Failure mode risk assessment method and system based on Internet review mining

Abstract

This invention discloses a failure mode risk assessment method and system based on internet review mining. The method includes the following steps: Step 1: Acquire and preprocess internet review data; Step 2: Extract product or service failure modes based on text features; Step 3: Calculate the severity, occurrence, and difficulty of detection of each failure mode; Step 4: Calculate the objective weights of severity, occurrence, and difficulty of detection; Step 5: Calculate the risk severity of each failure mode. This invention introduces external internet review factors into the FMEA analysis framework and, by mining internet review data, enables product or service failure mode identification.

Description

Technical Field

[0001] The present invention relates to a failure mode risk assessment method, in particular to a failure mode risk assessment method based on Internet comment mining, and also to a corresponding failure mode risk assessment system, belonging to the technical field of failure analysis. Background Art

[0002] Failure Mode and Effect Analysis (FMEA) is a systematic solution for improving product quality and reliability by analyzing each subsystem, component, and process step during the product and process design phases to identify all potential failure modes and analyze their possible consequences. This allows for proactive implementation of necessary measures. FMEA is a bottom-up analysis method. Analyzing specific failure types with particularly severe consequences is called a Criticality Analysis (CA). FMEA and CA are collectively referred to as FMECA (Failure Mode, Effects, and Criticality Analysis).

[0003] In a Chinese patent application with application number 202010000407.X, Shanghai University proposed an improved FMEA method based on the Pythagorean module and TOPSIS, which includes the following steps: S1. Calculate the collective risk assessment matrix; S2. Calculate the weighted risk assessment matrix; S3. Determine the positive and negative ideal solutions; S4. Calculate the distance between the failure mode and the positive and negative ideal solutions; S5. Calculate the relative proximity of each failure mode; S6. Determine the risk ranking of the failure modes. However, this improved FMEA method uses the FMEA team's subjective risk assessment matrix, which lacks objectivity. The risk weights for severity (S), occurrence (O), and difficulty of detection (D) are not calculated based on a basis, and the accuracy of the weight values ​​remains questionable.

[0004] In a separate Chinese patent application, application number 201710748331.7, the University of Electronic Science and Technology of China proposed an FMECA method for processing data using SVM text mining. The method includes the following steps: S1. Collecting fault data for FMECA analysis; S2. Deleting data that is not conducive to model building; S3. Processing the unstructured data into structured data suitable for SVM analysis; S4. Selecting a portion of the preprocessed data as training samples and the remaining data as test samples; S5. Establishing a predictive model: Using the training samples to establish a predictive model, the data is used to predict the output failure mode; S6. Verifying the model and modifying parameters: Using test samples to verify the accuracy of the established predictive model, and modifying parameters to optimize the model; S7. Refining the FMEA; S8. Generating an FMEA report; and S9. Generating a CA report. This technical solution is characterized by using SVM text mining to structure fault data. This eliminates the need for extensive human effort in data analysis and allows for the extraction of failure modes with simple setup, reducing analysis costs and significantly improving data utilization. Summary of the Invention

[0005] The primary technical problem to be solved by the present invention is to provide a failure mode risk assessment method based on Internet review mining.

[0006] Another technical problem to be solved by the present invention is to provide a failure mode risk assessment system based on Internet review mining.

[0007] In order to achieve the above object, the present invention adopts the following technical solutions:

[0008] According to a first aspect of an embodiment of the present invention, a failure mode risk assessment method based on Internet review mining is provided, comprising the following steps:

[0009] Step 1: Obtain internet review data and preprocess it;

[0010] Step 2: Extract failure modes of products or services based on text features;

[0011] Step 3: Calculate the severity, occurrence, and difficulty of detection of each failure mode;

[0012] Step 4: Calculate the objective weights of severity, occurrence, and difficulty of detection respectively;

[0013] Step 5: Calculate the risk severity of each failure mode.

[0014] Preferably, in step 1, the Internet review data is negative user evaluations in after-sales user reviews.

[0015] Preferably, in step 2, text features are extracted from the pre-processed text data, and the words that best describe the failure mode are selected as the failure mode of the product or service.

[0016] Preferably, in step 3, the severity is calculated using the following formula:

[0017]

[0018] in, Indicates failure mode f i The severity S, is the sentiment tendency value.

