Intelligent dynamic monitoring system for large well factory

Abstract

The invention discloses an intelligent dynamic monitoring system for a large well factory, and belongs to the technical field of oil and gas exploitation, and the system comprises a data management module which constructs a well group feature library containing single well and multi-well associated information based on the historical data of each well; the data acquisition module sets core detection data and dynamic monitoring frequency for each well according to the feature library, and deploys sensors to acquire data and store the data; the data analysis module receives the data, and breaks through the limitation of single-point monitoring through multi-source fusion and multi-well linkage analysis; parameters are normalized to eliminate well type differences, grouping is carried out according to reservoir connectivity, then linkage intimacy values are quantified according to parameter synchronism and response sensitivity indexes, a root is positioned in combination with a feature library and an anomaly rule library, and a system can capture associated anomalies in advance, shorten early warning and positioning time and reduce the risk of production interruption.

Description

Technical Field

[0001] This invention relates to the field of oil and gas extraction technology, specifically to an intelligent dynamic monitoring system for large-scale well factories. Background Technology

[0002] Large-scale well factories, as the core model of intensive oil and gas development, typically consist of dozens to hundreds of cluster wells, with strong coupling relationships between the production systems of each well, formation pressure fields, and fluid migration.

[0003] Existing monitoring methods mostly adopt a fixed single-point monitoring mode, that is, a limited number of monitoring indicators such as pressure and flow rate are preset for a single well, and the monitoring frequency remains constant.

[0004] When production parameters of a well fluctuate, traditional methods can only capture single-point abnormal signals and it is difficult to locate the root cause of the abnormality through multi-well data correlation analysis. This leads to delayed early warning of abnormalities, which are often only detected after the fault has expanded, resulting in serious production losses and increased operation and maintenance costs. Summary of the Invention

[0005] To address the problems of existing technologies, this invention provides an intelligent dynamic monitoring system for large-scale well factories, comprising: The data management module constructs a well group feature library that associates single wells with multiple wells based on the historical data of each well. The data acquisition module sets core detection data for each well based on the well group feature library, sets the detection frequency based on the core detection data, installs sensors on each well based on the detection frequency and the core detection data, collects data, and sends the data to the data management module for storage. The data analysis module is connected to both the data acquisition module and the data management module. The data analysis module is used to receive data sent by the sensors of each well, analyze the received data, determine whether there is any well with anomalies, and adjust the abnormal wells based on the remaining normal wells to bring the abnormal wells back to normal.

[0006] Furthermore, the well cluster feature database includes: single-well information and multi-well association information; The single-well information includes at least: well type, completion method, reservoir type, and production stage; The multi-well correlation information includes: well distance, reservoir connectivity data, and historical parameter correlation information; The reservoir connectivity data refers to whether there is connectivity in the underground storage; The historical parameter association information refers to the parameters that influence each other between the same well type and different well types; The information in the well cluster feature library is used as a reference standard for the data analysis module. When an anomaly occurs in a well, the data analysis module can adjust the parameters of the anomaly well based on the information in the well cluster feature library.

[0007] Furthermore, the core detection data set by the data acquisition module is configured differently according to the well type; the well types include: horizontal wells, vertical wells, and directional wells; The core detection data for the horizontal well include: pressure distribution in the horizontal section, fracture parameters, and tubing vibration frequency. The core monitoring data for the vertical well includes: bottom hole flowing pressure, casing pressure, and instantaneous wellhead production. The core detection data for the directional well includes: the pressure gradient of the inclined section, the wellbore trajectory offset, and the fluid water cut.

[0008] Furthermore, the detection frequency set by the data acquisition module is in dynamic adjustment mode, and the historical data in the data management module is updated in real time based on the data acquisition module.

[0009] Furthermore, the sensor includes: an optical fiber sensor, a high-precision pressure sensor, and a flow sensor, which are respectively connected to the data analysis unit and the data management module; The fiber optic sensor is used to collect data on tubing vibration and temperature distribution along the horizontal section of the well. The pressure sensors are strain gauge sensors and are installed at the wellhead, bottom of the well, and key manifold nodes, respectively. The flow sensor is installed on the wellhead production pipeline.

