Improved FMEA method based on intuitionistic fuzzy rough cloud

Abstract

The invention discloses an improved FMEA method based on intuitionistic fuzzy rough cloud. The method comprises the following steps: (1) identifying a failure mode of a target object; (2) obtaining an expert score and converting the expert score into an intuitionistic fuzzy rough cloud; (3) determining expert weights and summarizing information; (4) determining a risk factor weight; (5) sorting failure mode risk priorities; (6) carrying out grade layering by using a hierarchical clustering method; and (7) risk analysis and evaluation. According to the method, expert evaluation information is quantified by using an intuitionistic fuzzy rough cloud semantic set, expert weights are determined by using a method based on an index composite value function, risk factor weights are calculated in combination with a subjective weight method FUCOM and an objective weight method IDOCRIW, priority ranking of failure modes is carried out by using an MAIRCA method, and the failure modes are classified according to the priority ranking. The problem that a traditional RPN calculation method is not stable and unscientific is solved, the failure modes are classified by using a hierarchical clustering method, and corresponding prevention or solution measures can be conveniently formulated.

Description

Technical Field

[0001] The invention relates to a technical method for reliability and risk analysis, and in particular to an improved FMEA method based on intuitive fuzzy rough cloud. Background Art

[0002] Failure Mode and Effects Analysis (FMEA) is a widely used forward-looking reliability and risk analysis technology. Through the systematic analysis of failure modes and their possible causes and effects, FMEA can help identify the weaknesses and key links of the system, thereby providing a scientific basis for decision makers to formulate risk management activities and preventive measures with clear priorities. Currently, the FMEA method has been widely used in aerospace, equipment maintenance, medical and other fields, and has become an important tool in reliability and risk management research.

[0003] The traditional FMEA method mainly evaluates the severity (S), occurrence (O) and detection (D) of the failure mode through discrete numerical values ​​(1-10), and calculates the risk priority number (RPN) by multiplying these three values ​​to measure the risk level of the failure mode. By comparing the RPN values, decision makers can prioritize the failure modes, thus providing a basis for subsequent risk management. However, the rationality and robustness of the traditional FMEA method in practical applications have been questioned by many scholars. There are mainly three limitations: (1) In practice, the expert's evaluation of the failure mode is vague and uncertain, and it is difficult to express it with discrete values; (2) The cognitive differences between experts and the relative importance of risk factors are not considered; (3) The RPN calculation method is not robust and scientific. Different risk factor evaluation results may have the same RPN value but different meanings. At present, many improved FMEA methods have been proposed, among which the most common one is the improved FMEA method combining fuzzy semantic theory and multiple-criteria decision making (MCDM) theory. This type of method can express expert evaluation information more completely and accurately, and realize the weight allocation of risk factors and priority sorting of failure modes through the MCDM method, so as to achieve a more scientific and comprehensive reliability and risk evaluation, and overcome the limitations of traditional FMEA. Summary of the invention

[0004] Purpose of the invention: In view of the problems existing in the existing FMEA method, the purpose of the present invention is to provide an accurate, stable and scientific improved FMEA method based on intuitive fuzzy rough cloud.

[0005] The method of the present invention: first, the intuitive fuzzy rough cloud semantic set is used to express the expert evaluation information more completely and accurately; second, the expert weight is determined by a method based on an exponential composite value function, and the cognitive differences between experts are considered; third, the full consistency method (FUCOM) is extended to the intuitive fuzzy rough cloud to calculate the subjective weight of the risk factor; fourth, the objective weight of the risk factor is calculated by using the integrated determination of objective criteria weight (IDOCRIW), and the subjective and objective weights of the risk factor are further combined; fifth, the failure mode is prioritized by using the multi-attributive ideal-real comparative analysis (MAIRCA) to obtain a more scientific and comprehensive risk assessment result; finally, the hierarchical clustering method is used to hierarchically stratify the failure mode to assist the team in resource allocation and formulation of preventive measures. According to the present invention, the proposed improved FMEA method is implemented in the reliability and risk analysis of the digital twin system to verify the effectiveness and applicability of the proposed method.

