Method, device, equipment and medium for semantic-driven aero-engine component modeling and simulation code automatic generation

Abstract

Embodiments of the present application provide a kind of semantic driven aero-engine component modeling and simulation code automatic generation method, device, equipment and medium, involve modeling simulation technical field, wherein, including: by semantic analysis identification, generation structured flow path topology and rotor mechanical connection structure;Based on the preset section numbering system, automatically allocate section numbering for each component, generate the section numbering list of each component, determine the input and output parameters and calling sequence of calculation function;Based on the preset parameter naming system, generate the uniform parameter set required by model, and generate the balance equation and iteration variable for steady-state solution and construct nonlinear equation set;Based on flow path topology, uniform parameter set, nonlinear equation set and calling sequence, simulation solving program is automatically generated through preset program template.The scheme completes the whole process from engine natural language description to simulation code generation by semantic driving, improves modeling efficiency.

Description

Technical Field

[0001] This invention relates to the field of modeling and simulation technology, and in particular to a method, apparatus, equipment, and medium for automatic generation of semantically driven modeling and simulation codes for aero-engine components. Background Technology

[0002] The development of aero-engines is costly and challenging, and testing can cause additional damage. Model-based design, however, offers high efficiency and low trial-and-error costs. By establishing aerodynamic and thermodynamic models of core components such as the inlet, compressor, combustion chamber, turbine, and nozzle, development costs can be significantly reduced and test risks minimized. This is a crucial technical approach for overall engine design, cycle parameter optimization, and control system development. As aero-engines evolve towards high performance and multi-configuration designs, higher demands are placed on the efficiency, accuracy, and versatility of component-level modeling technology.

[0003] Existing component-level modeling methods for aero-engines mainly include the following three technical approaches:

[0004] (1) Modeling method based on manual programming: Engineers manually define flow path connection relationships, engine parameters, balance equations, etc. according to the characteristics of different engine types, and then complete the model building by manually writing component calculation programs. This method is highly flexible, but the modeling process relies heavily on the personal experience of engineers, with a large workload and long cycle. Moreover, the models built by different engineers lack consistency in parameter naming and structural organization, making them difficult to reuse and maintain.

[0005] (2) Methods based on graphical modeling environments: Engine models are constructed by dragging and dropping component modules and connecting air passages, such as Simulink. This method has a high degree of visualization, but it still requires manual definition of component interfaces and configuration of parameter mapping relationships. Furthermore, it is difficult to transfer models between platforms, which limits the versatility of the model.

[0006] (3) Template-based semi-automatic modeling method: Predefine model templates for typical engine configurations, and users can modify the template tools to achieve specific configuration modeling. This method is highly efficient under standard configurations, but the template applicability is poor when facing new or unconventional configurations, and more manual debugging and intervention are still required.

[0007] The existing methods described above have the following technical problems: First, the modeling process is highly dependent on manual operation. From flow path construction and parameter definition to code writing, everything needs to be done manually, which is inefficient and prone to human error. It also requires a high level of modeling experience from technical personnel. Second, the methods lack unified parameter naming and model structure specifications. Models built by different engineers or projects vary significantly in variable naming, section numbering, and parameter organization, making it difficult to share and reuse models among teams, increasing learning costs, collaboration costs, and maintenance difficulties. Finally, when faced with new configurations or iterative solutions, the existing methods require remodeling and reprogramming, which is difficult to meet the engineering needs of rapid design and verification. This is especially true in the overall conceptual design phase when a large number of candidate solutions need to be evaluated, where modeling efficiency becomes a limiting factor.

