Data and knowledge combined driven underwater manifold maintainability design scheme optimization system

Abstract

The invention belongs to the technical field of underwater production system information processing and intelligent operation and maintenance, and discloses a data and knowledge combined driven underwater manifold maintainability design scheme optimization system. The underwater manifold maintainability design scheme optimization system comprises four core modules, namely a data acquisition and preprocessing module, an expert knowledge management and data fusion module, a maintainability index weight calculation module and a maintainability design scheme evaluation and optimization module. According to the optimal selection system for the maintainability design scheme of the underwater manifold, the maintainability index weight is calculated by adopting a GRA-EWM method, and the maintainability design scheme is comprehensively evaluated and optimized by utilizing an MABAC method, so that scientific, intelligent and accurate maintainability scheme optimization is realized.

Description

Technical Field

[0001] The present invention belongs to the technical field of information processing and intelligent operation and maintenance of underwater production systems, and relates to a system for optimizing the maintainability design scheme of an underwater manifold driven jointly by data and knowledge. Background Art

[0002] The underwater manifold is a key component of the offshore oil and gas production system, undertaking important tasks such as fluid distribution, pipeline connection, valve control, etc. Its design and manufacture directly affect subsequent maintenance management and safe operation. Due to the limitations of the deep - water environment, restricted human intervention, and high maintenance costs, maintainability indicators such as the maintainability, modular interchangeability, disassembly and assembly, maintenance safety, and human factors engineering of the underwater manifold directly affect its maintainability and reliability throughout the life cycle. However, the existing methods for optimizing the maintainability design scheme have the following deficiencies:

[0003] (1) Over - reliance on expert experience: Current maintainability assessments usually rely on subjective judgments of experts, with strong personal biases and lack of objective data support, affecting the reliability of the assessment.

[0004] (2) The decision - making method is not intelligent enough: Traditional decision - making methods, such as the Analytic Hierarchy Process (AHP), Weighted Mean Method (WMA), etc., cannot fully utilize data - driven technologies for precise optimization and are difficult to cope with complex environments.

[0005] (3) The calculation method of maintainability index weights is single: Traditional methods usually adopt a single weight calculation method, such as the subjective weight method or the objective weight method, and it is difficult to balance the influence of data - driven and expert knowledge.

[0006] (4) Lack of an intelligent - optimized model for selecting the maintainability design scheme: Existing methods fail to fully combine data analysis, expert knowledge, and multi - criteria decision - making optimization during comprehensive maintainability evaluation, resulting in insufficient scientificity of the final optimized scheme.

[0007] To solve the above problems, the present invention proposes a system for optimizing the maintainability design scheme of an underwater manifold driven jointly by data and knowledge, which realizes the scientific evaluation and intelligent optimization of the maintainability design scheme of the underwater manifold through data - driven + expert knowledge fusion + intelligent optimization decision - making. Summary of the Invention

[0008] The object of the present invention is to provide a system for optimizing the maintainability design scheme of an underwater manifold driven jointly by data and knowledge to solve the problems existing in the prior art.

[0009] The technical solution of the present invention:

[0010] A system for optimizing the maintainability design scheme of an underwater manifold driven jointly by data and knowledge includes the following modules:

[0011] Module 1, Data Acquisition and Preprocessing Module;

[0012] The data acquisition and preprocessing module is used to collect, store and process objective data related to different underwater pipeline assembly design schemes, including but not limited to: component visual angle (degree), operable space (m 3 ), maintenance path length (m), replaceable component ratio (%), number of standard interfaces (pcs), number of disassembly and assembly steps (steps), number of required special tools (pcs), required man-hours (h), historical maintenance accident rate (%), environmental risk parameters during maintenance (temperature, pressure), operation space size (m 3 ), human-machine interface complexity score (0 - 10), etc.; normalize the above collected objective data, use the Min-Max normalization method to convert data with different dimensions to the [0, 1] interval; use an SQL database for structured storage;

[0013] Module 2, Expert Knowledge Management and Data Fusion Module;