[0006] Technical solution: The present invention provides an improved FMEA method based on intuitive fuzzy rough cloud, comprising the following steps:

[0007] (1) Identify the failure mode of the target object;

[0008] (2) Obtain expert scores and convert them into intuitionistic fuzzy rough clouds;

[0009] (3) Determine expert weights and information aggregation;

[0010] (4) Determine the risk factor weights;

[0011] (5) Failure mode risk priority ranking;

[0012] (6) Use hierarchical clustering to perform hierarchical stratification;

[0013] (7) Risk analysis and evaluation.

[0014] Furthermore, the method of step (1) is:

[0015] S1. Set up a FMEA team, determine the target object and its scope, and analyze the operation process of the target object;

[0016] S2. Determine the components of the target object, ensure that FMEA analysis covers every link, and identify the potential failure modes of the target object through literature review, brainstorming and expert interviews: FM1, FM2, ..., FMm , m is the total number of failure modes, FM i represents the i-th failure mode;

[0017] Furthermore, the step (2) method:

[0018] S3. Invite the expert team to score the S, O, and D of the failure mode and the importance of S, O, and D to the target object;

[0019] S4, establish intuitionistic fuzzy rough cloud semantic set: first, convert the expert ratings into intuitionistic fuzzy numbers, then combine all expert evaluations to construct intuitionistic fuzzy rough sets, and finally map the intuitionistic fuzzy rough sets to the cloud model;

[0020] Furthermore, the method of step (3) is:

[0021] S5. Combining the consensus of the expert group and the consistency of experts, the method based on the exponential compound value function is used to determine the expert weights, and the expert evaluation information is summarized to obtain the collective decision matrix;

[0022] Furthermore, the method of step (4) is:

[0023] S6. Extend FUCOM to the intuitionistic fuzzy rough cloud and impose complete consistency constraints on the three parameters of the semantic set to calculate the subjective weight of the risk factor;

[0024] S7. Combining the entropy weight method and the criterion shock loss, the objective weights of risk factors are calculated through IDOCRIW, and the subjective and objective weights of risk factors are further combined;

[0025] Furthermore, the method of step (5) is:

[0026] S8. Use MAIRCA to calculate the criterion function value of the failure mode and prioritize the failure mode;

[0027] Furthermore, the method of step (6) is:

[0028] S9. Use the hierarchical clustering method to stratify the failure modes. The FMEA team will formulate corresponding prevention and solution measures based on the stratification results.

[0029] Furthermore, the establishment of the intuitionistic fuzzy rough cloud semantic set in step S4 includes the following steps:

[0030] S4-1. Define intuitionistic fuzzy sets: Let X be a non-empty set, and the elements in X are denoted by x. Then the intuitionistic fuzzy set corresponding to X is A = ξ <x,μ A (x), v A (x)>|x∈X}, where μ A (x), vA (x) are the membership and non-membership of x in A, μ A : X→[0, 1], v A :X→[0,1], and satisfy the condition 0≤μ A (x)+v A (x)≤1.

[0031] S4-2. According to the semantic conversion table, the expert rating is converted into an intuitive fuzzy number, such as α = (μ α , v α );

[0032] S4-3. Construct intuitionistic fuzzy rough numbers. Let X be a non-empty set, the elements in X are denoted by x, and A is the intuitionistic fuzzy number set corresponding to X, which has n intuitionistic fuzzy numbers: α i =(μ i , ν i ), 1≤i≤n, then α i The lower approximation set and upper approximation set of are:

[0033] April (α i )=U{x∈X|A(x)≤α i}

[0034]

[0035] Among them, A(x) is the intuitionistic fuzzy number corresponding to x;

[0036] According to the construction of rough numbers, α i The lower and upper bounds of the intuitionistic fuzzy rough numbers can be defined as:

[0037] IFRN (α i )=IFA(A(x1),A(x2),...,A(x ML ))|x∈ April (α i )

[0038]

[0039] Among them, IFA is an average aggregation operator of intuitive fuzzy numbers, defined as: ML and MU are April (α i )and The number of elements in ;

