Gas field interval opening timing data management method based on a honkong system

Abstract

The application relates to the technical field of oil and gas field data management, and discloses an oil and gas field opening time sequence data management method based on a Hongmeng system. The method realizes real-time acquisition of opening time sequence data through a distributed data acquisition module of the Hongmeng system, calculates the data point density of a target area and compares the data point density with a safety threshold value; when the density exceeds the limit, a statistical analysis model is used to evaluate the interaction of data sources, and a data coupling effect hidden danger state is output; time sequence analysis is used to process environmental factor data, and a potential interference degree evaluation is generated; whether the overall data optimization process is triggered is determined according to the hidden danger state and the interference evaluation; when the overall data optimization process is triggered, individual path complexity evaluation is performed on each data source, an adjustment priority sequence is calculated, oil and gas field equipment is controlled according to the sequence, and dynamic management of opening time sequence data is realized. The method improves the real-time performance, accuracy and flexibility of data management, and adapts to the production requirements of modern oil and gas fields.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of oil and gas field data management, in particular to an oil and gas field interval opening time sequence data management method based on a Hongmeng system. BACKGROUND

[0002] In the development process of oil and gas fields, effective management of opening time sequence data plays a crucial role in production and operation. Traditional oil and gas field data management methods rely on a single centralized data collection and processing system. Such systems often show obvious limitations when faced with widely distributed and complex environmental oil and gas field areas.

[0003] The data collection module of the traditional system has strong independence, and the data of different regions and different equipment cannot be efficiently collected in a distributed manner, resulting in delays in data acquisition and the inability to reflect the dynamic production status of the oil and gas field in real time. Especially in large oil and gas field areas, equipment is distributed and data transmission distance is far, so centralized collection mode is prone to data congestion, loss and other problems, affecting the integrity and timeliness of the data.

[0004] For the collected opening time sequence data, the traditional management method lacks effective analysis and evaluation mechanism. In terms of data density monitoring, fixed threshold judgment is often used, which is difficult to dynamically adjust according to the production characteristics of different regions, and is prone to misjudgment or omission. At the same time, when analyzing the interaction between data sources and the interference of environmental factors on data, the traditional method relies on artificial experience or simple statistical models, which cannot accurately assess the data coupling effect hidden danger and the degree of environmental interference, making it difficult to quickly locate the problem source when data is abnormal.

[0005] When data needs to be optimized and adjusted due to abnormalities, the traditional management method lacks scientific priority sorting basis, often uses a unified adjustment strategy, and cannot differentiate between the importance and impact of different data sources, which not only affects the efficiency of data optimization, but also may cause unnecessary interference to normal production and operation. The existence of these problems makes it difficult for traditional oil and gas field opening time sequence data management methods to meet the needs of modern oil and gas field efficient, accurate and safe production. SUMMARY

[0006] The purpose of the present application is to provide an oil and gas field interval opening time sequence data management method based on a Hongmeng system to solve the problems raised in the background.

[0007] To achieve the above purpose, the present application provides an oil and gas field interval opening time sequence data management method based on a Hongmeng system, which comprises:

[0008] real-time acquisition of opening time sequence data of the oil and gas field through the distributed data collection module of the Hongmeng system;

[0009] Based on the opening time sequence data, the data point density in the target area is calculated, and it is judged whether the data point density exceeds the preset safety threshold;

[0010] When the data point density exceeds the safety threshold, the interaction between the data sources is evaluated using a statistical analysis model, and the data coupling effect hidden danger state is output;

[0011] The time series analysis method is used to process environmental factor data to generate an evaluation of the potential interference degree of environmental factors on the opening time sequence data change;

[0012] According to the data coupling effect hidden danger state and the potential interference degree evaluation, it is determined whether to trigger the overall data optimization process of the target area;

[0013] When the overall data optimization process is triggered, individual path complexity evaluation is performed for each data source, and an adjustment priority sequence is calculated based on the evaluation results;

[0014] According to the adjustment priority sequence, control instructions are executed on oil and gas field equipment to realize dynamic management of opening time sequence data.

[0015] Preferably, the opening time sequence data of the oil and gas field is obtained in real time through a distributed data acquisition module of the Hongmeng system, specifically including:

[0016] Collecting oil well pressure, flow and temperature parameters from multiple sensor nodes, each parameter being accompanied by an accurate time stamp;

[0017] The collected parameters are transmitted to the central processing unit of the Hongmeng system for data format unification and integrity verification;

[0018] Based on the verified data, the ratio of the number of data points in the target area to the area is calculated as the data point density value;

[0019] The data point density value is compared with the preset safety threshold, and if the data point density value is greater than the safety threshold, the target area is marked as a high-density state.

[0020] Preferably, when the data point density exceeds the safety threshold, the interaction between the data sources is evaluated using a statistical analysis model, and the data coupling effect hidden danger state is output, specifically including:

[0021] Extracting opening time sequence data in a high-density state, including oil well pressure, flow and temperature parameters;

[0022] Applying a regression analysis model to model the linear and nonlinear relationships between data sources and calculating an interaction strength indicator;

[0023] According to the interaction strength indicator, the data coupling effect hidden danger state is evaluated as normal or abnormal;

[0024] outputting the data coupling effect risk state to a subsequent environmental factor analysis process.

[0025] Preferably, the environmental factor data is processed using a time series analysis method to generate a potential interference degree assessment of the environmental factors on the time series data changes, specifically including:

[0026] obtaining environmental factor data in the target area, including surface vibration and climate conditions;

[0027] performing autocorrelation analysis on the environmental factor data to identify periodic change patterns and trends;

[0028] quantifying the potential interference degree of the environmental factors on the oil well pressure and flow parameters based on the periodic change patterns and trends;

[0029] transmitting the potential interference degree assessment to an overall optimization judgment process.

