Civil aircraft system function architecture generation method based on IGTA clustering algorithm

Abstract

The invention provides a civil aircraft system function architecture generation method based on an IGTA clustering algorithm, and the method comprises the steps: building a mathematical model of civil aircraft system function architecture design complexity, randomly selecting functions, calculating the adaptation degree between the functions and all function modules, updating the function modules according to the adaptation degree, and calculating the design complexity of the system function architecture. And generating a system function architecture scheme. According to the method, the deviation caused by only relying on the design experience of engineers under the condition that multiple functions are mutually associated in the complex system design is effectively avoided, and meanwhile, the function analysis modeling efficiency of the same type of systems under the model-driven forward design process is improved.

Description

Technical Field

[0001] The present invention relates to the field of civil aircraft system function analysis, and in particular to a function architecture generation method. Background Art

[0002] Today's civil aircraft production environment is a highly competitive, dynamic and uncertain process. Civil aircraft system design is under great test in the fierce competition of the global market and the ever-changing needs of stakeholders. On the one hand, the functional architecture of the system must be flexible and reusable to fully adapt to the needs of model serialization and diversification; on the other hand, the functional architecture must have a certain degree of resilience to cope with the changes in the needs of external stakeholders and the changes in requirements generated in the design.

[0003] In recent years, civil aircraft have gradually adopted a demand-driven forward design process in their design. The dynamic changes in market demand and the changes in demand caused by balancing conflicts and designs have put forward higher requirements for the top-level functional architecture design of civil aircraft. The system integration design method in the traditional system engineering method is difficult to provide sufficient resilience and flexibility to avoid duplication or redesign of civil aircraft systems caused by changes in demand.

[0004] In the traditional functional architecture design method, engineering designers use a top-down approach to establish scenario models and gradually analyze the relationship between requirements and use case models. After defining functions, they design the functional architecture based on their own professional knowledge, functional similarity, design knowledge and experience reserves of designers. With the trend of continuous refinement of civil aircraft system design granularity and increasing system complexity, it is difficult to intuitively quantify the complexity of internal interactions in the system to ensure the difficulty of reasonable interface design between subsystems and good maintainability of the system. The existing functional architecture design method adopts a model-based system engineering method and accurately describes the system functional architecture based on the system modeling language. However, the consideration of the intuitive coupling relationship between the functions of each system is still based on the design experience of the designer. When facing complex system design tasks, it is impossible to achieve rapid generation of functional architecture and provide civil aircraft designers with a reliable functional architecture decision-making assistance solution. Summary of the invention

[0005] In order to overcome the deficiencies of the prior art, the present invention provides a method for generating a civil aircraft system functional architecture based on an IGTA clustering algorithm.

[0006] The technical solution adopted by the present invention to solve the technical problem includes the following steps:

[0007] (1) Establishing a mathematical model for the complexity of the functional architecture design of civil aircraft systems

[0008] (2) Functional modules and parameter initialization

[0009] In order to reduce the impact of the initial functional module division on the system functional architecture, the number of initial functional modules in the system function set F is set to be consistent with the number of system functions. is the initialized function sequence. At this time, the function set F is divided into m1, m2, …, m n There are n functional modules, that is, each functional module contains only one function, and C(X 0 ) and make the lowest design complexity C min =C(X 0 ), set the iteration limit p max And let the number of iterations p = 1;

[0010] (3) Randomly select functions and calculate the degree of fit between them and all functional modules;

[0011] (4) Update functional modules according to adaptability

[0012] Based on the idea of ​​simulated annealing, the selected system function is added to the function module with the second highest fitness value with the function with a probability of 1 / 2n, otherwise the selected system function is added to the function module with the largest fitness value with the function; after the selected function is added to the function module, if there is a relationship of inclusion or inclusion between several function modules at this time, the included function module is deleted; at this time, the function sequence is recorded as X tep ;

[0013] (5) Design complexity of computing system functional architecture

[0014] The design complexity of the computing system functional architecture C(X tep ), if C(X tep ) <C min , then accept the updated function module and update C min =C(X tep ); if C(X tep )≥C min , then accept the updated functional module with a probability of 1 / 2n, otherwise go to step (3);