[0010] Furthermore, the data analysis module's data analysis process includes: A multi-well correlation analysis model is constructed, and the real-time acquired data is compared with the well group feature library in the data management module to obtain the correlation affinity value between the parameters of each individual well and the parameters of adjacent wells. Set a threshold, and when the real-time parameters exceed the threshold or the linkage value, the corresponding well is determined to be an abnormal well; The data analysis module includes an analysis module, which contains a controller. The analysis module is used to retrieve normal well data from the same reservoir unit as the abnormal well from the data management module, calculate the deviation between the abnormal parameters and the normal well parameters, and locate the root cause of the anomaly by combining the reservoir connectivity information.

[0011] Furthermore, the data analysis module adjusts based on the remaining normal wells by generating different control commands for different causes of anomalies. The different control failure commands include at least: pipeline blockage, fracture closure, and ground flow. If the pipeline is blocked, output a command to adjust the throttle valve opening parameter; If the fracture is closed, output an instruction to optimize the fracturing fluid injection volume; If it is a formation flow, output a command to coordinate the adjustment of production parameters of adjacent wells.

[0012] Furthermore, the process of obtaining the correlation coefficient values ​​between the parameters of each individual well and the parameters of adjacent wells includes: The core detection parameters of different types of wells are unified and normalized into general indicators to eliminate the dimensional differences of parameters for different well types; Based on parameter standardization, and according to reservoir connectivity data in the well group feature library, different well types that are in the same underground reservoir unit and have fluid migration and interoperability are divided into a linkage analysis group. Well types in different reservoir units are not included in the same analysis group. Within the linkage analysis group, the linkage affinity value between different well types is quantified through two types of indicators; One type of index is based on the synchronicity of parameter changes, which calculates the time difference and amplitude matching degree of parameter fluctuations in different well types. The higher the synchronicity, the greater the linkage affinity value. The second type is the parameter response sensitivity index, which monitors the change range and response time of corresponding parameters of other well types after the adjustment of parameters of a certain well type. The higher the sensitivity, the greater the linkage affinity value.

[0013] The beneficial effects of this invention are: This application overcomes the limitations of single-point monitoring by integrating multi-source data and conducting multi-well linkage analysis. First, it normalizes parameters of different well types into general indicators and divides linkage analysis groups according to reservoir connectivity, eliminating interference from well type differences and unrelated wells. Then, it quantifies the linkage affinity value using two types of indicators: parameter change synchronicity and response sensitivity. Combined with a well group feature library and anomaly pattern library, it achieves root cause localization from single-well anomaly identification to multi-well linkage verification to reservoir data verification. Compared to traditional methods, it can capture associated anomaly signals in advance, significantly shortening the early warning and root cause localization time and reducing the risk of production interruption. Attached Figure Description

[0014] Figure 1 A schematic block diagram illustrating the overall principle of this invention; Figure 2 A schematic block diagram illustrating the principle of the linkage process between abnormal wells and other wells provided by this invention. Detailed Implementation

[0015] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0016] Please see Figure 1-2 This invention provides an intelligent dynamic monitoring system for large-scale well factories, comprising: The data management module constructs a well group feature library that associates single wells with multiple wells based on the historical data of each well. The data acquisition module sets core detection data for each well based on the well group feature library, sets the detection frequency based on the core detection data, installs sensors on each well based on the detection frequency and the core detection data, collects data, and sends the data to the data management module for storage. The data analysis module is connected to both the data acquisition module and the data management module. The data analysis module is used to receive data sent by the sensors of each well, analyze the received data, determine whether there is any abnormality in the current well, and make adjustments to the abnormal wells based on the remaining wells to bring them back to normal.