[0019] Preferably, in step 3, the occurrence degree is calculated using the following formula:

[0020]

[0021] in, Indicates failure mode f i Occurrence rate O, is the number of occurrences of the failure mode.

[0022] Preferably, in step 3, the difficulty of detection is calculated using the following formula:

[0023]

[0024] Among them, D(f i ) represents the failure mode f i The difficulty of detection D, is the failure mode f i The number of negative reviews, t j (f i ) is the word frequency of each failure mode, i = 1, 2, ..., a, j = 1, 2, ..., m.

[0025] Preferably, step 4 includes the following sub-steps:

[0026] 4.1 Perform dimensionless processing on each evaluation index;

[0027] 4.2 Calculate indicator variability;

[0028] 4.3 Calculation of indicator conflicts;

[0029] 4.4 Calculate the information content of indicators;

[0030] 4.5 Calculate the objective weights and update the weighted decision matrix based on the objective weights.

[0031] Preferably, step 5 includes the following sub-steps:

[0032] 5.1 Obtaining the positive ideal solution and the negative ideal solution of the evaluation objective according to the weighted decision matrix;

[0033] 5.2 Calculate the distance between the solution value of each failure mode and the positive and negative ideal values;

[0034] 5.3 Calculate the relative progress between the solution value of each failure mode and the positive ideal solution, and rank the risk severity of each failure mode in descending order of relative progress.

[0035] According to a second aspect of an embodiment of the present invention, a failure mode risk assessment system based on Internet review mining is provided, comprising:

[0036] An information collection and storage server is used to collect external Internet review data, crawl negative user reviews of products or services, clean the text data of the negative user reviews, set a stop word dictionary and perform text segmentation, obtain pre-processed text data, and import it into the information database for storage;

[0037] The core data processing server includes comment text mining module, risk matrix calculation module, CRITIC weight calculation module, and TOPSIS risk severity assessment module; among them,

[0038] The review text mining module is used to obtain pre-processed text data from the information database, perform text feature extraction on the text data, and select the words that best describe the failure mode as the failure mode of the product or service;

[0039] The risk matrix calculation module is used to calculate the severity, occurrence and difficulty of detection of each failure mode;

[0040] The CRITIC weight calculation module is used to calculate the objective weights of severity, occurrence, and difficulty of detection;

[0041] The TOPSIS risk severity assessment module is used to calculate the risk severity of the failure mode.

[0042] Preferably, in the TOPSIS risk severity assessment module, the positive ideal solution and the negative ideal solution of the evaluation target are obtained according to the weighted decision matrix; the distance between the solution value of each failure mode and the positive and negative ideal values ​​is calculated; the relative progress between the solution value of each failure mode and the positive ideal solution is calculated, and the risk severity of each failure mode is ranked in descending order according to the relative progress.

[0043] Compared to existing technologies, this invention creatively incorporates external internet review factors into the FMEA analysis framework. By mining internet review data, it enables product or service failure mode identification. Furthermore, this invention employs a unified processing method for internet review data. Compared to the subjective expert scoring process in existing FMEA analysis frameworks, this solves the problem of insufficient data and a single source, increases the objectivity of failure mode analysis results, and effectively assists quality management personnel in quality improvement efforts. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 This is an overall flow chart of the failure mode risk assessment method provided by the present invention;

[0045] Figure 2 This is a schematic diagram of the structure of the failure mode risk assessment system provided by the present invention. DETAILED DESCRIPTION

[0046] The technical content of the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0047] In existing FMEA analysis methods, the definition of "user" extends beyond the "end user" to include engineers / design teams responsible for overall or higher-level assembly design, as well as production / process engineers responsible for production, assembly, and service activities. Therefore, existing FMEA analysis frameworks do not adequately prioritize user experience. On the other hand, users gain firsthand experience with a product or service's quality after actually using it. These experiences are often expressed in the form of online reviews. While individual differences exist, accumulating massive amounts of online review data can largely mitigate these differences, resulting in an objective evaluation of the product or service.

[0048] Therefore, the failure mode risk assessment method provided in the embodiment of the present invention creatively introduces external Internet review factors into the FMEA analysis framework. By mining, cleaning and using massive Internet review data, it achieves product or service quality improvement based on user review mining, improves the utilization of text data by enterprises and Internet platforms, and can well assist quality management personnel in quality improvement work.

