Methods and systems for assessing accident levels of transport aircraft weighing over 50 tons

Abstract

The present invention discloses a method and system for evaluating the accident grade of transport aircraft weighing more than 50 tons, and the method includes: S1, constructing an accident repair cost index parameter library, collecting the evaluation value corresponding to the transport aircraft, and statistically analyzing the historical accident repair costs of the transport aircraft according to the cost index of the accident repair cost index parameter library to obtain the repair costs corresponding to the transport aircraft; S2, constructing a K-Means clustering algorithm model, and dividing the initial repair cost ratio range corresponding to three categories based on the clustering results; S3, constructing a verification algorithm model based on a multi-objective comprehensive preference function, and based on the verification category range divided by the multi-objective comprehensive preference function, the verification category to which the new accident aircraft belongs is obtained. The present invention can realize the grade evaluation of transport aircraft accidents based on direct economic losses, realizes auxiliary accident evaluation based on data analysis and processing, and is convenient for scientific division and distinction of accident sizes.

Description

Technical Field

[0001] The present invention relates to the field of evaluating accident levels by using repair cost ratio, and in particular to an accident level evaluation method and system for transport aviation aircraft weighing more than 50 tons. Background Art

[0002] With the rising cost of transport aircraft, especially transport aircraft over 50 tons, which generally cost more than 600 million yuan, or even more than 1 billion yuan, general damage to transport aircraft over 50 tons can easily exceed the economic loss amount of the particularly serious accident level. For example, if a bird damages an engine of a B777 aircraft, the direct economic loss may exceed 100 million yuan, which will be classified as a particularly serious accident. Therefore, it is particularly important to construct a method to assist in evaluating the accident level of transport aircraft over 50 tons based on direct economic losses. Auxiliary classification and evaluation of the accident level of transport aircraft over 50 tons is convenient for distinguishing the size of accidents, which has practical evaluation significance. Summary of the invention

[0003] The purpose of the present invention is to solve the technical problems of the background technology and to provide a method and system for evaluating the accident level of transport aircraft weighing more than 50 tons. A historical clustering analysis of transport aircraft is performed through a K-Means clustering algorithm model to obtain a number of initial repair cost ratio ranges, and then data of transport aircraft under new accidents are collected to obtain validation categories through a verification algorithm model, thereby realizing a level evaluation of transport aircraft accidents based on direct economic losses, and realizing auxiliary accident evaluation based on data analysis and processing, which is convenient for scientific division and differentiation of accident sizes.

[0004] The purpose of the present invention is achieved through the following technical solutions:

[0005] A method for evaluating accident levels of transport aviation aircraft weighing more than 50 tons, the method comprising:

[0006] S1. Construct an accident repair cost index parameter library, which consists of multiple cost indicators divided by levels; collect the assessment value corresponding to the transport aircraft, and statistically analyze the historical accident repair costs of the transport aircraft according to the cost indicators in the accident repair cost index parameter library to obtain the repair costs corresponding to the transport aircraft, and obtain a transport aircraft repair cost database. The data in the transport aircraft repair cost database are stored in association with aircraft model, assessment value, key accident information, repair cost, repair cost ratio, and category.

[0007] S2. Construct a K-Means clustering algorithm model, set the number of clusters to three and three initial cluster centers, input the data in the transport aviation aircraft repair cost database into the K-Means clustering algorithm model, obtain clustering results under three categories, and divide the initial repair cost ratio ranges corresponding to the three categories based on the clustering results;

[0008] The initial repair cost ratio range for Category 1 is: P1≤initial repair cost ratio range <P2;

[0009] The initial repair cost ratio range for Category 2 is: P2 ≤ initial repair cost ratio range <P3;

[0010] The initial repair cost ratio range for Category 3 is: initial repair cost ratio range ≥ P3;