[0030] Preferably, based on the data coupling effect risk state and the potential interference degree assessment, it is determined whether to trigger an overall data optimization process for the target area, specifically including:

[0031] receiving the data coupling effect risk state and the potential interference degree assessment;

[0032] if the data coupling effect risk state is normal and the potential interference degree assessment is low interference, it is determined that the overall data optimization process does not need to be triggered;

[0033] otherwise, it is determined that the overall data optimization process is triggered, and an individual path complexity assessment module is started.

[0034] Preferably, when the overall data optimization process is triggered, individual path complexity assessments are performed for each data source, specifically including:

[0035] extracting time series data of each data source, including oil well pressure, flow and temperature parameters;

[0036] calculating data difference and time offset between each data source and other data sources;

[0037] identifying data conflict points, where the data difference exceeds a preset conflict threshold and the time offset is lower than a preset time window;

[0038] generating a path complexity score for each data source based on the number of data conflict points.

[0039] Preferably, an adjustment priority sequence is calculated based on the assessment results, specifically including:

[0040] receiving a path complexity score of each data source;

[0041] applying a weight distribution mechanism to assign a dynamic weight value to each data source, the dynamic weight value being based on a type parameter of the data source;

[0042] multiplying the path complexity score by the dynamic weight value to obtain an adjusted priority value;

[0043] ordering the adjusted priority values from high to low to generate an adjusted priority sequence.

[0044] Preferably, executing control instructions on oil and gas field equipment according to the adjusted priority sequence, specifically including:

[0045] obtaining the adjusted priority sequence;

[0046] extracting corresponding well locations and equipment identifiers for high priority data sources;

[0047] generating equipment control paths using a genetic algorithm to optimize the execution order of the control instructions;

[0048] sending the control instructions to an execution unit of the HOMOGENEUS system to drive the oil and gas field equipment to adjust operating parameters.

[0049] Preferably, during the execution of the control instructions according to the adjusted priority sequence, real-time correction of equipment operation is performed, specifically including:

[0050] monitoring changes in the opening timing sequence data after equipment execution to capture real-time environmental parameters;

[0051] comparing the real-time environmental parameters with historical baseline data to calculate a parameter deviation value;

[0052] if the parameter deviation value exceeds a correction threshold, activating a fuzzy inference mechanism to generate a correction instruction;

[0053] integrating the correction instruction into the control instruction stream to update the equipment operating parameters.

[0054] Preferably, dynamic management of the opening timing sequence data is implemented, specifically including:

[0055] continuously collecting optimized opening timing sequence data and environmental factor data;

[0056] applying a moving average filtering algorithm to process the opening timing sequence data to eliminate noise and extract long-term trends;

[0057] dynamically updating safety thresholds and control parameters based on the long-term trends and real-time feedback;

[0058] The updated parameters are distributed to the distributed data acquisition modules through a data bus of the Hongmeng system to form a closed-loop management cycle.

[0059] Compared with the prior art, the method has the following beneficial effects:

[0060] The method can obtain the opening time sequence data of the oil and gas field in real time with the help of the distributed data acquisition modules of the Hongmeng system, breaks the limitation of the traditional centralized acquisition mode, and makes the data acquisition more timely and complete, so as to timely reflect the dynamic changes of the oil and gas field. By comparing the data point density in the target area with the preset safety threshold, the abnormal situation of data distribution can be found in time, and the basis for subsequent data processing is provided. When the data point density exceeds the safety threshold, the statistical analysis model is used to evaluate the interaction between the data sources and output the data coupling effect hidden danger state, so that the correlation between the data sources can be understood in depth, and potential risks can be identified in advance. The time series analysis method is used to process the environmental factor data to generate the potential interference degree evaluation of the environmental factors on the opening time sequence data change, so that the influence of the environmental factors on the data can be clearly mastered, and the data misjudgment caused by environmental interference can be reduced. According to the data coupling effect hidden danger state and the potential interference degree evaluation, it is determined whether to trigger the overall data optimization process, unnecessary optimization operations can be avoided, and the pertinence and effectiveness of the optimization process are ensured. When the overall data optimization process is triggered, individual path complexity evaluation is performed for each data source, and adjustment priority sequence is calculated based on the evaluation result, so that scientific basis can be provided for data adjustment, and the adjustment operation is more reasonable. According to the adjustment priority sequence, control instructions are executed on the oil and gas field equipment to realize dynamic management of the opening time sequence data, so that the data can always be in a reasonable state and adapt to various changes in the oil and gas field production process.

[0061] The method deeply integrates the distributed advantages of the Hongmeng system and data management, realizes the whole-process cooperation from data acquisition, analysis to optimization adjustment, and improves the intelligent level of the opening time sequence data management of the oil and gas field. In the data acquisition stage, the use of distributed modules enhances the flexibility and coverage of data acquisition, and no matter how scattered the oil and gas field equipment is distributed, real-time data aggregation can be realized. In the data analysis link, the combination of multiple analysis models analyzes from multiple dimensions such as data density, data source interaction and environmental interference, so that the potential problems behind the data cannot escape. In the optimization adjustment stage, based on the priority sequence operation, resources can be reasonably allocated, important data sources can be adjusted in priority, and the stability of key data can be ensured.