[0017] The data management module constructs a well group feature library based on the historical production data and basic attribute data of all individual wells in the large well factory. This library contains basic information on individual well attributes and multi-well associations. The well group feature library is the core data support of the system, and its information is updated synchronously in real time. Based on the well cluster feature library, the data acquisition module determines the core monitoring dimensions of each individual well, establishes a dynamic adjustment model for monitoring frequency with production parameter volatility as input, sets the monitoring frequency corresponding to different volatility ranges, and sets distributed sensing nodes according to the determined monitoring dimensions and frequencies. The sensors collect multi-source data in real time, which is preprocessed by the edge computing node and then sent to the data management module for storage. The data analysis module receives preprocessed real-time data, compares and analyzes it with the well group feature library data to determine the abnormal state of the well group, locates the cause of the abnormality and generates differentiated control instructions. The closed-loop feedback module is connected to the data analysis module and the data management module respectively, and is used to collect production data after abnormal control to support system model iteration and data update.

[0018] In some embodiments, the well cluster feature library includes: single well information and multi-well association information; The single-well information includes at least: well type, completion method, reservoir type, and production stage; The multi-well correlation information includes: well distance, reservoir connectivity data, and historical parameter correlation information; The reservoir connectivity data refers to whether there is connectivity in the underground storage; The historical parameter correlation information refers to the parameters that influence each other between the same well type and different well types; the historical parameter correlation information includes: parameters that influence each other between the same well type (reservoir pressure related parameters, production and fluid property parameters, pipeline network and production equipment parameters), and parameters that influence each other between different well types (injection and production matching related parameters, reservoir connectivity feedback parameters, pipeline network system coupling parameters). The information in the well cluster feature library is used as a reference standard for the data analysis module. When an anomaly occurs in a well, the data analysis module can adjust the parameters of the anomaly well based on the information in the well cluster feature library.

[0019] When an anomaly occurs in a well, the data analysis module can adjust the parameters of the abnormal well based on the single-well attributes, multi-well correlation patterns, and historical case data in the well group feature library.

[0020] In some embodiments, the core detection data set by the data acquisition module is configured differently according to the well type; the well type includes: horizontal well, vertical well, and directional well. The core detection data for the horizontal well include: pressure distribution in the horizontal section, fracture parameters, and tubing vibration frequency. The core monitoring data for the vertical well includes: bottom hole flowing pressure, casing pressure, and instantaneous wellhead production. The core detection data for the directional well includes: the pressure gradient of the inclined section, the wellbore trajectory offset, and the fluid water cut.

[0021] The detection frequency set by the data acquisition module is in dynamic adjustment mode, and the historical data in the data management module is updated in real time based on the data acquisition module.

[0022] Among them, the adjustment rules of the dynamic adjustment mode are based on the real-time adaptation of the fluctuation rate of production parameters.

[0023] Specifically, when the fluctuation rate of production parameters is less than 5%, the monitoring frequency is set to once per hour; When the fluctuation rate of production parameters is between 5% and 15%, the monitoring frequency is set to once every 10 minutes. When the fluctuation rate of production parameters exceeds 15%, the monitoring frequency is set to once per minute; In some embodiments, the sensor includes: an optical fiber sensor, a high-precision pressure sensor, and a flow sensor, which are respectively connected to the data analysis unit and the data management module; The fiber optic sensor is used to collect data on tubing vibration and temperature distribution along the horizontal section of the well. The pressure sensors are strain gauge sensors and are installed at the wellhead, bottom of the well, and key manifold nodes, respectively. The flow sensor is installed on the wellhead production pipeline.

[0024] Among them, fiber optic sensors are deployed along the horizontal section of the horizontal well to collect data on tubing vibration and temperature distribution; The pressure sensors are strain gauge sensors, which are installed at the wellhead, bottom of the well and key manifold nodes to collect pressure parameters at each key location. The flow sensor is installed in the wellhead production pipeline to collect wellhead fluid flow data; The raw data collected by each sensor is preprocessed by the edge computing node. The preprocessing includes noise removal and deletion of data that deviates significantly from the normal range. The processed data is then transmitted to the data management module.

