents the environmental disturbance intensity of the wellbore state time sequence. Indicates normalization. This represents the maximum value of the wellbore condition time series. The minimum value of the wellbore condition time series is represented by T, where T represents the number of data points in the wellbore condition time series. This represents the value of the t-th data point. Indicates the first The value of each data point. Indicates the first The value of each data point. This represents a conditional function, when... When less than the constant 0, The value is a constant 1, otherwise the value is a constant 0.

3. The intelligent wellbore safety monitoring method based on multi-source data analysis according to claim 1, characterized in that, The step of obtaining the low-amplitude consistency index based on the continuous distribution characteristics and trend characteristics of the low-amplitude data in the wellbore condition time series includes: In the formula, Q represents the low-amplitude consistency index, T represents the number of data points in the wellbore condition time series, and S represents the quantization function. This represents the value of the t-th data point in the wellbore state time series. Indicates the first The values ​​of the data points, where E represents the preset low amplitude threshold. This indicates that the wellbore state time series satisfies the t-th and t-th conditions. The number of data points that are simultaneously less than the preset low amplitude threshold. This represents an exponential function with the natural constant as its base, and H represents the number of data points in the wellbore state time series that are less than a preset low amplitude threshold. This represents the value of the h-th low-amplitude data point. This represents the fitted value of the h-th low-amplitude data point after fitting the low-amplitude data in the wellbore condition time series. The wellbore state time sequence indicates that the following conditions are met. The quantity.

4. The intelligent wellbore safety monitoring method based on multi-source data analysis according to claim 1, characterized in that, The step of obtaining the low-amplitude co-index based on the co-correlation characteristics between low-amplitude data in the wellbore state time series and low-amplitude data in other wellbore state time series includes: In the formula, W represents the low-amplitude coordination index, and M represents the number of other wellbore state time series. The Pearson correlation coefficient represents the time series of the wellbore state and the time series of the m-th other wellbore state. Y represents the number of data points in the wellbore status time series, and Y represents the number of times when all wellbore status time series are simultaneously less than the preset low amplitude threshold at the same time.

5. The intelligent wellbore safety monitoring method based on multi-source data analysis according to claim 1, characterized in that, The step of obtaining the signal reliability of the wellbore state time series based on the low-amplitude consistency index and the low-amplitude coordination index includes: The signal reliability of the wellbore state time series is obtained by calculating the average of the low-amplitude consistency index and the low-amplitude coordination index.

6. The intelligent wellbore safety monitoring method based on multi-source data analysis according to claim 1, characterized in that, The step of obtaining the principal component decision factor based on the eigenvalues ​​of the environmental interference intensity, the signal reliability, and the wellbore state time series in the PCA algorithm includes: In the formula, D represents the principal component decision factor of the wellbore state time series, R represents the environmental disturbance intensity of the wellbore state time series, and A represents the eigenvalue of the wellbore state time series in the PCA algorithm. This represents the maximum eigenvalue of all wellbore state time series in the PCA algorithm. denoted by , represents the minimum eigenvalue of all wellbore state time series in the PCA algorithm, and F represents the signal reliability of the wellbore state time series.

7. The intelligent wellbore safety monitoring method based on multi-source data analysis according to claim 1, characterized in that, The step of obtaining principal components using the PCA algorithm based on principal component decision factors of different wellbore state time series includes: The principal components are sorted in descending order based on the principal component decision factors of all wellbore state time series, and the principal components are obtained using the PCA algorithm.

Images