[0040] Combined with the upper and lower limits, α i The intuitionistic fuzzy rough number is:

[0041] According to the above construction process, each expert's score is further converted into an intuitionistic fuzzy rough number;

[0042] S4-4. Convert the intuitionistic fuzzy rough number into a cloud model, and the corresponding lower limit is:

[0043] in,

[0044] The corresponding upper limit is:

[0045]

[0046] in,

[0047] Combined with the upper and lower limits, α i The intuitionistic fuzzy rough cloud is:

[0048] According to the above process, each expert score is further transformed into an intuitionistic fuzzy rough cloud.

[0049] Further, in step S5, determining the expert weight using a method based on an exponential compound value function includes the following steps:

[0050] S5-1. Assume that the intuitionistic fuzzy rough cloud matrix converted from the score of the t-th expert on the failure mode is 1≤t≤k, where m is the number of failure modes and k is the number of experts;

[0051] S5-2. Determine the minimum and maximum values ​​of the intuitive fuzzy rough cloud of each risk factor of the failure mode:

[0052] S5-3. The consensus of the expert group is measured by the sum of the relative divergences between the expert evaluation information. The expression is:

[0053]

[0054] Among them, the d function is the distance function of the intuitionistic fuzzy rough cloud, which is defined as:

[0055]

[0056] S5-4. Define the evaluation information of the tth expert and average aggregate information from all experts The similarity function is used to express the expert consistency information, and the expression is:

[0057]

[0058] in,

[0059] S5-5. Combining the two parameters of expert group consensus and expert consistency, a mathematical model based on exponential compound value function is constructed to solve the expert weight. The mathematical model is:

[0060]

[0061]

[0062] in, The expert weight corresponding to the t-th expert; the expert weight w is obtained by the Lagrange multiplier method t for:

[0063]

[0064] Furthermore, in step S6, the three parameters of the semantic set are fully constrained to be consistent, and FUCOM is extended to the intuitionistic fuzzy rough cloud. The intuitionistic fuzzy rough cloud corresponding to the score of the t-th expert on the importance of the j-th risk factor is set as The intuitionistic fuzzy rough cloud corresponding to the subjective weight of the t-th expert on the j-th risk factor is Then use FUCOM to solve The mathematical model is as follows:

[0065] min

[0066]

[0067] The subjective weights of risk factors corresponding to all expert evaluation information are averaged and aggregated to obtain the final subjective weights

[0068] Further, in step S7, using IDOCRIW to calculate the objective weight of the risk factor includes the following steps:

[0069] S7-1. Use the entropy weight method to calculate the objective weight of the risk factor q = (q1, q2, q3). Suppose the collective decision matrix obtained after summarizing the expert intuitive fuzzy rough cloud evaluation information is: The object of objective weight calculation is [c ij ] m×3 The first element of the intuitionistic fuzzy rough cloud right Perform normalization: Calculate the entropy value of each risk factor: Calculate the dispersion of the intrinsic information of each risk factor: DP j =1-Entropy j, then the objective weight of the risk factor obtained by the entropy weight method is j = 1, 2, 3;

[0070] S7-2. Use the CILOS method to calculate the objective weight of risk factors u = (u1, u2, u3), and transform The size of the direction, Assume z j for The number of rows corresponding to the maximum value in the jth column, constructing the matrix A = [a ij ] 3×3 , where a jj for The maximum value of the jth column, Construct the relative loss matrix R = [r ij ] 3×3 , The objective weights of risk factors corresponding to the CILOS method are obtained by solving the linear equation system Fu = 0, where:

[0071] S7-3. Combine the results of the entropy weight method and the CILOS method to calculate the final objective weight j=1,2,3, Extension to Intuitionistic Fuzzy Rough Clouds

[0072] Further, in step S8, using MAIRCA to prioritize the failure modes includes the following steps:

[0073] S8-1, for the collective decision matrix C m×3 =[c ij ] m×3 Perform normalization processing,

[0074] S8-2. Determine failure mode preference

[0075] S8-3. Constructing a Theoretical Evaluation Matrix Among them, ω j is the comprehensive weight of risk factors, ω j =0.5×ωs j +0.5×ωo j , j = 1, 2, 3;

[0076] S8-4. Constructing the actual evaluation matrix

[0077] S8-5. Obtain the total gap matrix G m×3=[g ij ] m×3 ,

[0078] S8-6. Calculate the criterion function value of each failure mode i=1,2,…,m,according to O i The value is used to prioritize the failure modes. i The smaller the value, the higher the failure mode priority ranking.