[0049] The following combination Figure 1 , the specific steps of the failure mode risk assessment method based on Internet review mining are described in detail:

[0050] Step 1: Obtain internet review data and preprocess it

[0051] First, we select a product or service and crawl its negative user reviews (e.g., negative reviews from various internet platforms) from its post-sale user reviews. We crawl negative user reviews here because positive user reviews are often stereotyped, and only negative reviews can directly promote the quality of the product or service.

[0052] After cleaning the internet comment data, we set up a stop word dictionary, segmented each text, and removed stop words to obtain preprocessed text data. The purpose of setting up a stop word dictionary is because internet comment data contains some completely meaningless words and symbols that have no impact on the meaning of the sentence, such as personal pronouns, modal particles, and punctuation marks. When processing internet comments, these stop words are often removed to improve the accuracy and efficiency of text processing.

[0053] Step 2: Extract failure modes of products or services based on text features

[0054] After preprocessing the text data, we first perform text feature extraction to obtain a number of highly critical words (i.e., keywords). From these words, we select the words that best describe the failure mode as the failure mode of the product or service. The importance of these highly critical words (i.e., keywords) is evaluated using the TF-IDF algorithm.

[0055] The TF-IDF algorithm is a commonly used weighting technique for information retrieval and data mining. TF stands for term frequency, and IDF stands for inverse document frequency. The TF-IDF index, calculated by multiplying TF and IDF, is used to analyze the importance of words in a document. For example, in a review text dataset, the high TF-IDF values ​​for the words "catering," "accommodation," and "transportation" indicate that consumers are particularly concerned about these three words in their reviews. These three words also happen to describe the source of failure modes. For example, in a negative review, the word "catering" may indicate that the food is not good or of poor quality, while "accommodation" may indicate that the accommodation conditions are poor. Therefore, for user reviews of travel services, words such as "catering," "accommodation," and "transportation" serve as failure modes.

[0056] Assume there are a total of failure modes, and record a failure modes as f i (i=1, 2, ..., a), where a is a positive integer.

[0057] Step 3: Calculate the severity (S), occurrence (O), and difficulty of detection (D) of each failure mode

[0058] 3.1: Calculation of severity (S)

[0059] In the FMEA analysis framework, S represents the severity of a failure mode. In principle, the greater the severity (S), the more severe the consequences of the failure mode. In traditional FMEA analysis, severity (S) is determined through expert scoring, which is susceptible to subjective judgment.

[0060] In the failure mode risk assessment method provided by the embodiment of the present invention, we use the user's dissatisfaction level to replace the severity level, and the user's dissatisfaction level can be described by the negative emotional level in the user's negative review. Specifically, we define the total number of failure mode categories as p, and record the failure mode as f i (i=1,2,…,p), assuming that failure mode f i User reviews of Here we assume that this failure mode occurs in all users times. We will use the failure mode f i The sentiment tendency values ​​of the comment texts are recorded as The calculation formula of severity (S) is given below:

[0061]

[0062] in, As a sentiment tendency value, it is a concept often used in the field of sentiment analysis. It represents the probability that a text tends to be positive or negative. Of course, the greater the probability, the higher the sentiment level. i The calculation idea of ​​severity (S) is to make an average of the sentiment tendency values ​​of all comments containing failure mode f to represent the average negative sentiment of consumers towards the failure mode. Represents failure mode f i The severity of the failure mode.

[0063] 3.2: Calculation of Occurrence (O)

[0064] The total number of all user comments is t, which can be regarded as a total of t times for this service. In the calculation of the severity (S) above, we mentioned that the failure mode f i There are 200 user comments We believe that this failure mode has occurred Therefore, we define the calculation formula of occurrence degree (O) as follows:

[0065]

[0066] Failure mode f iThe above formula can be understood as: we divide the number of occurrences of a failure mode by the number of service occurrences to calculate the occurrence degree.

[0067] 3.3: Calculation of Difficulty of Detection (D)

[0068] In the embodiment of the present invention, the calculation of the degree of difficulty of detection (D) applies the concept of information entropy. Information entropy solves the problem of quantitative measurement of information. The embodiment of the present invention provides the following calculation formula for information entropy:

[0069]

[0070] Among them, o(x i ) represents the random event x i Probability of occurrence.

