An automatic aircraft part assembly relationship carding method

Abstract

The application belongs to the technical field of aircraft assembly, and is an automatic aircraft part assembly relationship carding method. When screening aircraft part data, the data is corrected first, then the best first-level data of the maximum weight data is found according to the current aircraft engine part grade division, then the data of each level is screened in order from large to small according to the level, until all the data is screened, finally the confusion degree is evaluated to complete all the carding, the aircraft part weight data assembly relationship of each system and structure can be quickly and accurately carded, the errors in the provided data are pointed out, and the change of the assembly relationship confusion degree value before and after the data carding is given, which lays a data foundation for subsequent data calculation and analysis.

Description

Technical Field

[0001] This application belongs to the field of aircraft assembly technology, and specifically relates to an automatic sorting method for aircraft component assembly relationships. Background Technology

[0002] Currently, in China, the analysis of weight and center of gravity data for aircraft components primarily relies on engineers specializing in weight analysis to manually synthesize information such as weight, center of gravity, and drawing numbers to determine the assembly relationships between components. Due to the sheer volume of data, the assembly relationships derived from the weight and center of gravity data collected from various structures and systems are unreadable and not strictly accurate by computers. Furthermore, the weight and center of gravity data itself contains inherent errors. This data analysis requires a significant investment of human resources and yields numerous inaccuracies. Therefore, the weight analysis profession urgently needs a method that can quickly and accurately determine the assembly relationships between components. Summary of the Invention

[0003] The purpose of this application is to provide an automatic sorting method for aircraft component assembly relationships, so as to solve the problem of low efficiency caused by the need for manual sorting of component assembly relationships in the prior art.

[0004] The technical solution of this application is: an automatic method for sorting out the assembly relationships of aircraft parts, including:

[0005] Check whether all aircraft data meet the format requirements, locate the problematic data and correct it;

[0006] Based on the current classification of aero-engine components, obtain the best sub-level data with the largest weight among all the data to be sorted;

[0007] Iterate through each sub-level data and obtain the best sub-level data in each sub-level data;

[0008] Determine whether the last-level data of all sorted data contains a subset. If not, proceed to the next step. If so, repeat the traversal of each sub-level data until the best sub-level data that does not contain a subset is found.

[0009] Determine whether all data has been sorted. If yes, proceed to the next step; otherwise, repeat the process of traversing each sub-level data until the best sub-level data that does not contain a subset is found.

[0010] Assess the degree of disorder in the data assembly relationships and output the sorting results.

[0011] Preferably, the optimal sub-level data is obtained by calculating using a simulated annealing algorithm, and the calculation method is as follows:

[0012] Obtain the total number of rows R to be sorted, as well as the maximum weight W and minimum weight W to be sorted.

[0013] num max

[0014] Theoretical weight W min ;

[0015] The overall state energy E is measured by the degree of disorder in the assembly relationship of aircraft parts. The improved principal state energy E1 is measured by a principal objective function that is negatively correlated with the degree of alignment with the center of gravity. The secondary state energy E2 is represented by a secondary objective function that is negatively correlated with the degree of concentration of mass distribution. The initial temperature T0(R) is set. num Termination temperature T f Markov chain length L(R) num ), cooling rate a, subset element row limit R limit =R num -1 and the current temperature t=T0(R num );

[0016] Using the minimum weight to be sorted as the assembly, a first-level subset with no more than the number of rows of the subset elements is randomly generated. The main objective function is calculated, and it is determined whether the main objective function meets the design requirements.

[0017] If the design requirements are not met, calculate the increment ΔE1 of the main objective function. If ΔE1 < 0, accept it as a new subset; if ΔE1 > 0, then use probability... The random number is received as a new subset;

[0018] If the design requirements are met, calculate the secondary objective function ΔE2. If ΔE2 < 0, accept it as a new subset; if ΔE2 > 0, use probability... Accept the random number (0-1) as a new subset, and assign the number of rows of the new subset to the subset's row count constraint R. limit ;

[0019] Determine if the number of iterations has reached the length of the Markov chain. If yes, proceed to the next step; otherwise, recalculate the main objective function.

[0020] Set the current temperature t = t * a;

[0021] Determine if the current temperature has reached the minimum temperature T. f If so, output the current solution as the optimal solution; otherwise, recalculate the main objective function.

[0022] Preferably, the method for obtaining the optimal sub-level data is as follows: Input all data, given the total weight and center of gravity, first set an estimated number of parts, select a portion of the estimated number of parts from the data list, pick out the data columns with larger weights, remove the data columns with smaller weights and put them back into the original data list, calculate the difference between the selected data columns and the total weight; then select some or all of the data columns that meet the weight requirement, and select them once or multiple times; until the total weight requirement is met, calculate the center of gravity and moment of inertia of the selected data columns; repeat the above steps to find multiple sets of data columns that meet the weight requirement, and then calculate the set with the smallest difference between the center of gravity and the original center of gravity and the smallest moment of inertia, that is, find the optimal sub-set of data.