[0008] To address the aforementioned issues, a technical solution is needed that can autonomously generate standardized model structures and simulation code based on engine mechanism types, thereby improving modeling efficiency, ensuring model consistency, and supporting rapid modeling and simulation of complex engine mechanism types. Summary of the Invention

[0009] In view of this, embodiments of the present invention provide a semantically driven method for automatic generation of modeling and simulation code for aero-engine components, to solve the technical problems in the existing aero-engine component-level modeling process, which rely on manual experience, resulting in low modeling efficiency, inconsistent model structures and parameters, difficulty in model reuse, and high simulation code development costs. The method includes: By semantic parsing, the engine type and corresponding natural language description input by the user are identified, and the engine component type, airflow path relationship and mechanical connection relationship of rotating components are obtained. A structured flow path topology is generated based on the engine component type and the airflow path relationship, and a rotor mechanical connection structure is generated based on the mechanical connection relationship of rotating components. Based on the flow path topology, a cross section number is automatically assigned to each aero-engine component according to a preset cross section numbering system, generating a cross section number list for each aero-engine component. Based on the component type in the flow path topology, a matching calculation function is retrieved from a preset component general calculation function library, and the input and output parameters and calling order of the calculation function are determined. Based on the cross-section number list and input / output parameters of each component, a unified parameter set required for the model is generated according to a preset parameter naming system. Based on the flow path topology, the rotor mechanical connection structure, the cross-section number list and the unified parameter set, a balance equation and iterative variables for steady-state solution are generated according to a preset engine modeling system, and a nonlinear equation system is constructed. Based on the flow path topology, the unified parameter set, the nonlinear equation system, and the calling order, a simulation solution program is automatically generated using a preset program template.

[0010] This invention also provides a semantically driven device for automatically generating modeling and simulation code for aero-engine components, addressing the technical problems in existing aero-engine component-level modeling processes that rely on human experience, leading to low modeling efficiency, inconsistent model structures and parameters, difficulty in model reuse, and high simulation code development costs. The device includes: The topology generation module is used to identify the engine type and corresponding natural language description input by the user through semantic parsing, obtain the engine component type, airflow path relationship and mechanical connection relationship of rotating components, generate a structured flow path topology based on the engine component type and the airflow path relationship, and generate a rotor mechanical connection structure based on the mechanical connection relationship of rotating components. The section function determination module is used to automatically assign section numbers to each aero-engine component based on the flow path topology and a preset section numbering system, generate a section number list for each aero-engine component, retrieve matching calculation functions from a preset component general calculation function library according to the component type in the flow path topology, and determine the input and output parameters and calling order of the calculation functions; The equation construction module is used to generate a unified set of parameters required for the model based on the cross-section number list and the input and output parameters of each component, according to a preset parameter naming system. Based on the flow path topology, the rotor mechanical connection structure, the cross-section number list and the unified parameter set, it generates equilibrium equations and iterative variables for steady-state solution according to a preset engine modeling system, and constructs a nonlinear equation system. The code output module is used to automatically generate a simulation solution program based on the flow path topology, the unified parameter set, the nonlinear equation system, and the calling order, using a preset program template.

[0011] This invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the above-mentioned method for automatically generating modeling and simulation code for any semantically driven aero-engine components, thereby solving the technical problems in the prior art where reliance on human experience in aero-engine component-level modeling leads to low modeling efficiency, inconsistent model structures and parameters, difficulty in model reuse, and high simulation code development costs.

[0012] This invention also provides a computer-readable storage medium storing a computer program that executes any of the above-described semantically driven methods for automatically generating modeling and simulation codes for aero-engine components. This addresses the technical problems in the prior art where reliance on human experience during aero-engine component-level modeling leads to low modeling efficiency, inconsistent model structures and parameters, difficulty in model reuse, and high simulation code development costs.

[0013] Compared with the prior art, the beneficial effects that at least one technical solution adopted in the embodiments of this specification can achieve include at least: By using semantic-driven methods, the entire process of generating engine data—from natural language description to structured topology, parameter system, nonlinear equations, and simulation code—is automatically completed, significantly improving modeling efficiency and model consistency. Attached Figure Description

[0014] 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.

