Train control system credibility evaluation method based on fuzzy grey clustering

Abstract

The invention discloses a train control system credibility evaluation method based on fuzzy grey clustering. Comprising the following steps: establishing a train control system credibility evaluation hierarchical evaluation index system; respectively calculating subjective and objective weights of each index of the sub-index layers by adopting an AHP method and an entropy weight method, and calculating a comprehensive weight of each index through a combined weighting method; and determining a whitening weight function, calculating membership degree vectors of all indexes by adopting grey clustering, integrating the membership degree vectors into a fuzzy evaluation matrix, and calculating credibility grade membership degrees of the sub-target layers and the target layer by utilizing fuzzy comprehensive evaluation, so as to finally obtain the credibility grade of the train control system. According to the method, the subjectivity of the credibility evaluation of the train control system can be reduced, the objectivity, comprehensiveness and accuracy of the credibility evaluation of the train control system are improved, and a scientific and reasonable method means is provided for the credibility evaluation of the train control system.

Description

Technical Field

[0001] The present invention relates to the technical field of system trustworthiness evaluation and management, and particularly relates to a method for evaluating the trustworthiness of a train control system based on fuzzy grey clustering. Background Art

[0002] The train control system is one of the core technical equipments of rail transit, mainly responsible for commanding and controlling the safe and efficient operation of trains, and is usually called the "central nervous system" of rail transit. With the rapid development of technologies such as computers, wireless communications, satellite positioning, microelectronics, and intelligent control, and their in-depth application in the train control system, the train control system presents characteristics such as deep integration of software and hardware, large software scale, intensive safety requirements, and complex network structure.

[0003] Trustworthiness is a comprehensive index for measuring the performance of a train control system. Based on reliability, availability, maintainability, and safety, it extends more dimensional attributes such as integrity, recoverability, testability, confidentiality, and service maintainability. Its core connotation refers to that the dynamic behavior and results of the system conform to the expected goals, can adapt to certain environmental changes, have the ability to resist abnormal situations, are not damaged by accidental or malicious reasons, and the system can still operate safely and reliably when encountering external interference. Scientifically, accurately, and comprehensively evaluating the trustworthiness of a train control system is of great significance for ensuring the safe and reliable operation of the train control system in a complex environment.

[0004] At present, the train control system mainly conducts system guarantee and evaluation through means such as safety certification, system testing, and RAMS management, mainly focusing on reliability, availability, maintainability, and safety, and cannot comprehensively and accurately reflect the trustworthy operation ability of the train control system in a complex and dynamic environment. Especially when facing a large number of uncertain factors and fuzzy information, the accuracy and effectiveness of traditional methods have certain limitations. Therefore, there is an urgent need for a scientific, reasonable, and train control system trustworthiness evaluation method that can handle fuzzy and uncertain information to provide more accurate, objective, and comprehensive evaluation results, so as to provide strong guarantee for the safe and reliable operation of the train control system. Summary of the Invention

[0005] In view of the defects of the prior art, the present invention provides a method for evaluating the trustworthiness of a train control system based on fuzzy grey clustering. The aim is to establish a scientific, accurate, and comprehensive trustworthiness evaluation method through advanced technologies such as the comprehensive combination weighting method and fuzzy grey clustering, so as to provide strong support for ensuring the safe and reliable operation of the train control system.

[0006] In order to achieve the above invention purposes, the technical solutions adopted by the present invention are as follows:

[0007] A method for evaluating the trustworthiness of a train control system based on fuzzy grey clustering, comprising the following steps:

[0008] A1. Construct a hierarchical evaluation index system for the credibility of the train control system and determine the measurement criteria for the sub-index layer;

[0009] A2. Calculate the subjective and objective weights of each sub-index using the Analytic Hierarchy Process (AHP) and the entropy weight method respectively, and calculate the comprehensive weight using the combined weighting method based on the subjective and objective weights;

[0010] A3. Divide the credibility levels of the train control system, determine the whitening weight functions for each credibility level, and use grey clustering to obtain the membership degree vectors of each index in combination with the sub-index measurement values;

[0011] A4. Combine the membership degree vectors of each index to construct a fuzzy evaluation matrix for the index layer, and through secondary fuzzy synthesis, obtain the membership degree vectors of the target layer belonging to each credibility level;

[0012] A5. Use the principle of maximum membership degree to evaluate the credibility of the train control system.