[0011] S3. Collect the assessed value Q0 of the transport aircraft under the new accident, calculate the repair cost Q1 corresponding to the transport aircraft under the new accident according to the cost index of the accident repair cost index parameter library, and calculate the repair cost ratio P x ; Based on the initial repair cost ratio range in step S2, the initial category K1 to which the transport aviation aircraft belongs under the new accident is obtained;

[0012] S4. Construct a verification algorithm model based on a multi-objective comprehensive preference function. The formula of the multi-objective comprehensive preference function is as follows:

[0013]

[0014] Among them, g represents the target value of the repair cost Q1, g i (i=1, 2, 3) is the boundary point of the preference interval, i=1, 2, 3; P is the preference function;

[0015] g i (i=1, 2, 3) are calculated as follows:

[0016] g1=P1×Q0,

[0017] g2=P2×Q0,

[0018] g3=P3×Q0,

[0019] Thus, P(g1), P(g2), and P(g3) are calculated; the validation category range based on the multi-objective comprehensive preference function is:

[0020] The range of category 1 is: P(g1)≤P(g) <P(g2);

[0021] The range of category 2 is: P(g2)≤P(g) <P(g3);

[0022] The range of category three is: P(g) ≥ P(g3);

[0023] Substitute the repair cost Q1 as the target value g into the multi-objective comprehensive preference function formula to calculate P(g), and then obtain the effectiveness category K2 to which P(g) belongs according to the effectiveness category range.

[0024] In order to better implement the present invention, the K-Means clustering algorithm model calculates the Euclidean distance between the remaining data objects and the cluster center based on the initial cluster center, and distributes the data objects to the clusters corresponding to the cluster centers, and then calculates the average value of the data objects in each cluster as the new cluster center, and performs the next iteration until the cluster center no longer changes or the maximum number of iterations is reached; the Euclidean distance calculation formula between the data object and the cluster center is:

[0025]

[0026] Among them, x is the data object, C i is the i-th cluster center, m is the dimension of the data object, x j is the jth attribute value of x, C ij C i The j-th attribute value of .

[0027] Preferably, the number of iterations of the K-Means clustering algorithm model is constrained by the following SSE formula:

[0028]

[0029] Where k is the number of clusters, and its value is 3;

[0030] When the sum of squared errors (SSE) no longer changes, the iteration stops.

[0031] Preferably, part or all of the data in the transport aviation aircraft repair cost database is sample data, and the categories in the sample data are input category labels; in step S2, the three initial cluster centers are selected as sample data containing category labels.

[0032] Preferably, the validation category K2 obtained in step S4 and the corresponding transport aviation aircraft repair cost database are associated and stored as a piece of data in the transport aviation aircraft repair cost database.

[0033] Preferably, category one of the initial repair cost ratio range corresponds to category one of the multi-objective comprehensive preference function; category two of the initial repair cost ratio range corresponds to category two of the multi-objective comprehensive preference function; category three of the initial repair cost ratio range corresponds to category three of the multi-objective comprehensive preference function.

[0034] An accident level evaluation system for transport aviation aircraft weighing more than 50 tons includes an acquisition input module, a database system, an evaluation processing system and an output module. The database system internally constructs an accident repair cost index parameter library and a transport aviation aircraft repair cost database. The accident repair cost index parameter library is composed of multiple cost indicators divided by levels. The cost indicators divided by levels in the accident repair cost index parameter library are divided by levels according to national civil aviation regulations, such as Figure 2 As shown, some examples of this embodiment are as follows: the accident repair cost indicator parameter library includes three first-level indicators: aircraft repair cost, repair support cost, and other costs. The aircraft repair cost includes second-level indicators of labor time, equipment cost, and transportation cost. The repair support cost includes second-level indicators of factory facility rental fee and special equipment rental fee. Other costs include second-level indicators of aircraft moving fee, ground facility repair fee, site cleaning fee, and logistics support fee.