[0071] The higher the information entropy of a text, the more information it contains, and thus the more difficult it is to retrieve. This is because the amount of information is too much, making it more difficult to retrieve. Based on the above information entropy calculation formula, the present invention defines the difficulty of detecting a failure mode f as follows:

[0072] Assume that failure mode f i There are a total of negative reviews For the jth comment Count the word frequency of each failure mode, recorded as t j (f i )(i=1,2,…,a)(j=1,2,…,m). We i The information entropy of the negative reviews is averaged and used as the difficulty of detection (D) in the FMEA framework. The calculation formula for the difficulty of detection (D) is as follows:

[0073]

[0074] Step 4: Calculate the objective weights of severity (S), occurrence (O), and difficulty of detection (D) respectively

[0075] The CRITIC method is an objective weighting method. Its basic concept is to determine the objective weights of indicators based on two basic concepts: first, comparative strength, which represents the difference in the values ​​of different evaluation schemes for the same indicator, expressed as standard deviation. That is, the size of the standardized deviation indicates the difference in the values ​​of different schemes within the same indicator; the larger the standard deviation, the greater the difference in the values ​​of the different schemes; second, the conflict between evaluation indicators. The conflict between indicators is based on the correlation between the indicators. If there is a strong positive correlation between two indicators, it means that the conflict between the two indicators is low.

[0076] The failure mode risk assessment method provided by the embodiment of the present invention uses the CRITIC method to calculate the objective weights of severity (S), occurrence (O), and difficulty of detection (D), which serve as the basis for comparing the risk severity of each failure mode. The specific details are as follows:

[0077] Assuming that p failure modes are finally extracted (p is a positive integer), then according to the calculation method in step 3, the data matrix X of the original S, O, and D values ​​is set to:

[0078]

[0079] Among them, the first, second and third columns represent the severity (S), occurrence (O), and difficulty of detection (D) values ​​of each failure mode, respectively, and each row represents the S, O, and D values ​​of a failure mode.

[0080] 4.1 Index normalization

[0081] In order to eliminate the order of magnitude differences between different failure modes, it is necessary to perform dimensionless processing on each evaluation index. Assuming that the original data matrix after dimensionless processing is A, the specific processing method is:

[0082]

[0083] Wherein, j=1, 2, 3 (1 represents S, 2 represents O, 3 represents D), a ij It is the data after normalization.

[0084] 4.2 Calculation of indicator variability

[0085] The variability of indicators is mainly determined by the standard deviation of the data. The larger the standard deviation, the more information the indicator reflects and the greater the evaluation strength of the indicator itself. The formula for calculating indicator variability is as follows:

[0086]

[0087]

[0088] Among them, S j That is, the variability of the indicators representing severity (S), occurrence (O), and difficulty of detection (D), j = 1, 2, 3. S j The larger it is, the stronger the variability of the j-th indicator is, and the greater its impact on the evaluation results.

[0089] 4.3 Calculation of indicator conflicts

[0090]

[0091] Among them, r ijRepresents the correlation coefficient between indicators i and j, R j Represents the index conflict. The greater the index conflict, the greater the impact of the index on the evaluation results. ij The calculation method is as follows:

[0092] r ij =cov(X i , X j ) / S i S j

[0093] X i , X j Represents the column vectors of index i and index j respectively, cov(X i , X j ) represents X i X j The covariance of S i , S j Represent the variance of index i and index j respectively.

[0094] 4.4 Calculating the information content of indicators

[0095] C j =S j ×R j

[0096] C j Represents the amount of information of index j. C j The larger the value is, the greater the amount of information contained in the jth evaluation index. The greater the role of this index in the entire evaluation index system, the more weight is assigned to it.

[0097] 4.5 Calculating objective weights

[0098] The objective weight w of the jth evaluation index j for:

[0099]

[0100] At this point, the objective weight calculation of severity (S), occurrence (O), and difficulty of detection (D) is completed. Multiply the weight of each indicator by the original S, O, and D value data matrix, where F ij =w i ×a ij Represents the values ​​in the weighted matrix.