[0023] Preferably, the method for representing the degree of disorder in the data assembly relationship is as follows: a positive real number that can measure the disorder of the component assembly relationship is obtained by using the number of top images and the weight center relationship between the top image and its children; the larger the positive real number, the more disordered the assembly relationship.

[0024] This application discloses an automatic sorting method for aircraft component assembly relationships. When filtering aircraft component data, the method first corrects the data, then, according to the component classification of the current aero-engine, it first finds the optimal sub-level data with the largest weight. Then, it sequentially filters the data at each level from largest to smallest until all data is filtered. Finally, it evaluates the degree of disorder to complete the sorting. This method can quickly and accurately sort out the assembly relationships of component weight data for various aircraft systems and structures, point out errors in the provided data, and show the change in the degree of disorder of assembly relationships before and after data sorting, laying a data foundation for subsequent data calculation and analysis. Attached Figure Description

[0025] To more clearly illustrate the technical solutions provided in this application, the accompanying drawings will be briefly described below. Obviously, the drawings described below are merely some embodiments of this application.

[0026] Figure 1 This is a schematic diagram of the overall process of this application;

[0027] Figure 2 This is a schematic diagram of the simulated annealing algorithm in this application. Detailed Implementation

[0028] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions in the embodiments of this application will be described in more detail below with reference to the accompanying drawings.

[0029] An automatic sorting method for aircraft component assembly relationships was developed for a partial calculation task of the overall weight and center of gravity data of a certain model of aircraft in our institute. The data is shown in Appendix Table 1. The weight and center of gravity specialists processed the large amount of aircraft component weight and center of gravity data provided by various specialties, which contained errors, had disordered assembly relationships, and whose assembly relationships were unreadable by computers, using VBA programming.

[0030] Table 1

[0031]

[0032]

[0033] like Figure 1 As shown, it includes the following steps:

[0034] Step S100: For the data in Appendix 1, check whether each aircraft data meets the format requirements, locate the problematic data and correct it. Among the received aircraft data, there are problems such as non-numeric data, missing center of gravity, and data containing spaces. If these are not corrected, filtering errors may occur. Therefore, it is necessary to find and correct the problematic data first, and then sort out the data.

[0035] Step S200: According to the current classification of aero-engine components, obtain the best sub-level data of the data with the largest weight among all the data to be sorted;

[0036] Current aero engines are divided into multiple levels from the whole engine to components and parts. The different levels are represented by the assembly level method based on the data RGB values, as shown in Table 2.

[0037] Table 2

[0038]

[0039]

[0040] Each level of data corresponds to a level of data above it. This application uses a method of sorting data automatically from the largest level to the smallest level.

[0041] The data with the largest weight is also the highest-level data. To find the required data, this application employs two algorithms. The first is an optimized simulated annealing algorithm, such as... Figure 2 As shown, the specific steps include the following:

[0042] Step 1: Obtain the total number of rows of data to be processed (R) num And the maximum weight to be combed, W max Minimum weight to be combed W min ;

[0043] Step 2: The overall state energy E is measured using the degree of disorder in the assembly relationship of aircraft parts. The improved principal state energy E1 is measured by a principal objective function that is negatively correlated with the degree of alignment with the center of gravity. The secondary state energy E2 is represented by a secondary objective function that is negatively correlated with the degree of concentration of mass distribution. The initial temperature T0(R) is set. num Termination temperature T f Markov chain length L(R) num ), cooling rate a, subset element row limit R limit =R num -1 and the current temperature t=T0(R num );

[0044] Preferably, the overall state energy E is:

[0045]

[0046] Where k1 is the top graph coefficient, which is 1 in this example; p is the number of top graphs, including unclassifiable data; n is the number of parent-child relationship groups in the data; i is the current parent-child relationship; k2 is the error tolerance coefficient for a single parent-child relationship, which is 3,200,000 in this example; m i The weight of the parent data in the i-th parent-child relationship; j is the weight of the child data in the j-th row of the current parent-child relationship; m j Let be the weight of the j-th child element in the current parent-child relationship; k3 is the weight error tolerance coefficient, which is 0.001 in this example; u is the number of child elements in the current parent-child relationship; x j y j z j These are the coordinates of the j-th row of data in the current parent-child relationship; x i y i z i , , are the coordinates of the parent data in the i-th parent-child relationship; Δl is the base value of the centroid offset, which is taken as 0.1*10^6 meters here.

