A health degree evaluation method for multi-dimensional indexes of an oilfield flowmeter

Abstract

The application relates to an oil field flowmeter multi-dimension index health degree evaluation method and relates to the technical field of oil field production, and solves the problems of insufficient real-time performance, long required data acquisition period and large amount of required basic data of an existing flowmeter health state evaluation method. The method comprises the following steps: S1, collecting oil field production data collected by a flowmeter in real time through a PLC communication protocol to obtain PLC-based oil field flowmeter real-time data integration; S2, based on the oil field flowmeter real-time data integration, performing multi-dimension health degree index setting, health degree evaluation model creation, maintenance and instrument binding; S3, obtaining a health degree evaluation result of the oil field flowmeter through a fuzzy expert system; and S4, realizing self-adaptive adjustment of health degree evaluation index weight through health degree evaluation result feedback information. The oil field flowmeter multi-dimension index health degree evaluation method can improve the real-time performance and accuracy of evaluation, and reduce the requirement for a large amount of data and the acquisition period.

Description

Technical Field

[0001] This invention relates to the field of oilfield production technology, and in particular to a method for evaluating the health of multi-dimensional indicators of oilfield flow meters. Background Technology

[0002] In oilfield production, flow meters are critical equipment, and their health status directly affects the accuracy and stability of production. Currently, oilfields primarily rely on manual inspections and calculations to evaluate the health status of flow meters. This traditional method is inefficient. Furthermore, the reliability of the evaluation results varies due to factors such as the professional level and work attitude of personnel, lacking real-time accuracy and failing to provide timely feedback on the equipment's real-time operating status.

[0003] With the development of artificial intelligence technology, intelligent methods such as machine learning and deep learning are gradually being applied to the health evaluation of oilfield flow meters, which has solved problems such as low efficiency and untimely evaluation of manual inspections to some extent. However, these intelligent algorithms have high computational complexity and consume a lot of computing resources, resulting in weak real-time evaluation. At the same time, the use of intelligent algorithms alone makes the interpretability of the evaluation results low, and the generalization to complex production environments is poor, which affects the accuracy of the evaluation. Summary of the Invention

[0004] This invention addresses the problems of insufficient real-time performance, long data acquisition cycles, and large amounts of basic data required in existing flowmeter health status evaluation methods. It provides a multi-dimensional health evaluation method for oilfield flowmeters. This method improves the real-time performance and accuracy of the evaluation while reducing the need for large amounts of data and the associated acquisition cycle.

[0005] The present invention solves its problem through the following technical solution: a method for evaluating the health of multi-dimensional indicators of oilfield flow meters, comprising the following steps:

[0006] S1. Real-time acquisition of oilfield production data collected by flow meters through PLC communication protocol to obtain PLC-based real-time data integration of oilfield flow meters;

[0007] S2. Based on real-time data integration from oilfield flow meters, perform multi-dimensional health index setting, health evaluation model creation, maintenance, and instrument binding;

[0008] S3. Obtain the health evaluation results of oilfield flow meters through a fuzzy expert system;

[0009] S4. Through the feedback information of the health evaluation results, the weight of the health evaluation indicators can be adaptively adjusted.

[0010] Furthermore, the method for real-time data integration of oilfield flow meters based on PLC in step S1 is as follows:

[0011] The integrated system is coded in Python, and the code includes a PLC communication module and a data parsing module. It communicates with the PLC through the server's network card and Ethernet according to the PLC protocol to obtain real-time data from the flow meter collected by the PLC. Finally, the collected data is stored in the database.

[0012] In addition to acquiring data, the integrated system code is also responsible for converting the data into floating-point data.

[0013] Furthermore, the PLC communication module is responsible for communicating with different types of PLCs; by setting the PLC's IP address, port number, and device type parameters, the integrated system establishes connections with different PLCs.

[0014] After the integrated system establishes a connection with different PLCs, it sends a data request to the PLC via Ethernet according to the instrument point tags connected to the PLC; after receiving the request, the PLC returns the collected flow meter data in the agreed format.

[0015] The data parsing module is responsible for processing the data acquired from the PLC; the acquired data includes formats such as numbers, arrays, and byte strings; the data returned by the PLC includes multiple formats such as numbers, arrays, and byte strings, which need to be parsed and converted uniformly.

[0016] Furthermore, for byte string data, according to the PLC's data format and the flow meter's communication protocol, the struct block in Python is used to parse it into floating-point data;

[0017] For data in numeric and array formats, splitting is performed using the position parameters of point markers.

[0018] Furthermore, step S2 involves a method for setting multi-dimensional health indicators based on real-time data integration from oilfield flow meters.

[0019] Based on the experts' concerns regarding the health of oilfield flow meters, the following were set as health evaluation indicators: maintenance frequency, number of abnormal readings, working time, response time, abnormal cumulative flow meter readings, vibration amplitude, measuring range, and accuracy.