[0014] The expert knowledge management and data fusion module is used to collect, process and quantify the qualitative evaluation of experts on the maintainability design scheme of underwater pipeline assemblies, and fuse expert knowledge with objective data; invite domain experts to conduct fuzzy language evaluations on each maintainability index such as maintenance accessibility, modular interchangeability, disassembly and assembly, maintenance safety, and human factors engineering for each design scheme based on experience, and then convert it into quantitative data through fuzzy mathematics; among them, the fuzzy languages include very poor, poor, relatively poor, average, relatively good, good, very good; combine the quantitative data and the objective data output by the data acquisition and preprocessing module, calculate the score of each maintainability index, and thus construct an evaluation matrix of the maintainability design scheme;

[0015] Module 3, Maintainability Index Weight Calculation Module;

[0016] The maintainability index weight calculation module uses the Grey Relational Analysis - Entropy Weight Method (GRA - EWM) to calculate the weights of maintainability indexes; for the evaluation matrix obtained by the expert knowledge management and data fusion module, calculate the grey relational degree through Grey Relational Analysis (GRA) to obtain the corresponding weights, and then calculate the entropy values of each maintainability index through the Entropy Weight Method (EWM) to obtain the entropy weights; then take the average of the weights and entropy weights obtained by the two methods to obtain and output the weights of maintainability indexes;

[0017] Module 4, Maintainability Design Scheme Evaluation and Optimization Module;

[0018] The maintainability design scheme evaluation and optimization module uses the Multi-Attribute Border Approximation Area Comparison (MABAC) method to optimize the maintainability design scheme of the underwater pipeline assembly. According to the weights of the maintainability indicators output by the maintainability indicator weight calculation module, the evaluation matrix output by the expert knowledge management and data fusion module is weighted and standardized, and then the border approximation area is calculated, as well as the distances between each maintainability design scheme of the underwater pipeline assembly and the border approximation area. The maintainability design schemes of the underwater pipeline assembly are sorted according to the distance and the optimal maintainability design scheme of the underwater pipeline assembly is selected.

[0019] A brand-new underwater production system maintenance criticality analysis system includes four core modules: data collection and preprocessing, expert knowledge management and data fusion, maintainability indicator weight calculation, and maintainability design scheme evaluation and optimization module. The system uses the GRA-EWM method to calculate the weights of the maintainability indicators and uses the MABAC method for the comprehensive evaluation and optimization of the maintainability design scheme, so as to realize the scientific, intelligent, and accurate optimization of the maintainability scheme.

[0020] The beneficial effects of the present invention include:

[0021] (1) Deep integration of data and knowledge, improving the scientific nature of evaluation

[0022] Adopt the combination of expert fuzzy evaluation and objective data analysis to ensure the rationality and accuracy of the maintainability indicator evaluation.

[0023] (2) Use GRA-EWM to calculate the weights of maintainability indicators, improving the rationality of weight calculation

[0024] Grey relational analysis (GRA) determines the correlation degree of maintainability indicators with the ideal scheme and calculates the importance of indicators. The entropy weight method (EWM) calculates the information entropy of indicators and determines the objective weight driven by data. The combination of GRA-EWM makes the weight calculation more scientific and reasonable, avoiding the limitations of a single method.

[0025] (3) Use the MABAC method to optimize the maintainability scheme, improving the decision-making accuracy

[0026] The MABAC method can accurately calculate the distance of each scheme relative to the border approximation area, so as to scientifically evaluate the advantages and disadvantages of different maintainability schemes and finally determine the optimal scheme.

[0027] (4) Strong adaptability, suitable for complex underwater environments

[0028] This system can be applied to underwater pipeline assemblies with different sea areas and structures, improving the adaptability and generalization ability of the system. Description of the Drawings

[0029] Figure 1It is a schematic diagram of the optimal selection system for the maintainability design scheme of underwater manifolds jointly driven by data and knowledge provided by the present invention. Detailed implementation manners

[0030] The following uses specific embodiments and drawings to elaborate in detail on the specific structure and implementation process of this solution.

[0031] As Figure 1 shown, in an embodiment of the present invention, an optimal selection system for the maintainability design scheme of underwater manifolds jointly driven by data and knowledge is disclosed. The system adopts a modular design. The overall system includes five major modules: a data acquisition and preprocessing module, an expert knowledge management and data fusion module, a maintainability index weight calculation module, and a maintainability design scheme evaluation and optimization module. The various modules and functions of the system are described as follows:

[0032] Module 1: Data acquisition and preprocessing module

[0033] This module is used to collect, store, and process the objective data related to underwater manifolds to support subsequent maintainability evaluation and scheme optimization.