[0013] Furthermore, in step A1, the credibility evaluation index system of the train control system is divided into: target layer, sub-target layer, index layer and sub-index layer; measurement criteria are formulated for the sub-index layer in step A1.

[0014] Furthermore, in step A2, the calculation of three types of weights is completed, including:

[0015] Adopt the AHP method, use the 1-9 scale method to compare each index and sub-index pairwise, construct a judgment matrix, and then solve the subjective weights of the index and sub-index by the square root method; further, adopt the entropy weight method, combine multiple groups of sub-index measurement data, process and obtain the normalized matrix of sub-indexes, calculate the information entropy, and obtain the objective weight; finally, calculate the comprehensive weight using the combined weighting method for the subjective and objective weights, using the formula:

[0016]

[0017] In the formula: ε is the combined weight coefficient; represents the comprehensive weight of the m-th sub-index of the l-th index of the r-th sub-target; represents the subjective weight of the m-th sub-index of the l-th index of the r-th sub-target; represents the objective weight of the m-th sub-index of the l-th index of the r-th sub-target.

[0018] Furthermore, the formula for calculating the subjective weights of the index and sub-index by the AHP method is:

[0019]

[0020] where: w i is the subjective weight of the i-th judgment index; P is the judgment matrix; p ij is the element in the i-th row and j-th column of the judgment matrix; n is the order of the judgment matrix; W is the subjective weight vector of the judgment index; (W·P) i is the i-th component of W·P; λ max is the maximum eigenvalue of the judgment matrix.

[0021] Perform a consistency test on the obtained subjective weight vector. The consistency test formula is:

[0022]

[0023] where: λ max is the maximum eigenvalue of the judgment matrix; n is the order of the judgment matrix; CI is the general consistency deviation degree index; RI is the average random consistency index, used to eliminate the influence of the order; CR is used to evaluate the consistency degree of the judgment matrix. If CR < 0.1, the obtained subjective weight vector passes the consistency test. If CR ≥ 0.1, it is necessary to re-score according to the 1-9 scale method until the consistency test is passed.

[0024] Furthermore, the entropy weight method calculates the objective weight of the sub-index, and the calculation formula is:

[0025]

[0026]

[0027] where: r im represents the normalized data value of sub-index m on sample i; y im represents the proportion of r im ; d represents the number of sub-indices; β represents the information entropy coefficient; t is the number of data samples; s m represents the information entropy;.

[0028] The calculation formula of r im is as follows:

[0029]

[0030] where: h im represents the original data value of sub-index m on sample i; represents the minimum value of all samples of sub-index m, represents the maximum value of all samples of sub-index m; u m represents the optimal value of sub-index m.

[0031] Further, in step A3, the credibility of the train control system is divided into four levels: highly credible, relatively credible, weakly credible, and non-credible. The whitening weight functions for each credibility level are set, and combined with the average data values of the sub-indicators, the hybrid center point triangular whitening weight function is used to calculate the whitening values of each sub-indicator belonging to each grey class.

[0032] Further, the calculation of the whitening values of each sub-indicator belonging to each grey class by using the hybrid center point triangular whitening weight function is specifically as follows:

[0033]

[0034]

[0035] In the formula: is the average data value of the m-th sub-indicator; λ 1 , λ 2 , λ 3 , λ 4 are the center points of each grey class.

