Airborne system architecture design method and system based on constraint satisfaction problem

Abstract

The invention provides an airborne system architecture design method and system based on a constraint satisfaction problem, and belongs to the technical field of airborne system design, and the method specifically comprises the steps that an architecture decision model is constructed, the architecture decision model comprises a plurality of decision items and a plurality of alternative items corresponding to each decision item, and the alternative items are mutually independent; determining a measurement index of the decision item according to a development target of the airborne system; based on a constraint satisfaction problem model, in the architecture decision model, for each decision item, searching options conforming to a preset constraint rule in the alternative items, and obtaining an architecture scheme Pareto leading set; and performing multi-target tradeoff optimization on the architecture scheme Pareto leading-edge set to obtain an optimal architecture scheme. Through the processing scheme provided by the invention, the scientificity and efficiency of the architecture design of the airborne system are improved, and the influence of human factors on the decision result is also remarkably reduced.

Description

Technical Field

[0001] This application relates to the field of airborne system design technology, and in particular to an airborne system architecture design method and system based on constraint satisfaction problems. Background Technology

[0002] In recent years, the value of airborne systems in the overall aircraft has shown an increasing trend. Related literature indicates that the proportion of airborne systems in aviation has increased from 20% in the early days to over 50%. A normally operational aircraft typically consists of airborne systems, airframe structure, and engines. According to relevant standards, airborne systems are generally divided into 38 subsystems and 226 functional components and products, broadly categorized into three main types: avionics systems, electromechanical systems, and flight control systems. Airborne systems are a collective term in the aircraft field for systems such as flight control, engine control, hydraulic control, and avionics control. They enable the aircraft to complete various tasks, supporting flight, navigation, communication, control, detection, and safety assurance, as well as providing services to pilots and crew.

[0003] The R&D system for airborne systems is a forward-looking R&D system based on systems engineering. Model-based systems engineering (MBSE) is a new systems engineering management model proposed by the Society for Systems Engineering (SSE) to address the challenges posed by traditional document-based systems engineering. It represents a formal recognition of the application of modeling methods in systems engineering activities. By establishing relevant modeling standards and methodologies, and using a unified SysML language for forward design, it constructs requirement models, functional models, logical models, and physical models. This enables the decomposition, allocation, and trade-offs between requirements and functions to logical and physical architecture, thereby driving product design and iteration, and advancing product testing and comprehensive verification. However, as system complexity increases, designers cannot foresee the final architecture during the system design phase. System architectures cannot be generated overnight; therefore, appropriate decisions must be made. System architects are responsible for translating a series of development requirements and goals proposed by stakeholders into system architecture design requirements and reflecting them in the architectural process. During system architecture design, designers are responsible for making sound decisions to ultimately obtain a system with maximum value. Modern airborne system architecture trade-offs are gradually moving towards a model-driven, data-closed-loop, and multi-objective collaborative paradigm.

[0004] Currently, China lacks a mature and feasible technical roadmap for balancing architectural trade-offs in airborne system design. Instead, it still relies on historical design experience and joint expert reviews to validate airborne system architectures. While this approach ensures the feasibility of the airborne system architecture, it cannot guarantee optimal performance in terms of power consumption, efficiency, and other attributes. Summary of the Invention

[0005] In view of this, embodiments of this application provide an airborne system architecture design method and system based on constraint satisfaction problems, which at least partially solves the problem that existing airborne system architecture design methods cannot ensure the optimality of the architecture in terms of power consumption, efficiency and other attributes.

[0006] In a first aspect, embodiments of this application provide an airborne system architecture design method based on constraint satisfaction problems, the method comprising: Construct an architecture decision model, which includes multiple decision items and multiple alternatives corresponding to each decision item, with each alternative being independent of the others; Based on the development goals of the airborne system, determine the evaluation indicators for the decision-making items; Based on the constraint satisfaction problem model, in the architecture decision model, for each decision item, options that meet the preset constraint rules are searched among the alternatives to obtain the Pareto front set of architecture solutions; The Pareto front set of the proposed architecture is optimized by multi-objective trade-offs to obtain the optimal architecture.

[0007] According to a specific implementation of an embodiment of this application, the metrics include performance metrics, security metrics, reliability metrics, and cost metrics. Performance metrics include system response time, throughput, concurrent processing, and power consumption. Security metrics include data encryption strength and access control granularity. Reliability metrics include mean time between failures (MTBF) and mean time to recovery (MTBF). Cost metrics include initial development cost and operating cost.