[0025] In some embodiments, the data analysis module's data analysis process includes: A multi-well correlation analysis model is constructed, and the real-time acquired data is compared with the well group feature library in the data management module to obtain the correlation affinity value between the parameters of each individual well and the parameters of adjacent wells. Set a threshold, and when the real-time parameters exceed the threshold or the linkage value, the corresponding well is determined to be an abnormal well; The data analysis module includes an analysis module, which contains a controller. The analysis module is used to retrieve normal well data from the same reservoir unit as the abnormal well from the data management module, calculate the deviation between the abnormal parameters and the normal well parameters, and locate the root cause of the anomaly by combining the reservoir connectivity information. The data analysis module adjusts based on the remaining normal wells by generating different control commands for different causes of anomalies. The different control failure commands include at least: pipeline blockage, fracture closure, and ground flow. If the pipeline is blocked, output a command to adjust the throttle valve opening parameter; If the fracture is closed, output an instruction to optimize the fracturing fluid injection volume; If it is a formation flow, output a command to coordinate the adjustment of production parameters of adjacent wells.

[0026] The analysis process also includes: feature matrix construction; A multi-source data fusion model is constructed by inputting preprocessed data and well group feature library data. The parameters of different well types are unified into standardized indicators, such as the average pressure of horizontal wells and the pressure of deviated wells is converted into the equivalent pressure of vertical wells. Then, the well types are grouped according to the same underground reservoir, and the variation range of single well parameters within the group and the degree of synchronization of parameters between wells are calculated to generate a multi-well coupling correlation feature matrix. Anomaly detection and root cause location: By combining a preset threshold set of anomalies, real-time parameters and multi-well linkage values ​​are compared with the thresholds to identify abnormal wells; data from normal wells in the same reservoir unit are retrieved to calculate the deviation value of abnormal wells, and combined with reservoir connectivity information, the root cause of the anomaly is located. The root cause of the anomaly includes types such as formation crossflow, pipeline blockage, and fracture closure. The identified anomalies are classified and stored as cases according to anomaly type, well type combination, and reservoir type, and the parameter change characteristics and correlation of each well type are recorded.

[0027] In some embodiments, the process of obtaining the correlation coefficient between the parameters of each single well and the parameters of adjacent wells includes: The core detection parameters of different types of wells are unified and normalized into general indicators to eliminate the dimensional differences of parameters for different well types; Based on parameter standardization, and according to reservoir connectivity data in the well group feature library, different well types that are in the same underground reservoir unit and have fluid migration and interoperability are divided into a linkage analysis group. Well types in different reservoir units are not included in the same analysis group. Within the linkage analysis group, the linkage affinity value between different well types is quantified through two types of indicators; One type of index is based on the synchronicity of parameter changes, which calculates the time difference and amplitude matching degree of parameter fluctuations in different well types. The higher the synchronicity, the greater the linkage affinity value. The second type is the parameter response sensitivity index, which monitors the change range and response time of corresponding parameters of other well types after the adjustment of parameters of a certain well type. The higher the sensitivity, the greater the linkage affinity value.

[0028] Among them, the core detection parameters of different well types have natural dimensional differences. For example, the fracture parameter of a horizontal well is in mm, and the bottom hole flowing pressure of a vertical well is in MPa. Directly calculating the distance is meaningless. It is necessary to first eliminate the dimensional differences through normalization and map all parameters to the **[0,1] general index range** to provide a unified benchmark for subsequent distance calculation.

[0029] Normalized distance adaptation logic: Taking the core parameters of a single well as an example, let the original value range of a certain type of parameter be [X_min, X_max] (historical extreme values ​​are extracted from the well group feature library), and the real-time parameter value at a certain moment be X. Then the normalized general index X' is: X'=(X-X_min) / (X_max-X_min) (if X exceeds the historical range, it is forcibly normalized to 0 or 1 according to X_min or X_max). At this point, the parameters of different well types are all converted into unitless universal index values. The absolute distance (|X1'-X2'|) between the same type of parameters of any two wells can be directly calculated as the basis for distance quantification in subsequent linkage analysis.

[0030] Division of Linkage Analysis Groups: Screening Based on Reservoir Connectivity Distance The linkage affinity value is only applicable to well types with fluid migration correlation. It is necessary to first divide the analysis groups according to the reservoir connectivity distance. The smaller the reservoir connectivity distance, the stronger the fluid interconnection between wells, which is a prerequisite for linkage analysis.

[0031] Define reservoir connectivity distance: Based on reservoir connectivity data in the well cluster feature library, the reservoir is regarded as a three-dimensional spatial model. The connectivity distance is the shortest path length of fluid movement between two wells in the same reservoir unit, reflecting the physical accessibility of fluid exchange between wells.

