Safety assessment method for deep-sea mining surface support ship based on fmea-bn model

Abstract

This invention belongs to the field of deep-sea mining risk identification and risk assessment technology, specifically involving a safety assessment method for deep-sea mining surface support vessels based on the FMEA-BN model. The method includes the following steps: S1: Identifying potential risk factors for deep-sea mining surface support vessels; S2: Fuzzy rating of risk factors; S3: Constructing the FMEA-BN model; S4: Ranking the risk factors. This method evaluates risk factors even with limited data sources, thereby comprehensively and objectively reflecting the safety level of deep-sea mining surface support vessels and providing theoretical and technical support for the safety of deep-sea mining systems.

Description

Technical Field

[0001] This invention belongs to the field of deep-sea mining risk identification and risk assessment technology, specifically involving a method for assessing the safety of deep-sea mining surface support vessels based on the FMEA-BN model. Background Technology

[0002] Surface support vessels are fundamental equipment for deep-sea mineral resource development, primarily responsible for the deployment and retrieval of underwater equipment and the temporary storage and transfer of minerals. As a new type of vessel operation, surface support vessels differ significantly from traditional vessels in overall performance. They are structurally complex, highly integrated, and span multiple disciplines, resulting in diverse and heterogeneous performance data with small sample sizes. Furthermore, due to the coupled nature of multi-body mining equipment in deep-sea mining operations, surface support vessels are not only affected by wind, waves, and currents but also by the interconnected effects of other systems. Therefore, the failure modes of surface support vessels are complex, varied, and significantly correlated. The cascading effects triggered by a single failure mode can lead to system performance degradation or even loss of control, potentially causing accidents. To ensure the robustness and reliability of deep-sea mining operations, a comprehensive risk assessment of deep-sea mining surface support vessels is urgently needed. Currently, there is limited analysis on surface support vessels. The reliability and safety analysis of deep-sea mining surface support vessels by various countries is basically limited to stability, safety analysis, and overall longitudinal strength. There is little discussion on other key factors of safety and reliability, and there is a lack of multi-factor and systematic consideration. As a result, the reliability of deep-sea mining surface support vessels has not been fully verified, and the results have certain limitations. Summary of the Invention

[0003] The purpose of this invention is to provide a safety assessment method for deep-sea mining surface support vessels based on the FMEA-BN model, so as to solve the problems existing in the background art.

[0004] To achieve the above objectives, the present invention provides the following technical solution: a method for safety assessment of deep-sea mining surface support vessels based on the FMEA-BN model, comprising the following steps:

[0005] S1: Identifying potential risk factors for deep-sea mining surface support vessels; deep-sea mining surface support vessels undertake multiple tasks, including operations, navigation, and accommodation, and the complexity of their systems and the close interrelationships between these tasks increase the difficulty of analysis. Therefore, this invention combines multi-level comprehensive evaluation and failure mode and impact analysis to conduct a comprehensive risk identification of deep-sea mining surface support vessels. The factors considered include not only the surface support vessel system itself but also its operating environment and operational procedures, thereby constructing a risk assessment index system for deep-sea mining surface support vessels.

[0006] S2: Fuzzy Rating of Risk Factors; Given the relative lack of reliability and safety analysis of surface support vessels for deep-sea mining, risk data is somewhat inadequate. Therefore, based on the identified risk factors, a questionnaire was used to collect information related to surface support vessels for deep-sea mining. The questionnaire designed in this invention mainly consists of four parts: basic information of participants, a measure of the probability of risk factors, a measure of the severity of risk factors, and a measure of the detectability of risk factors. To obtain more comprehensive and reliable evaluation opinions, five experts with extensive experience in the field of deep-sea mining or related industries and research were invited.

[0007] S3: Constructing the FMEA-BN model; This invention constructs a Bayesian network model for deep-sea mining surface support vessels.

[0008] S4: Rank the risk factors.

[0009] Preferably, in S1, the Failure Mode and Effects Analysis (FMEA) method is used to identify the risks of deep-sea mining surface support vessels in operation. Based on the identified risk factors, a multi-index deep-sea mining surface support vessel evaluation system is established using a multi-level comprehensive evaluation method.

[0010] Preferably, in S2, a questionnaire is used to collect information related to surface support vessels for deep-sea mining. The questionnaire mainly consists of four parts: basic information of the participants, a measure of the probability of risk factors, a measure of the severity of risk factors, and a measure of the detectability of risk factors.