[0079] Compared with the prior art, the present invention has the following beneficial effects:

[0080] The intuitive fuzzy rough cloud semantic set is used to quantify the expert evaluation information, which is more scientific and accurate. The method based on the exponential composite value function is used to determine the expert weight, which takes into account the cognitive differences among experts. The subjective weight method FUCOM and the objective weight method IDOCRIW are combined to calculate the risk factor weights, which takes into account the relative importance of risk factors. The MAIRCA method is used to prioritize failure modes, which solves the problem that the traditional RPN calculation method is unrobust and unscientific. The hierarchical clustering method is used to hierarchically stratify the failure modes, which is convenient for formulating corresponding prevention or solution measures. BRIEF DESCRIPTION OF THE DRAWINGS

[0081] Figure 1 A flowchart of an improved FMEA method based on intuitionistic fuzzy rough cloud according to the present invention;

[0082] Figure 2 It is a summary of the intuitive fuzzy rough cloud evaluation of the failure mode of the digital twin system in the embodiment of the present invention;

[0083] Figure 3 This is the clustering result of the failure mode of the digital twin system in the embodiment of the present invention. DETAILED DESCRIPTION

[0084] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be further described below.

[0085] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0086] In this embodiment, the FMEA team analyzes the operation and composition of the industrial digital twin system and divides the system into 6 levels. The composition and function of each level are quite different. Therefore, the failure modes are identified and sorted out according to different levels. The final failure modes can be divided into 22 types. For the convenience of expression, each failure mode is represented by FMEA. i Indicates that i=1,2,…,22, as shown in Table 1.

[0087] Table 1 Failure mode table

[0088]

[0089]

[0090] An expert team was invited to score the failure modes, consisting of four people, including university scholars, digital twin system technical managers, digital twin system implementation consultants, and senior managers of information technology projects, denoted as {e1, e2, e3, e4}, and the scoring criteria are shown in Table 2. The experts scored the importance of risk factors, and the scoring criteria are shown in Table 3.

[0091] Table 2 Failure mode scoring criteria

[0092]

[0093] Table 3 Risk factor importance rating criteria

[0094]

[0095] The results of the failure mode scoring are shown in Table 4. The results of the risk factor importance scoring are shown in Table 5. The failure mode scores and risk factor importance scores are converted into intuitive fuzzy numbers, and the conversion relationship is shown in Table 6. Furthermore, the intuitive fuzzy rough cloud evaluation information is obtained. The intuitive fuzzy rough cloud evaluation results of the failure mode are shown in Table 6. Figure 2 The intuitive fuzzy rough cloud evaluation results of the importance of risk factors are shown in Table 7.

[0096] Table 4 Failure mode score summary (first five failure modes)

[0097]

[0098]

[0099] Table 5 Summary of risk factor importance scores

[0100]

[0101] Table 6 Intuitive fuzzy number conversion table

[0102]

[0103] Table 7 Intuitive fuzzy rough cloud evaluation results of risk factor importance

[0104]

[0105] The expert weight is determined using a method based on an exponential compound value function, where expert weight w = {w t |t=1,2,3,4}={0.2371,0.2450,0.2620,0.2559}.

[0106] According to the expert weights, the intuitionistic fuzzy rough cloud evaluation information of the failure mode of the four experts is weighted and aggregated. t The intuitionistic fuzzy rough cloud evaluation matrix is t=1,2,3,4, m is the number of failure modes, that is, m=22; suppose the collective decision matrix obtained after summarizing the expert intuition fuzzy rough cloud evaluation information is in, Some results of the collective decision-making matrix are shown in Table 8.