[0020] Furthermore, the maintenance frequency reflects the reliability of the flow meter. The higher the frequency, the more frequent the faults or abnormalities that occur during the operation of the flow meter, and the lower the reliability of the equipment.

[0021] The number of abnormal values ​​indicates the malfunction or interference; the longer the operating time, the more prone the flow meter is to fatigue and wear.

[0022] The response time is a key parameter for measuring the real-time response performance of a flow meter. Its value is negatively correlated with the speed at which the device responds to sudden changes in flow. That is, the shorter the response time, the better the instantaneous response capability of the flow meter to changes in operating conditions.

[0023] The abnormal cumulative values ​​of the flow meter indicate problems with the accuracy and reliability of the measurement; the vibration amplitude affects the measurement accuracy, and excessive vibration can easily lead to errors.

[0024] The measurement range and measurement accuracy are two key indicators characterizing the performance of a flow meter: the former directly determines the range of fluid parameters that the device can cover, while the latter reflects the degree to which the measurement results are close to the true values. Together, they affect the applicability and measurement reliability of the flow meter in specific engineering scenarios.

[0025] Furthermore, the method for maintaining the health assessment model and binding the instrument in step S2 is as follows:

[0026] The methods for maintaining a health assessment model include: first, creating a multi-dimensional health assessment model, with different health assessment models containing different numbers or types of assessment indicators, and then binding the instrument to the model; and then, according to the requirements, adding and deleting assessment indicators for each model.

[0027] The method of instrument binding is as follows: classify the flow meters according to their usage environment, location, and service life, and then bind the instruments to a suitable health assessment model to obtain the required evaluation indicators.

[0028] Furthermore, the process of creating the multi-dimensional indicator health evaluation model is as follows:

[0029] Extract key influencing factors of instrument health from actual working conditions and quantify them into independent indicators. Based on object-oriented programming, encapsulate the calculation logic of each indicator to generate member functions and simultaneously build a matching database table structure.

[0030] Analyze the health assessment needs of different instruments, create a special model subclass by inheriting the basic assessment class, and bind quantitative indicators with model parameters to form a complete assessment unit;

[0031] Each instrument is bound to the corresponding model according to its evaluation requirements to complete the preparation for health evaluation; the health evaluation model exists in the form of a class in object-oriented programming, and its creation process is equivalent to creating different health evaluation classes.

[0032] Furthermore, step S3 involves obtaining the health evaluation results of the oilfield flow meter through a fuzzy expert system. The specific steps are as follows:

[0033] S31. First, the basic instrument information and real-time oilfield production data are input into the fuzzy expert system; these data contain the parameters required for calculating the evaluation indicators.

[0034] S32. The data is processed using a fuzzy expert system to obtain the scores for each evaluation indicator;

[0035] The expert system comprises three main components: a knowledge base, a rule base, and an inference engine. The knowledge base and rule base support the inference engine, which performs inference based on the input data. The inference engine infers fuzzy index weights and evaluation rules based on the input data and the knowledge rules in the database. The fuzzy index weights are further analyzed and generated using the entropy weight method and the analytic hierarchy process, serving as the final index weights. The inference engine calculates the scores of each evaluation index based on the evaluation rules and the generated index weights, combined with the input data.

[0036] S33. The health evaluation result of the oilfield flow meter is obtained by comprehensively calculating the weighted scores of all indicators; the system outputs the health evaluation result of the oilfield flow meter.

[0037] Furthermore, step S32 processes the data using a fuzzy expert system to obtain scores for each evaluation index. The specific method includes:

[0038] First, it is necessary to build a health evaluation index database table and a model database table in the database, and establish a static data storage structure for the health evaluation index;

[0039] Secondly, the instrument to be evaluated is bound to the evaluation model. The fuzzy expert system will retrieve the bound model from the database based on the input instrument number and read the indicators contained in the model.

[0040] Then, based on the read indicator, the corresponding evaluation method is called, and according to the mapping relationship between the indicator and the static data related to the health of the instrument stored in the database, the corresponding data is read and the score of the indicator is calculated according to the evaluation method. If the data is similar to good, good, or poor and has a subjective evaluation intention, it is converted into the corresponding objective score according to the fuzzy rules in the evaluation method as the score of the indicator; if it is specific data, the score of the instrument for this indicator will be calculated by the calculation method recorded in the expert knowledge.

[0041] Finally, after obtaining the scores for each indicator, a weighted calculation is performed based on the weights of each indicator recorded in the health evaluation model to obtain the final health evaluation score of the instrument.

[0042] Furthermore, the indicator database is divided into fuzzy indicators and ordinary indicators; the model database associates indicator IDs with weight parameters through model IDs. The weights are generated by fuzzy weight analysis using entropy weight method and analytic hierarchy process after expert verification.