[0036] Further, the secondary fuzzy comprehensive evaluation in step A4 includes: integrating the membership degree vectors of each index into a fuzzy evaluation matrix, and performing fuzzy synthesis with the weight vector of the index layer to obtain the membership degree vector of the sub-goal layer; integrating the membership degree vector of the sub-goal layer into the sub-goal layer fuzzy evaluation matrix, and performing fuzzy synthesis with the weight vector of the sub-goal layer to obtain the membership degree vector of the goal layer.

[0037] Compared with the prior art, the advantages of the present invention are as follows:

[0038] 1. Improvement of the comprehensive weight calculation method: The present invention obtains the subjective weight of the sub-indicators by the AHP method, combines the entropy weight method to calculate the objective weight of the sub-indicators, and finally combines the subjective and objective weights through the combined weighting method to form the comprehensive weight. This method overcomes the limitations of single weight calculation in the traditional method, avoids the subjectivity of the evaluation results, reflects the influencing factors of each sub-indicator more objectively and comprehensively, and improves the accuracy and reliability of the credibility evaluation results.

[0039] 2. Innovative application of the fuzzy grey clustering method: The present invention combines the fuzzy comprehensive evaluation method with the grey clustering method, which can effectively solve the problems of information uncertainty, data scarcity, and clustering accuracy in the traditional fuzzy comprehensive evaluation method. Through fuzzy grey clustering, it can better process the fuzzy information with uncertainty, improve the accuracy of the clustering results, and thus provide more accurate evaluation results and credibility analysis for decision-makers.

[0040] 3. Multi-level credibility assessment of train control system: The present invention combines the grey clustering method with the fuzzy comprehensive evaluation method, and adopts a three-level evaluation method to hierarchically evaluate the train control system. It can not only comprehensively evaluate the credibility of the entire train control system, but also independently evaluate the credibility of each component device of the system. This hierarchical evaluation method helps to deeply understand the potential risks of each subsystem and device, timely discover and fix problems, effectively improve the safety and stability of the train control system, and ensure the efficient and reliable operation of the train control system. Brief Description of the Drawings

[0041] Figure 1 It is the flowchart of the credibility assessment of the train control system in the embodiment of the present invention;

[0042] Figure 2 It is the credibility assessment index system of the train control system in the embodiment of the present invention;

[0043] Figure 3 It is the mixed center point triangular whitenization weight function of the fuzzy grey clustering in the embodiment of the present invention. Detailed Embodiment

[0044] To make the purpose, technical solutions and advantages of the present invention clearer and more understandable, the following lists embodiments according to the drawings and further elaborates on the present invention in detail.

[0045] As Figure 1 shown, a method for assessing the credibility of a train control system based on fuzzy grey clustering includes the following steps:

[0046] A1. Construct a hierarchical evaluation index system for the credibility of the train control system and determine the measurement criteria for the sub-index layer.

[0047] The index system includes an objective layer, a sub-objective layer, an index layer and a sub-index layer. The objective layer is the train control system, denoted as X, and its credibility index system can be decomposed into a sub-objectives (i.e., each device), denoted as X r (r = 1, 2, 3,..., a). In this embodiment, on-vehicle equipment, radio block center, temporary speed limit server, and integrated train control and interlocking equipment are used as sub-objectives. Each sub-objective contains the same indexes, denoted as where n takes 5 in this embodiment, which are hardware credibility, software credibility, network credibility, system test result credibility and quality assurance process credibility. Then, corresponding sub-indexes are set for the indexes, denoted as Among them, d takes 3 or 4 in this embodiment. Hardware credibility includes availability, reliability, maintainability, and security; software credibility includes functionality, reliability, maintainability, and security; network credibility includes network performance, information security, and service quality credibility; system test result credibility includes the consistency between the test plan and execution, the comprehensiveness of the test content, the test coverage, as well as the test environment, methods, and tools; quality assurance process credibility includes process document credibility, process control credibility, environmental affinity, and process change credibility. The specific corresponding relationships are as Figure 2 shown.