[0035] The collection input module is used to input the assessment value corresponding to the transport aircraft (generally the construction cost or purchase price of the transport aircraft is selected), and input the historical accident repair costs of the transport aircraft according to the cost index of the accident repair cost index parameter library; the database system has a statistical calculation module inside, which calculates the historical accident repair costs of the transport aircraft according to the cost index of the accident repair cost index parameter library to obtain the repair costs corresponding to the transport aircraft and stores them in the transport aircraft repair cost database. The data in the transport aircraft repair cost database are stored in association with aircraft model, assessment value, key accident information, repair cost, repair cost ratio, and category;

[0036] The evaluation and processing system includes a K-Means clustering algorithm model, a validation algorithm model and a calculation module, inputs data in a transport aviation aircraft repair cost database into the K-Means clustering algorithm model, obtains clustering results under three categories, and divides the initial repair cost ratio ranges corresponding to the three categories based on the clustering results;

[0037] The acquisition input module is also used to collect the assessed value Q0 of the transport aircraft under the new accident. The calculation module calculates the repair cost Q1 corresponding to the transport aircraft under the new accident according to the cost index of the accident repair cost index parameter library, and calculates the repair cost ratio P x ; The K-Means clustering algorithm model obtains the initial category K1 of transport aviation aircraft under new accidents based on the initial repair cost ratio range;

[0038] The verification algorithm model is based on the multi-objective comprehensive preference function. The repair cost Q1 is substituted into the multi-objective comprehensive preference function formula as the target value g to calculate P(g). The verification category K2 to which P(g) belongs is obtained according to the verification category range.

[0039] The output module is used to output the validation category K2 of the transport aviation aircraft under the new accident.

[0040] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0041] (1) The present invention uses a K-Means clustering algorithm model to perform historical clustering analysis on transport aircraft to obtain several initial repair cost ratio ranges, and then collects data on transport aircraft under new accidents to obtain validation categories through a verification algorithm model, thereby achieving a level assessment of transport aircraft accidents based on direct economic losses, and realizing auxiliary accident assessment based on data analysis and processing, which is convenient for scientifically dividing and distinguishing the size of accidents.

[0042] (2) The present invention uses the K-Means clustering algorithm model to perform historical clustering analysis, fully considering the commonality of transport aircraft, and then collects data of transport aircraft under new accidents to obtain validation categories through the verification algorithm model, fully considering the specific reality of transport aircraft. The combination of the two realizes a scientific and reasonable accident size evaluation of transport aircraft over 50 tons. The present invention uses the K-Means clustering algorithm model and the verification algorithm model to constrain each other, continuously update data, and has more and more guiding significance under the support of a large amount of data. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 This is a method flow chart of the method for evaluating the accident level of transport aviation aircraft weighing more than 50 tons according to the present invention;

[0044] Figure 2 It is a schematic diagram of dividing the cost indicators of the accident repair cost indicator parameter library according to the level of the embodiment;

[0045] Figure 3 The present invention is a principle structural block diagram of the accident grade assessment system for transport aviation aircraft weighing more than 50 tons. DETAILED DESCRIPTION

[0046] The present invention is further described in detail below in conjunction with embodiments:

[0047] Example

[0048] like Figure 1 to Figure 3 As shown, a method for evaluating the accident level of transport aviation aircraft weighing more than 50 tons is provided, and the method comprises:

[0049] S1. Construct an accident repair cost index parameter library, which is composed of multiple cost indicators divided by levels; collect the assessment value corresponding to the transport aircraft, and statistically analyze the historical accident repair costs of the transport aircraft according to the cost indicators of the accident repair cost index parameter library to obtain the repair costs corresponding to the transport aircraft, and obtain a transport aircraft repair cost database. The data in the transport aircraft repair cost database are stored in association with aircraft models, assessment values, key accident information, repair costs, repair cost ratios, and categories (part or all of the data in the transport aircraft repair cost database are sample data, and the categories in the sample data are input category labels. In some embodiments, the input category labels can be determined by the expert survey method (Delphi method) after evaluation as the input category labels; in this embodiment, the categories of all data in the transport aircraft repair cost database are input category labels; of course, the categories of part of the data can also be the input category labels, and the categories of the remaining data can be blank, and the blank categories can be classified using the K-Means clustering algorithm model of the present invention).