[0011] Preferably, in S2, the three risk parameters of the risk factor, namely probability (O), severity (S), and detectability (D), are divided into five levels according to the risk matrix proposed by the American Bureau of Shipping (ABS) in Guidance Notes on Risk Assessment Applications for the Marine and Offshore Industries: very high (VH), high (H), medium (M), low (L), and very low (VL). These risk parameters serve as indicators for evaluating the risk factor.

[0012] Preferably, in S3, the specific construction method is as follows: 1) Calculate the prior probability of the leaf node. This invention uses a Gaussian function to transform the semantic information of the expert on the risk parameter into fuzzy numerical values. The specific level transformation formula is as follows:

[0013]

[0014] In the formula, S vnThe membership function represents the level n risk, σ represents the expert's uncertainty, and u represents the membership center corresponding to different levels;

[0015] The probability distribution matrix H of the five experts is determined based on the membership function in formula (1). m :

[0016]

[0017] In the formula, the elements in the matrix The probability of the m-th expert evaluating the risk parameter of the n-th level is represented by the matrix, and the sum of the elements in each row of the matrix is ​​1.

[0018] Take matrix H m One line With another line H j Multiplying them together yields a new matrix NM:

[0019]

[0020] Based on the improved DS evidence theory, the expert evaluation results are integrated by combining two pieces of evidence and recursively calculating. The improved synthesis formula is shown in formula (4), and the result is the prior probability of the leaf node.

[0021]

[0022] In the formula, It is the sum of all non-main diagonal elements of matrix NM, representing the degree of conflict k between the evidences. f(NM) = kq(NM) is the probability allocation function of the evidence conflict, that is, it allocates the degree of conflict k between the evidences to each element in the matrix.

[0023] 2) Constructing a conditional probability table; This invention, based on the confidence level of the fuzzy IF-THEN rule base, uses a proportional method to transform different combinations of risk parameter levels into conditional probabilities based on the relative weights of the three risk parameters. The specific rule transformation and paraphrasing are shown below:

[0024]

[0025] in, This indicates that the input meets the prerequisite. At that time, in rule K, I z This is considered the confidence distribution of the outcome, where z is the number of all possible outcomes.

[0026] Preferably, a utility value Reg is introduced. rlThe confidence distribution of risk status is converted into numerical values ​​for comparison. The Risk Number (RN) is used to quantitatively describe the value of the parent node. Scores are assigned based on the risk parameter level, with a range of [1, 5], where 1 point represents the least contribution to risk and 5 points represent the greatest contribution. Based on the above fuzzy rules, the utility score of the risk factor can be calculated according to formula (7):

[0027] Reg rl =RN(O n )×RN(S n )×RN(D n (7)

[0028] Among them, Reg r1 It is based on a combination of specific fuzzy rules and relevant risk scores;

[0029] Therefore, a new Risk Rank Number (RRN) can be introduced to calculate and rank risk factors. The specific formula is as follows:

[0030]

[0031] Where m(rl) is the probability of the risk state taking the i-th reference value.

[0032] The beneficial effects of this invention are as follows: This invention proposes a risk assessment method suitable for complex systems under conditions of incomplete data—a risk assessment method based on an improved FMEA-BN model using confidence rules: taking the assessment of the risk of surface support vessels for deep-sea mining as an example. This method combines subjective and objective data, calculates prior probabilities through evidence theory, then calculates node probabilities through a fuzzy rule base, and finally quantifies the risk level of risk factors. The above implementation example uses the risk assessment method proposed in this invention to conduct a comprehensive risk analysis of surface support vessels. The results show that this method can obtain reliable and accurate risk assessment results. This invention mainly utilizes an improved FMEA-BN model based on confidence rules to conduct risk assessment research on surface support vessels for deep-sea mining. It uses the probability of conflicting supporting evidence to fuse the evaluation opinions of the expert group, improving the ability to process multi-source uncertain information, while overcoming the shortcomings of traditional methods that rely too much on subjective judgment by expert scoring. Attached Figure Description

[0033] Figure 1 This is a technical approach for risk assessment of surface support vessels for deep-sea mining;

[0034] Figure 2 It is a Bayesian network for surface support vessels in deep-sea mining. Detailed Implementation

[0035] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings and preferred embodiments.

[0036] like Figure 1-2 As shown, a safety assessment method for deep-sea mining surface support vessels based on the FMEA-BN model is presented. Step 1: Identify risk factors.