[0048] The measurement criteria for the sub-index layer are shown in Table 1.

[0049] Table 1 Sub-index value judgment criteria

[0050]

[0051]

[0052] A2. Calculate the subjective and objective weights of the sub-indicators using the analytic hierarchy process and the entropy weight method respectively, and calculate the comprehensive weight using the combined weighting method based on the subjective and objective weights.

[0053] Based on the constructed credibility evaluation index system of the train control system, construct the judgment matrix of each level of indicators. Multiple experienced experts use the 1-9 scale method to compare the elements of the same level pairwise. The 1-9 scale method and its meaning are shown in Table 2.

[0054] Table 2 1-9 scale and its meaning

[0055]

[0056] For the sub-goal layer, in practice, it is considered that each device is equally important (all elements of the judgment matrix are 1), so the weight vector of the sub-goal layer is denoted as W 0 =(w r ) 1×a , where w r is the weight corresponding to the sub-goal X r and is equal to 1 / a, where a takes 4 in this embodiment. For the index layer, since the indicators under each sub-goal are the same, they can be denoted as the same index judgment matrix P r =(p lk ) n×n , where 1≤l,k≤n, n takes 5 in this embodiment; p lk >0 and p ll =1. Then use the square root method to solve the subjective weight of the index and the maximum eigenvalue As shown in formula (1). The order of the matrix and the differences in subjective thinking will affect the constructed judgment matrix. Therefore, it is necessary to conduct a consistency test on the obtained subjective weight vector, as shown in formula (2). If the consistency test fails, it is necessary to re-score according to the 1-9 scale method until the consistency test passes. Since multiple experts are involved, it is also necessary to average the subjective weights of each index obtained by multiple experts according to the AHP method to obtain the final subjective weight of the index Similarly, a judgment matrix P is constructed for each sub-index under each index l =(p mq ) d×d , where 1 ≤ m, q ≤ d, d is the number of sub-indices corresponding to each index, and d takes 3 or 4 in this embodiment; p mq > 0 and p mm = 1. The square root method is also used to solve the subjective weights of each sub-index corresponding to each index and the maximum eigenvalue of each sub-index judgment matrix As shown in formula (1). Then, a consistency test is conducted on the obtained subjective weight vector, as shown in formula (2). If the consistency test fails, it is necessary to re-score according to the 1-9 scale method until the consistency test passes. Finally, the subjective weights of each sub-index obtained by multiple experts according to the AHP method are averaged to obtain the final subjective weight of the sub-index

[0057]

[0058] In the formula: w i is the subjective weight of the i-th judgment index; P is the judgment matrix; p ij is the element in the i-th row and j-th column of the judgment matrix; n is the order of the judgment matrix; W is the subjective weight vector of the judgment index; (W·P) i is the i-th component of W·P; λ max is the maximum eigenvalue of the judgment matrix; CI is the general consistency deviation degree index; RI is the average random consistency index used to eliminate the influence of the order, and the values of RI are shown in Table 3; CR is used to evaluate the consistency degree of the judgment matrix. If CR < 0.1, the obtained subjective weight vector passes the consistency test. If CR ≥ 0.1, it is necessary to re-score according to the 1-9 scale method until the consistency test passes

[0059] Table 3 RI value table for 1-8 order judgment matrices

[0060]

[0061] For the train control system, the entropy weight method is used to calculate the objective weight. There are d sub-indicators under each indicator, and a total of t experts score each sub-indicator according to the measurement criteria, that is, there are t data samples in total, and the original data matrix of the indicator is denoted as where 1 ≤ i ≤ t, 1 ≤ m ≤ d, h im represents the original data value of the m-th sub-indicator on sample i, that is, the score value of the i-th expert for the m-th sub-indicator. Since the unit scale types of each sub-indicator are different, it is necessary to perform normalization preprocessing on h im . Select a suitable normalization formula according to the attributes of the sub-indicators, such as formula (3). After the above preprocessing, n normalized matrices are obtained. Then calculate the information entropy s m of the m-th sub-indicator, such as formula (4), and finally calculate the objective weight value of the sub-indicator as formula (5).