[0008] According to a specific implementation of an embodiment of this application, the constraint satisfaction problem model includes a variable set, a value range set, and a constraint set. In the architecture decision model, for each decision item, options that meet preset constraint rules are searched from the alternatives to obtain a Pareto front set of architecture solutions, including: Each decision item in the architecture decision model forms a variable in the variable set, and the multiple alternatives corresponding to each decision item form a value range in the value range set. A variable in the variable set corresponds to a value range in the value range set. Preset constraint rules are used as a constraint set, and the constraints in the constraint set act on the value range set. Each time, a variable is selected from the constraint satisfaction problem model, and all possible values ​​are assigned sequentially in the value range corresponding to that variable. Each selected variable is checked to see if it satisfies the constraint set. If it does not satisfy the constraint set, the process backtracks to the variable with available values ​​and continues to filter. If it satisfies the constraint set, the next variable is assigned and filtered, until all variables have been filtered and assigned values, and finally the Pareto front set of the architecture solution that satisfies the constraints is obtained.

[0009] According to a specific implementation of an embodiment of this application, the step of performing multi-objective trade-off optimization on the Pareto front set of the architecture scheme to obtain the optimal architecture scheme includes: Set an objective function for each decision item in the Pareto front set of the architecture scheme; Based on the objective function, the weight range of each decision item is obtained through weight sensitivity analysis. Based on the weight range of each decision item, weight combinations are performed to obtain multiple weight combinations; Based on each set of weights, the deviation matrix corresponding to multiple architecture schemes in the Pareto front set of architecture schemes is calculated respectively; The deviation values ​​between the architecture scheme and the ideal scheme under different weight combinations are calculated based on the deviation matrix. The optimal architecture solution is selected based on the deviation value.

[0010] According to a specific implementation of an embodiment of this application, the step of obtaining the weight range of each decision item based on the objective function and through a weight sensitivity analysis method includes: An initial weight is assigned to each decision item based on expert opinions or historical data, and the initial optimal solution and objective function value are calculated for each decision item. Each initial weight is perturbed with a preset step size, and the weight range of each decision item is obtained based on the rate of change of the objective function value.

[0011] According to a specific implementation of this application, the formula for calculating the rate of change of the objective function value is as follows: , Where k is the k-th decision term, Sensitivity k f is the rate of change of the objective function value. k The objective function value, w represents the change in the objective function value. k As weight, This represents the change in weight.

[0012] According to a specific implementation of an embodiment of this application, obtaining the weight range of each decision item based on the rate of change of the objective function value includes: When the rate of change of the objective function value When the objective function value is set, the weight range of the decision item corresponding to the objective function value is set within the first range of the initial weights; When the rate of change of the objective function value When the objective function value is set, the weight range of the decision term is set within the second range of the initial weight, and the span of the second range is greater than the span of the first range.

[0013] According to a specific implementation of an embodiment of this application, the step of calculating the deviation matrix corresponding to multiple architecture schemes in the Pareto front set of architecture schemes based on each set of weight combinations includes: A decision matrix is ​​formed based on multiple architectural schemes in the Pareto frontier set of architectural schemes. The indicator values ​​corresponding to the measurement indicators of each architectural scheme in the decision matrix are positiveized and standardized. The standardized decision matrix is ​​weighted based on a combination of weights to obtain a weighted standardized decision matrix. Based on the weighted standardized decision matrix, the positive and negative ideal solutions are calculated. Based on the positive and negative ideal solutions, the deviation matrix corresponding to each weight combination is obtained, and each deviation value d in the deviation matrix is... ij Representative architecture scheme F i In the measurement index S j The standardized distance between the above and the positive ideal solution is given. A deviation of 0 corresponds to the positive ideal solution, and a deviation of 1 corresponds to the negative ideal solution.

[0014] According to a specific implementation of an embodiment of this application, the forwarding expression is: , The standardized expression is: , The expression for the weighted standardized decision matrix is: , Among them, S ij S represents the value of the j-th metric for the i-th architecture scheme. j Let R be the value of the j-th metric. ij Let represent the standardized value of the j-th metric for the i-th architecture scheme, m represent the total number of architecture schemes, and w represent the total number of architecture schemes. j Let be the weight of the j-th metric, and T be the weighted standardized decision matrix.

[0015] Secondly, embodiments of this application also provide an airborne system architecture design system based on constraint satisfaction problems, used to implement the airborne system architecture design method based on constraint satisfaction problems as described in any embodiment of the first aspect, the system comprising: The model building module is used to build an architecture decision model, which includes multiple decision items and multiple alternatives corresponding to each decision item, and the alternatives are independent of each other. The metrics determination module is used to determine the metrics for decision items based on the development goals of the airborne system. The filtering module is used to search for options that meet preset constraint rules among the alternatives for each decision item in the architecture decision model based on the constraint satisfaction problem model, so as to obtain the Pareto front set of architecture solutions. The trade-off optimization module is used to perform multi-objective trade-off optimization on the Pareto front set of the architecture scheme to obtain the optimal architecture scheme.