[0015] Figure 1 This is a flowchart of a semantically driven method for automatically generating modeling and simulation code for aero-engine components, provided in an embodiment of the present invention. Figure 2 This is a schematic diagram of a system for automatically generating modeling and simulation code for aero-engine components using the above-described semantically driven method, provided by an embodiment of the present invention. Figure 3 This is a schematic diagram of data flow interaction between the components in an embodiment of the present invention; Figure 4 This is a schematic diagram of the calling order according to the airflow direction in an embodiment of the present invention; Figure 5 This is a structural block diagram of a computer device provided in an embodiment of the present invention; Figure 6 This is a structural block diagram of a semantically driven device for automatically generating modeling and simulation code for aero-engine components, provided in an embodiment of the present invention. Detailed Implementation

[0016] The embodiments of this application will now be described in detail with reference to the accompanying drawings.

[0017] 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.

[0018] In this embodiment of the invention, a semantically driven method for automatic generation of modeling and simulation code for aero-engine components is provided, such as... Figure 1 As shown, the method includes: Step S101: By semantic parsing, identify the engine type and corresponding natural language description input by the user, obtain the engine component type, airflow path relationship and mechanical connection relationship of rotating components, generate a structured flow path topology based on the engine component type and the airflow path relationship, and generate a rotor mechanical connection structure based on the mechanical connection relationship of rotating components; Step S102: Based on the flow path topology, automatically assign a cross section number to each aero-engine component according to a preset cross section numbering system, generate a cross section number list for each aero-engine component, and retrieve a matching calculation function from a preset component general calculation function library according to the component type in the flow path topology, and determine the input and output parameters and calling order of the calculation function; Step S103: Based on the cross-section number list and the input / output parameters of each component, generate a unified parameter set required for the model based on a preset parameter naming system. Based on the flow path topology, the rotor mechanical connection structure, the cross-section number list, and the unified parameter set, generate equilibrium equations and iterative variables for steady-state solution according to a preset engine modeling system, and construct a nonlinear equation system. Step S104: Based on the flow path topology, the unified parameter set, the nonlinear equation set, and the calling order, a simulation solution program is automatically generated using a preset program template.

[0019] In specific implementation, the following steps are used to automatically assign section numbers to each aero-engine component based on the flow path topology and a preset section numbering system, generate a section number list for each aero-engine component, and retrieve matching calculation functions from a preset component general calculation function library according to the component type in the flow path topology, and determine the input and output parameters and calling order of the calculation functions: Based on the arrangement order and airflow path direction of each aero-engine component in the flow path topology, an inlet section number, an outlet section number, and a throat section number are assigned to each aero-engine component according to the preset section numbering system, generating a section number list. Based on the type of engine component in the flow path topology, calculation functions corresponding to each aero-engine component are retrieved from the component general calculation function library. This library includes calculation functions for the inlet, compression component, combustion chamber, turbine component, bypass duct, mixing chamber, and nozzle. Each calculation function includes a function prototype, input and output parameters, and the characteristic structure information of the corresponding aero-engine component. Based on the airflow direction of the flow path topology, the calling order of the calculation functions for each aero-engine component is determined.

[0020] In specific implementation, the following steps are used to generate equilibrium equations and iterative variables for steady-state solution based on the flow path topology, the rotor mechanical connection structure, the cross-section number list, and the unified parameter set, according to the preset engine modeling system, and to construct a nonlinear equation system: Based on the airflow path relationship in the flow path topology and the coaxial connection relationship in the rotor mechanical connection structure, an equilibrium equation is constructed, which includes a mass flow rate continuity equation, a power balance equation, and a static pressure balance equation. According to a preset iterative variable selection method, iterative variables are determined from the unified parameter set, which include the relative physical rotational speed of the shaft, the pressure ratio of the compression component, the expansion ratio of the turbine component, and the ratio of the mass flow rate of the outer bypass duct to the inner bypass duct. The equilibrium equation and the iterative variables are combined to construct a system of nonlinear equations with the iterative variables as unknowns.