[0101] The matrix V is used as the input matrix for failure mode risk severity assessment:

[0102]

[0103] Step 5: Calculate the risk severity of each failure mode

[0104] TOPSIS (Technique for Order Preference by Similarity to an Ideal Solution) is a commonly used decision-making framework that selects candidate solutions by finding the points closest to the optimal ideal solution and farthest from the worst ideal solution. It fully utilizes raw data, fully reflects the differences between solutions, and objectively and truthfully reflects the actual situation. It is authentic, intuitive, and reliable, and has no special requirements for the dimensionality of the sample data.

[0105] In one embodiment of the present invention, the TOPSIS comprehensive evaluation method is used to evaluate the importance of each failure mode, and the extracted failure modes are rated according to risk severity, thereby providing merchants or Internet platforms with the content and order of resolving the failure modes. The specific steps are as follows:

[0106] 5.1 Obtaining the positive ideal solution Z of the failure mode based on the weighted decision matrix + and negative ideal solution Z -

[0107]

[0108]

[0109] Among them, J + It is a benefit-type indicator, that is, the larger the indicator, the better; J - It is a cost-based indicator, meaning the smaller the indicator, the better. Obviously, the severity (S), occurrence (O), and difficulty of detection (D) in the failure mode risk assessment method are indicators where the larger the value, the higher the risk severity. Therefore, they can all be classified as benefit-based indicators.

[0110] 5.2 Calculate the distance between the vectors of S, O, and D values ​​of each failure mode and the positive and negative ideal solutions

[0111]

[0112]

[0113] in, and are the distances between the vectors of S, O, and D values ​​of each failure mode and the positive and negative ideal solutions, respectively.

[0114] 5.3 Calculate the relative progress between the vector of S, O, and D values ​​of each failure mode and the positive ideal solution, and sort the risk severity of each failure mode in descending order according to the relative progress.

[0115]

[0116] Relative post progress γ i The value of is between 0 and 1. The closer it is to 1, the greater the risk severity of the failure mode.

[0117] Through the above calculation steps, the risk severity ranking of each failure mode can be ultimately obtained, thereby assisting quality management personnel in carrying out targeted quality improvement work. It should be noted that the relevant steps in the existing FMEA analysis framework can also be used to assist in implementing the failure mode risk assessment method provided in the embodiment of the present invention. This is a conventional technical choice that can be made by those skilled in the art and will not be explained in detail here.

[0118] Compared to existing technologies, this invention creatively incorporates external internet review factors into the FMEA analysis framework. By mining internet review data, it enables product or service failure mode identification. Furthermore, this invention employs a unified processing method for internet review data. Compared to the subjective expert scoring process in existing FMEA analysis frameworks, this solves the problem of insufficient data and a single source, increases the objectivity of failure mode analysis results, and effectively assists quality management personnel in quality improvement efforts.

[0119] Furthermore, the embodiment of the present invention also provides a failure mode risk assessment system based on Internet comment mining, which can be applied to smart devices such as PCs. Figure 2 As shown, the failure mode risk assessment system at least includes:

[0120] The information collection and storage server is used to collect external internet review data, crawl negative user reviews of products or services (such as negative reviews on various internet platforms), and then clean the text data of these negative user reviews, set a stop word dictionary and perform text segmentation to obtain pre-processed text data. This pre-processed text data is imported into the information database for storage;

[0121] The core data processing server includes comment text mining module, risk matrix calculation module, CRITIC weight calculation module, and TOPSIS risk severity assessment module; among them,

[0122] The review text mining module is used to obtain pre-processed text data from the information database, perform text feature extraction on the text data, obtain some highly critical words, and select the words that best describe the failure mode as the failure mode of the product or service;

[0123] The risk matrix calculation module is used to calculate the severity (S), occurrence (O), and difficulty of detection (D) of each failure mode;

[0124] The severity (S) is calculated using the following formula:

[0125]

[0126] The occurrence degree (O) is calculated by the following formula:

[0127]

[0128] The difficulty of detection (D) is calculated using the following formula:

[0129]

[0130] CRITIC weight calculation module, used to calculate the objective weights of severity (S), occurrence (O), and difficulty of detection (D);

[0131] The TOPSIS risk severity assessment module is used to calculate the risk severity of failure modes. It obtains the positive and negative ideal solutions of the evaluation target based on the weighted decision matrix. It calculates the distance between the solution value of each failure mode and the positive and negative ideal values ​​respectively. It calculates the relative progress between the solution value of each failure mode and the positive ideal solution, and ranks the risk severity of each failure mode in descending order of relative progress.

