Data-driven underwater production system fault mode and influence analysis system

Abstract

The invention belongs to the technical field of underwater production system information processing and analysis, and discloses a data-driven underwater production system fault mode and influence analysis system, which comprises four modules: firstly, a historical data collection and storage module is used for collecting, storing and checking historical faults in detail, and outputting a key fault mode; the historical data analysis module is used for determining data analysis indexes according to available data so as to form fault analysis data, specifically, four main data analysis indexes including the occurrence rate, the production loss, the average repair time and the fault interval time are focused on, but in practical application, the occurrence rate, the production loss, the average repair time and the fault interval time are not influenced. Other data analysis indexes can be included according to more comprehensive historical data; thirdly, the data analysis index weight calculation module obtains the weight of the data analysis index by using an objective weight calculation method; and the fault mode sorting and analyzing module is used for carrying out priority sorting on the fault modes by utilizing a multi-criterion decision algorithm, and finally outputting a result and recommending necessary prevention measures.

Description

Technical Field

[0001] The present invention belongs to the technical field of information processing and analysis of underwater production systems, and particularly relates to a data-driven fault mode and effect analysis system for underwater production systems. Background Art

[0002] Underwater production systems are the core facilities for offshore oil and gas exploitation, and their safety and reliability directly affect the production efficiency and operation cost of deepwater oil and gas fields. However, due to the extreme working environment (deep water, high pressure, low temperature, corrosion, etc.) and high system complexity, the failure frequency and failure risk of underwater production systems are relatively high. In particular, the failure of key equipment often leads to serious economic losses and environmental hazards. To ensure the safe operation of underwater production systems, it is particularly important to comprehensively analyze and scientifically evaluate their fault modes and effects.

[0003] Existing system fault analysis methods mainly rely on traditional failure mode and effect analysis (FMEA) and expert experience-based evaluation methods. Although these methods can initially identify fault modes, they have the following main problems:

[0004] (1) Insufficient data utilization: Existing methods rely more on qualitative analysis by experts and fail to fully utilize the large amount of historical data accumulated during the operation of underwater production systems.

[0005] (2) Strong subjectivity: Expert opinions dominate the analysis, and the analysis results are easily affected by the expert experience level and subjective judgment, resulting in insufficient scientificity and objectivity of the results.

[0006] (3) Lack of dynamics in priority ranking: Traditional methods for priority ranking of fault modes fail to fully consider the objective weights of multi-dimensional data (such as failure incidence rate, production loss, etc.), resulting in the ranking results being difficult to comprehensively reflect system risks.

[0007] (4) Lack of pertinence in preventive suggestions: Due to the lack of a data-driven analysis mechanism, the suggestions for risk control by existing methods are often not specific and targeted enough.

[0008] In view of the above deficiencies, the present invention proposes a data-driven fault mode and effect analysis system for underwater production systems. By combining technical means such as historical data collection, statistical analysis, weight calculation, and priority ranking, the present invention can comprehensively mine and utilize historical data, scientifically analyze the fault modes of the system, provide dynamic and objective priority rankings and specific preventive measures, thereby significantly improving the safety and reliability of underwater production systems. Summary of the Invention

[0009] The object of the present invention is to provide a data-driven fault mode and effect analysis system for underwater production systems to solve the problems existing in the prior art.

[0010] Technical solution of the present invention:

[0011] A data-driven underwater production system failure mode and effect analysis system, comprising the following modules:

[0012] Module 1, historical data collection and storage module;

[0013] Collect and store multi-source historical data, check historical faults, and determine the main fault modes to be analyzed;

[0014] Module 2, historical data analysis module;

[0015] Based on the available multi-source historical data in the historical data collection and storage module, determine the failure rate, production loss, mean time to repair, and mean time between failures of the main fault modes, and form failure analysis data;

[0016] Module 3, data analysis index weight calculation module;

[0017] Based on the failure analysis data determined by the historical data analysis module, obtain the weights of the failure rate, production loss, mean time to repair, and mean time between failures through an objective weight calculation method, namely the method based on the standard removal effect method (MEREC method);

[0018] Module 4, fault mode ranking and analysis module;

[0019] Based on the output data of the historical data analysis module and the data analysis index weight calculation module, perform priority ranking on the main fault modes through a multi-criteria decision-making algorithm, namely the combined compromise sorting method (CoCoSo method), display the results, and recommend preventive measures.