[0062] The method can dynamically adjust the management strategy according to the actual production state of the oil and gas field, and when the production environment or equipment state changes, the data management process can also be flexibly changed to adapt to the management needs in different scenes. Through this dynamic management mode, the influence of data anomalies on production can be reduced, and the continuity and stability of oil and gas field production can be maintained. At the same time, the close connection between each link forms a closed-loop management system, so that data can be analyzed in time after collection, and the analysis results can be quickly transformed into specific adjustment measures, improving the overall efficiency of data management. BRIEF DESCRIPTION OF DRAWINGS

[0063] Figure 1 A timing diagram of the oil and gas field opening timing data management method based on the Hongmeng system according to the present application;

[0064] Figure 2 A flowchart for data coupling effect hidden danger state evaluation;

[0065] Figure 3 A flowchart for environment factor potential interference degree evaluation;

[0066] Figure 4 A flowchart for individual path complexity evaluation;

[0067] Figure 5 A flowchart for real-time correction of equipment operation. DETAILED DESCRIPTION

[0068] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0069] Please refer to Figure 1 The present application provides an oil and gas field opening timing data management method based on the Hongmeng system, which comprises:

[0070] Real-time open-time data of the oil and gas field, including well pressure, flow rate, and temperature parameters, is acquired through a distributed data acquisition module. Based on the acquired data, the data point density within the target area is calculated and compared with a preset safety threshold. If the data point density exceeds the safety threshold, a statistical analysis model is used to assess the interaction between data sources, outputting the potential status of data coupling effects. Simultaneously, time series analysis methods are employed to process environmental factor data, generating an assessment of the potential interference level of environmental factors on the open-time data. Combining the potential data coupling effect status and the potential interference assessment, it is determined whether to trigger the overall data optimization process. If the optimization process is triggered, an individual path complexity assessment is performed for each data source, generating an adjustment priority sequence. Control commands are then executed on oil and gas field equipment based on this priority sequence, achieving dynamic management of the open-time data.

[0071] Example 1: See Figure 2 This embodiment details the distributed data acquisition and data coupling effect evaluation process based on the HarmonyOS system. A sensor network is deployed in the target area of ​​the oil and gas field, covering key nodes of the oil wells. Each sensor node is equipped with pressure, flow, and temperature monitoring modules, using piezoelectric pressure sensors, electromagnetic flowmeters, and platinum resistance thermometers to collect real-time parameters. Data acquisition is performed at a fixed frequency, generating a set of raw data records containing timestamps every second. The timestamps are calibrated using the HarmonyOS distributed clock synchronization protocol, achieving millisecond-level accuracy. The raw data is transmitted to the edge computing gateway via the LoRa wireless transmission protocol. The gateway has a built-in HarmonyOS data preprocessing module and implements an automatic retransmission mechanism for packet loss.

[0072] The central processing unit receives multi-source data streams forwarded by the gateway and performs data standardization operations. The standardization process comprises three levels: the first level unifies the timestamp format to Unix millisecond timestamps; the second level converts parameter units (pressure to MPa, flow rate to m³ / h, and temperature to ℃); the third level performs integrity verification, with verification rules including timestamp continuity checks and numerical range threshold filtering (pressure range 0-100MPa, flow rate range 0-5000 m³ / h, and temperature range -20℃-150℃). Missing data triggers a nearest-node compensation mechanism, using a spatial correlation algorithm to generate substitute values. Verified data is stored in a distributed database sharded and indexed by hash number partition.

[0073] The data point density calculation of the target area adopts a dynamic grid division method. The central processing unit loads the GIS geofence coordinates of the target area, which is divided into square grids with a side length of 50 meters. The number of sensors contained in each grid is counted in real time, and the data point density per unit area (points / meter2) is calculated. The density value is updated by a sliding window algorithm, and the window period is set to 5 minutes. The preset safety threshold is dynamically adjusted according to historical operation data, and the initial value is set to 0.08 data points per square meter. When the real-time density value exceeds the threshold, the system automatically generates a high-density state identifier, which includes the grid number, the time of exceeding the limit, and the exceeding limit ratio.

[0074] The high-density state triggers the coupling effect analysis engine. The engine extracts the latest 30-minute open time series data of all data sources in the exceeding grid, and constructs a multi-dimensional parameter matrix. The statistical analysis model adopts a hierarchical regression architecture: in the first stage, linear regression is performed to establish three basic relationship models of pressure-flow, flow-temperature, and pressure-temperature, and to calculate the determination coefficient R2 and the variance inflation factor VIF; in the second stage, nonlinear kernel regression is performed to fit the implicit relationship between parameters through radial basis function, and to calculate the interaction strength index. When the VIF value of the linear model is greater than 5 or the interaction index of the nonlinear model exceeds 0.7, the data coupling is marked as abnormal. The final output is a data coupling effect state code (0: normal / 1: abnormal), and a coupling heat map is generated in the abnormal state to mark the high correlation parameter pairs.

[0075] The data coupling analysis result and the environmental factor analysis module establish a hard real-time interface. The state code and the heat map are transmitted to the environmental assessment process through shared memory, and are written into a distributed message queue backup. To reduce transmission delay, the Hongmeng system enables zero-copy data transmission mechanism to avoid data copying from kernel space to user space. Thus, the closed-loop processing flow from data acquisition to coupling evaluation is completed, and the evaluation result enters the subsequent environmental disturbance analysis stage.

[0076] The time characteristics of the above process are guaranteed by the performance monitoring module. The end-to-end delay from the data acquisition end to the central processing end is controlled within 500 milliseconds, the density calculation period error is less than 50 milliseconds, and the processing time of the coupling analysis engine for 100 data sources is not more than 3 seconds. The system resource scheduling adopts a hybrid strategy: the density calculation task is assigned to the edge node for execution, and the regression analysis task is scheduled to the high-performance computing core of the central processor. All state transition events are recorded in the blockchain audit log to ensure that the operation process is traceable. The central processing unit updates the sensor sampling strategy in real time according to the coupling state, and the sampling frequency of the key parameters is increased to 2 times per second in the abnormal state to achieve higher precision dynamic monitoring.

[0077] Embodiment 2: see Figure 3The embodiment describes in detail the implementation process of environmental factor data processing and overall optimization triggering mechanism. The environmental monitoring network of the target area is independent of the production data acquisition system and includes three types of special sensing devices: the surface vibration monitoring station uses a three-axis acceleration sensor array, the climate monitoring station integrates a thermometer, a hygrometer, a barometer, and an ultrasonic anemometer, and the corrosion monitoring point is equipped with an electrochemical impedance probe. The surface vibration data is collected at a frequency of 4 times per second, covering the amplitude information of the 0.1-100Hz frequency band; the climate parameters are collected every 10 seconds, recording temperature, relative humidity, wind speed and direction data. All environmental sensors have built-in GPS timing modules, with a timestamp synchronization accuracy of ±5 milliseconds.