[0034] This system collects the following five categories of objective data according to the maintainability indexes. The data is sourced from the maintainability design scheme to be evaluated, as shown in the following table:

[0035] Table 1

[0036]

[0037] Through the formula process all the above data, and normalize it to [0, 1] for subsequent calculations.

[0038] Use an SQL database for structured storage to facilitate subsequent calculations and analyses.

[0039] Module 2: Expert knowledge management and data fusion module is as follows:

[0040] This module is used to collect, process, and quantify the qualitative evaluations of experts on the maintainability indexes of underwater manifolds, ensure that expert knowledge can be effectively converted into quantifiable data, and combine objective data for the optimization of maintainability design schemes.

[0041] (1) Expert knowledge acquisition;

[0042] Invite experts in the fields of offshore oil and gas engineering, mechanical design, and intelligent operation and maintenance. Based on years of engineering practice experience, evaluate the following maintainability indexes: maintainability accessibility, modular interchangeability, disassembly and assembly, maintenance safety, and human factors engineering;

[0043] Experts use the evaluation method of fuzzy language to evaluate each underwater pipeline repairability design scheme. The fuzzy language is divided into seven levels: very poor (VP), poor (P), relatively poor (RP), average (M), relatively good (RG), good (G), and very good (VG).

[0044] (2) Fuzzy mathematics quantification;

[0045] To convert the expert evaluation into computable quantitative data, the fuzzy mathematics method is adopted, that is, the fuzzy language level is converted into a triangular fuzzy number (TFN) as shown in the following table:

[0046]

[0047] (3) Fuzzy comprehensive evaluation model

[0048] (3.1) Establish the expert evaluation matrix

[0049] Suppose there are o experts evaluating m maintainability indicators, then the expert fuzzy evaluation matrix is constructed: S = [s kj o×m ;

[0050] where: s kj represents the evaluation of the jth maintainability indicator by the kth expert, expressed as a triangular fuzzy number;

[0051] (3.2) Calculate the fuzzy comprehensive score;

[0052] The weighted average method is used to calculate the fuzzy comprehensive evaluation value of each maintainability indicator:

[0053]

[0054] where: w k is the weight of the kth expert (by default, equal weights are used, and higher weights can be assigned to authoritative experts); s kj is the triangular fuzzy number provided by the expert;

[0055] (3.3) Defuzzification

[0056] The centroid method is used to convert the triangular fuzzy number into a clear value, and the calculation formula is as follows:

[0057]

[0058] where, (a, b, c) is the fuzzy comprehensive score of the maintainability indicator j; S j is the final score after defuzzification, ranging from [0, 1], and is respectively expressed as S 维修可达性 、S 模块化互换性 、S 拆卸装配性 、S 维修安全性 and S 人因工程 ;​

[0059] (4) Integration of expert knowledge and objective data;

[0060] For the scoring of each maintainability index, the objective data X and the calculated data S of the expert scoring are integrated, and the integration formula is as follows:

[0061] (4.1) Maintenance accessibility (Access):

[0062] X Access = 0.5·S 维修可达性 + 0.5·(0.4·X 可视角度 + 0.4·X 可操作空间 + 0.2·X 维修路径长度 )

[0063] Among them, X 可视角度 is the visible angle of the component, degree; X 可操作空间 is the operable space, m 3 ; X 维修路径长度 is the length of the maintenance path, m;

[0064] (4.2) Modular interchangeability (Modularity):

[0065] X Modularity = 0.5·S 模块化互换性 + 0.5·(0.6·X 可更换部件比例 + 0.4·X 标准接口数量 )

[0066] Among them, X 可更换部件比例 is the proportion of replaceable components, %; X 标准接口数量 is the number of standard interfaces, pieces;

[0067] (4.3) Disassembly and assembly (Disassembly):

[0068] X Disassembly = 0.5·S 拆卸装配性 + 0.5·(0.4·X 拆装步骤数量 + 0.3·X 专用工具数量 + 0.3·X 所需工时 )

[0069] Among them, X 拆装步骤数量 is the number of disassembly and assembly steps, steps; X 专用工具数量 is the number of required special tools, pieces; X 所需工时 is the required man-hours, h;

[0070] (4.4) Maintenance safety (Safety):