[0062]

[0063] In the formula: u m is the best value of the m-th sub-indicator; d represents the number of sub-indicators; β represents the information entropy coefficient; t is the number of data samples; h im represents the original data value of the sub-indicator m on sample i; r im represents the normalized data value of the m-th sub-indicator on sample i; y im represents the proportion of r im ; when y im = 0, let y im lny im = 0; And there is

[0064] The AHP method scores the index judgment matrix according to expert experience, which is easily interfered by subjective factors and ignores the information carried by the index data; the entropy weight method uses the information carried by the data to allocate the index weights, making the evaluation results objective. However, the entropy weight method cannot reflect the importance differences of each index of the system itself, and will also cause unreasonable weight allocation. Therefore, in order to comprehensively consider the influence of subjective and objective factors, the combined weighting method is used to eliminate the deficiencies of the above two weighting methods. Combining the subjective weight and the objective weight of each sub-indicator obtained by the AHP method and the entropy weight method respectively, calculate the comprehensive weight of the sub-indicator as formula (6).

[0065]

[0066] Where: ε is the combined weight coefficient, which is given by relevant experts based on experience. In this embodiment, ε = 0.7 is taken.

[0067] A3. Divide the credibility levels of the train control system, determine the whitenization weight functions of each credibility level, and combine the measurement values of the sub-indicators. Use grey clustering to obtain the membership degree vectors of each indicator.

[0068] Table 4 Credibility Evaluation Criteria

[0069]

[0070]

[0071] In this embodiment, the credibility levels of the train control system are divided into 4 levels: "high credibility, relatively credible, weakly credible, and non-credible". The specific evaluation criteria are described in Table 4. The set of credibility levels is expressed as V = [v 1 , v 2 , v 3 , v 4 .

[0072] Traditional fuzzy comprehensive evaluation methods usually require relatively complete and quantitative data to construct a fuzzy evaluation matrix. Introducing grey clustering analysis can effectively solve the problems of traditional fuzzy comprehensive evaluation methods in terms of information uncertainty, data scarcity, and clustering accuracy. Therefore, this embodiment selects fuzzy grey clustering for the credibility evaluation of the train control system.

[0073] First, establish the whitenization weight functions of each credibility level. The whitenization weight function, also known as the grey clustering function, is used to describe the degree to which an indicator belongs to each grey class, and its value is between [0, 1]. In this embodiment, a mixed-type center point triangular whitenization weight function is introduced. Four grey classes are set corresponding to the credibility levels, and the whitenization weight function of the center point of each grey class is 1. Since the measurement values of the sub-indicators in this example are all obtained through expert scoring, the boundary values of each grey class of the sub-indicators are the same as the measurement boundary values of each credibility level. The grey class ranges are shown in Table 5. For the previously determined index system, calculate the center points λ 1 , λ 2 , λ 3 , λ 4 of the 4 grey classes respectively. Take λ j = (y + z) / 2, where y and z are the boundary values at both ends of the credibility level v j . And combine the average data value of the sub-indicator to calculate the 4-element whitenization weight function value between the sub-indicator and the credibility level v j (j = 1, 2, 3, 4), as shown in formula (7-10). The mixed-type center point triangular whitenization weight function is as Figure 3 shown.

[0074] Table 5 Ranges of Each Grey Class

[0075]

[0076]

[0077] "Highly credible" level whitenization weight function

[0078]

[0079] "Relatively credible" level whitenization weight function

[0080]

[0081] "Weakly credible" level whitenization weight function

[0082]

[0083] "Uncredible" level whitenization weight function

[0084] In the formula: is the average data value of the m-th sub-index, obtained by averaging the t normalized data values r of the m-th sub-index im ; λ 1 , λ 2 , λ 3 , λ 4 are the center points of each grey class.

