Risk Assessment Method for Deep-Sea Mining Collector System Based on FMEA-IVIFS-MARCOS

Abstract

The present invention belongs to the technical field of deep - sea mining risk identification and risk assessment, and particularly relates to a risk assessment method for a deep - sea mining collector system based on FMEA - IVI FS - MARCOS, which is characterized by the following steps: S1: Construct a risk assessment index system for the deep - sea mining collector system; S2: Establish a risk evaluation index system and quantify evaluation information; S3: Establish a deep - sea mining risk evaluation model; S4: Conduct risk ranking based on the MARCOS method. The present invention provides a reasonable and feasible risk assessment method, provides important technical support for solving the safety hazards faced by the collector system in the deep - sea mining process, and improves the safety and reliability of the collector system. This method is also applicable to other reliable engineering fields of complex systems and has broad application scenarios.

Description

Technical Field

[0001] The present invention belongs to the technical field of deep - sea mining risk identification and risk assessment, and particularly relates to a risk assessment method for a deep - sea mining collector system based on FMEA - IVIFS - MARCOS. Background Technique

[0002] Deep - sea mining is an extremely challenging task. The deep - sea mining collector system is the core unit of the overall mining operation, undertaking the most complex and dangerous ore - collecting task on the seabed. The deep - sea mining collector system integrates multiple technologies such as machinery, hydraulics, and control, including three major parts: the ore - collecting system, the traveling mechanism, and the hydraulic system. It is a complex hydraulic - powered mechanical system that works underwater at a depth of several thousand meters. Its working environment is harsh, facing various extreme potential factors such as high pressure, low temperature, and component failure. Once a failure occurs, it may lead to the paralysis of the entire deep - sea mining system, bringing incalculable losses. The safe operation of the deep - sea mining collector system is crucial for the stable and reliable operation of deep - sea mining. Therefore, it is urgent to carry out accurate risk assessment of the deep - sea mining collector system under the combined action of multiple factors, reduce working risks, and improve safety and reliability. At present, the research on the safety assessment of the deep - sea mining collector system is relatively scarce, and there is insufficient understanding of the risk information of the collector system and the complex relationships between its internal structures. No comprehensive risk analysis method for the overlapping interaction of multiple risk factors has been carried out, resulting in relatively low reliability of its risk assessment results and making it difficult to eliminate or mitigate risks at the root. The risk assessment of the deep - sea mining collector system involves multiple risk - characterization parameters and belongs to a multi - criterion comprehensive evaluation problem. Given that the data sources of the deep - sea mining collector system are not comprehensive enough, it faces the challenges of dealing with insufficient information and complex factors. Traditional risk assessment methods require the establishment of accurate assessment models and have certain limitations in terms of application scope and limitations. Summary of the Invention

[0003] The purpose of the present invention is to provide a risk assessment method for a deep - sea mining collector system based on FMEA - IVIFS - MARCOS to solve the problems existing in the background technique.

[0004] To achieve the above - mentioned purpose, the present invention provides the following technical solution: A risk assessment method for a deep - sea mining collector system based on FMEA - IVIFS - MARCOS, including the following steps:

[0005] S1: Construct a risk assessment index system for deep-sea mining collector systems; for the complex internal structure and potential multi-source risk threats of deep-sea mining collector systems, according to the principle of structural hierarchy classification, adopt the Failure Mode and Effects Analysis (FMEA) method, decompose according to the overall structural framework of the deep-sea mining collector system layer by layer until the components at the lowest level, so as to conduct systematic and structured risk identification, summarize its functional characteristics and consequence impacts, determine potential failure modes and weak links, and construct a risk assessment index system for deep-sea mining collector systems.

[0006] S2: Establish a risk evaluation index system and quantify evaluation information; in view of the lack of relevant historical accident data, it is necessary to rely on expert evaluation to improve incomplete information. Therefore, invite experts with rich experience in related fields to form an expert group, and use risk characterization parameters, that is, the probability of occurrence of failure mode (O) and consequence severity (S) as evaluation indicators to evaluate the failure modes of deep-sea mining collector systems. Considering the characteristics of uncertainty and hesitation of evaluation information, according to the established risk characterization parameter rating standard, use interval intuitionistic fuzzy to obtain and express interval semantic evaluation information, and reduce the fuzziness and subjectivity of evaluation information.