[0015] (6) Determine whether the current solution is stable

[0016] Repeat steps (3) to (5). If C is not updated twice in a row, min , then the current solution is considered stable and go to step (7);

[0017] (7) Determine whether the algorithm has ended

[0018] When p <p max , go to step (3) and let p increase by 1. When p ≥ p maxWhen , the algorithm ends and goes to step (8);

[0019] (8) Generate system functional architecture plan

[0020] According to the historical lowest design complexity C min The corresponding function sequence And the functional module clustering results A functional architecture scheme can be generated based on the strength of the functional relationship between each system, providing auxiliary decision-making for civil aircraft designers when carrying out the functional architecture design of civil aircraft systems.

[0021] In the step (1),

[0022] Suppose the set of functions to be designed for a certain system of a civil aircraft is F and satisfies F = {f1,f2,…,f n}, the elements in the set F are the functional items to be designed for the system, F = {f1, f2, ..., f n} indicates that there are n system functions to be designed in the set, where f1, f2, …, f n In turn, they represent the n system functions of the system to be designed after functional definition;

[0023] Let X=(x1,x2,…,x n ) is any order of positive integers from 1 to n, let R(X)=(r ij ) X is the correlation matrix of civil aircraft system functions based on sequence X, r ij Civil aircraft system functions System Function The degree of influence, where 1≤i,j≤n and i,j are integers, 0≤r ij ≤5 and r ij is an integer, r ij The smaller the value, the better the system function. System Function The smaller the influence, the greater the impact. ij The value of is given by experts in the relevant field of the system to be designed based on theory and engineering design experience.

[0024] The system function architecture reflects the relationship between system functions. By modularizing and encapsulating several system functions, the complexity of system function architecture design is solved. The system function set F is combined into k functional modules, denoted as F = {m1, m2, …, m k}, the number of system functions in each functional module is recorded as n(m i ), where 1≤i≤k and i is an integer and The sequence corresponding to the jth functional module is Then the sequence X is expressed as:

[0025]

[0026] Define T(X)=(t ij ) X is the correlation matrix of the civil aircraft system function modules based on the sequence X, which is a k×k square matrix. The elements in T(X) in

[0027] Define the internal complexity C of the system functional module I (T(X)), as shown in formula (1):

[0028]

[0029] Among them, p I is the clustering penalty factor of the internal complexity of the functional module; when the number of system functions within the system functional module is too large, the internal complexity of the system functional module will increase exponentially, that is, when dividing the functional modules, it should be ensured that the number of system functions within a single functional module should not be too large.

[0030] Define the complexity of interaction between system functional modules C E (T(X)), as shown in formula (2):

[0031]

[0032] Among them, p E is the clustering penalty factor for the complexity of interaction between functional modules. When the influence between functional modules is too strong, the complexity of interaction between functional modules of the system will increase exponentially. That is, when dividing functional modules, the functional interaction between functional modules should be kept as small as possible to ensure the modular design of system functions.

[0033] Define the design complexity C(X) of the system functional architecture of the civil aircraft system function based on the sequence X, as shown in formula (3):

[0034] C(X)=C I (T(X))+C E (T(X)) (3)

[0035] The functions within each system function module are generally closely connected and have a high degree of mutual influence; while the influence between each system function module is relatively low. I (T(X)) is penalized based on the number of functions in the functional module, while the complexity of interaction between system functional modules C E(T(X)) is penalized based on the strength of the influence relationship between functional modules, that is, the number of each system functional module and the total number of system functional modules divided are not too large. In the mathematical model, the minimum value of the design complexity C(X) of the system functional architecture is obtained.

[0036] In step (3), randomly select an element x in the sequence X i , element x i Corresponding function f q , f q is the functional module m j Other functions in ; Calculate the element x i The corresponding function f q With function module m j The adaptability between j , the calculation method is shown in formula (4):

[0037]

[0038] Among them, p A is the fitness weight between the system function and the functional module; when the selected function f q With function module m j The greater the functional impact, the greater the functional module m j The number of internal functions n(m j ) is smaller, the function f q With function module m j The fitness of A j The bigger.