[0050] S2. Construct a K-Means clustering algorithm model, set the number of clusters to three and three initial clustering centers (the number of initial clustering centers is selected according to actual conditions. Since the categories of this application are defined as three types of categories, the use of three initial clustering centers can quickly and efficiently perform the clustering algorithm. Preferably, the three initial clustering centers are selected as sample data containing category labels; it is better to perform a preliminary division on the sample data containing category labels and determine the initial clustering centers, and then set them as the initial clustering centers), input the data in the transport aviation aircraft repair cost database into the K-Means clustering algorithm model, and obtain clustering results under three categories (this embodiment uses three categories to illustrate the technical principles of the present invention, and the technical ideas of the present invention are not limited to three categories, such as two categories, four categories, etc.; the three category definition rules of this embodiment are as follows: Category 1 corresponds to general accidents, Category 2 corresponds to major accidents, and Category 3 corresponds to major accidents), and divide the initial repair cost ratio ranges corresponding to the three categories based on the clustering results;

[0051] The initial repair cost ratio range for Category 1 is: P1≤initial repair cost ratio range <P2;

[0052] The initial repair cost ratio range for Category 2 is: P2 ≤ initial repair cost ratio range <P3;

[0053] The initial repair cost ratio range for Category 3 is: initial repair cost ratio range ≥ P3.

[0054] In some embodiments, the K-Means clustering algorithm model calculates the Euclidean distance between the remaining data objects and the cluster center based on the initial cluster center, and distributes the data objects to the cluster center corresponding to the cluster center (the cluster corresponds to one of the three categories), and then calculates the average value of the data objects in each cluster (the data objects correspond to the data in the transport aviation aircraft repair cost database, preferably, this embodiment uses the repair cost ratio as the data node with the highest weight and the most attention) as the new cluster center, and performs the next iteration until the cluster center no longer changes or the maximum number of iterations is reached; the Euclidean distance calculation formula between the data object and the cluster center is:

[0055]

[0056] Among them, x is the data object, C i is the i-th cluster center, m is the dimension of the data object, x j is the jth attribute value of x, C ij C i The j-th attribute value of .

[0057] In some embodiments, a more preferred technical solution for the iteration number constraint of the present invention is that the iteration number of the K-Means clustering algorithm model is constrained by the following square error sum SSE formula:

[0058]

[0059] Where k is the number of clusters (if the present invention is classified into K categories, the corresponding value is K), and the value of K in this embodiment is 3;

[0060] When the sum of squared errors (SSE) no longer changes, the iteration stops.

[0061] S3. Collect the assessed value Q0 of the transport aircraft under the new accident, calculate the repair cost Q1 corresponding to the transport aircraft under the new accident according to the cost index of the accident repair cost index parameter library, and calculate the repair cost ratio P x ; Based on the initial repair cost ratio range in step S2, the initial category K1 to which the transport aviation aircraft belongs under the new accident is obtained.

[0062] S4. Construct a verification algorithm model based on a multi-objective comprehensive preference function. The formula of the multi-objective comprehensive preference function is as follows:

[0063]

[0064] Among them, g represents the target value of the repair cost Q1, g i (i=1, 2, 3) is the boundary point of the preference interval, i = 1, 2, 3; P is the preference function. The value of i corresponds to the number of categories in the K-Means clustering algorithm model. When the number of categories in the K-Means clustering algorithm model is four, the above formula is as follows:

[0065]

[0066] In the three-category example of this embodiment, g i (i=1, 2, 3) are calculated as follows:

[0067] g1=P1×Q0,

[0068] g2=P2×Q0,

[0069] g3=P3×Q0,

[0070] Thus, P(g1), P(g2), and P(g3) are calculated; the validation category range based on the multi-objective comprehensive preference function is:

[0071] The range of category 1 is: P(g1)≤P(g) <P(g2);

[0072] The range of category 2 is: P(g2)≤P(g) <P(g3);

[0073] The range of category three is: P(g) ≥ P(g3);

[0074] Category one of the initial repair cost ratio range corresponds to category one of the multi-objective comprehensive preference function, category two of the initial repair cost ratio range corresponds to category two of the multi-objective comprehensive preference function; category three of the initial repair cost ratio range corresponds to category three of the multi-objective comprehensive preference function; for example, category one corresponds to general accidents, category two corresponds to major accidents, and category three corresponds to serious accidents.

[0075] Assuming that the K-Means clustering algorithm model outputs three categories, P1 is 10, P2 is 35, and P3 is 60. Taking the aircraft evaluation value of 658.4 million yuan, the economic loss of 588 million yuan, and the repair cost ratio of 0.8647 as an example, the calculated g1 = 65.84 million yuan, g2 = 230.44 million yuan, g3 = 395.04 million yuan, P (g1) = 0.37, P (g2) = 0.61, P (g3) = 1. Further calculations show that P (g) = 3.22> P (g3), which shows that the aircraft accident in the example is category three (i.e., a major accident).

[0076] Substitute the repair cost Q1 as the target value g into the multi-objective comprehensive preference function formula to calculate P(g), and then obtain the verification category K2 to which P(g) belongs according to the verification category range; verification category K2 is the evaluation level for accidents of transport aviation aircraft weighing more than 50 tons, and then output verification category K2.

[0077] The present invention uses the K-Means clustering algorithm model to perform historical clustering analysis, fully considering the commonalities of transport aircraft, and then collects data on transport aircraft under new accidents to derive validation categories through the verification algorithm model, fully considering the specific reality of transport aircraft. The combination of the two realizes a scientific and reasonable accident size evaluation for transport aircraft weighing more than 50 tons; the present invention mutually constrains the K-Means clustering algorithm model and the verification algorithm model, and continuously updates the data, so that the overall method has higher and higher division accuracy and more and more guiding significance under the support of a large amount of data, and also provides data support for subsequent adjustments to economic indicators.

[0078] In some embodiments, the verification category K2 obtained in step S4 and the corresponding transport aviation aircraft repair cost database are associated and stored as a piece of data in the transport aviation aircraft repair cost database, and the data of the transport aviation aircraft under the new accident evaluated is correspondingly stored in the transport aviation aircraft repair cost database.

[0079] like Figure 3 As shown, an accident grade assessment system for transport aircraft weighing more than 50 tons includes an acquisition input module, a database system, an assessment processing system and an output module. The database system internally constructs an accident repair cost index parameter library and a transport aircraft repair cost database, and the accident repair cost index parameter library is composed of a plurality of cost indicators divided by levels; the acquisition input module is used to input the assessment value corresponding to the transport aircraft, and input the historical accident repair costs of the transport aircraft according to the cost indicators of the accident repair cost index parameter library; the database system internally has a statistical calculation module, which performs statistics on the historical accident repair costs of the transport aircraft according to the cost indicators of the accident repair cost index parameter library to obtain the repair costs corresponding to the transport aircraft and stores them in the transport aircraft repair cost database, and the data in the transport aircraft repair cost database are stored in association with the aircraft model, assessment value, key accident information, repair cost, repair cost ratio, and category;

[0080] The evaluation and processing system includes a K-Means clustering algorithm model, a validation algorithm model and a calculation module, inputs data in a transport aviation aircraft repair cost database into the K-Means clustering algorithm model, obtains clustering results under three categories, and divides the initial repair cost ratio ranges corresponding to the three categories based on the clustering results;

[0081] The acquisition input module is also used to collect the assessed value Q0 of the transport aircraft under the new accident. The calculation module calculates the repair cost Q1 corresponding to the transport aircraft under the new accident according to the cost index of the accident repair cost index parameter library, and calculates the repair cost ratio P x ; The K-Means clustering algorithm model obtains the initial category K1 of transport aviation aircraft under new accidents based on the initial repair cost ratio range;

[0082] The verification algorithm model is based on the multi-objective comprehensive preference function. The repair cost Q1 is substituted into the multi-objective comprehensive preference function formula as the target value g to calculate P(g). The verification category K2 to which P(g) belongs is obtained according to the verification category range.