[0007] Table 1 Conversion relationship between risk characterization parameters and intuitionistic fuzzy numbers

[0008]

[0009] S3: Establish a risk evaluation model for deep-sea mining collection; after obtaining the scores of experts on failure modes, considering the fuzziness of risk characterization parameters themselves and the relationships between parameters, calculate the fuzzy number distance between O and S according to the deviation maximization weighting method to determine their objective weights. At the same time, measure the importance and support degree among experts through the Jousselme distance function to determine the objective weights of experts. To avoid simply weighting the evaluation results and solve the problem of information loss caused by evidence conflict, introduce D-S (Dempster-Shafer) evidence fusion expert evaluation, and combine the MARCOS method to determine the utility function of each failure mode given, so as to determine its risk level.

[0010] S4: Conduct risk ranking based on the MARCOS method.

[0011] Preferably, in S3, it includes the following steps:

[0012] 1) Determine the intuitionistic fuzzy evaluation matrix

[0013] For the multi-risk index evaluation problem of deep-sea mining collector systems, assume that there are / experts to form an expert set EX = {ex1, ex2,..., ex l}, there are m fault modes forming the evaluation set FF={ff1,ff2,…,ff m}, each failure mode has n risk characterization parameters forming the attribute set RF = {rf1, rf2, ..., rf n}, then / experts have a good understanding of the failure mode ff i The initial intuitionistic fuzzy evaluation matrix IA of the risk characterization parameter r is r :

[0014]

[0015] Where r = {O, S}, is the membership of the / th expert to the mth failure mode under the risk characterization parameter r, is the non-membership degree of the / th expert’s score on the mth failure mode under the risk characterization parameter r;

[0016] 2) Calculate the weights of risk characterization parameters

[0017] Integrate the evaluation information of each expert on the failure mode, use the maximum deviation principle of attributes, construct a nonlinear programming model for risk characterization parameters, calculate the difference between parameters, obtain the risk assessment index weight vector and determine the weight;

[0018]

[0019] Where λ represents the number of fuzzy numbers, Is an expert ex γ Risk characterization parameter RF j Lower failure mode FF i The deviation of the evaluation values from other failure modes, the greater the parameter value deviation, the greater the weight value assigned to it;

[0020] Introducing DS evidence theory, as shown in formula (3), the intuitive fuzzy number of the failure mode is converted into the basic probability distribution, and the basic probability distribution matrix of the failure mode with respect to the risk characterization parameter is obtained. ij ) m×l ,in is the basic probability distribution of the j-th expert's evaluation of the risk characterization parameters under the i-th failure mode.

[0021]

[0022] in, is the basic probability distribution of the identification framework;

[0023] 4) Calculate expert weights and weighted fusion expert evaluations

[0024] Measure the importance and ambiguity among expert information, and construct the mutual support degree among experts and obtain the weights based on the similarity of experts to be fused according to the Jousselme distance formula, as follows:

[0025]

[0026] Among them, Cred(ex l ) represents the credibility of expert ex l , that is, the expert weight, represents the difference in the basic belief assignment function of experts ex i and ex j for the m-th fault mode. is the Jaccard coefficient, which represents their similarity degree by the ratio of the intersection and union of the focal elements ex i and ex j ;

[0027] Then, substitute the calculated expert weights into formula (6) to perform weighted correction on the basic probability distribution of the fault mode. Finally, fuse the corrected values l - 1 times according to the D-S evidence theory to obtain the fused result. The specific correction and fusion method is shown in formulas (6) - (8):

[0028]

[0029] Among them, m MAE (EXM) is the basic probability distribution after expert weighted correction, is called the orthogonal sum, which is the result of fusing the information of / experts. K is the normalization constant, representing the conflict degree between two experts. The larger the K value, the higher the conflict degree between the evidences.

[0030] Preferably, in S4, by defining the ideal solution (AI) and the anti-ideal solution (AAI), the result of the information after expert fusion is extended to obtain the extended evaluation matrix R':

[0031]

[0032] Among them, the fuzzy negative ideal solution scheme AAI is the scheme with the worst characteristics, while AI is the alternative scheme with the best characteristics.