[0020] The present invention discloses a brand-new underwater production system data processing and analysis system, which is completely data-driven and its function is to realize the failure mode analysis of the underwater production system. The proposed system includes four modules and main functional steps. First, the first module is used to collect, store, and detailly check historical faults, and output key fault modes; the second module is to determine data analysis indexes according to the available data, so as to form failure analysis data. Specifically, the present invention focuses on four main data analysis indexes: failure rate, production loss, mean time to repair, and mean time between failures. However, in actual applications, other data analysis indexes can be included according to more comprehensive historical data; next, module 3 uses an objective weight calculation method to obtain the weights of the data analysis indexes; module 4 then uses a multi-criteria decision-making algorithm to prioritize the fault modes, and finally outputs the results and recommends necessary preventive measures.

[0021] Advantages of the present invention:

[0022] (1) Comprehensive data mining and utilization

[0023] The present invention constructs a historical data collection and analysis module to comprehensively collect the operation data and fault records of the underwater production system, fully excavate and utilize the effective information in the data, and overcome the limitation of insufficient data utilization in the existing methods.

[0024] (2) Objectivity of weight calculation

[0025] The present invention calculates the weights of data analysis indicators (such as failure rate, production loss, etc.) using the standard removal effect method (MEREC method), significantly improving the scientificity and objectivity of system evaluation and avoiding the influence of expert subjective judgment in traditional methods.

[0026] (3) Multi-dimensional priority ranking

[0027] Using the combined compromise sorting method (CoCoSo method), the present invention realizes the priority ranking under multi-criteria decision-making, can comprehensively consider multiple key indicators, dynamically and scientifically determine the priority of failure modes, and improves the accuracy of evaluation and decision-making efficiency.

[0028] (4) Modular design, strong applicability

[0029] The system module of the present invention is designed flexibly, easy to expand and apply, can adjust the analysis indicators and methods according to different production scenarios, and is applicable to the failure mode analysis and safety management of various underwater production systems. Brief description of the drawings

[0030] Figure 1 is a schematic diagram of the data-driven underwater production system failure mode and effect analysis system provided by the present invention. Detailed implementation manners

[0031] The following details the specific structure and implementation process of the present solution through specific embodiments and drawings.

[0032] As Figure 1 shown, in an embodiment of the present invention, a data-driven underwater production system failure mode and effect analysis system is disclosed, which includes four components: a historical data collection and storage module, a historical data analysis module, a data analysis indicator weight calculation module, and a failure mode ranking and analysis module. The realization of the system functions has the following steps:

[0033] Module 1: The historical data collection and storage module is as follows:

[0034] (1) Data collection;

[0035] Collect multi-source historical data of the underwater production system, including but not limited to the following data types:

[0036] (1.1) Fault log data: including the fault occurrence time, fault description, involved equipment, and repair time;

[0037] (1.2) Production loss data: records of production interruption losses directly related to faults;

[0038] (1.3) System operation data: including the operation parameters of equipment (such as temperature, pressure, flow rate) and maintenance history records;

[0039] (2) Data storage;

[0040] The collected multi-source historical data is stored in a distributed database after preprocessing to ensure data integrity, consistency, and traceability;

[0041] (2.1) Data preprocessing includes data cleaning (removing redundant or incomplete data), standardization (unifying data formats), and denoising (filtering out abnormal data);

[0042] (2.2) Data storage is implemented using a NoSQL database, supporting large-scale data parallel processing and query optimization;

[0043] (3) Output results;

[0044] (3.1) Output the cleaned multi-source historical data to provide basic data support for subsequent modules;

[0045] (3.2) Output the main fault modes of the underwater production system determined after preliminary analysis of the available multi-source historical data.

[0046] Module 2: Historical data analysis module is as follows:

[0047] (1) Data index extraction;

[0048] Based on the available multi-source historical data stored in the historical data collection and storage module, for each fault mode to be analyzed, the following four core indicators are extracted:

[0049] (1.1) Fault occurrence rate: the ratio of the total number of fault occurrences to time, used to reflect the frequency of faults;

[0050] (1.2) Production loss: the economic loss directly caused by faults, expressed in monetary units;

[0051] (1.3) Mean time to repair: the average time required to repair each fault;

[0052] (1.4) Mean time between failures: the time interval between two consecutive faults, used to evaluate the reliability of system operation;

[0053] (2) Data analysis methods;

[0054] Data analysis is carried out according to the above four core indicators, and the specific steps are as follows:

[0055] (2.1) Failure rate OR: By counting the number of failures and the running time within a set time period (one year), calculate the failure rate of each failure mode as follows:

[0056]

[0057] (2.2) Production loss PL: According to the production interruption situation in the historical records, count the economic losses caused by each failure and calculate the average loss;

[0058] (2.3) Mean repair time MRT: Count the total repair time of all failures and divide it by the number of failures to obtain the mean repair time as follows:

[0059]

[0060] (2.4) Time between failures TBF: Count the time intervals between all adjacent failures and calculate the average value of these intervals as follows:

[0061]

[0062] (3) Data characteristic analysis;

[0063] (3.1) By analyzing the distribution characteristics of the time between failures, evaluate the reliability trend of the underwater production system operation;

[0064] (3.2) Combine visualization tools to generate histograms, trend charts and box plots to visually display the statistical characteristics of the indicators;

[0065] (4) Output results;

[0066] The output of the historical data analysis module includes four core indicators and the data analysis results under each failure mode, providing a quantitative basis for the weight calculation of the subsequent data analysis indicator weight calculation module and the failure ranking analysis of the failure mode ranking and analysis module; the generated analysis report can also be directly used for the performance evaluation of the underwater production system and the optimization of maintenance strategies.

[0067] Module 3: The data analysis indicator weight calculation module is as follows:

[0068] (1) Weight calculation method;

[0069] The data analysis index weight calculation module calculates the weights of each core index using the standard removal effect method based on the four core indexes output by the historical data analysis module; the standard removal effect method (MEREC method) determines the importance of an index by calculating the degree of influence on the overall evaluation result of the underwater production system after each core index is removed; the greater the influence, the higher the importance of the index; the specific steps are as follows:

[0070] (1.1) Based on the output result of the historical data analysis module, construct a decision matrix X = [x ij m×n , where x ij represents the evaluation value of the i-th failure mode under the j-th index. There are m failure modes and n indexes in total, and n = 4;

[0071] (1.2) Since the dimensions of each index may be different, standardize the decision matrix. The formula is:

[0072]

[0073] Among them, r ij represents the value of the i-th failure mode under the j-th index after standardization;

[0074] (1.3) Calculate the removal effect of each index; for each index, calculate the influence on the overall evaluation result of the underwater production system after it is removed, which is the removal effect E j . Its calculation formula is:

[0075]

[0076] Among them, E j is the removal effect of the j-th index;

[0077] (2) Weight calculation formula;

[0078] According to the removal effect E j determine the weight w j of the j-th index:

[0079]

[0080] (3) Output result;

[0081] Generate a weight distribution table for each core index, intuitively showing the influence weight of each core index on the failure of the underwater production system.

[0082] Module 4: Failure mode sorting and analysis module is as follows:

[0083] (1) Sorting method;

[0084] ​The failure mode ranking and analysis module, based on the output data of the historical data analysis module and the data analysis index weight calculation module, uses the combined compromise sorting method (CoCoSo method) to prioritize the main failure modes of the underwater production system:

[0085] For each failure mode, calculate its weighted sum S i and weighted geometric mean G i :

[0086]

[0087] where S i is the weighted sum of the i-th failure mode, reflecting its overall performance on all indicators; G i is the weighted geometric mean of the i-th failure mode, providing another perspective on overall evaluation;

[0088] Introduce three comprehensive scores representing three different combination strategies:

[0089]

[0090]

[0091] where and represent the score values based on the arithmetic mean, the score value relative to the worst, and the score value based on a balanced attitude, respectively; the value range of λ is 0 ≤ λ ≤ 1, and it is 0.5;

[0092] Based on the above scoring results, calculate the final score, and the formula is:

[0093]

[0094] where Φ i is the final score of the i-th failure mode. Rank the failure modes according to Φ i The larger the value, the higher the priority;

[0095] (2) Result display;

[0096] The priority ranking results are output in the form of tables and visual charts, showing the priorities of the failure modes and their comprehensive scores;

[0097] (3) Recommendation generation

[0098] Generate specific control recommendations based on the ranking results.

[0099] System operation instructions

[0100] 1. Data input

[0101] The user imports historical data into the system in a standardized format, and the system automatically completes data preprocessing and storage.

[0102] 2. Inter-module collaborative work

[0103] (1) Module 1 provides data input and storage.

[0104] (2) Module 2 extracts and analyzes data metrics.