[0021] In specific implementation, the following steps are used to construct the equilibrium equation based on the airflow path relationship in the flow path topology and the coaxial connection relationship in the rotor mechanical connection structure: Based on the airflow path relationships in the flow path topology, a mass flow continuity equation is constructed between two adjacent aero-engine components whose outlet flow can be calculated independently. If there is additional mass injection or extraction, the mass flow continuity equation is explicitly modified. The aero-engine components that can independently calculate outlet flow include the compressor, turbine, and nozzle. Based on the coaxial connection relationship in the rotor mechanical connection structure, a power balance equation is constructed between the compressor and turbine components on the same shaft. If there is extraction of power, the extraction term is explicitly subtracted from the power balance equation. Based on the airflow path relationships in the flow path topology, when a mixing structure between the outer bypass duct and the core duct, implemented by the mixing chamber, is identified, a static pressure balance equation is constructed at the inlet of the mixing chamber.

[0022] In specific implementation, the simulation solution program is automatically generated using a preset program template based on the flow path topology, the unified parameter set, the nonlinear equation system, and the calling order through the following steps: According to the calling order, flow path calculation functions are generated that sequentially call each of the aero-engine components along the airflow direction. Each flow path calculation function includes statements for passing input and output parameters between the unified parameter set and the local variables of the flow path calculation function. The flow path calculation function is used to take the current value of the iteration variable as input and output the parameters of each section. Based on the equilibrium equations in the nonlinear equation system, a residual calculation function is constructed. This residual calculation function is used to calculate the residual value of each equilibrium equation under the current value of the iteration variable. Based on the iteration variables in the nonlinear equation system, an iterative solver interface is generated. This iterative solver interface is used to call the Newton-Raphson iteration method or the fixed-point iteration method to solve the nonlinear equation system. The flow path calculation function, the residual calculation function, and the iterative solver interface are assembled according to a preset program template to generate a simulation solution program. This program template includes a main simulation loop framework, parameter initialization code, and result output code.

[0023] In practice, logical consistency checks are performed through the following steps: Based on the physical mechanism of aero-engines, a logical consistency check is performed on the results of semantic parsing to calculate a confidence level. This logical consistency check includes detecting the existence of isolated components, whether the rotor power chain is closed, and whether the flow path is continuous. If the confidence level is below a preset threshold, a visual flow path topology diagram is generated, and the aero-engine components or connections that failed the logical consistency check are displayed in the graphical interface of the diagram. Users can modify the flow path topology and / or the rotor mechanical connection structure through a graphical correction interface, and update the flow path topology and rotor mechanical connection structure. User correction records are saved to a system library and used for iterative optimization of the semantic parsing.

[0024] In practice, the following steps are used to perform a logical consistency check on the semantic parsing results based on the physical mechanism of aero-engines, and to calculate the confidence level evaluation: Based on the principles of aero-engine aero-thermodynamics, the flow path topology is checked for isolated components. If isolated components are found, the logic consistency is deemed unqualified. An isolated component is defined as one with no upstream airflow inflow and no downstream airflow outflow. Based on the rotor power chain closure principle in aero-engine physics, the power chain in the rotor mechanical connection structure is checked for closure. If an unclosed power chain exists, the logic consistency is deemed unqualified. A closed power chain means that each shaft has at least one compression component and one turbine component, and the power transmission path is complete. Based on the flow path continuity principle in aero-engine aero-thermodynamics, the airflow path in the flow path topology is checked for continuity. If discontinuous flow paths exist, the logic consistency is deemed unqualified. Flow path continuity means that the inlet and outlet sections of adjacent components on each airflow branch from the inlet to the nozzle are sequentially connected without any breaks. Based on the results of each logic consistency check, a confidence level assessment corresponding to the semantic parsing result is generated.

[0025] This invention provides a semantically driven automatic code generation system for modeling and simulation of aero-engine components.

[0026] The overall system structure is as follows Figure 2 As shown, it includes the following components: Semantic parsing and flow path topology generation components, section definition components, component function management components, parameter construction and naming components, equilibrium equation and iterative variable generation components, and code generation components. For example... Figure 3 As shown, the above components are coupled through data flow to achieve automatic modeling of component-level models and generation of simulation code.

[0027] The system inputs include: natural language text describing the engine structure or function, a general component calculation function library, a section numbering system, a parameter naming system, an engine modeling system, program templates, and the physical mechanisms of aero-engines. User input can be in natural language or structured format, used to describe the engine configuration and its component characteristics.