[0043] The fuzzy index stores the mapping relationship between text descriptions and scores; the ordinary index stores calculation formulas based on instrument data.

[0044] Furthermore, the data representing "good," "fair," and "poor" with subjective evaluation intentions are converted into corresponding objective scores as follows:

[0045] A good objective score is ≥80 points;

[0046] 60 points < Good corresponds to an objective score < 80 points;

[0047] The objective score corresponding to the difference is ≤60 points.

[0048] Furthermore, the method for adaptively adjusting the weights of the health evaluation indicators in step S4 includes:

[0049] After generating the instrument health evaluation results, the system automatically generates a health evaluation result feedback survey report containing the final evaluation score, the scores of each indicator and their weights, and the user can adjust the results in the system according to actual needs.

[0050] After receiving user feedback, the system uses the PSO optimization algorithm with MSE as the optimal function to recalculate appropriate fuzzy weights, forming a dynamically optimized evaluation model to achieve real-time evaluation of the health of flow meters in oilfield production.

[0051] Compared with the above-mentioned background technology, the present invention has the following beneficial effects:

[0052] This invention provides a method for evaluating the health of oilfield flow meters based on multiple dimensions. By integrating multiple intelligent functions, it offers the following beneficial effects:

[0053] (1) Improved real-time performance: This invention utilizes a PLC oilfield production data integration method and a PLC communication protocol to collect oilfield production data from flow meters in real time, enabling real-time health assessment. Compared to existing technologies, this invention significantly improves data acquisition capabilities and accuracy, thereby enhancing the real-time performance and accuracy of the assessment.

[0054] (2) Reduced data requirements: Existing technologies rely on large amounts of historical and real-time data as training samples for machine learning algorithms. This invention, however, utilizes a fuzzy expert system to construct a fuzzy rule base based on expert knowledge and experience, thus reducing the need for large amounts of data. Even with limited data, effective health status assessments of oilfield flow meters can still be performed, solving the problem of high data requirements.

[0055] (3) Reduced algorithm complexity: Machine learning algorithms typically have high complexity, long computation time, and high computational resource consumption. This invention uses a fuzzy expert system, which performs reasoning based on fuzzy logic. The algorithm is simple and efficient, significantly reducing computation time and computational resource consumption. At the same time, the fuzzy expert system has high processing efficiency and is more suitable for evaluating the health status of hundreds of thousands of flow meters in the diverse oilfield production environment.

[0056] (4) Improved interpretability: The decision-making process of machine learning algorithms is often difficult to interpret, while fuzzy expert systems, based on expert knowledge and fuzzy rules, make their working principles and decision-making processes easier to understand and explain. Users can clearly understand the basis for each step of the judgment, which increases users' trust and acceptance of the oilfield flow meter model.

[0057] (5) Enhanced flexibility and adaptability: The multi-dimensional index integrated calculation model for oilfield flowmeters in this invention can create multiple evaluation models based on information such as the brand of flowmeter used in oilfield production, the operating environment, and fluid characteristics, and can flexibly set the required health evaluation indicators. Through an adaptive adjustment mechanism, the evaluation model can continuously optimize the fuzzy rule base and weight allocation based on feedback from actual applications, dynamically adapting to different production environments and operating conditions, preventing overfitting and underfitting, and improving the flexibility and adaptability of the method. Attached Figure Description

[0058] Figure 1 This is a flowchart of the health evaluation method for multi-dimensional indicators of oilfield flow meters according to the present invention;

[0059] Figure 2 This is a flowchart of the PLC oilfield flow meter data integration module according to an embodiment of the present invention;

[0060] Figure 3 This is a schematic diagram of the configuration of the multi-dimensional index integrated health evaluation model for oilfield flow meters according to an embodiment of the present invention;

[0061] Figure 4 This is a flowchart of the fuzzy expert system according to an embodiment of the present invention. Detailed Implementation

[0062] To make the objectives, technical solutions, and advantages of the present invention clearer, the embodiments of the present invention will be described in further detail below with reference to the accompanying drawings.

[0063] Appendix Figure 1 This is the overall system architecture diagram of the present invention, illustrating the process of the multi-dimensional indicator integrated health evaluation technology for oilfield flow meters; (Attached) Figure 2This is a detailed flowchart of the PLC oilfield flow meter data integration module, demonstrating the specific steps involved in real-time acquisition and processing of oilfield production data read from the flow meter via the PLC communication protocol; (Attached) Figure 3 This is a schematic diagram of the configuration of a multi-dimensional indicator integrated health evaluation model for oilfield flow meters, showing the structure of multiple evaluation models, including examples of the binding of evaluation indicators and flow meters; (Attached) Figure 4 This is a flowchart of a fuzzy expert system, illustrating the input variables, reasoning process, and evaluation result calculation process of the fuzzy expert system.