[0033] Then, perform normalization processing on the matrix R' to obtain the normalized matrix N. Multiply the normalized matrix N by the weight coefficient of the risk characterization parameter to obtain the weighted matrix V = [v ij m×n , v ij represents the evaluation value of the j-th risk characterization parameter of the i-th fault mode;

[0034] ​

[0035] Calculate the utility function of the failure mode through formulas (11) to (13), and sort the failure modes according to the values of the utility function. The higher the value of the utility function means the higher the risk level of the risk factor;

[0036]

[0037] Among them, S i represents the sum of the elements of the weighted matrix V, where S ai represents the sum of the elements in the row where the ideal solution ai is located in the weighted matrix V; S aai represents the sum of the elements in the row where the negative ideal solution aai is located in the weight matrix V; f(K i + ) and f(K i - ) represent the utility functions related to the ideal solution and the negative ideal solution respectively.

[0038] The beneficial effects of the present invention are as follows: The present invention provides a multi-attribute decision-making method for determining positive and negative ideal solutions based on interval intuitionistic fuzzy sets, which combines the reference point ranking method and the ratio method to handle the uncertainty and hesitation of evaluation information under different weights, so as to accurately evaluate the failure mode. Provide short-term risk forecasts for the deep-sea mining collector system, and provide a scientific basis for the elimination, reduction and control of risks, thereby optimizing the operation and maintenance plan of the deep-sea mining collector system and improving the reliability and stability of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 is the technical roadmap of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0040] The following further describes the specific embodiments of the present invention in detail with reference to the accompanying drawings and preferred embodiments.

[0041] The present invention proposes a risk assessment method for a deep - sea mining collector system based on the Failure Mode and Effects Analysis - Interval - valued intuitionistic fuzzy numbers - Measurement of Alternative and Ranking to Compromise Solution (FMEA - IVIFS - MARCOS). The FMEA method is used to identify the failure modes of the deep - sea mining collector system and construct a comprehensive evaluation index system. The IVIFS is used to reflect the risk preferences of experts, and quantify the membership and non - membership tendencies of experts towards failure modes. The deviation maximization weighting method and Jousselme distance are used to determine the risk characterization parameters and the objective weights of experts respectively. On this basis, the D - S (Dempster - Shafer) evidence theory is used to weight and fuse the expert evaluation information, and by combining the Measurement of Alternative and Ranking to Compromise Solution (MARCOS) method, a risk assessment model of the deep - sea mining collector system based on FMEA - IVIFS - MARCOS is established to calculate the utility values of each failure mode, so as to rank the risk values of the failure modes. Taking 15 failure modes of the deep - sea mining collector system as an example, the specific technical route is as Figure 1 shown, including the following steps:

[0042] Step 1: Failure mode identification

[0043] Aiming at the complex internal structure characteristics of the deep - sea mining collector system and the threat of potential multi - source risks, based on the structural - hierarchical classification, the FMEA method is adopted. The overall structure of the deep - sea mining collector system is decomposed hierarchically until the components at the lowest level, so as to conduct systematic and structured risk identification, determine the potential failure modes and weak links, and identify 15 main failure modes, as shown in Table 2.

[0044] Table 2 Failure modes of the deep - sea mining collector system

[0045]

[0046] Step 2: Establish a risk evaluation index system and quantify the evaluation information

[0047] Invite 5 researchers and experts in the field of deep - sea mining engineering as evaluators. Combining their own experience and knowledge with relevant literature and standards of deep - sea mining, use linguistic variables to evaluate O and S of failure modes. The risk evaluation levels are divided into very high, high, medium, low, and very low, and the risk hesitation degree π is divided into 0.1, 0.2, 0.3, 0.4. Given that most experts have rich experience, hesitation degrees higher than 0.4 are not considered. The failure mode evaluation form is shown in Table 3 as follows.

[0048] Table 3 Semantic evaluation information of the expert group on failure modes

[0049]

[0050] Step 3: Quantify the evaluation results and evaluate the failure mode risks based on the IVIFS - MARCOS model

[0051] 1) Quantify the failure mode

[0052] The present invention adopts interval - valued intuitionistic fuzzy numbers to transform the expert evaluation results, dealing with the subjective uncertainty of experts' own evaluations. According to the transformation criteria in Table 1, the initial intuitionistic fuzzy evaluation matrix IA of the risk characterization parameters O and S is obtained. r As follows:

[0053]

[0054] 2) Calculate the weights of risk characterization parameters

[0055] The deviation maximization weight - assignment method is used to calculate the fuzzy number deviations of O and S under each failure mode. According to formula (2), take the maximum deviation to determine the weights of the two as 0.50921 and 0.49079 respectively.