[0028] I. Semantic parsing and flow path topology generation component.

[0029] This system is designed to automatically identify the relationship between engine component types and airflow paths from user-input natural language information such as engine configuration descriptions and airflow characteristics. Its implementation includes: using natural language processing techniques based on large language models to perform semantic analysis on the input text, identifying component categories, airflow branches, and mechanical connection information, and generating an ordered component data structure according to the airflow direction; when parallel branches exist, generating a parallel topology according to the semantic structure. Furthermore, it identifies the shaft system and generates the mechanical connection relationship between the compression and turbine components. The system outputs a structured airflow topology description and rotor mechanical connection structure to drive subsequent modeling steps.

[0030] Semantic parsing result confidence assessment and interactive correction: After generating the flow path topology and rotor mechanical connection structure, this component also performs a logical consistency check on the parsing results based on the physical mechanisms of aero-engines (including rules for isolated component detection, rotor power chain closure detection, and flow path continuity detection). A confidence evaluation is generated based on the satisfaction of each rule in the check results. If the confidence evaluation is lower than a preset threshold, a visual flow path topology diagram is generated, and components or connections that failed the check are highlighted in the graphical interface. The system provides a graphical correction interface, accepting user drag-and-drop, addition, deletion, or parameter modification operations on the flow path topology and / or the rotor mechanical connection structure, updating the flow path topology and rotor mechanical connection structure according to user operations. User correction records are stored in the rule base (system library) for subsequent iterative optimization of semantic parsing.

[0031] II. Section Definition Components.

[0032] This component assigns inlet, outlet, and throat sections (if applicable) to each component according to a preset section numbering system (preferably the SAEAS755G standardized numbering system for aerospace propulsion systems), and outputs a table defining the key sections of each component. These numbers are used for naming constraint parameters and transmitting data, ensuring consistency in the model structure.

[0033] In practice, based on the arrangement order of components in the flow path topology and the direction of the airflow path, an inlet cross-section number, an outlet cross-section number, and a throat cross-section number are assigned to each component according to the cross-section numbering system, generating a component cross-section number list. This list is used for subsequent parameter construction.

[0034] III. Component Function Management Component.

[0035] The component function management component retrieves functions and interfaces corresponding to the components from a general component calculation function library. This library contains general calculation function definitions for components such as the intake duct, compression component, combustion chamber, turbine component, bypass duct, mixing chamber, and nozzle. Each function includes a function prototype, a set of input parameters, a set of output parameters, and characteristic structure information of the corresponding component. The system automatically retrieves the corresponding function based on the component type and determines the calling order of the calculation functions for each component according to the airflow direction of the flow path topology, ensuring that component-level aerodynamic and thermodynamic calculations are executed sequentially along the airflow direction. Preferably, this component uses a parallel execution mode to search for general component calculation functions and supports the reuse of general component calculation functions.

[0036] IV. Parameter Construction and Naming Components.

[0037] By combining the already determined key engine section numbers and the function input and output parameters of each component, necessary parameters such as mass flow rate, total pressure, total temperature, static temperature, static pressure, component characteristics, component power consumption / generation, and cross-sectional area are generated according to a preset parameter naming system, achieving unified naming and organization. The naming system defines the composition of parameter identifiers, including parameter type identifiers, section identifiers, component characteristic identifiers, and design parameter identifiers, thereby ensuring that various parameters have a consistent semantic structure and resolvability under different engine configurations.

[0038] Specifically, parameters related to key cross-section locations in the flow path are named using a combination of "parameter type + cross-section number". For example, mass flow rate is represented by "W" followed by the cross-section number (e.g., W030), total pressure by "Pt" followed by the cross-section number (e.g., Pt030), total temperature by "Tt" followed by the cross-section number (e.g., Tt030), static pressure by "Ps" followed by the cross-section number (e.g., Ps050), and static temperature by "Ts" followed by the cross-section number. For component characteristic parameters, a "characteristic identifier + component abbreviation" method is used. For example, the pressure ratio of a compression component is represented by "PR" followed by the component abbreviation (e.g., PRFan, PRHPC), the expansion ratio of a turbine component is also represented by "PR" followed by the component abbreviation, and power is represented by "Power" followed by the component abbreviation (e.g., PowerFan). Additionally, a distinguishing marker between total and static parameters is added to the parameter identifier (e.g., "t" for total parameters, "s" for static parameters). For design point parameters, a design parameter marker (e.g., "_ds") is added after the characteristic identifier to differentiate them from operating point parameters.