[0064] The principle of this invention, a multi-dimensional health evaluation method for oilfield flow meters:

[0065] First, based on the PLC protocol, real-time oilfield production data acquired by flow meters is collected. Second, multiple evaluation indicators and calculation methods are selected based on experts' concerns regarding the health of oilfield flow meters. Third, a multi-dimensional integrated health evaluation model for oilfield flow meters is created, and different dimensions of indicators are selected for the integrated health evaluation model, which is then linked to the instruments that need to be evaluated in oilfield production. Next, the selected multi-dimensional indicators are used as input data and fed into a fuzzy expert system. The expert system infers the fuzzy indicator weights and evaluation rules. The fuzzy indicator weights are further obtained using the entropy weight method and the analytic hierarchy process. Finally, the indicator weights are weighted and calculated with the evaluation indicator parameters to obtain the health evaluation results of the oilfield flow meters. Finally, the weights in the knowledge base are updated based on user feedback.

[0066] As attached Figure 1 As shown, this invention discloses a method for evaluating the health of multi-dimensional indicators of oilfield flow meters. The invention mainly consists of five parts: PLC-based oilfield production data integration, multi-dimensional indicator selection, maintenance and instrument binding of the integrated health evaluation model for multi-dimensional indicators of oilfield flow meters, a fuzzy expert system, and an adaptive adjustment mechanism. Specifically, it includes the following steps:

[0067] Step 1: Real-time acquisition of oilfield production data collected by flow meters via PLC communication protocol to obtain PLC-based real-time data integration of oilfield flow meters.

[0068] PLC-based real-time data integration of oilfield flow meters is shown in the attached figure. Figure 2 As shown, the integrated system code includes a PLC communication module and a data parsing module. Through the server's network card and Ethernet, it communicates with the PLC according to the PLC protocol to obtain real-time data from the flow meter collected by the PLC. The acquired data includes various formats (numbers, arrays, and byte strings). In addition to acquiring the data, the integrated system code is also responsible for converting the data into floating-point data.

[0069] The implementation process utilized Python for coding, primarily comprising a PLC communication module and a data parsing module. The PLC communication module handles communication with different types of PLCs. By setting parameters such as the PLC's IP address, port number, and device type, the system can flexibly establish connections with PLCs from various brands, including AB, Siemens, Schneider Electric, and Omron. After connection establishment, data requests are sent to the PLC via Ethernet based on the instrument point labels connected to the PLC. Upon receiving the request, the PLC returns the collected flowmeter data in a pre-defined format. The data parsing module processes the data acquired from the PLC. Since the data returned by the PLC includes various formats such as numbers, arrays, and byte strings, unified parsing and conversion are required. For byte string data, the `struct` module in Python is used to parse it into floating-point data, based on the PLC's data format and the flowmeter's communication protocol. For number and array data, the data is split using the position parameters of the point labels. Finally, the collected data is stored in the database.

[0070] Step 2: Setting up multi-dimensional indicators, maintaining the health assessment model, and binding instruments.

[0071] The configuration of the multi-dimensional health evaluation model for oilfield flow meters is attached. Figure 3 As shown. Multi-dimensional evaluation index settings: Based on the experts' concerns about the health of oilfield flow meters, health evaluation indicators are set from multiple dimensions, including maintenance frequency, number of abnormal values, working time, response time, cumulative abnormal values ​​of flow meters, vibration amplitude, measuring range, and accuracy.

[0072] The maintenance frequency reflects the reliability of the flow meter. The higher the frequency, the more frequent the faults or abnormalities that occur during the operation of the flow meter, and the lower the reliability of the equipment.

[0073] The number of abnormal values ​​indicates the malfunction or interference; the longer the operating time, the more prone the flow meter is to fatigue and wear.

[0074] The response time is a key parameter for measuring the real-time response performance of a flow meter. Its value is negatively correlated with the speed at which the device reacts to sudden changes in flow. That is, the shorter the response time, the better the instantaneous response capability of the flow meter to changes in operating conditions.

[0075] The abnormal cumulative values ​​of the flow meter indicate problems with the accuracy and reliability of the measurement; the vibration amplitude affects the measurement accuracy, and excessive vibration can easily lead to errors.

[0076] The measurement range and measurement accuracy are two key indicators characterizing the performance of a flow meter: the former directly determines the range of fluid parameters that the device can cover, while the latter reflects the degree to which the measurement results are close to the true values. Together, they affect the applicability and measurement reliability of the flow meter in specific engineering scenarios.