[0085] After obtaining the whitenization values of the sub-index belonging to each grey class through the whitenization weight function, the clustering coefficient of the index belonging to the j-th grey class is calculated by combining the sub-index weight as shown in formula (11). Thus, the clustering vector of the l-th index can be obtained That is, the membership degree vector of the l-th index belonging to each credibility level. That is, the membership degree vector of the l-th index belonging to each credibility level.

[0086]

[0087] In the formula: is the whitenization weight function value of the sub-index belonging to the j-th grey class.

[0088] A4. Combine the membership degree vectors of each index to construct the fuzzy evaluation matrix of the index layer. Through secondary fuzzy synthesis, obtain the membership degree vector of the target layer belonging to each credibility level.

[0089] Integrate the membership degree vectors of each index belonging to each credibility level into a matrix A r and call this matrix the fuzzy evaluation matrix of the index layer. Combine the index layer weights calculated by the analytic hierarchy process with the fuzzy evaluation matrix A rPerform fuzzy synthesis to obtain the membership degree vector B of the r-th sub-goal belonging to each credibility level r , as shown in formula (12). Complete the above calculation steps for all sub-goals, and integrate the calculated membership degree vectors of the a sub-goals belonging to each credibility level into the sub-goal layer fuzzy evaluation matrix A 0 . Then, the sub-goal layer weight vector W 0 and the sub-goal layer fuzzy evaluation matrix A 0 are subjected to fuzzy synthesis to obtain the membership degree vector B of the goal belonging to each credibility level 0 , as shown in formula (12).

[0090]

[0091] In the formula, is the "product-sum" fuzzy operator

[0092] A5. Use the maximum membership degree principle to conduct credibility assessment on the train control system.

[0093] The method according to the present invention can be implemented in hardware, firmware, or be implemented as software or computer code that can be stored in a recording medium (such as CD ROM, RAM, floppy disk, hard disk, or magneto-optical disk), or be implemented as computer code originally stored in a remote recording medium or a non-transitory machine-readable medium and downloaded through a network and to be stored in a local recording medium, so that the method described herein can be stored in such software processing on a recording medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware (such as ASIC or FPGA). It can be understood that a computer, a processor, a microprocessor controller, or programmable hardware includes a storage component (such as RAM, ROM, flash memory, etc.) that can store or receive software or computer code. When the software or computer code is accessed and executed by the computer, the processor, or the hardware, the credibility assessment method of the train control system described herein is implemented. In addition, when a general-purpose computer accesses the code for implementing the processing shown herein, the execution of the code converts the general-purpose computer into a dedicated computer for executing the processing shown herein.

[0094] Those of ordinary skill in the art will realize that the embodiments described herein are for helping readers understand the implementation method of the present invention, and it should be understood that the protection scope of the present invention is not limited to such specific statements and embodiments. Those of ordinary skill in the art can make various other specific deformations and combinations that do not deviate from the essence of the present invention based on the technical revelations disclosed in the present invention, and these deformations and combinations are still within the protection scope of the present invention.

Claims

1. A train control system credibility assessment method based on fuzzy grey clustering, characterized in that: The following steps are involved: A1. Construct a hierarchical evaluation index system for the credibility of the train control system and determine the metric standards for the sub-index layer; A2. Calculate the subjective and objective weights of each sub-indicator using the analytic hierarchy process (AHP) and entropy weight method respectively, and calculate the comprehensive weight based on the subjective and objective weights using the combined weighting method; A3. Divide the credibility level of the train control system, determine the whitening weight function of each credibility level, combine the sub-indicator measurement values, and use grey clustering to obtain the membership vector of each indicator; A4. Combine the index membership vectors to construct the index layer fuzzy evaluation matrix, and obtain the target layer membership vectors of each credibility level through secondary fuzzy synthesis; A5. Use the maximum membership principle to conduct credibility assessment on the train control system.