[0037] Referring to relevant reliability guidelines for deep-sea mining surface support vessels, and considering floating production storage and offloading (FPSO) vessels with similar operating cycles and ore carriers with similar structures, Failure Mode and Effects Analysis (FMEA) was used to identify risks in deep-sea mining surface support vessels during operations. Examples of the results are shown in Table 1. Based on the identified risk factors, a multi-level comprehensive evaluation method was used to establish a multi-index assessment system for deep-sea mining surface support vessels, as shown in Table 2, identifying a total of 35 fourth-level indicators. Based on this, a Bayesian network model was established, as follows... Figure 2 As shown, it can be divided into 35 root nodes.

[0038] Table 1. Results of Risk Factor Identification

[0039]

[0040] Table 2 Evaluation System for Surface Support Vessels for Deep-Sea Mining

[0041]

[0042] Step 2: Calculate the prior probabilities of parameters related to risk factors

[0043] Five experts with extensive experience in the relevant fields were invited to evaluate the risk parameters of the four-level indicators using fuzzy semantic terms, and the results were converted into fuzzy numbers using a Gaussian function. The evaluation results were then fused using an improved evidence theory. Since the five experts involved in this invention have similar qualifications, they were given equal weight when their evaluation results were synthesized. Taking the risk parameter S of the risk factor "real-time monitoring and control system failure" as an example, the evaluation results, probability distribution, and probability values ​​after data fusion by the five experts are shown in Table 3-5.

[0044] Table 3 Evaluation results of fault modes in the real-time monitoring and control system

[0045]

[0046] Table 4. Distribution of Fault Mode Severity Probability Values ​​in Real-Time Monitoring and Control Systems

[0047]

[0048] Table 5 Results after merging the severity of risk factors

[0049]

[0050] Step 3: Establish a confidence rule base for risk factor assessment

[0051] Based on "IF-THEN", and combined with the relevant risk status of surface support risk factors in deep-sea mining, a confidence rule base is constructed, containing a total of 125 rules (5×5×5). The extended confidence rules are defined as follows:

[0052] 1) IF: Occurrence Level (VL), Severity Level (VL), Detectability Level (VH)

[0053] THEN risk state {(Low,1),(Med i um,0),( h i gh,0)};

[0054] 2) IF: Occurrence M, Severity M, Detectability M,

[0055] THEN risk state {(Low,0),(Med i um,1),( h i gh,0)};

[0056] 3) IF: Occurrence (VH), Severity (VH), Detectability (VL)

[0057] THEN risk state {(Low,0),(Med i um,0),( h i gh,1)};

[0058] By using Bayesian network modeling, risk parameters are transformed into parent nodes, and risk states are transformed into child nodes. The rules in the confidence rule base are converted into conditional probabilities, ultimately transforming the confidence rule base into a Bayesian network. The specific conditional probability table is shown in Table 6, and the risk inference results are as follows: Figure 2 As shown.

[0059] Table 6 Conditional Probability Table

[0060]

[0061] Step 4: Quantify the risk value

[0062] According to formula (7), the confidence distribution of risk factors under different risk states is calculated as follows:

[0063] Reg r1 =RN(O1)×RN(S1)×RN(D1)=1 3 =1

[0064] Reg r2 =RN(O2)×RN(S2)×RN(D2)=2 3 =8

[0065] Reg r2=RN(O3)×RN(S3)×RN(D3)=3 3 =27

[0066] Reg r2 =RN(O4)×RN(S4)×RN(D4)=4 3 =64

[0067] Reg r3 =RN(O5)×RN(S5)×RN(D5)=5 3 =125

[0068] The risk ranking number is calculated using formula (8), and the RRNs of all risk factors are shown in Table 7. During the calculation process, updates to the parent node data will cause changes in the distribution of child nodes, thus allowing for real-time updates and better aligning with practical applications.

[0069] Table 7 Calculation of Risk Factor RRN

[0070]

[0071]

[0072] The results in Table 7 indicate that the top five risk factors affecting the safety of deep-sea mining surface support vessels are: X29 low operator skill level, X23 excessively close alongside, X30 high sea state, X15 tower and base malfunction, and X26 wire rope breakage. These five factors have a greater impact on deep-sea mining surface support vessels and are serious risk factors that require special attention; therefore, it is necessary to strengthen their monitoring and management.