[0078] The environmental data preprocessing is performed on the edge node of the Hongmeng system. The vibration raw data is first baseline corrected to eliminate the device's own drift error; then a band-pass filter is applied to retain the 0.5-80Hz engineering effective frequency band. The climate data performs outlier rejection, using the sliding quartile range method to identify values deviating from the normal range. The preprocessed environmental data is classified according to the spatial grid, and a mapping relationship is established with the oil and gas data grid in Embodiment 1. Each 50x50m grid is associated with the nearest environmental monitoring point data.

[0079] The time series analysis engine starts the autocorrelation processing flow. The lag analysis is performed on the vibration data time series, with a lag step set to the engineering vibration characteristic period (0.5 seconds to 2 minutes). The autocorrelation coefficient of each lag point is calculated, and when the coefficient peak value appears at the periodic interval point, the main cycle value and confidence are recorded. The trend component of the climate data is extracted by seasonal decomposition, using a moving average window to separate long-term trends and short-term fluctuations, with the window width dynamically adjusted according to the data seasonal period (e.g. temperature data set the window width according to the 24-hour period).

[0080] The potential interference degree quantification adopts a multi-level evaluation strategy. The first level analysis establishes a time shift correlation model between environmental parameters and production parameters: taking the oil well flow data as the reference, the environmental parameter sequence is time shifted (-30 minutes to +30 minutes), and the Pearson correlation coefficient at each time shift position is calculated. The second level analysis constructs an interference intensity matrix, with rows corresponding to environmental parameter types (vibration X / Y / Z axis, temperature and humidity, wind speed) and columns corresponding to affected production parameters (pressure, flow). The matrix element value is the weighted product of the maximum correlation coefficient absolute value and the effective cycle confidence, and the weight coefficient is generated according to the historical accident data training. The final output is the interference degree level code (L1: 0-0.3 low interference; L2: 0.3-0.6 medium interference; L3 >0.6 high interference).

[0081] The trigger decision module receives two-way input: data coupling state code (C0 normal / C1 abnormal) from embodiment 1 and interference level code (L1 / L2 / L3) of this embodiment. The decision logic is implemented by a state machine: when the coupling state is C0 and the interference level is L1, the output is the optimization flag OFF; the remaining 13 combinations all output the optimization flag ON. The state transition conditions are stored in the form of a two-dimensional lookup table, which supports online updating. The decision result is broadcast through the event bus of the Hongmeng system, and at the same time it is written into the optimization decision log.

[0082] The data interface layer realizes cross-module collaboration. The environmental raw data is stored in the time series database, and the compression uses the floating point run-length encoding technology to reduce the storage overhead. The analysis intermediate result is cached in the distributed memory pool, and the survival time is set to 3 analysis periods. The decision result transmission uses a priority message queue, and the optimization trigger instruction is set to the highest priority QoS level, and the end-to-end transmission delay is not more than 200 milliseconds.

[0083] The resource scheduling system implements dynamic allocation of environmental analysis tasks. Under normal circumstances, allocate 2 CPU cores to execute analysis tasks; when the vibration data frequency exceeds the threshold or the weather warning is issued, automatically expand to 4 cores and enable GPU acceleration. The task execution state is monitored in real time, and the single analysis timeout threshold is set to 8 seconds. When the timeout triggers the analysis degradation mechanism (simplify algorithm complexity or narrow analysis time window).

[0084] The feedback control loop connects the device execution layer. After each decision result is generated, the system automatically evaluates the last optimization effect: extract the standard deviation change rate of production parameters before and after the start of this environmental analysis. When the change rate does not reach the expected target, the interference level decision threshold is adjusted adaptively, and the adjustment amplitude follows the PID control law. At the same time, update the sensor configuration strategy: the vibration sampling frequency is increased to 10 times per second in high interference state, and the meteorological data sampling interval is shortened to 5 seconds. The device calibration instruction is issued synchronously, and the vibration sensor performs automatic zero-point calibration every 24 hours, and the barometer of the weather station is checked three times a day.

[0085] The spatio-temporal correlation engine enhances the positioning accuracy. Each environmental data packet is attached with geographical location metadata, and the improved Delaunay triangulation algorithm is used to establish the topological relationship of the monitoring point positions. When a specific grid triggers optimization, the system automatically expands the environmental data of the surrounding 8 adjacent grids. The spatial expansion coefficient is calculated based on the geological structure map: the expansion coefficient in the fault zone area is increased by 40%, and the homogeneous rock layer area maintains the basic expansion mode. A lag influence model is established in the time dimension, and the environmental data within 30 minutes after a strong vibration event is marked with a special processing identifier.

[0086] An emergency response plan is activated in stages. If an earthquake early warning signal (P-wave detection) or a wind speed exceeding the limit is detected during environmental analysis, the regular analysis process is immediately suspended, and the system switches to emergency mode. In emergency mode, a simplified interference assessment report is generated, containing only the status of key parameters exceeding the limit and emergency optimization suggestions. Simultaneously, a data backup mechanism is activated, with all original environmental data stored in duplicate on edge nodes at different physical locations. When the system returns to normal operation, interruption task compensation calculations are automatically performed to ensure the continuity of the analysis cycle.

[0087] This implementation process adapts to fluctuations in the field environment through a multi-layered buffer design. The data input stage incorporates a flow control valve to automatically downsample high-frequency vibration data when communication bandwidth is limited. The processing stage deploys flexible time windows, allowing for scalable analysis duration (reduced from the standard 30 minutes to a minimum of 10 minutes) in extreme weather conditions. The output stage includes a result verification channel, requiring critical judgments to be cross-validated by two independent computing nodes before taking effect. The entire process uses the PTP precision time protocol for clock synchronization, ensuring time deviations at each node are controlled within 1 millisecond, guaranteeing the accuracy of time series analysis.