[0047] The improved main state energy E1 is:

[0048]

[0049] The secondary state energy E2 is:

[0050]

[0051] In one specific embodiment, an initial temperature T0(R) is set. num ) = 20R num Termination temperature T f =3, set the Markov chain length L(R) num ) = 100R numSet the cooling rate a = 0.9, and set the limit R for the number of rows of subset elements. limit =R num -1, set the current temperature t = T0(R) num ).

[0052] Step 3: Use the minimum weight to be combed, W max For an assembly, randomly generate a first-level subset with no more rows than the limit of the number of rows of elements in the subset, calculate the main objective function, and determine whether the main objective function meets the design requirements;

[0053] Step 4: If the design requirements are not met, calculate the increment ΔE1 of the main objective function. If ΔE1 < 0, the receiver is taken as a new subset; if ΔE1 > 0, the probability is used to determine the new subset. (A random number between 0 and 1) is received as a new subset, and then Step 6 is executed;

[0054] Step 5: If the design requirements are met, calculate the secondary objective function ΔE2. If ΔE2 < 0, accept it as a new subset; if ΔE2 > 0, then use probability... Accept the random number (0-1) as a new subset, and assign the number of rows of the new subset to the subset's row count constraint R. limit ;

[0055] Step 6: Determine if the number of iterations has reached the length of the Markov chain. If yes, proceed to the next step; otherwise, recalculate the main objective function.

[0056] Step 7: Set the current temperature t = t * a;

[0057] Step 8: Determine if the current temperature has reached the minimum temperature T. f If so, output the current solution as the optimal solution; otherwise, recalculate the main objective function, which means executing Step 3.

[0058] This application employs an improved simulated annealing algorithm to find the optimal sub-level data with the largest weight. It uses random number generation to find the assembly with the smallest unprocessed weight. Based on the weight data, it can determine parameters such as initial temperature, termination temperature, cooling rate, and number of iterations. By continuously calculating the increment of the objective function value, it obtains the optimal solution, thus accurately and efficiently finding the required optimal sub-level data. The found data is then removed from the aircraft's internal data.

[0059] The improvement to this search method mainly involves changing the original simulated annealing algorithm from one that doesn't limit the number of data points involved in the calculation to one where, after the primary objective function value meets the requirement for the first time, the number of data points involved in the calculation decreases continuously as the secondary objective function value decreases, thus continuously improving search efficiency. When the primary objective function value meets the requirement for the first time, it signifies that the first subset of the assembly has found that satisfies the weight-center-of-gravity relationship. Since the secondary objective function value reflects the concentration of the assembly's mass distribution, the number of elements in the search for a better subset should not exceed the number of elements in the current subset. Therefore, the number of data points involved in the calculation is continuously reduced in each round of the search.

[0060] The second method uses estimation to filter the required data, specifically including:

[0061] Input all the data, given the total weight and center of gravity. First, set an estimated number of parts. Select a portion of the estimated number of parts from the data list, pick out the data columns with larger weights, remove the data columns with smaller weights and put them back into the original data list. Calculate the difference between the selected data columns and the total weight. Then, select some or all of the data columns that meet the weight requirement, and repeat this process once or multiple times until the total weight requirement is met. Calculate the center of gravity and moment of inertia of the selected data columns.

[0062] Repeat the above steps to find multiple sets of data that meet the weight requirements. Then calculate the set with the smallest difference between the center of gravity and the original center of gravity, and the smallest moment of inertia. This is the optimal subset of data.

[0063] By adopting this method, less data needs to be processed, and the optimal sub-level data can be found more quickly.

[0064] Step S300: Traverse each sub-level data and obtain the best sub-level data in each sub-level data; for data with at least ten levels, all data of the previous level must be filtered before the data of the next level can be filtered; the filtering of each level of data is carried out in accordance with the method in step S200 until all levels of data are filtered.

[0065] Step S400: Determine whether the last-level data of all sorted data contains a subset. If not, proceed to the next step. If so, repeat the traversal of each sub-level data until the best sub-level data that does not contain a subset is found.

[0066] Step S500: Determine whether all data has been sorted. If yes, proceed to the next step. If no, repeat the traversal of each sub-level data until the best sub-level data that does not contain a subset is found.

[0067] Step S600: Assess the degree of disorder in the data assembly relationship and output the sorting results.

[0068] Preferably, the method for representing the degree of disorder in data assembly relationships is as follows: A positive real number is obtained by using the number of top-level diagrams and the weight center relationship between the top-level diagram and its children to measure the disorder in component assembly relationships; the larger the positive real number, the more disordered the assembly relationships. The final sorting results are shown in Table 3.