[0071] X Safety = 0.5·S 维修安全性+0.5·(0.6·X 历史事故率 +0.4·X 环境风险评分 )

[0072] where X 历史事故率 is the historical maintenance accident rate, %; X 环境风险评分 is the environmental risk score during the maintenance process, ranging from 0 to 10;

[0073] (4.5) Human Factors:

[0074] X Human Factors = 0.5·S 人因工程 +0.5·(0.6·X 操作空间大小 +0.4·X 人机界面复杂度评分 )

[0075] where X 操作空间大小 is the operation space size, m 3 ; X 人机界面复杂度评分 is the complexity score of the human-machine interface, ranging from 0 to 10. Module 3: Maintainability Index Weight Calculation Module is as follows:

[0076] (1) Calculate the index weights using GRA;

[0077] (1.1) Construct the original data matrix;

[0078] Suppose there are n maintainability design schemes, and each maintainability design scheme has m maintainability indicators. According to the evaluation matrix finally output by the expert knowledge management and data fusion module, construct the original decision matrix:

[0079] X = [x ij n×m

[0080] where x ij represents the score of the i-th maintainability design scheme on the j-th maintainability indicator;

[0081] (1.2) Data normalization:

[0082]

[0083] (1.3) Determine the ideal reference sequence;

[0084] To calculate the correlation degree of each maintainability indicator with the ideal scheme, it is necessary to construct the ideal reference sequence:

[0085] X 0 ′ = {x 01 ′, x 02 ′,..., x 0m ′} ​

[0086] Among them, represents the optimal value of the j-th maintainability index;

[0087] (1.4) Calculate the grey correlation coefficient;

[0088] The grey correlation coefficient is used to measure the closeness of a certain scheme to the ideal scheme, and the calculation formula is as follows:

[0089]

[0090] Among them: Δ ij = ∣X ij ′ - x 0j ′∣ is the absolute difference;

[0091] (1.5) Calculate the grey correlation degree:

[0092]

[0093] Among them: γ j represents the grey correlation degree of the maintainability index j; ξ ij is the grey correlation coefficient of the maintainability design scheme i on the maintainability index j;

[0094] (1.6) Calculate the GRA weight

[0095]

[0096] (2) EWM calculates the objective weight;

[0097] (2.1) Calculate the entropy value:

[0098]

[0099] Among them, is the normalized probability value;

[0100] (2.2) Calculate the entropy weight:

[0101]

[0102] Among them, represents the weight calculated by the entropy weight method, indicating the amount of information of the index;

[0103] (2.3) Calculate the final comprehensive weight:

[0104]

[0105] This module uses the Grey Relational Analysis-Entropy Weight Method (GRA-EWM) to calculate the weights of maintainability indicators, ensuring that the weight calculation can not only incorporate data-driven (EWM entropy weight method) but also reflect the relative relevance between indicators (GRA grey relational analysis).

[0106] Module 4: The evaluation and optimization module for maintainability design solutions is as follows:

[0107] This module uses the Multi-Attribute Border Approximation Area Comparison (MABAC) method to optimize maintainability design solutions. The specific steps are as follows:

[0108] (1) Construct a standardized decision matrix;

[0109] Using the weights calculated by GRA-EWM in the maintainability indicator weight calculation module, standardize all maintainability design solutions to obtain a weighted standardized decision matrix:

[0110] V = [v ij n×m , v ij = W j ·X ij ′

[0111] (2) Calculate the border approximation area matrix;

[0112] The border approximation area (BA area) is defined as:

[0113]

[0114] where G j represents the mean of maintainability indicator j, used to determine the optimal solution;

[0115] (3) Calculate the deviation value of each solution relative to the BA area;

[0116]

[0117] where D i > 0 indicates that solution i is superior to the BA area and is a better solution; D i = 0 indicates that solution i is close to the BA area and is an acceptable solution; D i < 0 indicates that solution i is inferior to the BA area and is a poorer solution;

[0118] (4) Solution ranking and optimal selection

[0119] Rank the solutions according to the magnitude of D i and output the recommended opinion on the optimal maintainability design solution.

[0120] 4. Solution ranking and optimal selection

[0121] ​Sort the solutions according to the size of D i and output the recommended opinions on the optimal maintainability design solution.