[0032] Set a connectivity distance threshold D0: extract an empirical value from historical production data in the well group feature library (e.g., D0 = 500m). When the reservoir connectivity distance between two wells is ≤ D0, they are considered to have fluid migration interoperability; when it is > D0, they are considered to have no substantial connection and are not included in the same analysis group.

[0033] Within the same reservoir unit, different well types with a connection distance ≤ D0 form a linkage analysis group, and the well types within the group meet the conditions for calculating the linkage intimacy value.

[0034] The intimacy value is determined on the one hand based on the synchronization distance of parameter changes: Synchronicity index focuses on the degree of matching of fluctuations in parameters of different well types under the same time dimension. It inversely maps the strength of synchronicity by calculating the comprehensive distance of time difference distance and amplitude difference distance.

[0035] The parameter change synchronization distance refers to the two-dimensional comprehensive distance between the fluctuation time deviation and fluctuation amplitude deviation of any two wells within the linkage analysis group, such as well A being a horizontal well and well B being a vertical well, during the fluctuation of the core parameters. It is denoted as S_dist.

[0036] Calculation steps: For the normalized parameter sequence, such as the values ​​every minute within an hour, perform trend analysis to identify the peak points of fluctuation; let the peak time of fluctuation of well A be T_A, and the peak amplitude be A_A (the normalized value); let the peak time of fluctuation of well B be T_B, and the peak amplitude be A_B.

[0037] Calculate the distance between individual items; Time difference distance: ΔT_dist=|T_A-T_B| (unit: minutes), reflects the time synchronicity of fluctuations. The smaller ΔT_dist is, the closer the time correlation is.

[0038] Amplitude difference distance: ΔA_dist=|A_A-A_B| (normalized absolute distance), reflects the matching degree of fluctuation amplitude. The smaller ΔA_dist is, the closer the amplitude correlation.

[0039] Calculate the overall synchronization distance S_dist; use weighted Euclidean distance, with weights set based on reservoir connectivity; the stronger the connectivity, the higher the weight of ΔA_dist. S_dist=√[(w_T×ΔT_dist) 2 +(w_A×ΔA_dist) 2 ]; Where w_T+w_A=1 (e.g., when the connection distance is ≤200m, w_A=0.7, w_T=0.3; when the connection distance is 200-500m, w_A=0.5, w_T=0.5).

[0040] In the formula, S_dist is the comprehensive synchronization distance, which represents the weighted Euclidean distance of the "time difference distance" and "amplitude difference distance" of the core parameter fluctuations of different well types (such as horizontal wells and vertical wells) within the linkage analysis group. It is a comprehensive index for quantifying the synchronization of parameter fluctuations. The smaller the value, the stronger the synchronization of parameter fluctuations between wells. w_T is the weighting coefficient for the time difference distance, used to adjust the contribution of "time difference distance (ΔT_dist)" in the comprehensive synchronization distance calculation, satisfying w_T+w_A=1; the weight value is set based on reservoir connectivity (the smaller the reservoir connectivity distance, the smaller the value of w_T, such as w_T=0.3 when the connectivity distance is ≤200m, and w_T=0.5 when it is 200-500m). ΔT_dist is the time difference distance, representing the absolute value of the difference between the "peak time" of the core parameter fluctuations of two wells in the linkage analysis group, in "minutes"; for example, the peak time of fluctuation of well A is T_A, and that of well B is T_B; w_A is the weighting coefficient for amplitude difference distance, used to adjust the contribution of "amplitude difference distance (ΔA_dist)" in the calculation of comprehensive synchronization distance, satisfying w_T+w_A=1; the weight value is positively correlated with reservoir connectivity (the smaller the reservoir connectivity distance, the stronger the connectivity, and the larger the value of w_A, such as w_A=0.7 when the connectivity distance is ≤200m, and w_A=0.5 when it is 200-500m), thereby highlighting the influence of amplitude matching degree on synchronization; ΔA_dist is the amplitude difference distance, representing the absolute value of the difference between the "peak amplitude" of the core parameter fluctuation of two wells in the linkage analysis group (which has been normalized to a general index in the range of [0,1]); for example, the peak amplitude of the fluctuation of well A (after normalization) is A_A, and that of well B is A_B; The correlation with the linkage intimacy value; the intimacy value corresponding to synchronicity is S_score = 1 / (1+S_dist) (normalized to [0,1]). When S_dist→0, that is, when both the time difference and amplitude difference approach 0, S_score→1, representing extremely strong synchronicity; when S_dist increases, S_score approaches 0, representing extremely weak synchronicity.