[0056] 3) Transform intuitionistic fuzzy numbers

[0057] Define that 15 failure modes form a finite non - empty set, which is the identification framework Θ. The basic probability assignment value on the identification framework is the support degree for each failure mode. In order to utilize the D - S evidence theory, it is necessary to convert the interval - valued intuitionistic fuzzy numbers into basic probability values according to formula (3), as shown in Table 4.

[0058] Table 4 Basic probability assignment of failure modes

[0059]

[0060] 4) Calculate the expert weights

[0061] When calculating the weights of risk characterization parameters, the experts' hesitation and uncertainty regarding their evaluation information are already taken into account. Therefore, when calculating expert weights, the degree of conflict between expert evaluation information also needs to be considered. This paper uses the Josselme distance to measure the similarity and support of expert evaluation information. The weights of the five experts under O and S are calculated according to formulas (4) to (5). Taking the occurrence degree as an example, the specific expert evaluation information distances are shown in Table 5. At the same time, based on the risk characterization parameter weights, the expert weights are recombined to obtain 0.20032, 0.19827, 0.20052, 0.20072, and 0.20017.

[0062] Table 5 Expert evaluation information distance (occurrence degree)

[0063]

[0064] 5) Weighted fusion of expert evaluation to form an extended initial matrix

[0065] According to the obtained expert weights, the basic probability distribution of the fault mode is weighted and fused through formulas (6) to (8) to obtain the basic probability distribution of the fault mode O and S. Since the fault mode identification framework covers all fault modes, but these fault modes are mutually exclusive during the evaluation process, the basic probability distribution of each fault mode after fusion is renormalized and the results are aggregated into a multi-criteria decision evaluation matrix. The ideal solution and anti-ideal solution are introduced through formula (9) for expansion. The ideal solution takes the maximum value of O and S, and the anti-ideal solution takes the minimum value of O and S. The multi-criteria aggregate evaluation of the fault mode is obtained as follows:

[0066] Table 6 Multi-standard aggregation evaluation table

[0067]

[0068] Then, through formulas (10) to (13), the analysis data of the failure mode under multiple evaluation indicators are obtained, as shown in Table 7.

[0069] Table 7 Multi-criteria analysis results of failure modes

[0070]

[0071] The analysis in Table 7 shows that the top three failure modes are motor failure, oil leakage, and damage to the plug-in self-locking mechanism. These three failure modes have a high risk level and are more likely to affect the system, thereby causing failures in the deep-sea mining system. We should focus on these failure modes and take risk control measures such as repairs or replacements and upgrades in a timely manner to reduce their occurrence.

[0072] It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention, and these improvements and modifications should also be considered as the scope of protection of the present invention.