[0122] Through the above implementation manners, the present invention combines data-driven, expert knowledge fusion, and intelligent optimization to improve the scientificity of the maintainability assessment of underwater manifolds, optimize the design solution, improve the operation and maintenance efficiency, reduce the maintenance cost, and has broad engineering application value.

[0123] The present implementation manner has the following beneficial effects compared with the prior art:

[0124] 1. Fusion of data and knowledge to improve the scientificity of assessment

[0125] Existing maintainability assessment methods usually rely on the subjective experience of experts and are easily affected by personal cognition, resulting in insufficient objectivity of the assessment results. The present invention comprehensively combines objective data (historical maintenance records, operation monitoring data, design parameters) with expert knowledge (fuzzy language evaluation), quantifies through fuzzy mathematics methods, realizes the combination of qualitative and quantitative, and ensures the scientificity and consistency of the assessment.

[0126] 2. Adopt the GRA-EWM method to optimize the calculation of maintainability index weights

[0127] Traditional weight calculation methods, such as the analytic hierarchy process (AHP), have problems such as weight allocation relying on expert judgment and low calculation accuracy. The present invention adopts the grey relational analysis-entropy weight method (GRA-EWM). When calculating the weights, it not only considers the relative correlation between indicators (GRA), but also combines the objective variability of data (EWM), thereby improving the rationality and accuracy of weight calculation. This method makes the weight allocation of maintainability indicators more scientific and avoids the limitations of a single calculation method.

[0128] 3. Adopt the MABAC method to improve the accuracy of optimal selection of maintainability design solutions

[0129] Existing methods for optimal selection of maintainability design solutions (such as TOPSIS, VIKOR) often only consider the distance between the solution and the ideal solution during calculation and fail to fully utilize data and boundary information for optimization. The present invention adopts the multi-attribute boundary approximation area comparison method (MABAC). By calculating the distance between each solution and the boundary approximation area (BA), it ensures that the optimal solution is not only globally optimal but also stable under boundary conditions, improving the reliability and accuracy of decision-making. This method is more robust than traditional methods and is applicable to complex underwater manifold maintainability optimization problems.

[0130] 4. Strong adaptability and applicable to various underwater working conditions

[0131] The present invention can optimize the maintainability design of underwater pipe racks under various working conditions such as different water depths, environmental pressures, temperatures, and fluid characteristics. By adopting a modular architecture, it can be adjusted according to different application scenarios, enabling the system to have good scalability and adaptability.

[0132] 5. Improve the efficiency of maintainability optimization and reduce operation and maintenance costs

[0133] The traditional process of optimizing the maintainability of underwater pipe racks usually requires a large amount of manual calculation and expert demonstration, resulting in a long time, high cost, and strong subjectivity of the optimization scheme. The present invention greatly improves the efficiency of selecting the optimal maintainability scheme through automated calculation and intelligent optimization, reduces the dependence on experts, and makes the maintainability optimization process faster and more accurate. By using the MABAC method, the optimal maintainability design scheme can be screened out in a short time, which helps to reduce the maintenance cycle and operation and maintenance costs.

[0134] 6. Ensure the safety and reliability of underwater pipe rack operation and maintenance

[0135] By comprehensively considering key indicators such as historical failure data, environmental risk factors (temperature, pressure), maintenance accessibility, and maintenance safety, the present invention can provide a more reliable maintainability optimization scheme and improve the safety of underwater pipe racks. By adopting an optimization method based on objective data and intelligent decision-making, it reduces the possible safety risks during the maintenance process and improves the long-term operation stability of underwater pipe racks.

[0136] In summary, the present invention is significantly superior to the prior art in terms of data and knowledge integration, weight calculation optimization, accuracy of scheme selection, system adaptability, optimization efficiency, and safety. It provides a scientific, efficient, and intelligent solution for the maintainability optimization of underwater pipe racks and has important engineering application value.