[0077] Health assessment model maintenance and instrument binding: The main idea is to first build the model and then bind the instruments to it. The health assessment model includes modules for assessment indicator maintenance, model creation, and indicator binding. Health assessment model maintenance refers to first creating the health assessment model and then binding the indicators. Different health assessment models contain different numbers or types of assessment indicators. Health assessment model maintenance allows for the addition and deletion of assessment indicators for each model. Instrument binding involves categorizing flowmeters according to their usage environment, location, and age, and then binding them to the appropriate model to obtain the assessment indicators needed for subsequent calculations, improving the flexibility of flowmeter health assessment. (See attached...) Figure 3 As shown, flow meters that have been in operation for more than 10 years can be bound to multi-dimensional indicator health evaluation model 1, flow meters that have been in operation for less than 3 years can be bound to multi-dimensional indicator health evaluation model 2, flow meters that operate outdoors can be bound to multi-dimensional indicator health evaluation model 3, and flow meters that operate on branch lines can be bound to multi-dimensional indicator health evaluation model 4.

[0078] Step 3: Obtain the health evaluation results of the oilfield flow meter using a fuzzy expert system.

[0079] The flowchart of the fuzzy expert system is attached. Figure 4 As shown, when evaluating the health of oilfield flow meters, the initial inputs are basic instrument information and real-time oilfield production data, which contain the parameters required for calculating evaluation indicators. The expert system comprises three main components: a knowledge base, a rule base, and an inference engine. The knowledge base and rule base support the inference engine, which performs inference based on the input data. The inference engine, based on the input data and the knowledge rules in the database, infers fuzzy indicator weights and evaluation rules. These fuzzy indicator weights are further analyzed and generated using the entropy weighting method and the analytic hierarchy process (AHP), serving as the final indicator weights. The inference engine, based on the evaluation rules and the generated indicator weights, combined with the input data, calculates the scores for each evaluation indicator. Finally, the system outputs the health evaluation results of the oilfield flow meters. These results are derived from the weighted scores of all indicators, effectively assessing the health status of the oilfield flow meters and providing a scientifically sound evaluation conclusion.

[0080] The knowledge base stores relevant knowledge about the health evaluation of oilfield flow meters, including the evaluation indicators, evaluation methods, and expert-given fuzzy weights required for different oilfield flow meters during the evaluation process. The rule base contains a series of rules formulated based on expert experience and knowledge to guide the system in reasoning and decision-making based on input data. The inference engine is the core component of the expert system; it performs logical reasoning and calculations based on the input data and the rules in the rule base.

[0081] The input data includes basic instrument information (such as flow meter model, installation location, and service life) and real-time oilfield production data (such as flow rate, pressure, and temperature), primarily used to calculate scores for each evaluation indicator. The inference engine analyzes the data using the entropy weighting method and the analytic hierarchy process (AHP) to generate fuzzy indicator weights. Simultaneously, based on the evaluation rules and the generated indicator weights, the inference engine performs weighted summation and calculation to derive the health evaluation result of the oilfield flow meter.

[0082] The implementation process first requires constructing a health evaluation indicator database and a model database to establish a static data storage structure for the health evaluation indicators. The indicator database is divided into fuzzy indicators and ordinary indicators: fuzzy indicators store the mapping relationship between textual descriptions and scoring, while ordinary indicators store calculation formulas based on instrument data. The model database associates indicator IDs with weight parameters through model IDs. The weights are generated by fuzzy weight analysis using entropy weighting and analytic hierarchy process (AHP) after expert validation. Next, a fuzzy expert system is built, and a database interface is developed to connect the indicator database and the model database, enabling the fuzzy expert system to read health evaluation indicators from the health evaluation indicator database. After reading an indicator, the expert system reads the corresponding data based on the mapping relationship between the indicator and the static health-related data of the instruments stored in the database. If the data is subjective, such as "good," "fair," or "poor," it is converted into a corresponding objective score according to fuzzy rules. If the data is specific data, the score for that indicator is calculated using calculation methods recorded in the expert knowledge. Finally, after obtaining the scores for each indicator, a weighted calculation is performed based on the weights of each indicator recorded in the health evaluation model to obtain the final health evaluation score of the instrument.

[0083] The objective scores corresponding to the data with subjective evaluation intentions, such as "good," "fair," and "poor," are:

[0084] "Good" corresponds to an objective score of ≥80 points;

[0085] 60 points < the objective score corresponding to "Good" < 80 points;

[0086] The objective score corresponding to "poor" is ≤60 points.

[0087] Step 4: Adaptively adjust the weights of health assessment indicators based on feedback information from the health assessment results.

[0088] In practical applications, after generating the instrument health evaluation results, the system automatically generates a health evaluation result feedback survey report containing the final evaluation score, the scores of each indicator, and their weights. Users can adjust the results in the system according to their actual needs. After receiving user feedback, the system uses the PSO optimization algorithm, with MSE as the optimal function, to recalculate appropriate fuzzy weights, forming a dynamically optimized evaluation model. This enables real-time, accurate, and flexible evaluation of the health of flow meters in oilfield production, thereby improving the safety and efficiency of the production process.