[0105] (3) Module 3 calculates the weights of data analysis metrics.

[0106] (4) Module 4 outputs sorting results and control suggestions based on data and weights.

[0107] 3. Application scenarios

[0108] The present invention can be widely applied to the failure mode analysis and management of underwater production systems, such as deepwater oil and gas production, submarine facility operation and maintenance, and other scenarios.

[0109] Through the comprehensive analysis of historical data and the data-driven priority sorting method, the present invention innovatively realizes the automation and scientificization of the failure mode and effects analysis of underwater production systems. The four modules cooperate closely, which can provide efficient and reliable decision-making support for risk managers and significantly improve the safety and operation and maintenance efficiency of underwater production systems.

[0110] The present embodiment has the following beneficial effects compared with the prior art:

[0111] (1) Data-driven intelligent analysis

[0112] The prior art mainly relies on expert experience, while the present invention realizes the identification and analysis of failure modes in a data-driven manner, which can more comprehensively reflect the actual operation status of the system and reduce the interference of subjective factors at the same time.

[0113] (2) Dynamic weight calculation and sorting

[0114] Through the MEREC method and the CoCoSo method, the present invention is more dynamic and scientific in the weight allocation of data analysis metrics and the priority sorting of failure modes, overcoming the deficiency of single weight allocation in traditional methods.

[0115] (3) Comprehensive index coverage and flexibility

[0116] The present invention focuses on the analysis of four core metrics (failure incidence rate, production loss, mean time to repair, and mean time between failures), and at the same time supports the extension of other metrics, and can flexibly adjust the analysis content according to user needs.

[0117] (4) Visualization and intuitiveness of results

[0118] The present invention generates intuitive charts and analysis reports, presenting the sorting results of failure modes and their impact degrees, providing clear decision-making basis for users.

[0119] In summary, the present invention innovatively integrates data-driven analysis methods and intelligent priority sorting technologies, providing a brand-new technical solution for the safe operation and maintenance of underwater production systems, with significant academic value and engineering application value.

[0120] At this point, those skilled in the art should recognize that although multiple exemplary embodiments of the present invention have been shown and described in detail herein, many other variations or modifications that conform to the principles of the present invention can still be directly determined or derived based on the content disclosed in the present invention without departing from the spirit and scope of the present invention. Therefore, the scope of the present invention should be understood and determined to cover all these other variations or modifications.

Claims

1. A data-driven subsea production system failure mode and effect analysis system, characterized in that: The subsea production system failure mode and effect analysis system includes the following modules: Module 1, historical data collection and storage module; Collect and store multi-source historical data, examine historical failures, and identify the main failure modes to be analyzed; Module 2, historical data analysis module; Determine the failure occurrence rate, production loss, mean time to repair and time between failures of major failure modes based on the available multi-source historical data in the historical data collection and storage module to form failure analysis data; Module 3, data analysis indicator weight calculation module; According to the fault analysis data determined by the historical data analysis module, the weights of the occurrence rate, production loss, mean repair time and time between failures are obtained through an objective weight calculation method, i.e., based on the standard removal effect method; Module 4, failure mode sorting and analysis module; Based on the output data of the historical data analysis module and the data analysis index weight calculation module, the main failure modes are prioritized through a multi-criteria decision-making algorithm, namely the combined compromise ranking method, the results are displayed, and preventive measures are recommended.

2. The underwater production system failure mode and effect analysis system according to claim 1, characterized in that: Module 1: Historical data collection and storage module is as follows: (1) Data collection; Collect multi-source historical data of subsea production systems, including but not limited to the following data types: (1.1) Fault log data: including fault occurrence time, fault description, equipment involved, and repair time; (1.2) Production loss data: records of production interruption losses directly related to the failure; (1.3) System operation data: including equipment operating parameters and maintenance history records; (2) Data storage; The collected multi-source historical data is pre-processed and stored in a distributed database to ensure the integrity, consistency and traceability of the data; (2.1) Data preprocessing includes data cleaning, standardization and denoising; (2.2) Data storage is implemented using NoSQL databases, which support large-scale data parallel processing and query optimization; (3) Output results; (3.1) Output cleaned multi-source historical data to provide basic data support for subsequent modules; (3.2) Output The main failure modes of the subsea production system are determined after preliminary analysis of the available multi-source historical data.