2. The train control system credibility assessment method according to claim 1, characterized in that: In the step A1, the train control system credibility evaluation index system is divided into: target layer, sub-target layer, index layer and sub-index layer; and a metric standard is established for the sub-index layer in the step A1.

3. The train control system credibility assessment method according to claim 1, characterized in that: The calculation of three types of weights is completed in step A2, including: The AHP method is used to compare the indicators and sub-indicators pairwise using the 1-9 scaling method to construct a judgment matrix, and then the subjective weights of the indicators and sub-indicators are solved by the square root method; further, the entropy weight method is used to combine multiple groups of sub-indicator measurement data, process and obtain the sub-indicator normalized matrix, calculate the information entropy, and obtain the objective weight; finally, the combined weighting method is used to calculate the comprehensive weight of the subjective and objective weights, using the formula: Where: ε is the combined weight coefficient; represents the comprehensive weight of the mth sub-indicator of the lth indicator of the rth sub-goal; represents the subjective weight of the mth sub-indicator of the lth indicator of the rth sub-goal; It represents the objective weight of the mth sub-indicator of the lth indicator of the rth sub-goal.

4. The train control system credibility assessment method according to claim 3 is characterized by: The AHP method calculates the subjective weights of indicators and sub-indicators using the following formula: Where: w i is the subjective weight of the i-th judgment indicator; P is the judgment matrix; p ij is the element in the i-th row and j-th column of the judgment matrix; n is the order of the judgment matrix; W is the subjective weight vector of the judgment index; (W·P) i is the i-th component of W·P; max is the maximum eigenvalue of the judgment matrix.

5. The train control system credibility assessment method according to claim 4 is characterized by: The obtained subjective weight vector is subjected to consistency check, and the consistency check formula is: Where: max is the maximum eigenvalue of the judgment matrix; n is the order of the judgment matrix; CI is the general consistency deviation index; RI is the average random consistency index, which is used to eliminate the influence of the order; CR is used to evaluate the consistency of the judgment matrix. If CR is less than 0.1, the obtained subjective weight vector passes the consistency test. If CR is greater than or equal to 0.1, it needs to be re-scored according to the 1-9 scaling method until it passes the consistency test.

6. The train control system credibility assessment method according to claim 3 is characterized by: The entropy weight method described above calculates the objective weight of the sub-indicator, and the calculation formula is: Where: r im represents the normalized data value of sub-indicator m on sample i; y im Represents r im The proportion of; d represents the number of sub-indicators; β represents the information entropy coefficient; t represents the number of data samples; s m represents information entropy; r im The calculation formula is as follows: Where: h im Represents the original data value of sub-indicator m on sample i; Represents the minimum value of all samples of sub-index m, Indicates the maximum value of all samples of sub-index m; u m Represents the optimal value of sub-index m.

7. The train control system credibility assessment method according to claim 1, characterized in that: In step A3, the credibility of the train control system is divided into four levels: highly credible, relatively credible, weakly credible and uncredible; The whitening weight function of each credibility level is set, and the whitening value of each sub-indicator belonging to each gray class is calculated using a hybrid center point triangle whitening weight function combined with the average data value of the sub-indicator.

8. The train control system credibility assessment method according to claim 7 is characterized by: The hybrid center point triangle whitening weight function is used to calculate the whitening value of each sub-index belonging to each gray class. The specific formula is as follows: Where: is the average data value of the mth sub-indicator; λ1, λ2, λ3, λ4 are the center points of each gray class.

9. The train control system credibility assessment method according to claim 1, characterized in that: The secondary fuzzy synthesis in step A4 includes: integrating the index membership vectors into a fuzzy judgment matrix, and performing fuzzy synthesis with the weight vector of the index layer to obtain the sub-target layer membership vector; integrating the sub-target layer membership vectors into a sub-target layer fuzzy judgment matrix, and performing fuzzy synthesis with the weight vector of the sub-target layer to obtain the target layer membership vector.

Images