[0089] Existing methods for evaluating instrument health suffer from poor interpretability. This invention addresses this issue by employing manually customized indicators to achieve a scientific and reasonable evaluation of instrument health. This method effectively improves the interpretability of evaluation results, providing a more convincing basis for relevant decisions. The evaluation results include not only the overall evaluation but also the specific evaluation results of each indicator, significantly enhancing interpretability. Regarding the issue of poor adaptability to complex production environments, this invention uses an indicator-model binding method and a model-instrument binding method, allowing for adjustments to the evaluation method as needed, significantly improving the adaptability of the instrument health evaluation method to complex working environments. Addressing the issue of poor real-time performance in instrument health evaluation, this invention utilizes mathematical statistics methods such as variance calculation and threshold analysis. While ensuring the validity of the results, it eliminates the need for training neural networks, greatly improving the real-time performance of the evaluation results. This invention improves the real-time performance and accuracy of the evaluation, reduces the need for large amounts of data and the data collection cycle, reduces computational resource consumption, and enhances interpretability and adaptability to complex production environments.

[0090] Example 1

[0091] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail with reference to the accompanying drawings, taking the West-East Three Injection Station as an example.

[0092] (I) Real-time data integration of oilfield flow meters based on PLC

[0093] First, the PLC communication module of the integrated system code is configured according to the model and communication protocol of the field equipment, enabling the system to communicate with the PLCs in the West-East Three Injection Station. Then, the PLC communication module, carrying point tags, initiates a data read request to the PLC device via Ethernet and sends the request. The PLC device locates the corresponding data based on the point tags and returns it. Finally, the acquired data is split into floating-point data according to the point type to obtain the data value of the corresponding flowmeter at a specific moment. Considering that the PLC collects instrument data every 10 minutes, the integrated system code is also set to collect data every 10 minutes. This system will obtain real-time data from all flowmeters in the West-East Three Injection Station based on different point tags.

[0094] (II) Setting up multi-dimensional indicators, maintaining the health assessment model and binding instruments

[0095] First, implement multi-dimensional evaluation index settings: Based on the experts' concerns about the health of oilfield flow meters, health evaluation indexes are set from multiple dimensions, including maintenance frequency, number of abnormal values, working time, response time, abnormal cumulative values ​​of flow meters, vibration amplitude, range and accuracy.

[0096] Secondly, the flow meters of the three injection stations in the west and east are classified according to their usage environment, location, and service life.

[0097] Then, multiple health assessment models covering multi-dimensional indicators are built using code, and each model is given a unique and easily identifiable name. The indicators used for evaluation are edited to ensure their accuracy and effectiveness.

[0098] Finally, the edited and optimized evaluation indicators are bound to the corresponding health assessment models. These steps are primarily implemented through code and a database. The health assessment model maintenance function allows for the addition and deletion of evaluation indicators for each model. (See attached document.) Figure 3 As shown, flow meters that have been in operation for more than 10 years (oilfield flow meter 1*) can be bound to multi-dimensional indicator health evaluation model 1; flow meters that have been in operation for less than 3 years (oilfield flow meter 2*) can be bound to multi-dimensional indicator health evaluation model 2; flow meters that operate outdoors (oilfield flow meter 3*) can be bound to multi-dimensional indicator health evaluation model 3; and flow meters that operate on branch lines (oilfield flow meter 4*) can be bound to multi-dimensional indicator health evaluation model 4.

[0099] (III) Fuzzy Expert Systems

[0100] For the Xidong No. 3 injection station, flow meter No. 01030202XX23112127, installed outdoors, was selected. Five indicators were used: service life, maintenance frequency, operating environment, data fluctuation amplitude, and number of data fluctuations. The initial fuzzy weights for each indicator were 0.5, 0.5, 0.25, 0.5, and 0.5, respectively. First, a hierarchical structure was constructed including service life, maintenance frequency, operating environment, data fluctuation amplitude, and number of data fluctuations. The system then used a 1-9 scale method to perform pairwise comparisons to form a judgment matrix.

[0101]

[0102] Calculate the eigenvectors and normalize them to obtain the subjective weights w. AHP = [0.300.300.150.100.15]; then process the original data matrix.

[0103]

[0104] The extreme value method is used for standardization, the information entropy E = [0.95, 0.94, 0.96, 0.97, 0.95] is calculated, and the objective weight w is obtained. 熵权 = [0.22, 0.24, 0.18, 0.16, 0.20];

[0105] Finally, the subjective and objective weights are combined using a fusion coefficient of α = 0.6, resulting in: Final weight = 0.6w AHP+ 0.4w 熵权 =[0.28,0.30,0.17,0.12,0.17].

[0106] The manufacturing date is 2022-11-01, the current date is 2024-11-04, the current date minus the manufacturing date equals 24 months, and the service life score is 100-24=76 (points).

[0107] The number of repairs is 0, so the repair frequency score is 100 - 0 = 100 (points).