[0088] Example 3: See Figure 4 This embodiment details the implementation process of individual path complexity assessment and priority sequence generation. After triggering the overall data optimization process, the system initiates a multi-dimensional data source analysis mechanism. The identification information of each data source includes three parts: well location code, equipment type, and grid area, forming a structured index tree. The data extraction module loads the latest open time-series data of the target data source and its associated data sources from the distributed database according to the index. The time window is fixed at 60 minutes before the optimization trigger time, and the data granularity maintains the same second-level sampling rate as the acquisition end.

[0089] The core of path complexity assessment lies in quantifying the degree of conflict between data sources. For any two related data sources... and Define data variability The calculation method is as follows:

[0090]

[0091] in: The number of valid sampling points within the time window. , , Representing data sources respectively In the Normalized values ​​of pressure, flow rate, and temperature at each sampling point (normalization range 0-1). Time offset. Solving this problem using the dynamic time warping algorithm, first construct... and The cumulative distance matrix of the parameter sequence is then backtracked to find the minimum cost path, and the final offset is the average of the absolute values of the time differences of the points on the path.

[0092] The conflict point detection adopts a dual-criterion mechanism. The first criterion requires exceeding the dynamic conflict threshold , which is set based on the 75th percentile of historical operation data of the target area; the second criterion limits less than the time window , with the window value dynamically adjusted according to the response speed of the equipment (5 seconds for fast response equipment and 30 seconds for slow equipment). Time periods that satisfy both conditions are marked as conflict periods, and the total duration of each data source participating in the conflict is calculated as the original complexity score .

[0093] The weight allocation system runs in an independent service container and includes a type parameter parser and a dynamic weight calculator. The type parameters are divided into three levels: the first-level parameter is the basic type of the equipment (production well / injection well / monitoring well), the second-level parameter is the service life segment of the equipment (0-5 years / 5-10 years / over 10 years), and the third-level parameter is the recent maintenance status (maintenance within 7 days / maintenance within 30 days / no maintenance record). The weight calculation adopts a hierarchical accumulation strategy, with the first-level basic weight set to 0.6, the second-level year limit weight increment range ±0.2, and the third-level maintenance weight increment range ±0.1. The final dynamic weight value is generated by the following formula:

[0094]

[0095] where: is the year limit coefficient (new equipment takes +0.2, old equipment takes -0.2), is the maintenance coefficient (recent maintenance takes +0.1, no maintenance takes -0.1), and are binary indicator variables.

[0096] The priority sequence generator receives the pairs of each data source and performs standardization processing. After Z-score standardization, it is converted to to avoid differences in scoring scales in different areas. The priority value is calculated using a weighted product model:

[0097]

[0098] This model ensures that even if the score of a data source is low, it can still obtain appropriate priority improvement when the weight is high. The sequence sorting uses an improved heap sorting algorithm, and the sorting time of 1000 data sources is controlled within 50 milliseconds.

[0099] The system is implemented in microservice architecture, and the path evaluation, weight allocation, and priority sorting modules interact through RESTful API. The data cache uses LRU strategy, and the most recently used associated data sources are retained in memory, with a cache hit rate of over 85%. The fault tolerance mechanism includes three levels of fallback: when the dynamic time warping calculation times out (> 500ms), automatically switch to the fast DTW approximation algorithm; when the weight service is unavailable, enable the latest weight value cached locally; when the sorting service fails, degrade to simple weighted summation sorting.

[0100] The real-time monitoring interface displays a conflict heat map, with gradient colors indicating the comprehensive conflict index of each grid area. The operator can manually adjust two parameters: the conflict threshold 's relaxation coefficient (0.8-1.2 times the baseline value), and the time window 's scaling factor (0.5-2 times the baseline value). The adjustment takes effect immediately and triggers a re-evaluation. All manual operations are recorded in the audit log, including operator ID, adjusted parameters, and timestamp.

[0101] The evaluation results are output in a hierarchical data structure. The top layer is a regional summary, including the grid number that triggered optimization, the total number of involved data sources, and the highest priority value; the middle layer is a data source-level detail, listing each source's 、 、 and conflict details; and the bottom layer is a raw data snapshot, saving the 60-minute time series segment used for evaluation. The data is serialized in ProtocolBuffers format, reducing the transmission bandwidth by 60% compared to JSON.

[0102] The interface design with the device control system takes into account the real-time requirements. The priority sequence is updated every 30 seconds, and the update event is pushed to the control instruction generation module through the message queue. The identification information of high-priority ( top 20% of value) data sources is marked with a red warning, and the control module needs to start the response process within 5 seconds after receiving it. Medium and low-priority data sources are allowed to be queued for processing, but the maximum waiting time does not exceed the optimization period (default 300 seconds).

[0103] The historical data analysis service runs offline regularly, performing two tasks: one is to calculate the historical conflict frequency of each data source and generate a device health report; the other is to analyze the correlation between weight parameters and final optimization results, and automatically calibrate the coefficients in the weight calculation formula. The analysis period is set to once a week, and the MapReduce framework is used to process the full historical data, with a single analysis time of about 2 hours.

[0104] This implementation method ensures processing efficiency through multi-level caching and parallel computing. During the data loading phase, time series data from associated data sources are prefetched, and SSD caching accelerates I / O. In the evaluation phase, a thread pool is used to process data source pairs in parallel, with the number of threads dynamically adjusted based on the number of CPU cores. In the result generation phase, copy-on-write technology is employed to avoid lock contention. The entire process achieves an average processing time of 8 seconds per thousand data sources on a 16-core server, meeting the real-time optimization needs of oil and gas fields.