[0016] Beneficial effects: The airborne system architecture design method and system based on constraint satisfaction problems in this application quantifies the decision-making problems encountered in the model-based architecture design process and forms the optimal solution decision result through a reasonable reasoning process. This can assist designers in conducting more reasonable modeling and analysis of complex systems. Within the constrained architecture solution space, the solution with the best functionality and performance is selected. Furthermore, the architectural trade-offs between conflicting or interacting design objectives must be comprehensively considered to obtain the most satisfactory architecture solution with superior overall performance and higher value. This method and system not only improve the scientific rigor and efficiency of airborne system architecture design but also significantly reduce the impact of human factors on the decision-making results. In practical applications, this system can provide strong support for the research and development of airborne systems, especially in scenarios with high complexity and multiple constraints. For example, in the design of flight control systems, this method can be used to quickly evaluate multiple alternative solutions and find the optimal architecture that meets real-time requirements while also considering energy consumption and safety. This data-driven and model-reasoning-based design approach not only improves the overall performance of airborne systems but also reserves ample space for future technological upgrades. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a schematic diagram of an architecture decision model according to an embodiment of the present invention; Figure 2 This is a schematic diagram representing a constraint satisfaction problem according to an embodiment of the present invention; Figure 3 This is a schematic diagram of a decision-making scheme selection process based on a constraint satisfaction problem according to an embodiment of the present invention; Figure 4 This is a schematic diagram of the Pareto front solution set according to an embodiment of the present invention; Figure 5This is a schematic diagram of weight sensitivity analysis and architecture optimization screening according to an embodiment of the present invention; Figure 6 This is a schematic diagram of decision matrix optimization analysis according to an embodiment of the present invention. Detailed Implementation

[0019] The embodiments of this application will now be described in detail with reference to the accompanying drawings.

[0020] The following specific examples illustrate the implementation of this application. Those skilled in the art can easily understand other advantages and effects of this application from the content disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. This application can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of this application. It should be noted that, in the absence of conflict, the following embodiments and features in the embodiments can be combined with each other. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0021] It should be noted that various aspects of embodiments within the scope of the appended claims are described below. It will be apparent that the aspects described herein can be embodied in a wide variety of forms, and any particular structure and / or function described herein is merely illustrative. Based on this application, those skilled in the art will understand that one aspect described herein can be implemented independently of any other aspect, and two or more of these aspects can be combined in various ways. For example, any number of aspects set forth herein can be used to implement the device and / or practice the method. Additionally, this device and / or method can be implemented using structures and / or functionalities other than one or more of the aspects set forth herein.

[0022] It should also be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of this application. The illustrations only show the components related to this application and are not drawn according to the number, shape and size of the components in actual implementation. In actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0023] Furthermore, specific details are provided in the following description to facilitate a thorough understanding of the examples. However, those skilled in the art will understand that the described aspects can be practiced without these specific details.

[0024] For complex systems, the purpose of architecture design is not only to achieve system functions but also to meet non-functional requirements, such as security and reliability goals and constraints. However, when multiple different system architecture solutions exist, architects find it difficult to make a reasonable decision directly. This is because selecting the architecture with the best functional performance from the space of architecture solutions that meet the constraints requires comprehensively considering the trade-offs between conflicting or interacting design goals to obtain the most satisfactory architecture solution with better overall performance and higher value.

[0025] In the model-based system architecture design process, this invention extracts constraints and requirements related to the system from the document content to form decision constraints. At each stage of system design, a decision matrix is ​​formed based on the decision items and their alternative objectives. Using a constraint satisfaction problem approach, all possible system architecture spaces corresponding to the decisions are filtered to ultimately obtain the optimal airborne system architecture design scheme.

[0026] In a first aspect, embodiments of this application provide an airborne system architecture design method based on constraint satisfaction problems, the method comprising: Construct an architecture decision model, which includes multiple decision items and multiple alternatives corresponding to each decision item, with each alternative being independent of the others; Based on the development goals of the airborne system, determine the evaluation indicators for the decision-making items; Based on the constraint satisfaction problem model, in the architecture decision model, for each decision item, options that meet the preset constraint rules are searched among the alternatives to obtain the Pareto front set of architecture solutions; The Pareto front set of the proposed architecture is optimized by multi-objective trade-offs to obtain the optimal architecture.

[0027] In this embodiment, by structurally modeling decision items and alternatives, the independence of each decision item and the diversity of its alternatives are ensured. This method effectively reduces redundant calculations in the system architecture design process and improves design efficiency. Simultaneously, through strict definition and screening mechanisms of constraint rules, it ensures that the generated architecture scheme meets actual needs and achieves balanced optimization among multiple objectives. This method is not only applicable to the design of airborne systems but can also be extended to the architecture design of other complex systems, demonstrating strong versatility and applicability.