[0107] Table 8 Collective decision matrix (first five failure modes)

[0108]

[0109] Use FUCOM to calculate the subjective weight of risk factors, risk factor subjective weight ωs = {ωs j |j=1,2,3}={(0.4208,0.0121,0.0012),(0.3139,0.0002,0.0001),(0.2653,0.0108,0.0009)}. IDOCRIW is used to calculate the objective weight of risk factors. The objective weight of risk factors ωo={ωo j |j=1,2,3}={(0.1785,0.0000,0.0000),(0.5710,0.0000,0.0000),(0.2506,0.0000,0.0000)}. Further, combining the subjective and objective weights of risk factors, we get the final weight ω={ω j |j=1,2,3}={(0.2996,0.0000,0.0000),(0.4424,0.0000,0.0000),(0.2579,0.0000,0.0000)}.

[0110] Use MAIRCA to prioritize failure modes. First, the collective decision matrix C m×3=[c ij ] m×3 After normalization, the normalized matrix is Some of the results are shown in Table 9. Failure mode preference i=1,2,…,22. Theoretical evaluation matrix Further obtain the actual evaluation matrix Some results are shown in Table 10.

[0111] Table 9 Normalized matrix of collective decision making

[0112]

[0113] Table 10 Actual evaluation matrix

[0114]

[0115] Further calculation to obtain the total gap matrix G m×3 =[g ij ] m×3 , and calculate the criterion function value O of each failure mode i , according to O i The failure modes are prioritized by the size of the values, and the results are shown in Table 11.

[0116] Table 11 Calculation and ranking of failure mode criterion function values

[0117]

[0118] Furthermore, a two-dimensional feature data is constructed as the input of the hierarchical clustering method to hierarchically classify the failure modes. The first feature value is the ranking score. Let s i For FM i Sorting, definition The sorting order is about to be reversed, and further Normalize it, FM i The ranking score of the second eigenvalue is the normalized RPN. The clustering results of the hierarchical clustering method are as follows Figure 3 shown.

[0119] Depend on Figure 3 It can be seen that the high-risk failure modes are FM1 (the device fails to accurately execute the control command), FM8 (data parsing error), FM 13 (Simulation optimization is imperfect), FM 18(System decision error), for high-risk failure modes, the most resources should be allocated for prevention or correction. For example, FM1 (the equipment fails to accurately execute control commands), the processing equipment is responsible for the operation of the physical layer of the digital twin system and the generation of basic data. Once there is a problem with the execution of the processing equipment, it will greatly affect the entire system. Therefore, more resources must be invested in monitoring all aspects of the equipment operation to ensure that the equipment accurately executes control commands. Failure modes of medium risk include FM4 (data collection failure), FM6 (slow data upload), FM 10 (Environmental configuration error) and other eight failure modes. These eight failure modes have a certain impact on the operation of the digital twin system, and there is no need to invest too much resources for monitoring and prevention. The low-risk failure modes include FM2 (equipment operation shutdown), FM3 (safety issues), FM5 (data storage loss) and other ten failure modes. These ten failure modes have little impact on the operation of the digital twin system, and appropriate resources can be invested for monitoring and prevention.

[0120] Through example analysis, the present invention is applicable and effective in reliability and risk analysis.

[0121] The above is only a preferred embodiment of the present invention and does not limit the present invention in any way. Any technician in the relevant technical field, without departing from the scope of the technical solution of the present invention, makes any form of equivalent replacement or modification to the technical solution and technical content disclosed in the present invention, which does not depart from the content of the technical solution of the present invention and still falls within the protection scope of the present invention.

Claims

1. An improved FMEA method based on intuitionistic fuzzy rough cloud, characterized in that: The following steps are involved: (1) Identify the failure mode of the target object; (2) Obtain expert scores and convert them into intuitionistic fuzzy rough clouds; (3) Determine expert weights and information aggregation; (4) Determine the risk factor weights; (5) Failure mode risk priority ranking; (6) Use hierarchical clustering to perform hierarchical stratification; (7) Risk analysis and evaluation.