[0132] It should be noted that the above-mentioned information collection and storage server and core data processing server can be implemented by a PC or a dedicated server, and each module therein can be at least partially implemented as a hardware circuit, such as a field programmable gate array (FPGA), a programmable logic array (PLA), a system on a chip, a system on a substrate, a system on a package, an application-specific integrated circuit (ASIC), or can be implemented by hardware or firmware in any other reasonable way of integrating or packaging the circuit, or implemented in any one of the three implementation methods of software, hardware and firmware or in an appropriate combination of any of them. Alternatively, one or more of the various modules in the embodiments of the present invention can be at least partially implemented as a computer program module, which can perform the corresponding function when the computer program module is run. Alternatively, at least part of the functions of one or more of these modules can be combined with at least part of the functions of other modules and implemented in one module.

[0133] The above describes in detail the failure mode risk assessment method and system based on internet review mining provided by the present invention. Any obvious modifications made to the present invention by a person skilled in the art without departing from the essence of the present invention will fall within the scope of protection of the present invention.

Claims

1. A failure mode risk assessment method based on Internet review mining, characterized by The steps include: Step 1: Obtain negative user reviews from after-sales user comments and perform preprocessing; set a stop word dictionary and perform word segmentation on each text, remove stop words, and obtain preprocessed text data; Step 2: Extract text features from the preprocessed text data and evaluate their importance based on the TF-IDF algorithm to obtain keywords. Select the words that best describe the failure mode as the failure mode of the product or service. Step 3: Calculate the severity, occurrence, and difficulty of detection of each failure mode; Calculate the average value of the information entropy of the negative reviews and use this value as the difficulty of detection; The severity is calculated using the following formula: ＝－ / ， ； in, Indicates failure mode The severity S, is the sentiment tendency value; The occurrence degree is calculated by the following formula: ； in, Indicates failure mode Occurrence rate O, is the number of occurrences of the failure mode; The difficulty of detection is calculated by the following formula: ； in, Indicates failure mode The difficulty of detection D, Is the failure mode The number of negative reviews, is the word frequency of each failure mode, ; Step 4: Calculate the objective weights of severity, occurrence, and difficulty of detection using the CRITIC method, which includes the following sub-steps: dimensionless processing of each evaluation indicator; calculating indicator variability; calculating indicator conflict; calculating indicator information content; calculating objective weights, and updating the weighted decision matrix based on the objective weights; Step 5: Calculate the risk severity of each failure mode.

2. The failure mode risk assessment method according to claim 1, characterized in that The step 5 includes the following sub-steps: Obtaining a positive ideal solution and a negative ideal solution of the evaluation target according to the weighted decision matrix; Calculate the distance between the solution value of each failure mode and the positive and negative ideal values; Calculate the relative progress between the solution value of each failure mode and the positive ideal solution, and rank the risk severity of each failure mode in descending order according to the relative progress.

3. A failure mode risk assessment system based on Internet review mining, used to implement the failure mode risk assessment method according to claim 1 or 2, characterized in that include: An information collection and storage server is used to collect external Internet review data, crawl negative user reviews of products or services, clean the text data of the negative user reviews, set a stop word dictionary and perform text segmentation, obtain pre-processed text data, and import it into the information database for storage; The core data processing server includes comment text mining module, risk matrix calculation module, CRITIC weight calculation module, and TOPSIS risk severity assessment module; among them, The review text mining module is used to obtain pre-processed text data from the information database, perform text feature extraction on the text data, and select the words that best describe the failure mode as the failure mode of the product or service; The risk matrix calculation module is used to calculate the severity, occurrence and difficulty of detection of each failure mode; The CRITIC weight calculation module is used to calculate the objective weights of severity, occurrence, and difficulty of detection; The TOPSIS risk severity assessment module is used to calculate the risk severity of the failure mode.

4. The failure mode risk assessment system according to claim 3, wherein: In the TOPSIS risk severity assessment module, the positive ideal solution and negative ideal solution of the evaluation target are obtained according to the weighted decision matrix; the distance between the solution value of each failure mode and the positive and negative ideal values ​​is calculated; the relative progress between the solution value of each failure mode and the positive ideal solution is calculated, and the risk severity of each failure mode is ranked in descending order according to the relative progress.

Images