[0083] The formula of multi-objective comprehensive preference function is as follows:

[0084]

[0085] Among them, g represents the target value of the repair cost Q1, g i (i=1, 2, 3) is the boundary point of the preference interval, i=1, 2, 3; P is the preference function;

[0086] g i (i=1, 2, 3) are calculated as follows:

[0087] g1=P1×Q0,

[0088] g2=P2×Q0,

[0089] g3=P3×Q0,

[0090] Thus, P(g1), P(g2), and P(g3) are calculated; the validation category range based on the multi-objective comprehensive preference function is:

[0091] The range of category 1 is: P(g1)≤P(g) <P(g2);

[0092] The range of category 2 is: P(g2)≤P(g) <P(g3);

[0093] The range of category three is: P(g) ≥ P(g3);

[0094] Substitute the repair cost Q1 as the target value g into the multi-objective comprehensive preference function formula to calculate P(g), and then obtain the effectiveness category K2 to which P(g) belongs according to the effectiveness category range.

[0095] The output module is used to output the validation category K2 of the transport aviation aircraft under the new accident.

[0096] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for evaluating accident levels of transport aircraft weighing more than 50 tons, characterized in that: The methods include: S1. Construct an accident repair cost index parameter library, which consists of multiple cost indicators divided by levels; collect the assessment value corresponding to the transport aircraft, and statistically analyze the historical accident repair costs of the transport aircraft according to the cost indicators in the accident repair cost index parameter library to obtain the repair costs corresponding to the transport aircraft, and obtain a transport aircraft repair cost database. The data in the transport aircraft repair cost database are stored in association with aircraft model, assessment value, key accident information, repair cost, repair cost ratio, and category. S2. Construct a K-Means clustering algorithm model, set the number of clusters to three and three initial cluster centers, input the data in the transport aviation aircraft repair cost database into the K-Means clustering algorithm model, obtain clustering results under three categories, and divide the initial repair cost ratio ranges corresponding to the three categories based on the clustering results; The initial repair cost ratio range for Category 1 is: P1≤initial repair cost ratio range <P2; The initial repair cost ratio range for Category 2 is: P2 ≤ initial repair cost ratio range <P a ; The initial repair cost ratio range for Category 3 is: initial repair cost ratio range ≥ P3; S3. Collect the assessed value Q0 of the transport aircraft under the new accident, calculate the repair cost Q1 corresponding to the transport aircraft under the new accident according to the cost index of the accident repair cost index parameter library, and calculate the repair cost ratio P x ; Based on the initial repair cost ratio range in step S2, the initial category K1 to which the transport aviation aircraft belongs under the new accident is obtained; S4. Construct a verification algorithm model based on a multi-objective comprehensive preference function. The formula of the multi-objective comprehensive preference function is as follows: Among them, g represents the target value of the repair cost Q1, g i (i=1, 2, 3) is the boundary point of the preference interval, P is the preference function; g i (i=1, 2, 3) are calculated as follows: g1=P1×Q0, g2=P2×Q0, g3=P3×Q0, Thus, P(g1), P(g2), and P(g3) are calculated; the validation category range based on the multi-objective comprehensive preference function is: The range of category 1 is: P(g1)≤P(g) <P(g2); The range of category 2 is: P(g2)≤P(g) <P(g3); The range of category three is: P(g) ≥ P(g3); Substitute the repair cost Q1 as the target value g into the multi-objective comprehensive preference function formula to calculate P(g), and then obtain the effectiveness category K2 to which P(g) belongs according to the effectiveness category range.