[0039] An electronic device, comprising: one or more processors; a memory; and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs are configured to execute the method as described in any one of claim 1.

[0040] A computer-readable storage medium stores program code, wherein the program code can be called by a processor to execute the method as claimed in any one of claims 1.

[0041] The beneficial effect of the present invention is that it quantitatively describes the complex correlation between the functions of civil aircraft systems, and by clustering and combining closely related system functions, it proposes quantitative indicators of the design complexity of the internal and external functional modules, establishes a mathematical model of the design complexity of the civil aircraft system functional architecture, and solves the model through the IGTA clustering algorithm to generate a civil aircraft system functional architecture scheme that is divided based on the strength of the correlation between the functions of each system. The invention can rely on the engineering design experience of experts in the field to provide civil aircraft designers with a modular and highly reusable system architecture design scheme, effectively avoiding the deviation caused by relying solely on the design experience of engineers when many functions are interrelated in complex system design, and at the same time improves the efficiency of functional analysis modeling of the same type of system under the model-driven forward design process. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 It is a flow chart of the functional architecture generation method of the IGTA clustering algorithm of the present invention.

[0043] Figure 2 It is the function association matrix of the civil aircraft flight control system of the present invention.

[0044] Figure 3 It is a visualization example of the functional modules of the flight control system of the present invention. DETAILED DESCRIPTION

[0045] The present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0046] (1) Establish a mathematical model for the design complexity of the functional architecture of a civil aircraft flight control system;

[0047] Assume that the set of functions to be designed for the civil aircraft flight control system is F, and it satisfies F = {f1,f2,…,f 36}, where f1,f2,…,f 36 It represents the 36 system functions of the system to be designed after functional definition.

[0048] X=(1,2,…,36) is a sequence of positive integers from 1 to 36, R(X)=(r ij ) X is the correlation matrix of civil aircraft system functions based on sequence X, r ij Civil aircraft system functions System Function The degree of influence, where 1≤i,j≤n and i,j are integers, 0≤r ij ≤5 and r ij is an integer, r ij The smaller the value, the better the system function. System Function The smaller the impact, the greater the impact.ij The value of is given by experts in the relevant field of the system to be designed based on theory and engineering design experience. In this example, the function correlation matrix R(X) of the civil aircraft flight control system is as follows: Figure 2 shown.

[0049] The system function architecture reflects the relationship between system functions. By modularizing and encapsulating several system functions, the complexity of system function architecture design can be effectively solved. The system function set F is combined into k functional modules, denoted as F = {m1, m2, …, m k}, the number of system functions in each functional module is recorded as n(m i ), where 1≤i≤k and i is an integer and The sequence corresponding to the jth functional module is Then the sequence X is expressed as:

[0050]

[0051] Define T(X)=(t ij ) X is the correlation matrix of the civil aircraft system function modules based on the sequence X, which is a k×k square matrix. The elements in T(X) in

[0052] When calculating the design complexity of the civil aircraft flight control system, we take p I =1, p E =1, then the complexity of the internal functional module of the civil aircraft flight control system is C I (T(X)) and the complexity of interaction between functional modules C E (T(X)), as shown in formula (5) and formula (6) respectively:

[0053]

[0054] The design complexity C(X) of the functional architecture of the civil aircraft flight control system is shown in formula (7):

[0055]

[0056] (2) Functional modules and parameter initialization

[0057] The number of initial function modules in the system function set F is set to be the same as the number of system functions, which is 36. That is, each function module contains only one system function. 0 =(1,2,…,36) is the function sequence after initialization. Calculate C(X 0 ) is 565088, and the lowest design complexity C min =C(X 0)=565088. Set the iteration limit p max =10 and let the number of iterations p=1.