Claims

1. A risk assessment method for a deep - sea mining collector system based on FMEA - IVIFS - MARCOS, characterized in that : It includes the following steps: S1: Construct a risk assessment index system for deep-sea mining collector systems; S2: Establish a risk evaluation index system and quantify evaluation information; S3: Establish a risk evaluation model for deep-sea mining collectors; S4: Conduct risk ranking based on the MARCOS method; In S2, invite experts with rich experience in relevant fields to form an expert group. Using risk characterization parameters, namely the likelihood of failure mode occurrence (O) and consequence severity (S), as evaluation indicators, evaluate the failure modes of deep-sea mining collector systems. Considering the uncertainty and hesitancy of evaluation information, etc., according to the established risk characterization parameter rating standard, use interval intuitionistic fuzzy to obtain and express interval semantic evaluation information, reducing the fuzziness and subjectivity of evaluation information; In S3, it includes the following steps: 1) Determine the intuitionistic fuzzy evaluation matrix For the multi-risk index evaluation problem of deep-sea mining collector systems, assume there is an expert set EX = {ex1, ex2, …, ex L} composed of L experts, and an evaluation set FF = {ff1, ff2, …, ff m} composed of m failure modes. Under each failure mode, there is an attribute set RF = {rf1, rf2, …, rf n} composed of n risk characterization parameters. Then the initial intuitionistic fuzzy evaluation matrix IA i of the L experts for the risk characterization parameter r under the failure mode ff r is as follows: Among them, the risk characterization parameter r comes from the risk attribute set RF and is used to measure the characteristic attributes of the failure mode under different risk dimensions. It is defined that r = {O, S}, representing the occurrence probability and severity respectively. is the membership degree of the L-th expert's score for the m-th failure mode under the risk characterization parameter r. is the non-membership degree of the L-th expert's score for the m-th failure mode under the risk characterization parameter r. 2) Calculate the weights of risk characterization parameters Integrate the evaluation information of each expert on the failure mode. Using the maximum deviation principle of attributes, construct a nonlinear programming model for risk characterization parameters, calculate the differences between parameters, and obtain the weight vector of risk evaluation indicators to determine the weights; where λ represents the number of fuzzy numbers, is the expert ex γ for the risk characterization parameter RF j the deviation of the evaluation value between the failure mode FF i and other failure modes. The larger the parameter value deviation, the greater the weight value assigned to it; Introduce the D-S evidence theory. As shown in formula (3), convert the intuitionistic fuzzy numbers of the failure modes into basic probability distributions to obtain the basic probability assignment matrix R=(r ij ) m×l , where is the basic probability distribution of the j-th expert's evaluation of the risk characterization parameter under the i-th failure mode; Among them, is the basic probability distribution of the identification framework; 4) Calculate expert weights and perform weighted fusion of expert evaluations Measure the importance and fuzziness between expert information, and based on the similarity of experts to be fused, according to the Jousselme distance formula, construct the mutual support degree between experts and obtain the weights as follows: Among them, Cred(ex l ) represents the credibility of expert ex l , that is, the expert weight, represents the difference in the basic belief assignment function of expert ex i and ex j for the m-th failure mode, is the Jaccard coefficient, which represents their similarity degree by the ratio of the intersection and union of the focal elements ex i and ex j ; Then, substitute the calculated expert weights into formula (6) to perform weighted correction on the basic probability distribution of the failure mode. Finally, fuse the corrected values L - 1 times according to the D - S evidence theory to obtain the fused result. The specific correction and fusion methods are shown in formulas (6) - (8): where EXM represents the evaluation result after the integration of multiple experts, and m MAE (EXM) is the basic probability distribution after the weighted correction by experts, which is called the orthogonal sum and is the result of the information fusion of L experts. K is the normalization constant, representing the conflict degree between two experts. The larger the value of K, the higher the conflict degree between the evidences.

2. The risk assessment method for the deep-sea mining collector system based on FMEA-IVIFS-MARCOS according to claim 1, wherein: In S1, adopt the Failure Mode and Effects Analysis (FMEA) method. Decompose hierarchically according to the overall structural framework of the deep-sea mining collector system until the components at the lowest level, so as to conduct systematic and structured risk identification, summarize its functional characteristics and consequence impacts, determine potential failure modes and weak links, and construct a risk assessment index system for deep-sea mining collector systems.

3. The risk assessment method for deep-sea mining collector system based on FMEA-IVIFS-MARCOS according to claim 1, wherein: In S4, by defining the ideal solution (AI) and the anti-ideal solution (AAI), expand the result of the information after expert fusion to obtain the expanded evaluation matrix R′: Among them, the fuzzy negative ideal solution scheme AAI is the scheme with the worst characteristics, while AI is the alternative with the best characteristics. Represents the basic probability assignment after correction of the scheme with the worst characteristics under the risk characterization parameter O. Represents the basic probability assignment after correction of the scheme with the worst characteristics under the risk characterization parameter S. Then, perform normalization processing on the matrix R' to obtain the normalized matrix N, and multiply the normalized matrix N by the weight coefficient of the risk characterization parameter to obtain the weighted matrix V = [v ij m×n , v ij represents the evaluation value of the j-th risk characterization parameter of the i-th failure mode;​ Calculate the utility function of the failure mode through formulas (11) - (13), and rank the failure modes according to the value of the utility function. The higher the value of the utility function, the higher the risk level of the risk factor; Among them, S i represents the sum of the elements of the weighted matrix V, where S ai represents the sum of the elements in the row where the ideal solution ai is located in the weighted matrix V; S aai represents the sum of the elements in the row where the negative ideal solution aai is located in the weight matrix V; f(K i + ), f(K i - ) represent the utility functions related to the ideal solution and the negative ideal solution respectively.

Images