[0137] At this point, those skilled in the art should recognize that although the present invention has been shown and described in detail with respect to multiple exemplary embodiments herein, many other variations or modifications that conform to the principles of the present invention can still be directly determined or derived from the disclosed content of 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 and knowledge-driven underwater manifold maintainability design scheme optimization system, characterized in that: Includes the following modules: Module 1, data acquisition and preprocessing module; The data acquisition and preprocessing module is used to collect, store and process objective data related to different underwater manifold design schemes, including but not limited to: component viewing angle, operable space, maintenance path length, proportion of replaceable components, number of standard interfaces, number of disassembly and assembly steps, number of required special tools, required working hours, historical maintenance accident rate, environmental risk parameters during maintenance, size of operating space, and human-machine interface complexity score; wherein, environmental risk parameters include temperature and pressure; human-machine interface complexity score is 0-10; the above collected objective data are normalized, and the Min-Max normalization method is used to convert data of different dimensions to the [0,1] interval; SQL database is used for structured storage; Module 2, expert knowledge management and data fusion module; The expert knowledge management and data fusion module is used to collect, process and quantify the qualitative evaluation of the experts on the underwater manifold maintainability design scheme, and integrate the expert knowledge with the objective data; invite domain experts to conduct fuzzy language evaluation on the maintainability indicators of maintainability accessibility, modular interchangeability, disassembly and assembly, maintenance safety, and ergonomics for each design scheme based on experience, and then convert them into quantitative data through fuzzy mathematics; among them, fuzzy language includes very poor, poor, relatively poor, average, relatively good, good, and very good; combine the quantitative data with the objective data output by the data acquisition and preprocessing module to calculate the score of each maintainability indicator, so as to construct the evaluation matrix of the maintainability design scheme; Module 3, maintainability index weight calculation module; The maintainability index weight calculation module uses the grey correlation analysis-entropy weight method to calculate the weight of the maintainability index; for the evaluation matrix obtained by the expert knowledge management and data fusion module, the grey correlation degree is calculated through grey correlation analysis to obtain the corresponding weight, and then the entropy value of each maintainability index is calculated through the entropy weight method to obtain the entropy weight; then the weights and entropy weights obtained by the two methods are averaged to obtain and output the weight of the maintainability index; Module 4, maintainability design scheme evaluation and optimization module; The maintainability design scheme evaluation and optimization module adopts the multi-attribute boundary approximate area comparison method to optimize the underwater manifold maintainability design scheme; according to the weight of the maintainability index output by the maintainability index weight calculation module, the evaluation matrix output by the expert knowledge management and data fusion module is weighted and standardized, and then the boundary approximation area and the distance between each underwater manifold maintainability design scheme and the boundary approximation area are calculated. The underwater manifold maintainability design schemes are sorted according to the distance and the optimal underwater manifold maintainability design scheme is selected.

2. The underwater manifold maintainability design scheme optimization system according to claim 1 is characterized in that: Module 2: Expert Knowledge Management and Data Fusion The details of the module are as follows: (1) Expert knowledge collection; Experts in offshore oil and gas engineering, mechanical design and intelligent operation and maintenance were invited to evaluate the following maintainability indicators based on years of engineering practice experience: maintainability accessibility, modular interchangeability, disassembly and assembly, maintainability safety, and human factors engineering; Experts use fuzzy language evaluation methods to evaluate each subsea manifold maintainability design scheme. The fuzzy language is divided into seven levels: very poor VP, poor P, relatively poor RP, average M, relatively good RG, good G, and very good VG. (2) Fuzzy mathematical quantification; In order to transform the expert evaluation into computable quantitative data, the fuzzy mathematics method is used, that is, the fuzzy language level is converted into triangular fuzzy number TFN, as shown in the following table: (3) Fuzzy comprehensive evaluation model (3.1) Establishing an expert evaluation matrix Assume that o experts evaluate m maintainability indicators, then construct the expert fuzzy evaluation matrix: S = [s kj ] o×m ; Where: s kj represents the evaluation of the kth expert on the jth maintainability index, expressed by triangular fuzzy numbers; (3.2) Calculate the fuzzy comprehensive score; The weighted average method is used to calculate the fuzzy comprehensive evaluation value of each maintainability index: Where: w k is the weight of the kth expert; s kj Triangular fuzzy numbers for experts; (3.3) Defuzzification The centroid method is used to convert triangular fuzzy numbers into clear values. The calculation formula is as follows: Among them, (a, b, c) is the fuzzy comprehensive score of maintainability index j; S j is the final score after defuzzification, ranging from [0,1], denoted as S 维修可达性 , S 模块化互换性 , S 拆卸装配性 , S 维修安全性 and S 人因工程 ; (4) Integration of expert knowledge and objective data; For the score of each maintainability index, the objective data X and the calculated data S of the expert score are fused, and the fusion formula is as follows: (4.1) Maintenance accessibility: X Access =0.5·S 维修可达性 +0.5·(0.4·X 可视角度 +0.4·X 可操作空间 +0.2·X 维修路径长度 ) Among them, X 可视角度 is the component viewing angle, degree; X 可操作空间 is the operable space, m 3 ;X 维修路径长度 is the maintenance path length, m; (4.2) Modular interchangeability: X Modularity =0.5·S 模块化互换性 +0.5·(0.6·X 可更换部件比例 +0.4·X 标准接口数量 ) Among them, X 可更换部件比例 is the proportion of replaceable parts, %; X 标准接口数量 is the number of standard interfaces, pcs; (4.3) Disassembly and assembly: X Disassembly =0.5·S 拆卸装配性 +0.5·(0.4·X 拆装步骤数量 +0.3·X 专用工具数量 +0.3·X 所需工时 ) Among them, X 拆装步骤数量 is the number of disassembly and assembly steps, steps; X 专用工具数量 is the number of special tools required, pieces; X 所需工时 is the required working hours, h; (4.4) Maintenance safety: X Safety =0.5·S 维修安全性 +0.5·(0.6·X 历史事故率 +0.4·X 环境风险评分 ) Among them, X 历史事故率 is the historical maintenance accident rate, %; X 环境风险评分 Score the environmental risk during the maintenance process on a scale of 0-10; (4.5) Human Factors Engineering: X Human Factors =0.5·S 人因工程 +0.5·(0.6·X 操作空间大小 +0.4·X 人机界面复杂度评分 ) Among them, X 操作空间大小 is the size of the operating space, m 3 ;X 人机界面复杂度评分 Score the complexity of the human-computer interface on a scale of 0-10.