[0105] Example 4: See Figure 5 This embodiment takes the optimized control process of Block A in an oil and gas field as an example to explain in detail the implementation method of generating and executing equipment control commands. The block includes 12 oil wells (numbered W01-W12), 3 water injection stations (numbered I01-I03), and 8 monitoring points (numbered M01-M08). The priority sequence generation results are shown in Table 1.

[0106] Table 1: Priority sequence generation results.

[0107]

[0108] After reading the data in the table above, the control command generation system first extracts the devices marked as "urgent" and "high" priority. Real-time parameters for well W09 show that its wellhead pressure fluctuation is 1.8 times the standard value, and the flow deviation exceeds the allowable range by 15%. The system calls the genetic algorithm optimization module, initializing the population size to 50 control scheme individuals. Each individual's code contains three types of operation commands: adjusting the pumping unit stroke rate (range 4-12 strokes / minute), adjusting the electric pump frequency (range 30-50Hz), and modifying the water injection ratio (range 0-100%). The fitness function comprehensively considers three objectives: pressure stability, flow deviation recovery speed, and energy consumption change rate.

[0109] After 15 generations of iterative calculations, the optimal control scheme was determined as follows: reducing the stroke rate of the W09 pumping unit from 8 strokes / minute to 6 strokes / minute, and adjusting the electric pump frequency from 42Hz to 38Hz; simultaneously increasing the opening of the No. 3 water injection valve of the associated water injection station I02 from 65% to 72%. The command sequence was transmitted through the secure channel of the HarmonyOS system, using AES-256 encryption and CRC32 verification during transmission. Upon receiving the command, the field RTU controller first verified the digital signature, and then immediately executed the robotic arm positioning and valve adjustment. The W09 pumping unit frequency converter completed parameter updates within 3 seconds, and the water injection valve actuator took 8 seconds to operate.

[0110] The real-time correction system initiates closed-loop monitoring after command execution. Data from the triaxial vibration sensor deployed at the W09 wellhead shows that the pressure fluctuation amplitude is reduced by 57% within 5 minutes after adjustment, but the flow recovery does not meet expectations. The system automatically activates the fuzzy inference engine, input parameters include: current flow difference from target (-12%), pressure drop rate (0.4 MPa / min), and environmental temperature change (+2℃). The fuzzy rule base contains 32 expert experience rules, such as "if the flow negative difference is large and the pressure drop rate is moderate, then fine-tune the electric pump frequency slightly." The inference result generates a supplementary command: fine-tune the electric pump frequency from 38Hz to 39.5Hz. This command is superimposed on the original control flow in an incremental manner to avoid significant fluctuations.

[0111] The device operation monitoring interface shows that the flow of W09 gradually returns to the normal range within the next 15 minutes after the supplementary adjustment. The system records the key time nodes of the entire optimization process: from priority sequence generation to the first control command issuance, which takes 9 seconds, the first adjustment effect evaluation period is 5 minutes, and the fuzzy correction decision takes 800 milliseconds. All time stamps are synchronized to the Beidou satellite timing system, with a clock deviation of less than 1 millisecond.

[0112] For devices with a "medium" priority label such as W11, the system uses a batch processing mode for optimization. W11 is combined with W05 and M03 in the same region to form a control group, and genetic algorithms are used to optimize the parameters of the three simultaneously. The optimization scheme chooses a compromise strategy: W05 wellhead pressure is prioritized for stability, W11 flow is prioritized for recovery, and M03 maintains monitoring accuracy. The final generated time-sharing control command sequence is as follows: adjust the W05 pumping unit parameters for the first 10 minutes, adjust the W11 electric pump frequency for the next 8 minutes, and calibrate the M03 sensor reference value for the last 5 minutes. The execution process is managed through a task queue, and a 2-minute buffer period is set between each device operation to avoid sudden changes in power grid load.

[0113] The historical data comparison module starts analysis after optimization is complete. The W09's records of handling similar events in the past 30 days are retrieved, and this optimization shortens the pressure recovery time by 22% and controls the energy consumption increase to within 5%. These data are input into the learning module of the weight allocation system to dynamically adjust the conflict duration proportion calculation coefficient of the production wells. At the same time, the device health archives are updated, and a "high-frequency adjustment sensitivity" note tag is added to W09, and subsequent optimization will limit its single frequency adjustment amplitude to no more than 3Hz.

[0114] The abnormal handling mechanism is triggered in the following three scenarios: device response timeout, the system automatically resends the command and marks the device communication as abnormal; parameter reverse fluctuation, immediately suspend the current control flow and start root cause analysis; multiple device chain abnormality, switch to the preset emergency pressure reduction mode. All abnormal events generate independent reports, including on-site sensor reading screenshots, control command copies, environmental parameter snapshots, and other information.

[0115] The version management of control instructions adopts a branching strategy. Each optimization generates a main version number (e.g., V2.1.5), and the incremental adjustment record generated by the fuzzy correction is a sub-version (V2.1.5.1). The version information is embedded in the device control log, supporting the operation sequence to be traced back in time. When the version rollback function is triggered, the system can accurately restore the device parameter combination at any historical moment.

[0116] This embodiment guarantees control safety through multi-level verification. All instructions need to pass through the simulator test before being issued, the simulator loads the digital twin model of the target device to predict the response effect; key operations are implemented by double-checking, such as adjustments to the water injection valve opening exceeding 5% need to be confirmed by the on-site engineer; when the execution result deviates from the expectation by more than 10%, a video consultation is automatically convened to connect the remote expert system. The control instruction database is incrementally backed up daily, with a retention period of 180 days, and major operation records are permanently archived.

[0117] Example 5: This embodiment describes the closed-loop implementation process of open timing data dynamic management. After completing the device control instruction execution, the system enters the cycle phase of continuous optimization and parameter update. The distributed data acquisition module maintains the original sampling frequency, and real-time acquisition of core parameters such as oil well pressure, flow rate, and temperature is performed, while continuously receiving environmental monitoring station ground vibration and climate data. All newly collected data packets are attached with time stamps and location labels, and transmitted to the central processing node through an encrypted channel.