[0073] In summary, this invention proposes a risk assessment method for complex systems under conditions of incomplete data—a risk assessment method based on the FMEA-BN improved model using confidence rules: taking the assessment of the risk of surface support vessels for deep-sea mining as an example. This method combines subjective and objective data, calculates prior probabilities through evidence theory, then calculates node probabilities using a fuzzy rule base, and finally quantifies the risk level of risk factors. The above implementation example uses the risk assessment method proposed in this invention to conduct a comprehensive risk analysis of surface support vessels. The results show that this method can obtain reliable and accurate risk assessment results.

[0074] It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of this invention, and these improvements and modifications should also be considered within the scope of protection of this invention.

Claims

1. A method for safety assessment of deep-sea mining surface support vessels based on the FMEA-BN model, characterized in that... Includes the following steps: S1: Identify potential risk factors for surface support vessels used in deep-sea mining; S2: Fuzzy rating of risk factors; S3: Construct the FMEA-BN model; S4: Rank the risk factors; In S3, the specific construction method is as follows: 1) Calculate the prior probability of the leaf nodes. By using a Gaussian function, the semantic information of the expert on the risk parameters is transformed into fuzzy values. The specific level transformation formula is shown below: (1) In the formula, S vn Representing the n Membership function of risk level, σ This indicates the uncertainty of the experts. u Indicates the membership centers corresponding to different levels; The probability distribution matrix of the five experts is determined based on the membership function in formula (1). H m : (2) In the formula, the elements in the matrix Representing the m The expert's evaluation of the first n The probability of the risk parameter of the level is such that the sum of the elements in each row of the matrix is ​​1; Take matrix H m One line With another line Multiplication yields a new matrix NM : (3) Based on the improved DS evidence theory, the expert evaluation results are integrated by combining two pieces of evidence and recursively calculating. The improved synthesis formula is shown in formula (4), and the result is the prior probability of the leaf node. (4) (5) In the formula, It is a matrix NM The sum of all non-main diagonal elements represents the degree of conflict between pieces of evidence. k , f(NM) = kq(NM) It is a probability allocation function for conflicting evidence, that is, it assigns the degree of conflict between pieces of evidence. k Each element is assigned to the matrix; 2) Construct a conditional probability table; based on the confidence level of the fuzzy IF-THEN rule base, and on the relative weights of the three risk parameters, use the proportional method to transform different combinations of risk parameter levels into conditional probabilities. The specific rule transformation and paraphrasing are shown below: (6) in, This indicates that the input meets the prerequisite. At that time, the first K In the rules I z This is considered to be the confidence distribution of the results. z It is the number of all possible outcomes; Introducing utility values Reg rl The confidence distribution of risk status is converted into numerical values ​​for comparison. The RiskNumber is used to quantitatively describe the value of the parent node. Scores are assigned according to the level of the risk parameter, with a value range of [1,5]. A score of 1 indicates the least contribution to the risk, and a score of 5 indicates the greatest contribution to the risk. Based on fuzzy rules, the utility value score of the risk factor can be calculated according to formula (7): (7) in, Reg r1 It is based on a combination of specific fuzzy rules and relevant risk scores; Therefore, a new risk ranking number (RRN) can be introduced to calculate and rank risk factors. The specific formula is as follows: (8) in, It represents the probability of taking the reference value for a risky state.

2. The method for safety assessment of deep-sea mining surface support vessels based on the FMEA-BN model according to claim 1, characterized in that: In S1, the Failure Mode and Effects Analysis (FMEA) method is used to identify risks of deep-sea mining surface support vessels in operation. Based on the identified risk factors, a multi-index assessment system for deep-sea mining surface support vessels is established using a multi-level comprehensive evaluation method.

3. The method for safety assessment of deep-sea mining surface support vessels based on the FMEA-BN model according to claim 1, characterized in that: In S2, a questionnaire was used to collect information related to surface support vessels for deep-sea mining. The questionnaire consisted of four parts: information on participants, a measure of the likelihood of risk factors, a measure of the severity of risk factors, and a measure of the detectability of risk factors.

4. The method for safety assessment of deep-sea mining surface support vessels based on the FMEA-BN model according to claim 1, characterized in that: In S2, the three risk parameters of the risk factors, namely probability... O Severity S and detectability D Based on the risk matrix proposed by the American Bureau of Shipping (ABS) in Guidance Notes on Risk Assessment Applications for the Marine and Offshore Industries, it is divided into 5 levels: very high VH, high H, medium M, low L, and very low VL. These risk parameters serve as indicators for evaluating risk factors.

Images