[0039] Through the above steps, a unified set of parameters required for the model is defined, realizing a structured and unified expression of the model parameters.

[0040] V. Equilibrium Equations and Iterative Variable Generation Components.

[0041] Based on the preset engine modeling system and the established component connection relationships, this component automatically generates equilibrium equations and iterative variables for steady-state solutions, and constructs a system of nonlinear equations.

[0042] 5.1 Construction of equilibrium equations.

[0043] Based on the airflow path relationships in the flow path topology and the coaxial connection relationships in the rotor mechanical connection structure, the following three types of equilibrium equations are constructed: (1) Mass flow continuity equation: Constructed only between two adjacent components whose outlet flow can be calculated independently. Components whose outlet flow can be calculated independently include compression components (such as fans and compressors), turbine components, and nozzles. If there is additional mass injection (such as fuel injection) or extraction (such as venting), explicit corrections are made in the equation. For example, in a dual-rotor mixed exhaust turbofan engine, the mass flow continuity equation is constructed as follows: Fan outlet flow W025 = High-pressure compressor outlet flow W030 + Bypass duct inlet flow W125; High-pressure compressor outlet flow + Fuel flow = High-pressure turbine outlet flow W045; High-pressure turbine outlet flow = Low-pressure turbine outlet flow W050; Low-pressure turbine outlet flow + Bypass duct outlet flow = Nozzle outlet flow W090.

[0044] (2) Power balance equation: Based on the coaxial connection relationship in the rotor mechanical connection structure, the power balance equation is constructed between the compression component and the turbine component on the same shaft. For example, on the low-pressure rotor, the power consumed by the fan is PowerFan = the power generated by the low-pressure turbine is PowerLPT; on the high-pressure rotor, the power consumed by the high-pressure compressor is PowerHPC = the power generated by the high-pressure turbine is PowerHPT. If there is power extraction (such as drive accessories), the power extraction term is explicitly deducted from the equation.

[0045] (3) Static pressure balance equation: According to the flow path topology, when the existence of the outer bypass and core channel mixing structure realized by the mixing chamber is identified, the static pressure balance equation is constructed at the inlet of the mixing chamber, that is, the static pressure Ps160 at the outlet of the outer bypass is equal to the static pressure Ps060 at the outlet of the core channel, so as to realize the static pressure matching of the two airflows.

[0046] 5.2 Selection of Iteration Variables.

[0047] According to the preset method for selecting iterative variables, the set of iterative variables is determined from the unified parameter set, including: low-pressure rotor relative physical speed PNL, high-pressure rotor relative physical speed PNH, fan pressure ratio PRFan, high-pressure compressor pressure ratio PRHPC, high-pressure turbine expansion ratio PRHPT, low-pressure turbine expansion ratio PRLPT, and bypass ratio BPR (the ratio of the mass flow rate of the outer bypass duct to the inner bypass duct).

[0048] 5.3 Extrapolation of component characteristic diagrams.

[0049] While generating the nonlinear equations, the system also generates code snippets for extrapolation during the solution process. Specifically, when any iterative variable (such as pressure ratio or rotational speed) in the set of iterative variables exceeds the data boundary of the preset component general characteristic diagram, the system automatically constructs an extrapolation model based on similarity criteria. It calculates the extrapolated value based on the extension trend of the isochronous speed lines and physical limits (blockage boundary, stall boundary), and uses this extrapolated value as the alternative output value when the component cannot be interpolated from the characteristic diagram at the current operating point. Simultaneously, a code snippet for a warning message output interface is generated to output a warning message containing the component name, the type of error, and the extrapolation method during simulation. The extrapolation processing logic and the warning interface code snippet are passed to the code generation component along with the nonlinear equations.