2. The improved FMEA method based on intuitionistic fuzzy rough cloud according to claim 1 is characterized in that: The method of step (1) comprises: S1. Set up a FMEA team, determine the target object and its scope, and analyze the operation process of the target object; S2. Determine the components of the target object, ensure that FMEA analysis covers every link, and identify the potential failure modes of the target object through literature review, brainstorming and expert interviews: FM1, FM2, ..., FM m , m is the total number of failure modes, FM i represents the ith failure mode.

3. The improved FMEA method based on intuitionistic fuzzy rough cloud according to claim 2 is characterized in that: The step (2) method comprises: S3. Invite the expert team to score the S, O, and D of the failure mode and the importance of S, O, and D to the target object; S4. Establish an intuitionistic fuzzy rough cloud semantic set: first, convert the expert scores into intuitionistic fuzzy numbers, then combine all expert evaluations to construct an intuitionistic fuzzy rough set, and finally map the intuitionistic fuzzy rough set to the cloud model.

4. The improved FMEA method based on intuitionistic fuzzy rough cloud according to claim 3 is characterized in that: The establishment of the intuitionistic fuzzy rough cloud semantic set in step S4 includes the following steps: S4-1. Define intuitionistic fuzzy sets: Let X be a non-empty set, and the elements in X are denoted by x. Then the intuitionistic fuzzy set corresponding to X is A = { <x,μ A (x),ν A (x)>|x∈X}, where μ A (x), ν A (x) are the membership and non-membership of x in A, μ A :X→[0,1],ν A :X→[0,1], and satisfy the condition 0≤μ A (x)+v A (x)≤1; S4-2. According to the semantic conversion table, the expert rating is converted into an intuitive fuzzy number, such as α = (μ α ,ν α ); S4-3. Construct intuitionistic fuzzy rough numbers. Let X be a non-empty set, the elements in X are denoted by x, and A is the intuitionistic fuzzy number set corresponding to X, which has n intuitionistic fuzzy numbers: α i =(μ i ,ν i ), 1≤i≤n, then α i The lower approximation set and upper approximation set of are: April (a i )=∪{x∈X|A(x)≤α i } Among them, A(x) is the intuitionistic fuzzy number corresponding to x; According to the construction of rough numbers, α i The lower and upper bounds of the intuitionistic fuzzy rough numbers can be defined as: IFRN (a i )=IFA(A(x1),A(x2),…,A(x ML ))|x∈ April (a i ) Among them, IFA is an average operator of intuitive fuzzy numbers, defined as: ML and MU are April (α i )and The number of elements in ; Combined with the upper and lower limits, α i The intuitionistic fuzzy rough number is: According to the above construction process, each expert's score is further converted into an intuitionistic fuzzy rough number; S4-4. Convert the intuitionistic fuzzy rough number into a cloud model, and the corresponding lower limit is: in, The corresponding upper limit is: in, Combined with the upper and lower limits, α i The intuitionistic fuzzy rough cloud is: According to the above process, each expert score is further transformed into an intuitionistic fuzzy rough cloud.

5. The improved FMEA method based on intuitionistic fuzzy rough cloud according to claim 4 is characterized in that: The step (3) method comprises: S5. Combining the consensus of the expert group and the consistency of experts, the method based on the exponential compound value function is used to determine the expert weights, and the expert evaluation information is summarized to obtain the collective decision-making matrix.