[0108] Located indoors, the usage environment score is 90 (points);

[0109] The variance of the instrument's data over the past 100 readings is s = 0.10257269, while the average variance of the same type of instrument within the same station over the past 100 readings is s² = 0.085763. The data fluctuation range is rated as follows:

[0110] The average value of the instrument's data over the past 100 times. The value is 0.10279, and the data obtained by the instrument in the i-th time is x. i The data fluctuation frequency score is:

[0111] Total health rating score = Years of service score * 0.28 + Maintenance frequency score * 0.30 + Usage environment score * 0.17 + Data fluctuation range score * 0.12 + Number of data fluctuations score * 0.17 = 60.07 (points);

[0112] Because the health score is less than 70 (60 < health rating score), the instrument's rating level is sub-healthy.

[0113] (iv) Adaptive adjustment mechanism

[0114] At the Xidong No. 3 injection station, after using the health assessment system for a period of time, operators discovered a discrepancy between the health score of flowmeter No. 1030202XX23112127 and its actual condition. The operators submitted a survey report based on the health assessment results. The system, using the PSO algorithm with MSE as the optimal function, recalculated appropriate fuzzy weights based on the feedback, achieving adaptive adjustment of the weights for the instrument health assessment indicators.

[0115] In May 2024, a health assessment was conducted on five flow meter wells located at the West-East Three Injection Stations, and the results were tracked and recorded. The specific health assessment results and performance are shown in Table 1 below:

[0116] Table 1

[0117]

[0118] Based on the experimental data in Table 1, the oilfield flow meter health evaluation method based on multi-dimensional indicator integrated health evaluation technology can meet the real-time acquisition of oilfield production data. The configuration of the multi-dimensional indicator integrated health evaluation model improves the flexibility of oilfield flow meter evaluation, and the selection of indicators from multiple dimensions in the model improves interpretability. The evaluation time can be controlled within 1.5 minutes, improving computational efficiency. At the same time, for the calculation of some evaluation indicators, only real-time data within 24 hours is required, reducing the amount of data required.

[0119] Those skilled in the art will recognize that the embodiments described herein are intended to help the reader understand the implementation methods of the present invention, and should be understood that the scope of protection of the present invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific modifications and combinations based on the technical teachings disclosed in this invention without departing from the spirit of the invention, and these modifications and combinations are still within the scope of protection of the present invention.

Claims

1. A method for evaluating the health of oilfield flow meters based on multiple dimensions, characterized in that: Includes the following steps: S1. Real-time collection of oilfield production data from flow meters via PLC communication protocol to obtain PLC-based real-time data integration of oilfield flow meters; S2. Based on real-time data integration from oilfield flow meters, perform multi-dimensional health index setting, health evaluation model creation, maintenance, and instrument binding; S3. Obtain the health evaluation results of oilfield flow meters through a fuzzy expert system; S4. Through the feedback information of the health evaluation results, the weight of the health evaluation indicators can be adaptively adjusted.

2. The method for evaluating the health of multi-dimensional indicators of oilfield flow meters according to claim 1, characterized in that: Step S1, the method for real-time data integration of oilfield flow meters based on PLC, is as follows: The integrated system is coded in Python, and the code includes a PLC communication module and a data parsing module. It communicates with the PLC through the server's network card and Ethernet according to the PLC protocol to obtain real-time data from the flow meter collected by the PLC. Finally, the collected data is stored in the database. The integrated system code is also responsible for converting data into floating-point data.

3. The method for evaluating the health of multi-dimensional indicators of oilfield flow meters according to claim 2, characterized in that: The PLC communication module is responsible for communicating with different types of PLCs; by setting the PLC's IP address, port number, and device type parameters, the integrated system establishes connections with different PLCs. After the integrated system establishes a connection with different PLCs, it sends a data request to the PLC via Ethernet according to the instrument point tags connected to the PLC; after receiving the request, the PLC returns the collected flow meter data in the agreed format. The data returned by the PLC includes numbers, arrays, and byte strings. The data parsing module is responsible for processing the data acquired from the PLC; the acquired data includes numbers, arrays, and byte strings; the acquired data needs to be parsed and converted in a unified manner.

4. The method for evaluating the health of multi-dimensional indicators of oilfield flow meters according to claim 3, characterized in that: For byte string data, based on the specific data format of the PLC and the communication protocol followed by the flow meter, the struct module in Python is used to parse it into floating-point data to ensure that the data is presented in an appropriate numerical form. For data in numerical and array formats, the data is split into meaningful sub-data units based on the position parameters of the point markers, which facilitates subsequent analysis and processing.

5. The method for evaluating the health of multi-dimensional indicators of oilfield flow meters according to claim 1, characterized in that: The multi-dimensional health indicators in step S2 include: Maintenance frequency, number of abnormal values, working time, response time, abnormal cumulative values ​​of the flow meter, vibration amplitude, measuring range and accuracy.