[0050] 5.4 Formation of nonlinear equation systems.

[0051] By combining the above equilibrium equations and the set of iteration variables, a system of nonlinear equations with the iteration variables as unknowns is constructed.

[0052] VI. Code generation component.

[0053] The code generation component uses a confirmed flow path topology, a unified parameter set, a system of nonlinear equations, a call order, and extrapolation code snippets to generate a simulation solver based on a preset program template. The generated simulation solver is written in C language and includes: (1) Flow path calculation function: According to the calling order, a flow path calculation function is generated that calls the calculation functions of each component sequentially along the airflow direction. It includes the transfer statements between the input parameters and output parameters of each component in a unified parameter set and the local variables of the flow path calculation function. The flow path calculation function is designed to accept the current value of the iterative variable set as input and output the parameters of each section (such as the flow rate, total pressure, total temperature, static pressure, power, etc. of each section).

[0054] (2) Residual calculation function: Generate residual calculation function based on the equilibrium equations in the nonlinear equation system. Internally call the flow path calculation function to obtain the parameters of each section and calculate the residual value of each equilibrium equation under the current iteration variable value (e.g.: high pressure turbine and low pressure turbine flow residual = high pressure turbine outlet flow / low pressure turbine outlet flow - 1; low pressure rotor power residual = fan power consumption / low pressure turbine power generation).

[0055] (3) Iterative Solver Interface: An iterative solver interface is generated based on the iterative variables in the nonlinear equation system. During the solution process, the residual calculation function is repeatedly called and used to call the Newton-Raphson iteration method or the fixed-point iteration method to solve the nonlinear equation system. Optionally, the convergence weights or solution order of each equilibrium equation can be adjusted through priority scheduling instructions during the solution process.

[0056] (4) Assembly and Output: Assemble the flow path calculation function, residual calculation function, and iterative solver interface according to the preset program template to generate a complete simulation solution program. The program template includes the main simulation loop framework, parameter initialization code, and result output code.

[0057] The following uses a dual-rotor hybrid exhaust turbofan engine as an example to illustrate the implementation process of the present invention.

[0058] The user inputs a natural language description: "Modeling a dual-rotor hybrid exhaust turbofan engine." The system then performs the following steps: 1. Semantic parsing and topology generation.

[0059] The semantic parsing and flow path topology generation component identifies the following components: intake, fan, high-pressure compressor, combustion chamber, high-pressure turbine, low-pressure turbine, bypass duct, mixing chamber, and nozzle. Based on the airflow path, a flow path topology is generated: intake → fan → {bypass duct; high-pressure compressor → combustion chamber → high-pressure turbine → low-pressure turbine} → mixing chamber → nozzle. Based on the mechanical connection relationships of the rotating components, a rotor mechanical connection structure is generated: the fan and low-pressure turbine are coaxial, and the high-pressure compressor and high-pressure turbine are coaxial.

[0060] Confidence check: The system performs a logical consistency check on the parsing results and finds that there are no isolated components, the rotor power chain is closed, and the flow path is continuous. The semantic parsing result meets the physical consistency requirements of the aero-engine structure and can proceed directly to the next step.

[0061] The confidence check specifically includes the following steps: Based on the principles of aero-engine aero-thermodynamics, check whether there are isolated components in the flow path topology. An isolated component is a component that has neither upstream airflow in nor downstream airflow out. For example, a compressor or turbine that has neither an inlet connection nor an exhaust connection. If an isolated component is found, this check is deemed unqualified. Based on the rotor power chain closure principle in the physical mechanism of aero-engines, the power chain in the rotor mechanical connection structure is checked for closure. A closed power chain means that each shaft has at least one compression component (such as a fan or compressor) and one turbine component, and the power transmission path is complete (i.e., the power consumed by the compression component and the power generated by the turbine are balanced on the same shaft). If a shaft has only a compression component without a turbine component, or only a turbine component without a compression component, then this check is deemed unqualified. Based on the principle of flow path continuity in the aerodynamics and thermodynamics of aero-engines, the flow path topology is checked for continuity. Flow path continuity means that on each airflow branch from the inlet to the nozzle, the inlet and outlet sections of adjacent components are sequentially connected without any breaks. For example, if there is no connection between the outlet section of a compressor and the inlet section of the combustion chamber, or if the connection between the bypass duct and the mixing chamber is missing, this check is deemed unqualified. The system generates a corresponding confidence score based on the above logical consistency check results, characterizing the degree to which the current semantic parsing result satisfies the physical constraints of the aero-engine. When the confidence score is lower than a preset threshold, the system issues an anomaly warning and provides a visual correction process. In some implementations, the rationality of the engine structure can be evaluated or supplemented based on a large language model.