3. The underwater production system failure mode and effect analysis system according to claim 2, characterized in that: Module 2: Historical data analysis module is as follows: (1) Extraction of data indicators; Based on the available multi-source historical data stored in the historical data collection and storage module, the following four core indicators are extracted for each failure mode to be analyzed: (1.1) Fault occurrence rate: the ratio of the total number of fault occurrences to the time, used to reflect the frequency of faults; (1.2) Production loss: economic loss directly caused by the failure, expressed in monetary units; (1.3) Mean time to repair: the average time required to repair each fault; (1.4) Failure interval: The time interval between two consecutive failures, used to evaluate the reliability of system operation; (2) Data analysis methods; According to the above four core indicators, the specific steps for data analysis are as follows: (2.1) Fault Occurrence Rate OR: By counting the number of faults and the operating time within a set time period, the fault incidence rate of each fault mode is calculated as follows: (2.2) Production loss PL: Based on the production interruption records in the history, the economic loss caused by each failure is counted and the average loss is calculated; (2.3) Mean repair time (MRT): The sum of all fault repair times is calculated and divided by the number of faults to obtain the mean repair time, as shown in the following formula: (2.4) Time Between Failures TBF: Count the time intervals between all two adjacent failures and calculate the average value of these time intervals, as shown in the following formula: (3) Data characteristics analysis; (3.1) Evaluate the reliability trend of the subsea production system by analyzing the distribution characteristics of the failure interval time; (3.2) Combine visualization tools to generate histograms, trend graphs, and box plots to intuitively display the statistical characteristics of indicators; (4) Output the results; The output of the historical data analysis module includes four core indicators and their data analysis results under each fault mode, which provide a quantitative basis for the weight calculation of the subsequent data analysis indicator weight calculation module and the fault sorting analysis of the fault mode sorting and analysis module; The generated analysis reports can also be directly used for subsea production system performance evaluation and maintenance strategy optimization.

4. The underwater production system failure mode and effect analysis system according to claim 1, characterized in that: Module 3: Data analysis indicator weight calculation module is as follows: (1) Weight calculation method; The data analysis indicator weight calculation module is based on the four core indicators output by the historical data analysis module and uses the standard removal effect method to calculate the weight of each core indicator; The standard removal effect method determines the importance of each core indicator by calculating the impact of each core indicator removed on the overall evaluation result of the subsea production system; the greater the impact, the higher the importance of the indicator; the specific steps are as follows: (1.1) Based on the output results of the historical data analysis module, the decision matrix X = [x ij ] m×n , where x ij It represents the evaluation value of the i-th fault mode under the j-th index. There are m fault modes and n indexes in total, n=4; (1.2) Since the dimensions of each indicator may be different, the decision matrix is ​​standardized and the formula is: Among them, r ij It represents the value of the i-th fault mode under the j-th index after standardization; (1.3) Calculate the removal effect of each indicator; for each indicator, calculate the impact of its removal on the overall evaluation result of the subsea production system, which is the removal effect E j , and its calculation formula is: Among them, E j is the removal effect of the jth indicator; (2) Weight calculation formula; According to the removal effect E j Determine the weight w of the jth indicator j : (3) Output results; Generate a weight distribution table for each core indicator to intuitively display the impact weight of each core indicator on the failure of the underwater production system.

5. The underwater production system failure mode and effect analysis system according to claim 1, characterized in that: Module 4: Failure Mode Sorting and Analysis Module is as follows: (1) Sorting method; The failure mode ranking and analysis module uses the combined compromise ranking method to prioritize the main failure modes of the underwater production system based on the output data of the historical data analysis module and the data analysis index weight calculation module: For each failure mode, calculate its weighted sum S i and weighted geometric mean G i : Among them, S i is the weighted sum of the i-th fault mode, reflecting its comprehensive performance in all indicators; G i It is the weighted geometric mean of the i-th failure mode, providing another comprehensive evaluation perspective; Introducing three comprehensive scores Three different combination strategies are represented: in, and They represent the score based on arithmetic mean, the score relative to the worst, and the score based on a balanced attitude; the value range of λ is 0≤λ≤1, which is 0.5; Based on the above scoring results, the final score is calculated using the formula: Among them, Φ i is the final score of the i-th fault mode, according to Φ i Sort the failure modes, the larger the value, the higher the priority; (2) Results presentation; The priority ranking results are output in the form of tables and visual charts, showing the priority of the failure mode and its comprehensive score; (3) Suggestion generation Generate specific management and control recommendations based on the sorting results.

Images