[0028] In practical implementation, during system architecture modeling, decisions will be required. These decisions are typically made explicit by constructing a decision model for comparison and analysis by the architect. In this embodiment, the decision items and alternatives are represented as a matrix, referring to... Figure 1As shown, each option in the decision matrix is ​​a discrete choice, and each decision item represents a decision type. Each decision requires selecting one and only one option from the alternatives, and each option is independent of the others. Designers can use rules to constrain and eliminate unreasonable combinations. Through the representation of the decision matrix, the system architecture can demonstrate the problem to be decided and the available options for each problem, providing assistance for further architectural analysis.

[0029] In one embodiment, the metrics include performance metrics, security metrics, reliability metrics, and cost metrics. Performance metrics include system response time, throughput, concurrent processing, and power consumption. Security metrics include data encryption strength and access control granularity. Reliability metrics include mean time between failures (MTBF) and mean time to recovery (MTBF). Cost metrics include initial development cost and operating cost.

[0030] In this embodiment, the analysis of metrics for architectural decision items is a key step in evaluating and comparing different architectural solutions. When evaluating the quality of an architecture, it is necessary to propose metrics that can measure performance, functionality, cost, risk, and other aspects.

[0031] In practice, the selection of these metrics must be tailored to the specific application scenarios and requirements of the airborne system. For example, in the architecture design of a flight control system, performance metrics may focus more on real-time performance and low latency, while security metrics need to emphasize data encryption strength and access control management. Furthermore, for reliability metrics, the system may be further refined into hardware reliability and software reliability dimensions to comprehensively assess its operational stability. Cost metrics need to comprehensively consider the cost distribution throughout the entire lifecycle, including costs incurred during development, deployment, maintenance, and upgrade phases. By quantifying these multi-dimensional metrics and incorporating them into the decision matrix, the actual performance of each alternative architecture can be more accurately reflected, providing a scientific basis for subsequent trade-offs and optimizations.

[0032] To ensure the rationality and effectiveness of metrics, it is typically necessary to combine the opinions of domain experts and experience data from historical projects when defining metrics. At the same time, the priorities of different metrics must be ranked to address potential conflicts. For example, improving system performance may lead to increased costs, while strengthening security may reduce system responsiveness. Therefore, when making trade-off optimizations, it is essential to clearly define which metrics are rigid constraints and which metrics can be compromised by adjusting their weights. This flexible yet rigorous approach helps architects find optimal solutions in complex design spaces, thereby meeting the diverse needs of airborne systems.

[0033] In one embodiment, during the system architecture design process, the established decision items are organized and summarized to obtain a sizable space of architectural solutions. However, not all of these solutions meet the feasibility requirements. Therefore, the architect needs to set specific constraint rules to filter out a large number of infeasible architectural solutions, thereby narrowing the space of architectural solutions and selecting a small number of decision solutions that meet the constraints. By constructing the Pareto front, the optimal decision solution for each objective can be obtained.

[0034] In one embodiment, the constraint satisfaction problem model includes a set of variables, a set of value ranges, and a set of constraints. In the architecture decision model, for each decision item, options that meet preset constraint rules are searched from the alternatives to obtain a Pareto front set of architecture solutions, including: Each decision item in the architecture decision model forms a variable in the variable set, and the multiple alternatives corresponding to each decision item form a value range in the value range set. A variable in the variable set corresponds to a value range in the value range set. Preset constraint rules are used as a constraint set, and the constraints in the constraint set act on the value range set. Each time, a variable is selected from the constraint satisfaction problem model, and all possible values ​​are assigned sequentially in the value range corresponding to that variable. Each selected variable is checked to see if it satisfies the constraint set. If it does not satisfy the constraint set, the process backtracks to the variable with available values ​​and continues to filter. If it satisfies the constraint set, the next variable is assigned and filtered, until all variables have been filtered and assigned values, and finally the Pareto front set of the architecture solution that satisfies the constraints is obtained.