2. The accident level assessment method for transport aircraft weighing more than 50 tons according to claim 1, characterized in that: The K-Means clustering algorithm model calculates the Euclidean distance between the remaining data objects and the cluster center based on the initial cluster center, and assigns the data objects to the cluster center corresponding to the cluster center. Then, the average value of the data objects in each cluster is calculated as the new cluster center, and the next iteration is performed until the cluster center no longer changes or the maximum number of iterations is reached. The Euclidean distance calculation formula between the data object and the cluster center is: Among them, x is the data object, C i is the i-th cluster center, m is the dimension of the data object, x j is the jth attribute value of x, C ij C i The j-th attribute value of .

3. The accident level assessment method for transport aircraft weighing more than 50 tons according to claim 2, characterized in that: The number of iterations of the K-Means clustering algorithm model is constrained by the following square error SSE formula: Where k is the number of clusters, and its value is 3; When the sum of squared errors (SSE) no longer changes, the iteration stops.

4. The accident level assessment method for transport aircraft weighing more than 50 tons according to claim 1, characterized in that: Part or all of the data in the transport aviation aircraft repair cost database is sample data, and the categories in the sample data are input category labels; in step S2, the three initial cluster centers are selected as sample data containing category labels.

5. The accident level assessment method for transport aircraft weighing more than 50 tons according to claim 1, characterized in that: The validation category K2 obtained in step S4 and the corresponding transport aviation aircraft repair cost database are associated and stored as a piece of data in the transport aviation aircraft repair cost database.

6. The accident level assessment method for transport aircraft weighing more than 50 tons according to claim 1, characterized in that: Category one of the initial repair cost ratio range corresponds to category one of the multi-objective comprehensive preference function; category two of the initial repair cost ratio range corresponds to category two of the multi-objective comprehensive preference function; category three of the initial repair cost ratio range corresponds to category three of the multi-objective comprehensive preference function.

7. A system for implementing the accident level assessment method for transport aircraft weighing more than 50 tons as described in any one of claims 1 to 6, characterized in that: It includes an acquisition input module, a database system, an evaluation processing system and an output module. The database system has an accident repair cost index parameter library and a transport aircraft repair cost database, and the accident repair cost index parameter library is composed of multiple cost indicators divided by levels; the acquisition input module is used to input the assessment value corresponding to the transport aircraft, and input the historical accident repair costs of the transport aircraft according to the cost indicators of the accident repair cost index parameter library; the database system has a statistical calculation module, which calculates the historical accident repair costs of the transport aircraft according to the cost indicators of the accident repair cost index parameter library to obtain the corresponding repair costs of the transport aircraft and stores them in the transport aircraft repair cost database. The data in the transport aircraft repair cost database are stored in association with aircraft model, assessment value, key accident information, repair cost, repair cost ratio, and category; The evaluation and processing system includes a K-Means clustering algorithm model, a validation algorithm model and a calculation module, inputs data in a transport aviation aircraft repair cost database into the K-Means clustering algorithm model, obtains clustering results under three categories, and divides the initial repair cost ratio ranges corresponding to the three categories based on the clustering results; The acquisition input module is also used to collect the assessed value Q0 of the transport aircraft under the new accident. The calculation module calculates the repair cost Q1 corresponding to the transport aircraft under the new accident according to the cost index of the accident repair cost index parameter library, and calculates the repair cost ratio P x ; The K-Means clustering algorithm model obtains the initial category K1 of transport aviation aircraft under new accidents based on the initial repair cost ratio range; The verification algorithm model is based on the multi-objective comprehensive preference function. The repair cost Q1 is substituted into the multi-objective comprehensive preference function formula as the target value g to calculate P(g). The verification category K2 to which P(g) belongs is obtained according to the verification category range. The output module is used to output the validation category K2 of the transport aviation aircraft under the new accident.

Images