[0058] (3) Randomly select elements and calculate the degree of fit between them and all functional modules

[0059] Randomly select an element x from the sequence X i , element x i Corresponding function f q , f q is the functional module m j Other functions in ; Calculate the element x i The corresponding function f q With function module m j The adaptability between j , the calculation method is shown in formula (8) (take the fitness weight p between the system function and the functional module A =4):

[0060]

[0061] (4) Update functional modules according to adaptability

[0062] Based on the idea of ​​simulated annealing, the selected system function is added to the function module with the second highest fitness value with a probability of 1 / 2n=1 / 72, otherwise the selected system function is added to the function module with the largest fitness value. After the selected function is added to the function module, if there is a relationship of inclusion or inclusion between several function modules at this time, the included function module is deleted; at this time, the function sequence is recorded as X tep .

[0063] (5) Design complexity of computing functional architecture

[0064] Calculate the design complexity of the functional architecture C(X tep ). If C(X tep ) <C min , then accept the updated function module and update C min =C(X tep ); if C(X tep )≥C min , then accept the updated functional module with a probability of 1 / 2n=1 / 72, otherwise go to step (3).

[0065] (6) Determine whether the current solution is stable

[0066] Repeat steps (3) to (5). If C is not updated twice in a row, min , then the current solution is considered stable and go to step (7);

[0067] (7) Determine whether the algorithm has ended

[0068] When p <p max , go to step (3) and set p = p + 1. When p ≥ p max =10, the algorithm ends and goes to step (8);

[0069] (8) Generate system functional architecture plan

[0070] In the example simulation, after 194 algorithm cycles, the lowest design complexity C was calculated. min =211335, Functional sequence And functional module clustering results As shown in Table 1, the 36 functions to be designed in the civil aircraft flight control system in this example are divided into 10 functional modules according to the mutual influence relationship between the functions. A functional architecture scheme based on the strength of the functional relationship between each system can be obtained, which can provide reference and decision-making for civil aircraft designers when designing the functional architecture of civil aircraft flight control systems. Figure 3 shown.

[0071] Table 1 Example simulation results

[0072]

[0073] The present invention can effectively quantify and integrate functional associations, comprehensively consider the complex coupling and associations between system functions under a unified complex system, and modularize the system functions of a civil aircraft flight control system under complex functional coupling relationships by establishing a mathematical model and an improved IGTA clustering algorithm. This avoids the influence of too many subjective factors caused by manual construction of functional architecture in fine-grained forward design, provides a reference for the work of functional analysis engineers in the functional analysis stage of civil aircraft design, and ensures that the current model-driven civil aircraft forward design process can be better implemented and carried out.

Claims

1. A method for generating a civil aircraft system functional architecture based on the IGTA clustering algorithm, characterized in that The steps include: (1) Establish a mathematical model for the complexity of the functional architecture design of civil aircraft systems; (2) Functional modules and parameter initialization; In order to reduce the impact of the initial functional module division on the system functional architecture, the number of initial functional modules in the system function set F is set to be consistent with the number of system functions. is the initialized function sequence. At this time, the function set F is divided into m1, m2, …, m n There are n functional modules, that is, each functional module contains only one function, and C(X 0 ) and make the lowest design complexity C min =C(X 0 ), set the iteration limit p max And let the number of iterations p = 1; (3) Randomly select functions and calculate the degree of fit between them and all functional modules; (4) Update functional modules according to adaptability Based on the idea of ​​simulated annealing, the selected system function is added to the function module with the second highest fitness value with the function with a probability of 1 / 2n, otherwise the selected system function is added to the function module with the largest fitness value with the function; after the selected function is added to the function module, if there is a relationship of inclusion or inclusion between several function modules at this time, the included function module is deleted; at this time, the function sequence is recorded as X tep ; (5) Design complexity of computing system functional architecture The design complexity of the computing system functional architecture C(X tep ), if C(X tep ) <C min , then accept the updated functional module and update C min =C(X tep ); if C(X tep )≥C min , then accept the updated functional module with a probability of 1 / 2n, otherwise go to step (3); (6) Determine whether the current solution is stable Repeat steps (3) to (5). If C is not updated twice in a row, min , then the current solution is considered stable and go to step (7); (7) Determine whether the algorithm has ended When p <p max , go to step (3) and let p increase by 1. When p ≥ p max When , the algorithm ends and goes to step (8); (8) Generate system functional architecture plan According to the historical lowest design complexity C min The corresponding function sequence And the functional module clustering results A functional architecture scheme can be generated based on the strength of the functional relationship between each system, providing auxiliary decision-making for civil aircraft designers when carrying out the functional architecture design of civil aircraft systems.