[0035] In specific implementation, refer to Figure 2 The Constraint Satisfaction Problem (CSP) method mainly consists of three elements: a set of variables V, a set of values ​​D, and a set of constraints C. Each variable in set V corresponds to a value in set D, and the constraints in set C act on V, restricting each value in V. In the system architecture design process, variable V is the decision item in the decision matrix, the value range D is the option corresponding to the decision item, and the constraint C is the relationship between the options. The constraint C is defined by the designer and determined according to the constraints, standards, and design specifications between the decision items in the architecture. A CSP model can be constructed using the decision matrix and domain knowledge, ultimately obtaining an architecture solution space that satisfies all constraints. For the three components of CSP, variable V and value range D can be obtained from the decision matrix described above, such as... Figure 2 As shown in the example, in system architecture decisions, variable V can be represented as a set of multiple variable values, where n represents the number of decision items, and V is the set of values. xThe decision items are defined in the architecture. The value range D is the set of all alternatives in the decision matrix. In Figure 2, the set of variables V is (V1, V2, V3, V4), which has four variables; the corresponding value range D contains (D1, D2, D3, D4), where V1 takes values ​​in D1={H1,H2,H3,...,H6}, and V2, V3, and V4 are similarly defined, meaning that each variable V1 has a corresponding value range. After constraint modeling using the CSP method, the system architecture is described in a standardized way, and further optimization and solution of the solution space are then performed. The method flow is as follows: Figure 3 As shown, each time a variable V is selected from the CSP model i In its corresponding value range D i Assign all possible values ​​to D in sequence. ij Indicates the range D i The j-th value is selected. Each selected variable is filtered to see if it meets the requirements of constraint C. If it does not meet the requirements of constraint C, the process backtracks to variables with available values ​​and continues filtering. If it meets the requirements of constraint C, the next variable is assigned and filtered, and this process is repeated until all variables have been filtered and assigned values. The final result is the set of Pareto fronts that satisfy the constraints. Figure 4 The diagram shows the Pareto front solution set for the three decision indicators.

[0036] In another embodiment, the Falcon-ASN (Critical Architecture Space Navigation) method based on CSP (Constraint Satisfaction Problem) can also be used for constraint screening. This method decomposes decision items into multiple independent orthogonal spaces based on functional independence. For each space, all possible options are listed. Only the two seemingly best basic options are selected in each dimension. Then, a specific analysis is conducted on how combinations of options from different dimensions can produce new effects or new problems. Based on this analysis, approximately 10% of innovative combinations are added. Simultaneously, some key constraints can be reversed to design solutions that can counteract these constraints, resulting in an initial "solution pool" containing various possible combinations. Based on the known constraints, an architecture-constraint traceability matrix is ​​constructed. Each solution in the solution pool is examined to see if it fully covers these constraints, eliminating those that do not meet the constraints. Finally, a Pareto front set of architecture solutions that satisfy the constraints is obtained.

[0037] In this embodiment, the Falcon-ASN (Critical Architecture Space Navigation) method based on CSP (Constraint Satisfaction Problem) significantly improves the efficiency and quality of architecture design. By decomposing complex decision items into multiple orthogonal spaces, it not only reduces the complexity of the problem but also makes the analysis of each dimension more focused and in-depth. Simultaneously, the design approach of selecting basic options and combining them with innovative combinations ensures the feasibility of the solutions while injecting flexibility and creativity into the system architecture. Furthermore, reverse design using key constraints effectively addresses potential risks and challenges, ensuring that the generated architecture solutions have stronger robustness and adaptability. The final "solution pool," after rigorous constraint screening, yields a Pareto front set that not only satisfies all preset conditions but also achieves a good balance among multiple objectives, thus providing architects with a scientific and practical basis for decision-making. This method is particularly suitable for airborne system design in highly complex, multi-constraint scenarios and also has the potential to be extended to complex systems in other fields, demonstrating high engineering application value.

[0038] In one embodiment, the step of performing multi-objective trade-off optimization on the Pareto front set of the architecture schemes to obtain the optimal architecture scheme includes: Set an objective function for each decision item in the Pareto front set of the architecture scheme; Based on the objective function, the weight range of each decision item is obtained through weight sensitivity analysis. Based on the weight range of each decision item, weight combinations are performed to obtain multiple weight combinations; Based on each set of weights, the deviation matrix corresponding to multiple architecture schemes in the Pareto front set of architecture schemes is calculated respectively; The deviation values ​​between the architecture scheme and the ideal scheme under different weight combinations are calculated based on the deviation matrix. The optimal architecture solution is selected based on the deviation value.

[0039] Furthermore, based on the objective function, the weight range of each decision item is obtained through weight sensitivity analysis, including: An initial weight is assigned to each decision item based on expert opinions or historical data, and the initial optimal solution and objective function value are calculated for each decision item. Each initial weight is perturbed with a preset step size, and the weight range of each decision item is obtained based on the rate of change of the objective function value.

[0040] Furthermore, the formula for calculating the rate of change of the objective function value is as follows: , Where k is the k-th decision term, Sensitivity kf is the rate of change of the objective function value. k The objective function value, w represents the change in the objective function value. k As weight, This represents the change in weight.

[0041] Furthermore, obtaining the weight range for each decision item based on the rate of change of the objective function value includes: When the rate of change of the objective function value When the objective function value is set, the weight range of the decision item corresponding to the objective function value is set within the first range of the initial weights; When the rate of change of the objective function value When the objective function value is set, the weight range of the decision term is set within the second range of the initial weight, and the span of the second range is greater than the span of the first range.