[0041] The intimacy value is determined on the other hand based on the parameter response sensitivity distance: The sensitivity index focuses on the correlation of parameter responses of different well types under regulatory intervention. By calculating the comprehensive distance of the time distance and amplitude deviation distance of the regulatory response, the sensitivity strength is inversely mapped.

[0042] The parameter response sensitivity distance refers to the two-dimensional combined distance of **response delay distance and response amplitude deviation distance**, denoted as R_dist, when the corresponding parameters of other well types (such as well B) are responded to after parameter adjustment of a certain well type (such as well A) within the linkage analysis group.

[0043] Calculation steps: Implement control and data acquisition; control the core parameters of well A (such as adjusting the fracturing fluid injection volume of horizontal wells) and record the control time as T0; simultaneously acquire the target parameters of well B (such as the instantaneous production at the wellhead of vertical wells), identify the time T1 when the parameters begin to change significantly (the change amplitude is greater than 0.05 after normalization) and the response amplitude A_R after stabilization.

[0044] Calculate the single-item distance; response delay distance: ΔT_Rdist=|T1-T0| (unit: minutes), which reflects the timeliness of the response. The smaller ΔT_Rdist is, the higher the sensitivity.

[0045] Amplitude deviation distance: ΔA_Rdist=|A_R-A_Exp| (A_Exp is the normalized value of the theoretical response amplitude predicted based on historical cases in the well cluster feature library), reflects the matching accuracy of the response. The smaller ΔA_Rdist is, the higher the sensitivity.

[0046] Calculate the overall response distance R_dist; also using weighted Euclidean distance (with higher weighting for response timeliness, w_T1=0.6, w_A1=0.4): R_dist=√[(w_T1×ΔT_Rdist) 2 +(w_A1×ΔA_Rdist) 2 ] Relationship with intimacy level: The sensitivity-corresponding affinity value R_score = 1 / (1+R_dist) (normalized to [0,1]). When R_dist→0 (no response delay, perfect amplitude matching), R_score→1, indicating extremely strong sensitivity of the inter-well parameter response; when R_dist increases, R_score approaches 0, indicating extremely weak sensitivity.

[0047] Final Synthesis of Intimacy Values: The linkage intimacy value is a weighted composite of the synchronicity intimacy value (S_score) and the sensitivity intimacy value (R_score). The weights can be adjusted based on the reservoir type (e.g., synchronicity weight 0.6 for shale reservoirs, sensitivity weight 0.6 for sandstone reservoirs). The final formula is: Linkage Intimacy Value = w_S × S_score + w_R × R_score S_score and R_score are both obtained by inverse distance mapping. They are quantified by three core distances: time, amplitude, and response. This transforms the abstract inter-well linkage relationship into a calculable numerical degree of intimacy. The smaller the distance, the larger the intimacy value, which can directly reflect the parameter correlation strength between different well types.

[0048] The above description is only a preferred 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 within the protection scope of the present invention.

Claims

1. A large-scale intelligent dynamic monitoring system for well factories, characterized in that, include: The data management module constructs a well group feature library that associates single wells with multiple wells based on the historical data of each well. The data acquisition module sets core detection data for each well based on the well group feature library, sets the detection frequency based on the core detection data, installs sensors on each well based on the detection frequency and the core detection data, collects data, and sends the data to the data management module for storage. The data analysis module is connected to both the data acquisition module and the data management module. The data analysis module is used to receive data sent by the sensors of each well, analyze the received data, determine whether there is any well with anomalies, and adjust the abnormal wells based on the remaining normal wells to bring the abnormal wells back to normal.