6. The improved FMEA method based on intuitionistic fuzzy rough cloud according to claim 5 is characterized in that: Determining the expert weight using the method based on the exponential compound value function in step S5 comprises the following steps: S5-1. Assume that the intuitionistic fuzzy rough cloud matrix converted from the score of the t-th expert on the failure mode is 1≤t≤k, where m is the number of failure modes and k is the number of experts; S5-2. Determine the minimum and maximum values ​​of the intuitive fuzzy rough cloud of each risk factor of the failure mode: S5-3. The consensus of the expert group is measured by the sum of the relative divergences between the expert evaluation information. The expression is: Among them, the d function is the distance function of the intuitionistic fuzzy rough cloud, which is defined as: S5-4. Define the evaluation information of the tth expert and average aggregate information from all experts The similarity function is used to express the expert consistency information, and the expression is: in, S5-5. Combining the two parameters of expert group consensus and expert consistency, a mathematical model based on exponential compound value function is constructed to solve the expert weight. The mathematical model is: in, The expert weight corresponding to the t-th expert; the expert weight w is obtained by the Lagrange multiplier method t for:

7. The improved FMEA method based on intuitionistic fuzzy rough cloud according to claim 6 is characterized in that: The step (4) method comprises: S6. Extend FUCOM to the intuitionistic fuzzy rough cloud and impose complete consistency constraints on the three parameters of the semantic set to calculate the subjective weight of the risk factor; S7. Combining the entropy weight method and criterion shock loss, the objective weights of risk factors are calculated through IDOCRIW, and the subjective and objective weights of risk factors are further combined.

8. The improved FMEA method based on intuitionistic fuzzy rough cloud according to claim 7 is characterized in that: In step S6, the three parameters of the semantic set are fully constrained to be consistent, and FUCOM is extended to the intuitionistic fuzzy rough cloud. The intuitionistic fuzzy rough cloud corresponding to the score of the t-th expert on the importance of the j-th risk factor is assumed to be The intuitionistic fuzzy rough cloud corresponding to the subjective weight of the t-th expert on the j-th risk factor is Then use FUCOM to solve The mathematical model is as follows: min The subjective weights of risk factors corresponding to all expert evaluation information are averaged and aggregated to obtain the final subjective weights The step S7 in which IDOCRIW is used to calculate the objective weight of the risk factor includes the following steps: S7-1, using the entropy weight method to calculate the objective weight of the risk factor q = (q1, q2, q3), assuming that the collective decision matrix obtained after summarizing the expert intuitive fuzzy rough cloud evaluation information is The object of objective weight calculation is [c ij ] m×3 The first element of the intuitionistic fuzzy rough cloud right Perform normalization: Calculate the entropy value of each risk factor: Calculate the dispersion of the intrinsic information of each risk factor: DP j =1-Entropy j , then the objective weight of the risk factor obtained by the entropy weight method is j = 1, 2, 3; S7-2. Use the CILOS method to calculate the objective weight of risk factors u = (u1, u2, u3), and transform The size of the direction, Assume z j for The number of rows corresponding to the maximum value in the jth column, constructing the matrix A = [a ij ] 3×3 , where a jj for The maximum value of the jth column, Construct the relative loss matrix R = [r ij ] 3×3 , The objective weights of risk factors corresponding to the CILOS method are obtained by solving the linear equation system Fu = 0, where: S7-3. Combine the results of the entropy weight method and the CILOS method to calculate the final objective weight j=1,2,3, Extension to Intuitionistic Fuzzy Rough Clouds 9. The improved FMEA method based on intuitionistic fuzzy rough cloud according to claim 8 is characterized in that: The step (5) method comprises: S8. Use MAIRCA to calculate the criterion function value of the failure mode and prioritize the failure mode.

10. The improved FMEA method based on intuitionistic fuzzy rough cloud according to claim 9 is characterized in that: The step S8 of using MAIRCA to prioritize failure modes includes the following steps: S8-1, for the collective decision matrix C m×3 =[c ij ] m×3 Perform normalization processing, S8-2. Determine failure mode preference S8-3. Constructing a Theoretical Evaluation Matrix Among them, ω j is the comprehensive weight of risk factors, ω j =0.5×ωs j +0.5×ωo j , j = 1, 2, 3; S8-4. Constructing the actual evaluation matrix S8-5. Obtain the total gap matrix G m×3 =[g ij ] m×3 , S8-6. Calculate the criterion function value of each failure mode i=1,2,…,m,according to O i The value is used to prioritize the failure modes. i The smaller the value, the higher the failure mode priority ranking.

Images