[0042] In specific implementation, refer to Figure 5 For multi-objective trade-off optimization, the first step is to analyze the reasonable weights of each objective function based on the preferences of different stakeholders regarding the system architecture design. Here, the objective function refers to the function used for analysis of each decision item. Initially, based on expert opinions or historical data, initial weights W0 = (w1, w2, ..., w...) are set. n Solve for the initial optimal solution x. * and the values ​​of each objective function f i (x * Then, for one of the weights w k Perturb by a step size Δw (e.g., 5%), solve the optimization problem again, and record the rate of change of the objective value: ,like If the k-th decision objective is highly sensitive, the disturbance will cause the optimal solution to change abruptly, indicating that the range of the weight value is within a small range of the initial value.

[0043] Furthermore, a global weight scan can be performed using the Monte Carlo method, setting the weight value range step size to 0.1, and sampling in the weight space to satisfy... , The Pareto optimal solution set is calculated by analyzing each set of weights. The clustering regions of optimal solutions in the target space are observed, and a "weight-solution set" correlation matrix is ​​constructed. In each iteration, the decision matrix is ​​updated by incorporating the weight combinations. The optimal architecture under this weight combination is found using the ideal point ranking method, and the deviation from the system's expected target is calculated. By traversing all design scenarios, the deviation matrix is ​​obtained. Combined with the weight matrix analysis, the weight range required by the system design is determined. Finally, the decision schemes are ranked and analyzed based on the ideal point ranking method.

[0044] In one embodiment, calculating the deviation matrix corresponding to multiple architectural schemes in the Pareto front set for each set of weight combinations includes: A decision matrix is ​​formed based on multiple architectural schemes in the Pareto frontier set of architectural schemes. The indicator values ​​corresponding to the measurement indicators of each architectural scheme in the decision matrix are positiveized and standardized. The standardized decision matrix is ​​weighted based on a combination of weights to obtain a weighted standardized decision matrix. Based on the weighted standardized decision matrix, the positive and negative ideal solutions are calculated. Based on the positive and negative ideal solutions, the deviation matrix corresponding to each weight combination is obtained, and each deviation value d in the deviation matrix is... ij Representative architecture scheme F i In the measurement index S j The standardized distance between the above and the positive ideal solution is given. A deviation of 0 corresponds to the positive ideal solution, and a deviation of 1 corresponds to the negative ideal solution.

[0045] Furthermore, the expression for the forwarding is: (1), The standardized expression is: (2), The expression for the weighted standardized decision matrix is: (3), (4), Among them, S ij S represents the value of the j-th metric for the i-th architecture scheme. j Let R be the value of the j-th metric. ij Let represent the standardized value of the j-th metric for the i-th architecture scheme, m represent the total number of architecture schemes, and w represent the total number of architecture schemes. j Let I be the weight of the j-th metric, and T be the weighted standardized decision matrix. + I- represents a positive ideal solution, and I- represents a negative ideal solution.

[0046] In specific implementation, refer to Figure 6 Assume there are m possible architectures for the Pareto front after multi-objective optimization, denoted as m. Suppose there are three indicators for selecting the architecture scheme, denoted as: Therefore, each architecture has a corresponding metric value. Figure 6 The value on the right is its standardized value, denoted as . The index S is positiveized through a specific formula (1) and standardized through formula (2).

[0047] Assumption There are n scenarios designed for designers. Taking scenario W1 as an example, assume its weight combination is... ,satisfy The decision matrix is ​​further weighted using formula (3), and the positive and negative ideal solutions are calculated using formula (4). The final selected non-dominated solution has the shortest distance to the determined positive ideal solution and the greatest distance to the determined negative ideal solution.

[0048] Taking the weighted combination W1 as an example, its design deviation (d) 11 ,d 12 ,d 13 The calculation method for indicator S1 is explained as follows: When S1 is an extremely large indicator, its deviation is calculated as follows: , When indicator S1 is a very small indicator: , In the above formula, when the architecture F j Indicator value S 1j When optimal, the deviation value d 11 =0, when its index value is at its worst, d 11 =1, thus making constraint d =1, thereby constraining d 1j Within the range [0,1]. Deviation d 12 and d 13 The calculation method is similar. By calculating the deviation, the deviation value under the weight combination W1 can be obtained, that is... The deviation values ​​for other cases are calculated sequentially to obtain the deviation matrix D. d : , By calculating the deviation matrix, the deviation values ​​between the proposed scheme and the ideal scheme under different weight combinations are obtained, and finally the most satisfactory scheme that is closest to the ideal scheme is selected.