2. The method for generating a civil aircraft system functional architecture based on the IGTA clustering algorithm according to claim 1, characterized in that: In the step (1), Suppose the set of functions to be designed for a certain system of a civil aircraft is F and satisfies F = {f1,f2,…,f n }, the elements in the set F are the functional items to be designed for the system, F = {f1,f2,…,f n } indicates that there are n system functions to be designed in the set, where f1, f2, …, f n In turn, they represent the n system functions of the system to be designed after functional definition; Let X=(x1,x2,…,x n ) is any order of positive integers from 1 to n, let R(X)=(r ij ) X is the correlation matrix of civil aircraft system functions based on sequence X, r ij For civil aircraft system functions xi For system function f xj The degree of influence, where 1≤i,j≤n and i,j are integers, 0≤r ij ≤5 and r ij is an integer, r ij The smaller the value, the greater the system function f xi For system function f xj The smaller the influence, the greater the impact. ij The value of is given by experts in the relevant field of the system to be designed based on theory and engineering design experience; The system function architecture reflects the relationship between system functions. By modularizing and encapsulating several system functions, the complexity of system function architecture design is solved. The system function set F is combined into k functional modules, denoted as F = {m1, m2, …, m k }, the number of system functions in each functional module is recorded as n(m i ), where 1≤i≤k and i is an integer and The sequence corresponding to the jth functional module is Then the sequence X is expressed as: Define T(X)=(t ij ) X is the correlation matrix of the civil aircraft system function modules based on the sequence X, which is a k×k square matrix. The elements in T(X) in Define the internal complexity C of the system functional module I (T(X)), as shown in formula (1): Among them, p I is the clustering penalty factor of the internal complexity of the functional module; when the number of system functions within the system functional module is too large, the internal complexity of the system functional module will increase exponentially, that is, when dividing the functional modules, it should be ensured that the number of system functions within a single functional module should not be too large; Define the complexity of interaction between system functional modules C E (T(X)), as shown in formula (2): Among them, p E is the clustering penalty factor for the complexity of interaction between functional modules. When the influence between functional modules is too strong, the complexity of interaction between functional modules of the system will increase exponentially. That is, when dividing functional modules, the functional interaction between functional modules should be kept as small as possible to ensure the modular design of system functions. Define the design complexity C(X) of the system functional architecture of the civil aircraft system function based on the sequence X, as shown in formula (3): C(X)=C I (T(X))+C E (T(X)) (3) Internal complexity of system function module C I (T(X)) is penalized based on the number of functions in the functional module, while the complexity of interaction between system functional modules C E (T(X)) is penalized based on the strength of the influence relationship between functional modules, that is, the number of each system functional module and the total number of system functional modules divided are not too large. In the mathematical model, the minimum value of the design complexity C(X) of the system functional architecture is obtained.

3. The method for generating a civil aircraft system functional architecture based on the IGTA clustering algorithm according to claim 1, characterized in that: In step (3), randomly select an element x in the sequence X i , element x i Corresponding function f q , f q is the functional module m j Other functions in; calculate the element x i The corresponding function f q With function module m j The adaptability between j , the calculation method is shown in formula (4): Among them, p A is the fitness weight between the system function and the functional module; when the selected function f q With function module m j The greater the functional impact, the greater the functional module m j The smaller the number of internal functions n(mj), the smaller the function f q With function module m j The fitness of A j The bigger.

4. An electronic device, characterized in that: include: one or more processors; Memory; One or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs are configured to execute the method according to any one of claims 1 to 3.

5. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores program codes, and the program codes can be called by a processor to execute the method according to any one of claims 1 to 3.

Images