[0069] Table 3

[0070]

[0071]

[0072] In screening aircraft component data, this application first corrects the data, then, according to the current classification of aero-engine components, first finds the best sub-level data with the largest weight, and then filters the data at each level in descending order of weight until all data is screened. Finally, it evaluates the degree of disorder to complete the sorting. This method can quickly and accurately sort out the assembly relationships of component weight data of various systems and structures of the aircraft, point out errors in the provided data, and show the change in the degree of disorder of assembly relationships before and after data sorting, laying a data foundation for subsequent data calculation and analysis.

[0073] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for automatically identifying assembly relationships of aircraft parts, characterized in that, include: Check whether all aircraft data meet the format requirements, locate the problematic data and correct it; Based on the current classification of aero-engine components, obtain the best sub-level data with the largest weight among all the data to be sorted; Iterate through each sub-level data and obtain the best sub-level data in each sub-level data; Determine whether the last-level data of all sorted data contains a subset. If not, proceed to the next step. If so, repeat the traversal of each sub-level data until the best sub-level data that does not contain a subset is found. Determine whether all data has been sorted. If yes, proceed to the next step; otherwise, repeat the process of traversing each sub-level data until the best sub-level data that does not contain a subset is found. Assess the degree of disorder in the data assembly and output the sorting results; The optimal sub-level data is obtained by calculating using the simulated annealing algorithm. The calculation method is as follows: Get all rows of data to be sorted and the maximum weight to be combed Minimum weight to be combed ; Overall state energy Using the standard of disorder in the assembly relationship of aircraft parts, the improved main state energy The energy of the secondary state is measured by the main objective function, which is negatively correlated with the degree of alignment with the center of gravity. The objective function is negatively correlated with the degree of concentration of mass distribution, and the initial temperature is set. Termination temperature Markov chain length Cooling rate (a), Subset element row limit and current temperature ; Minimum weight to be combed For an assembly, randomly generate a first-level subset with no more rows than the limit of the number of rows of elements in the subset, calculate the main objective function, and determine whether the main objective function meets the design requirements; If the design requirements are not met, calculate the increment of the main objective function. ,like If < 0, then accept it as a new subset. If the value is greater than 0, then the probability is... Receive it as a new subset; If the design requirements are met, then calculate the secondary objective function. ,like If <0, then accept it as a new subset. >0, then with probability Accept it as a new subset, and assign the number of rows of the new subset to the limit of the number of rows of the subset. ; Determine if the number of iterations has reached the length of the Markov chain. If yes, proceed to the next step; otherwise, recalculate the main objective function. Set the current temperature ; Determine if the current temperature has reached the minimum temperature. If so, output the current solution as the optimal solution; otherwise, recalculate the main objective function.

2. The automatic sorting method for aircraft component assembly relationships as described in claim 1, characterized in that, The overall state energy E is: in, For the top graph coefficients; This refers to the number of top-level images; This represents the number of parent-child relationships in the data. The current father-son relationship; This represents the error tolerance coefficient for a single parent-child relationship. For the first The weight of the parent data in a parent-child relationship; In the current father-son relationship, the first Row-level data; In the current father-son relationship, the first The weight of the row-level; This is the weight tolerance coefficient; This represents the number of child nodes in the current parent-child relationship. , , These are the first children in the current parent-child relationship. The coordinates of the row data; , , The first The coordinates of the parent data in a parent-child relationship; This serves as the base value for the center of gravity offset. Improved main state energy for: Secondary state energy for: 。 3. The automatic assembly relationship sorting method for aircraft parts as described in claim 1, characterized in that: The method for obtaining the optimal sub-level data is as follows: Input all the data, given the total weight and center of gravity, first set an estimated number of parts, select a partial data column of the estimated number of parts from the data list, pick out the data column with the larger weight, remove the data column with the smaller weight and put it back into the original data list, calculate the difference between the selected data column and the total weight; then select some or all of the data columns that meet the weight requirement, and select them in one or more batches; until the total weight requirement is met, calculate the center of gravity and moment of inertia of the selected data column; Repeat the above steps to find multiple sets of data that meet the weight requirements. Then calculate the set with the smallest difference between the center of gravity and the original center of gravity, and the smallest moment of inertia. This is the optimal subset of data.

4. The automatic assembly relationship sorting method for aircraft parts as described in claim 1, characterized in that, The method for representing the degree of disorder in the data assembly relationship is as follows: positive real numbers that can measure the disorder of the component assembly relationship are obtained by using the number of top images and the weight center relationship between the top image and its children; the larger the positive real number, the more disordered the assembly relationship.

Images