2. The intelligent dynamic monitoring system for large-scale well factories according to claim 1, characterized in that, The well cluster feature database includes: single well information and multi-well association information; The single-well information includes at least: well type, completion method, reservoir type, and production stage; The multi-well correlation information includes: well distance, reservoir connectivity data, and historical parameter correlation information; The reservoir connectivity data refers to whether there is connectivity in the underground storage; The historical parameter association information refers to the parameters that influence each other between the same well type and different well types; The information in the well cluster feature library is used as a reference standard for the data analysis module. When an anomaly occurs in a well, the data analysis module can adjust the parameters of the anomaly well based on the information in the well cluster feature library.

3. The intelligent dynamic monitoring system for large-scale well factories according to claim 2, characterized in that, The core detection data set by the data acquisition module is configured differently according to the well type. The well types include: horizontal wells, vertical wells, and directional wells; The core detection data for the horizontal well include: pressure distribution in the horizontal section, fracture parameters, and tubing vibration frequency. The core monitoring data for the vertical well includes: bottom hole flowing pressure, casing pressure, and instantaneous wellhead production. The core detection data for the directional well includes: the pressure gradient of the inclined section, the wellbore trajectory offset, and the fluid water cut.

4. The intelligent dynamic monitoring system for large-scale well factories according to claim 1, characterized in that, The detection frequency set by the data acquisition module is in dynamic adjustment mode, and the historical data in the data management module is updated in real time based on the data acquisition module.

5. The intelligent dynamic monitoring system for large-scale well factories according to claim 1, characterized in that, The sensors include: an optical fiber sensor, a high-precision pressure sensor, and a flow sensor, which are respectively connected to the data analysis unit and the data management module; The fiber optic sensor is used to collect data on tubing vibration and temperature distribution along the horizontal section of the well. The pressure sensors are strain gauge sensors and are installed at the wellhead, bottom of the well, and key manifold nodes, respectively. The flow sensor is installed on the wellhead production pipeline.

6. The intelligent dynamic monitoring system for large-scale well factories according to claim 1, characterized in that, The data analysis module performs the following data analysis process: A multi-well correlation analysis model is constructed, and the real-time acquired data is compared with the well group feature library in the data management module to obtain the correlation affinity value between the parameters of each individual well and the parameters of adjacent wells. Set a threshold, and when the real-time parameters exceed the threshold or the linkage value, the corresponding well is determined to be an abnormal well; The data analysis module includes an analysis module, which contains a controller. The analysis module is used to retrieve normal well data from the same reservoir unit as the abnormal well from the data management module, calculate the deviation between the abnormal parameters and the normal well parameters, and locate the root cause of the anomaly by combining the reservoir connectivity information.

7. The intelligent dynamic monitoring system for large-scale well factories according to claim 6, characterized in that, The data analysis module adjusts based on the remaining normal wells by generating different control commands for different causes of anomalies. If the pipeline is blocked, output a command to adjust the throttle valve opening parameter; If the fracture is closed, output an instruction to optimize the fracturing fluid injection volume; If it is a formation flow, output a command to coordinate the adjustment of production parameters of adjacent wells.

8. The intelligent dynamic monitoring system for large-scale well factories according to claim 6, characterized in that, The process of obtaining the correlation coefficient between the parameters of each individual well and the parameters of adjacent wells includes: The core detection parameters of different types of wells are unified and normalized into general indicators to eliminate the dimensional differences of parameters for different well types; Based on parameter standardization, and according to reservoir connectivity data in the well group feature library, different well types that are in the same underground reservoir unit and have fluid migration and interoperability are divided into a linkage analysis group. Well types in different reservoir units are not included in the same analysis group. Within the linkage analysis group, the linkage affinity value between different well types is quantified through two types of indicators; One type of index is based on the synchronicity of parameter changes, which calculates the time difference and amplitude matching degree of parameter fluctuations in different well types. The higher the synchronicity, the greater the linkage affinity value. The second type is the parameter response sensitivity index, which monitors the change range and response time of corresponding parameters of other well types after the adjustment of parameters of a certain well type. The higher the sensitivity, the greater the linkage affinity value.

Images