[0049] In this embodiment, by analyzing the deviation matrix, the performance of each architecture scheme under different weight combinations can be further clarified. Specifically, the closer the deviation value is to 0, the closer the architecture scheme is to the positive ideal solution on the corresponding indicator; while the closer the deviation value is to 1, the closer it is to the negative ideal solution. Based on this principle, all architecture schemes are ranked, and the scheme with the smallest comprehensive deviation value is selected as the recommended scheme. To improve the efficiency of screening and the reliability of the results, visualization tools can also be introduced to assist decision-making. For example, the data in the deviation matrix can be displayed in the form of a heatmap, and the color intensity can intuitively reflect the deviation of each scheme on different indicators. At the same time, the distribution characteristics of the architecture schemes in the multi-dimensional target space can be presented using a three-dimensional scatter plot or parallel coordinate plot to help designers quickly identify potential candidate schemes. This method not only improves the transparency of decision-making but also facilitates communication and negotiation among stakeholders, thereby reaching a consensus.

[0050] In practical applications, the proposed optimization method can be embedded into the entire design process of airborne systems, forming a closed-loop feedback mechanism. After each iteration, the weight combination is adjusted according to new constraints or objective functions, the deviation matrix is ​​recalculated, and the optimal solution is updated. This dynamic adjustment capability enables the system design to flexibly respond to environmental changes or demand adjustments, significantly enhancing the adaptability and robustness of the design process. Ultimately, through multiple rounds of iteration and optimization, not only can the optimal architecture solution for the current scenario be obtained, but also a large amount of historical data can be accumulated, providing valuable experience support for similar projects in the future.

[0051] Secondly, embodiments of this application also provide an airborne system architecture design system based on constraint satisfaction problems, used to implement the airborne system architecture design method based on constraint satisfaction problems as described in any embodiment of the first aspect, the system comprising: The model building module is used to build an architecture decision model, which includes multiple decision items and multiple alternatives corresponding to each decision item, and the alternatives are independent of each other. The metrics determination module is used to determine the metrics for decision items based on the development goals of the airborne system. The filtering module is used to search for options that meet preset constraint rules among the alternatives for each decision item in the architecture decision model based on the constraint satisfaction problem model, so as to obtain the Pareto front set of architecture solutions. The trade-off optimization module is used to perform multi-objective trade-off optimization on the Pareto front set of the architecture scheme to obtain the optimal architecture scheme.

[0052] The embodiments provided by this invention quantify the decision-making problems encountered in the model-based architecture design process and, through a reasonable reasoning process, form the optimal solution decision result. This can assist designers in conducting more reasonable modeling and analysis of complex systems. Within a constrained architectural solution space, the solution with the best functionality and performance is selected. Furthermore, the architectural trade-offs between conflicting or interacting design objectives must be comprehensively considered to obtain the most satisfactory architectural solution with superior overall performance and higher value. This method and system not only improve the scientific rigor and efficiency of airborne system architecture design but also significantly reduce the impact of human factors on decision-making results. In practical applications, this system can provide strong support for the research and development of airborne systems, especially in scenarios with high complexity and multiple constraints. For example, in the design of flight control systems, this method can quickly evaluate multiple alternative solutions to find the optimal architecture that meets real-time requirements while also considering energy consumption and safety. This data-driven and model-reasoning-based design approach not only improves the overall performance of airborne systems but also reserves ample space for future technological upgrades.

[0053] 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. An airborne system architecture design method based on constraint satisfaction problems, characterized in that, The method includes: Construct an architecture decision model, which includes multiple decision items and multiple alternatives corresponding to each decision item, with each alternative being independent of the others; Based on the development goals of the airborne system, determine the evaluation indicators for the decision-making items; Based on the constraint satisfaction problem model, in the architecture decision model, for each decision item, options that meet the preset constraint rules are searched among the alternatives to obtain the Pareto front set of architecture solutions; The Pareto front set of the proposed architecture is optimized by multi-objective trade-offs to obtain the optimal architecture.

2. The airborne system architecture design method based on constraint satisfaction problem according to claim 1, characterized in that, The metrics include performance metrics, security metrics, reliability metrics, and cost metrics. Performance metrics include system response time, throughput, concurrent processing, and power consumption. Security metrics include data encryption strength and access control granularity. Reliability metrics include mean time between failures (MTBF) and mean time to recovery (MTBF). Cost metrics include initial development cost and operating cost.