[0118] The data preprocessing link adopts a moving average filtering algorithm to process real-time stream data. The width of the filtering window is dynamically adjusted according to the parameter characteristics: a shorter 5-minute window is used for pressure data to capture rapid fluctuations, a 15-minute window is used for flow data to smooth transient disturbances, and a 30-minute window is selected for temperature data to reflect slow-changing trends. The algorithm implementation adopts a ring buffer structure, and when a new data point arrives, the oldest data is automatically discarded to maintain the freshness of the data in the window. The filtered data retains the long-term trend component and eliminates short-term noise caused by device vibration or signal interference.

[0119] The dynamic updating mechanism of safety thresholds is based on historical data analysis. The system maintains a sliding time window, with a default setting of the last 7 days of operation data, and calculates the statistical characteristics of each parameter in the window every hour. The pressure threshold is set to 1.5 times the standard deviation of the moving average, and the flow threshold is set according to the day-night mode, with the average value of the peak period during the day and 1.2 times the value of the trough period at night. The temperature threshold considers seasonal factors, with different baseline values in summer and winter. Before each threshold update, cross-validation is performed, and the new threshold must pass the consistency test of adjacent three nodes' data to take effect.

[0120] The adjustment of control parameters follows a gradual principle. The system compares the deviation of the current operating state from the ideal model and generates parameter adjustment suggestions. For example, when the oil well pressure continues to be higher than the model predicted value, the system suggests gradually reducing the pumping unit stroke, with each adjustment not exceeding 5% of the current value. The adjustment instruction is distributed through the control bus with an effective timestamp to avoid system oscillation caused by simultaneous execution of multiple nodes. Key parameter modification records version numbers, supporting rollback to any of the last three stable versions as needed.

[0121] The closed-loop feedback mechanism includes multi-level monitoring. The bottom layer sensors report real-time readings every 10 seconds, the intermediate layer analysis nodes summarize regional data features every minute, and the top layer control center evaluates the overall operating state every 5 minutes. Abnormal detection uses a composite strategy: short-term jumps are identified by adjacent sampling point differences, medium-term deviations are judged by moving average line deviation, and long-term abnormalities rely on seasonal decomposition residual analysis. When an anomaly is detected, the system automatically increases the sampling frequency of the relevant area and triggers the diagnostic process to determine whether it is a device failure or a normal operating condition fluctuation.

[0122] The data distribution system uses a subscription and publication mode. The central processing node classifies the processed data by topic, such as pressure topic, flow topic, and environment topic. Edge nodes subscribe to relevant topics according to their responsibilities, such as water injection station controllers only receiving pressure data associated with oil wells. Data transmission uses differential compression technology, only sending the difference from the previous value, reducing network load. The subscriber sends an acknowledgment after receiving the data, and unconfirmed data packets are transferred to the offline queue after three retries for manual intervention.

[0123] The system health status self-check program runs regularly. A comprehensive diagnosis is started during the low load period in the early morning, checking items including data collection integrity, analysis process timeliness, and control instruction execution success rate. The diagnosis result generates a health score, and below the threshold triggers an early warning and automatically allocates repair resources. For example, when the data packet loss rate of a certain area exceeds 3% for two consecutive hours, the system automatically switches to the backup communication channel and retransmits the lost data packets. At the same time, the learning module analyzes the causes of the anomaly and updates the fault prediction model parameters.

[0124] Historical data archiving uses a hot and cold layered storage strategy. The last 7 days of data are saved in a high-speed storage cluster for real-time analysis; 7 days to 3 months of data are transferred to a distributed file system; and 3 months of data are compressed and stored in a tape library. Archival data is indexed by well number and date, supporting multi-dimensional retrieval by time range, geographic area, and device type. The data lifecycle management strategy automatically cleans up expired data, but the data 30 minutes before and after key events are permanently retained.

[0125] The clock synchronization network maintains the consistency of the system time. The master clock source accesses the Beidou and GPS dual-satellite signals, and secondary clock servers are deployed in each region. The data acquisition equipment synchronizes the time every 5 minutes, and the control node uses a hardware clock card to ensure microsecond-level precision. The time zone and leap second information are included in all log timestamps, and cross-region operations are executed after being converted to standard time. The clock offset detection program runs every hour, and automatically calibrates nodes with a deviation of more than 50 milliseconds.

[0126] The user interface provides a dynamic management panoramic view. The geographic information system displays real-time parameters and health status of each well, with color coding to distinguish between normal, warning, and fault states. Trend charts can freely select time spans, supporting multi-parameter linkage analysis such as pressure-flow. Operators can manually adjust system parameters, such as modifying the filter window width or temporarily freezing the automatic optimization of a certain region. All manual operations require two-factor authentication, and operation logs and corresponding system state snapshots are automatically associated and stored.

[0127] The disaster recovery system is designed to handle extreme situations. The core node uses a dual-machine hot standby configuration, with a fault switching time controlled within 30 seconds. Each edge node stores a local data copy for the last 8 hours, maintaining basic operation in the event of network disruption. Disaster recovery plans are rehearsed every quarter, simulating system emergency responses in scenarios such as earthquakes and floods. Key configuration parameters are backed up daily to off-site data centers, retaining the last 7 versions for emergency recovery options.

[0128] This dynamic management implementation enables the system to adapt to the dynamic changes of oil and gas field production through continuous data-driven optimization. From data acquisition, processing, and analysis to control execution, the closed-loop process realizes autonomous optimization and adjustment of parameters under the premise of ensuring production safety. Time-sensitive operations at each stage are guaranteed by precise clock synchronization for time consistency, and spatial coordination relies on a unified geographic information coding system. The system continuously accumulates experience data in long-term operation, gradually improving prediction models and adjustment strategies for various parameters.