6. The method for evaluating the health of multi-dimensional indicators of oilfield flow meters according to claim 5, characterized in that: The maintenance frequency reflects the reliability of the flow meter. The higher the frequency, the more frequent the faults or abnormalities that occur during the operation of the flow meter, and the lower the reliability of the equipment. The number of abnormal values ​​indicates the malfunction or interference; the longer the operating time, the more prone the flow meter is to fatigue and wear. The response time is negatively correlated with the speed at which the device responds to sudden changes in flow rate; that is, the shorter the response time, the better the instantaneous response capability of the flow meter to changes in operating conditions. The abnormal cumulative values ​​of the flow meter indicate problems with the accuracy and reliability of the measurement; the vibration amplitude affects the measurement accuracy, and excessive vibration can easily lead to errors. The measurement range and measurement accuracy are two indicators characterizing the performance of a flow meter. The measurement range directly determines the range of fluid parameters that the device can cover, while the measurement accuracy reflects how close the measurement result is to the true value.

7. The method for evaluating the health of multi-dimensional indicators of oilfield flow meters according to claim 1, characterized in that: The method for maintaining the health assessment model and binding instruments in step S2 is as follows: The methods for maintaining a health assessment model include: first, creating a multi-dimensional health assessment model, where different health assessment models contain different numbers or types of assessment indicators; then, binding the instrument to the model; and finally, adding and deleting assessment indicators for each model as needed. The method of instrument binding is as follows: classify the flow meters according to their usage environment, location, and service life, and then bind the instruments to a suitable health assessment model to obtain the required evaluation indicators.

8. The method for evaluating the health of multi-dimensional indicators of oilfield flow meters according to claim 7, characterized in that: The process of creating the multi-dimensional health evaluation model is as follows: Extract key influencing factors of instrument health from actual working conditions and quantify them into independent indicators. Based on object-oriented programming, encapsulate the calculation logic of each indicator to generate member functions and simultaneously build a matching database table structure. Analyze the health assessment needs of different instruments, create a special model subclass by inheriting the basic assessment class, and bind quantitative indicators with model parameters to form a complete assessment unit; Each instrument is bound to the corresponding health assessment model according to its evaluation requirements, thus completing the preparation for the health assessment; The health assessment model exists in the form of classes in object-oriented programming, and its creation process is equivalent to creating different health assessment classes.

9. The method for evaluating the health of multi-dimensional indicators of oilfield flow meters according to claim 1, characterized in that: Step S3 involves obtaining the health evaluation results of the oilfield flow meter using a fuzzy expert system. The specific steps are as follows: S31. The instrument basic information and real-time oilfield production data are fed into the fuzzy expert system as input data; the input data includes the parameters required for calculating the evaluation indicators. S32. The data is processed using a fuzzy expert system to obtain the scores for each evaluation indicator; S33. The health evaluation result of the oilfield flow meter is obtained by comprehensively calculating the weighted scores of all indicators; the system outputs the health evaluation result of the oilfield flow meter.

10. The method for evaluating the health of multi-dimensional indicators of oilfield flow meters according to claim 9, characterized in that: Step S32 involves processing the data using a fuzzy expert system to obtain scores for each evaluation indicator. The specific method includes: First, it is necessary to build a health evaluation index database table and a model database table in the database, and establish a static data storage structure for the health evaluation index; Secondly, the instrument to be evaluated is bound to the health evaluation model. The fuzzy expert system will retrieve the health evaluation model bound to it from the database based on the input instrument number and read the indicators contained in the model. Then, based on the read indicator, the corresponding evaluation method is called, and according to the mapping relationship between the indicator and the static data related to the health of the instrument stored in the database, the corresponding data is read and the score of the indicator is calculated according to the evaluation method. Finally, after obtaining the scores for each indicator, a weighted calculation is performed based on the weights of each indicator recorded in the health evaluation model to obtain the final health evaluation score of the instrument.

11. The method for evaluating the health of multi-dimensional indicators of oilfield flow meters according to claim 10, characterized in that: The index database is divided into fuzzy indicators and ordinary indicators; the model database associates the index ID with the weight parameters through the model ID. The weights are generated by fuzzy weight analysis using the entropy weight method and the analytic hierarchy process after expert verification. The fuzzy index stores the mapping relationship between text descriptions and scores; The common indicators store calculation formulas based on instrument data.

12. The method for evaluating the health of multi-dimensional indicators of oilfield flow meters according to claim 1, characterized in that: Step S4 involves the following method for adaptively adjusting the weights of health assessment indicators: After generating the instrument health evaluation results, the system automatically generates a health evaluation result feedback survey report containing the final evaluation score, the scores of each indicator and their weights, and the user can adjust the results in the system according to actual needs. After receiving user feedback, the system uses the PSO optimization algorithm with MSE as the optimal function to recalculate appropriate fuzzy weights, forming a dynamically optimized evaluation model to achieve real-time evaluation of the health of flow meters in oilfield production.

Images