3. The underwater manifold maintainability design scheme optimization system according to claim 2 is characterized in that: The maintainability index weight calculation module is as follows: (1) GRA calculates indicator weights; (1.1) Construct the original data matrix; Assuming there are n maintainability design schemes, each maintainability design scheme has m maintainability indicators. According to the evaluation matrix finally output by the expert knowledge management and data fusion module, the original decision matrix is ​​constructed: X=[x ij ] n×m Among them, x ij represents the score of the i-th maintainability design scheme on the j-th maintainability index; (1.2) Data normalization: (1.3) Determine the ideal reference sequence; In order to calculate the relevance of each maintainability index to the ideal solution, it is necessary to construct an ideal reference sequence: X0′={x 01 ′,x 02 ′,...,x 0m ′} in, represents the optimal value of the j-th maintainability index; (1.4) Calculate the grey correlation coefficient; The grey correlation coefficient is used to measure the closeness of a solution to the ideal solution. The calculation formula is as follows: Where: Δ ij =|X ij ′-x 0j ′∣ is the absolute difference; (1.5) Calculate the grey relational degree: Where: γ j Represents the grey relational degree of maintainability index j; ξ ij is the grey correlation coefficient of maintainability design scheme i on maintainability index j; (1.6) Calculate GRA weight (2) EWM calculates objective weights; (2.1) Calculate the entropy value: in, is the normalized probability value; (2.2) Calculate entropy weight: in, Represents the weight calculated by the entropy weight method, indicating the amount of information in the indicator; (2.3) Calculate the final comprehensive weight:

4. The underwater manifold maintainability design scheme optimization system according to claim 5 is characterized in that: The maintainability design scheme evaluation and optimization modules are as follows: (1) Construct a standardized decision matrix; The weights calculated by GRA-EWM in the maintainability index weight calculation module are used to standardize all maintainability design schemes to obtain the weighted standardized decision matrix: V=[v ij ] n×m ,v ij =W j ·X ij ′ (2) Calculate the boundary approximation region, i.e., the BA region matrix; The boundary approach region is defined as: Among them, G j represents the mean value of the maintainability index j, which is used to determine the optimal solution; (3) Calculate the offset value of each scheme relative to the BA area; Among them, D i >0 means that solution i is better than area BA and is a better solution; D i =0 means that solution i is close to the BA area and is an acceptable solution; D i <0 means that solution i is inferior to the BA region and is a poor solution; (4) Scheme ranking and optimal selection According to D i The schemes are sorted according to their size and the optimal maintainability design scheme recommendation is output.

Images