[0129] It should be noted that, in this text, relational terms such as first and second are used only to distinguish one entity or operation from another, and do not necessarily require or imply any actual such relationship or order between the entities or operations. Moreover, the terms "include", "contain" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device that includes a series of elements not only includes those elements, but also includes other elements not explicitly listed or inherent to such a process, method, article or device.

[0130] While embodiments of the application have been shown and described, it is to be understood that the embodiments described are merely exemplary of the principles and application of the present application. Numerous modifications and adaptions can be effected without departing from the spirit and scope of the present application, which is not limited to the exact construction and arrangement described. It is intended, therefore, to cover all modifications and adaptions that fall within the scope of the claims and their equivalents.

Claims

1. A method for managing oil and gas field opening timing data based on a Hongmeng system, characterized in that, The method comprises the following steps: real-time acquisition of the opening time sequence data of the oil and gas field by a distributed data acquisition module of the Hongmeng system; based on the opening time sequence data, calculating the data point density in the target area, and judging whether the data point density exceeds a preset safety threshold; when the data point density exceeds the safety threshold, using a statistical analysis model to evaluate the interaction between data sources and output the data coupling effect hidden danger state; using a time series analysis method to process environmental factor data to generate an evaluation of the potential interference degree of environmental factors on the opening time sequence data change; determining whether to trigger the overall data optimization process of the target area according to the data coupling effect hidden danger state and the potential interference degree evaluation; when the overall data optimization process is triggered, performing individual path complexity evaluation for each data source, and calculating an adjustment priority sequence based on the evaluation results; according to the adjustment priority sequence, executing control instructions on the oil and gas field equipment to realize dynamic management of the opening time sequence data; when the overall data optimization process is triggered, the individual path complexity evaluation for each data source specifically includes: extracting the opening time sequence data of each data source, including oil well pressure, flow rate and temperature parameters; calculating the data difference degree and time offset between each data source and other data sources; identifying data conflict points, which are data difference degrees exceeding a preset conflict threshold and time offsets below a preset time window; based on the number of data conflict points, generating a path complexity score for each data source; based on the evaluation results, calculating an adjustment priority sequence specifically includes: receiving the path complexity score of each data source; applying a weight allocation mechanism to assign a dynamic weight value to each data source, the dynamic weight value being based on the type parameters of the data source; multiplying the path complexity score by the dynamic weight value to obtain an adjustment priority value; sorting the adjustment priority values from high to low to generate an adjustment priority sequence; according to the adjustment priority sequence, executing control instructions on the oil and gas field equipment specifically includes: obtaining the adjustment priority sequence; for high-priority data sources, extracting the corresponding oil well location and equipment identifier; using a genetic algorithm to generate a device control path to optimize the execution order of the control instructions; sending the control instructions to the execution unit of the Hongmeng system to drive the oil and gas field equipment to adjust the operating parameters. 2.The method of claim 1, wherein, The opening time sequence data of the oil and gas field is acquired in real time by the distributed data acquisition module of the Hongmeng system, specifically including: collecting oil well pressure, flow rate and temperature parameters from multiple sensor nodes, each parameter being accompanied by an accurate time stamp; transmitting the collected parameters to the central processing unit of the Hongmeng system for data format unification and integrity checking; based on the checked data, calculating the ratio of the number of data points in the target area to the area, as the data point density value; comparing the data point density value with the preset safety threshold, if the data point density value is greater than the safety threshold, marking the target area as a high-density state. 3.The method according to claim 1, characterized in that, when the data point density exceeds the safety threshold, using a statistical analysis model to evaluate the interaction between data sources and output the data coupling effect hidden danger state, specifically including: Extracting open-time series data in high-density state, including oil well pressure, flow rate and temperature parameters; Applying regression analysis model to model linear and nonlinear relationships between data sources and calculate interaction strength indicators; Evaluating data coupling effect hidden state as normal or abnormal according to the interaction strength indicators; Outputting the data coupling effect hidden state to subsequent environmental factor analysis process.

4. The method according to claim 1, wherein, Processing environmental factor data using time series analysis method to generate potential interference degree assessment of environmental factors on open-time series data changes, specifically including: Obtaining environmental factor data in target area, including surface vibration and climate conditions; Performing autocorrelation analysis on the environmental factor data to identify periodic change patterns and trends; Quantifying potential interference degree of environmental factors on oil well pressure and flow rate parameters based on the periodic change patterns and trends; Transmitting the potential interference degree assessment to overall optimization judgment process.

5. The method according to claim 1, wherein, Determining whether to trigger overall data optimization process of target area according to the data coupling effect hidden state and the potential interference degree assessment, specifically including: Receiving the data coupling effect hidden state and the potential interference degree assessment; If the data coupling effect hidden state is normal and the potential interference degree assessment is low interference, determining that overall data optimization process does not need to be triggered; Otherwise, determining to trigger overall data optimization process and starting individual path complexity evaluation module.

6. The method according to claim 5, wherein, In the process of executing control instructions according to the adjustment priority sequence, real-time correction of device operation is performed, specifically including: Monitoring open-time series data changes after device execution to capture real-time environmental parameters; Comparing the real-time environmental parameters with historical baseline data to calculate parameter deviation value; If the parameter deviation value exceeds correction threshold, activating fuzzy reasoning mechanism to generate correction instructions; Integrating the correction instructions into control instruction stream to update device operation parameters. 7.The method according to claim 1, wherein, Implementing dynamic management of open-time series data, specifically including: Continuously collecting optimized open-time series data and environmental factor data; Applying moving average filtering algorithm to process the open-time series data to eliminate noise and extract long-term trend; Dynamically updating safety threshold and control parameters according to the long-term trend and real-time feedback; Distributing updated parameters to distributed data collection modules through data bus of Hongmeng system to form closed-loop management cycle.

Images