3. The airborne system architecture design method based on constraint satisfaction problem according to claim 1, characterized in that, The constraint satisfaction problem model includes a set of variables, a set of value ranges, and a set of constraints. In the architecture decision model, for each decision item, options that meet preset constraint rules are searched from the alternatives to obtain the Pareto front set of architecture solutions, including: Each decision item in the architecture decision model forms a variable in the variable set, and the multiple alternatives corresponding to each decision item form a value range in the value range set. A variable in the variable set corresponds to a value range in the value range set. Preset constraint rules are used as a constraint set, and the constraints in the constraint set act on the value range set. Each time, a variable is selected from the constraint satisfaction problem model, and all possible values ​​are assigned sequentially in the value range corresponding to that variable. Each selected variable is checked to see if it satisfies the constraint set. If it does not satisfy the constraint set, the process backtracks to the variable with available values ​​and continues to filter. If it satisfies the constraint set, the next variable is assigned and filtered, until all variables have been filtered and assigned values, and finally the Pareto front set of the architecture solution that satisfies the constraints is obtained.

4. The airborne system architecture design method based on constraint satisfaction problem according to claim 1, characterized in that, The step of performing multi-objective trade-off optimization on the Pareto front set of the architecture scheme to obtain the optimal architecture scheme includes: Set an objective function for each decision item in the Pareto front set of the architecture scheme; Based on the objective function, the weight range of each decision item is obtained through weight sensitivity analysis. Based on the weight range of each decision item, weight combinations are performed to obtain multiple weight combinations; Based on each set of weights, the deviation matrix corresponding to multiple architecture schemes in the Pareto front set of architecture schemes is calculated respectively; The deviation values ​​between the architecture scheme and the ideal scheme under different weight combinations are calculated based on the deviation matrix. The optimal architecture solution is selected based on the deviation value.

5. The airborne system architecture design method based on constraint satisfaction problem according to claim 4, characterized in that, The weight range for each decision item, obtained based on the objective function and through weight sensitivity analysis, includes: An initial weight is assigned to each decision item based on expert opinions or historical data, and the initial optimal solution and objective function value are calculated for each decision item. Each initial weight is perturbed with a preset step size, and the weight range of each decision item is obtained based on the rate of change of the objective function value.

6. The airborne system architecture design method based on constraint satisfaction problem according to claim 5, characterized in that, The formula for calculating the rate of change of the objective function value is: ， Where k is the k-th decision term, Sensitivity k f is the rate of change of the objective function value. k The objective function value, w represents the change in the objective function value. k As weight, This represents the change in weight.

7. The airborne system architecture design method based on constraint satisfaction problem according to claim 6, characterized in that, The weight range for each decision item, obtained based on the rate of change of the objective function value, includes: When the rate of change of the objective function value When the objective function value is set, the weight range of the decision item corresponding to the objective function value is set within the first range of the initial weights; When the rate of change of the objective function value When the objective function value is set, the weight range of the decision term is set within the second range of the initial weight, and the span of the second range is greater than the span of the first range.

8. The airborne system architecture design method based on constraint satisfaction problem according to claim 4, characterized in that, The step of calculating the deviation matrix corresponding to multiple architecture schemes in the Pareto front set for each set of weight combinations includes: A decision matrix is ​​formed based on multiple architectural schemes in the Pareto frontier set of architectural schemes. The indicator values ​​corresponding to the measurement indicators of each architectural scheme in the decision matrix are positiveized and standardized. The standardized decision matrix is ​​weighted based on a combination of weights to obtain a weighted standardized decision matrix. Based on the weighted standardized decision matrix, the positive and negative ideal solutions are calculated. Based on the positive and negative ideal solutions, the deviation matrix corresponding to each weight combination is obtained, and each deviation value d in the deviation matrix is... ij Representative architecture scheme F i In the measurement index S j The standardized distance between the above and the positive ideal solution is given. A deviation of 0 corresponds to the positive ideal solution, and a deviation of 1 corresponds to the negative ideal solution.

9. The airborne system architecture design method based on constraint satisfaction problem according to claim 8, characterized in that, The forwarding expression is: ， The standardized expression is: ， The expression for the weighted standardized decision matrix is: ， Among them, S ij S represents the value of the j-th metric for the i-th architecture scheme. j Let R be the value of the j-th metric. ij Let represent the standardized value of the j-th metric for the i-th architecture scheme, m represent the total number of architecture schemes, and w represent the total number of architecture schemes. j Let be the weight of the j-th metric, and T be the weighted standardized decision matrix.

10. An airborne system architecture design system based on constraint satisfaction problems, used to implement the airborne system architecture design method based on constraint satisfaction problems as described in any one of claims 1-9, characterized in that, The system includes: The model building module is used to build an architecture decision model, which includes multiple decision items and multiple alternatives corresponding to each decision item, and the alternatives are independent of each other. The metrics determination module is used to determine the metrics for decision items based on the development goals of the airborne system. The filtering module is used to search for options that meet preset constraint rules among the alternatives for each decision item in the architecture decision model based on the constraint satisfaction problem model, so as to obtain the Pareto front set of architecture solutions. The trade-off optimization module is used to perform multi-objective trade-off optimization on the Pareto front set of the architecture scheme to obtain the optimal architecture scheme.

Images