[0062] 2. Section numbering assignment.

[0063] The section definition component assigns section numbers according to the SAE AS755G system, as shown in Table 1. A list of section numbers is generated, for example: Inlet 010, Outlet 020; Fan Outlet 025; High-Pressure Compressor Inlet 025, Outlet 030; Combustion Chamber Outlet 040; High-Pressure Turbine Outlet 045; Low-Pressure Turbine Outlet 050; Bypass Duct Inlet 125, Outlet 160; Mixing Chamber Outlet 064; Nozzle Inlet 070, Outlet 090; Nozzle Throat 080.

[0064] Table 1. Definition of key sections for dual-rotor hybrid exhaust turbofan engines

[0065] 3. Component function retrieval.

[0066] The component function management component retrieves calculation functions for each component from the function library. For example, the input parameters for the high-pressure compressor function include guide vane angle, relative physical speed, pressure ratio, inlet total pressure, inlet total temperature, and design point parameters; the output parameters include outlet flow rate, outlet total pressure, outlet total temperature, power consumption, efficiency, and surge margin; the characteristic structure contains characteristic plot data.

[0067] like Figure 4 As shown, the calling order is in the direction of airflow: intake duct → fan → {outer bypass duct; high-pressure compressor → combustion chamber → high-pressure turbine → low-pressure turbine} → mixing chamber → nozzle.

[0068] 4. Parameter set generation.

[0069] The parameter construction and naming component generates a unified parameter set, such as: W025, Pt025, Tt025, PowerFan, PRFan, W030, Pt030, Tt030, PowerHPC, W125, W050, Ps050, Ps160, etc., and distinguishes between total parameters (Pt, Tt) and static parameters (Ps), and design point parameters (adding _ds as a suffix).

[0070] 5. Equilibrium equations and iterative variables.

[0071] The equilibrium equation and iterative variable generation component generates the following equilibrium equations: Flow balance equation: W025 = W030 + W125; W030 + WFB = W045; W045 = W050; W050 + W160 = W090.

[0072] Power balance equation: PowerFan = PowerLPT; PowerHPC = PowerHPT.

[0073] Static pressure balance equation: Ps160 = Ps060; Iteration variables: PNL, PNH, PRFan, PRHPC, PRHPT, PRLPT, BPR.

[0074] Priority rule: If a conflict occurs in the solution where the same node simultaneously satisfies both flow continuity and static pressure balance (e.g., at the inlet of the mixing chamber), the system prioritizes satisfying the flow continuity equation and then adjusts the weight of static pressure balance.

[0075] Extrapolation processing: When the pressure ratio of a component exceeds the range of the characteristic diagram during the solution process, the system automatically extrapolates the efficiency, flow rate, etc. corresponding to the pressure ratio based on the trend of the constant speed line and issues a warning.

[0076] 6. Code generation.

[0077] The code generation component generates a complete C language simulation program, including: The flow path calculation function calls the calculation functions of each component in sequence, passing parameters and characteristic structures.

[0078] The residuals of each equilibrium equation are calculated using the residual calculation function.

[0079] The iterative solver's main loop calls the Newton-Raphson method to solve the nonlinear equation system and applies extrapolation during the solution process.

[0080] The generated code can be compiled and run directly for engine steady-state simulation.

[0081] In this embodiment, a computer device is provided, such as... Figure 5 As shown, it includes a memory 501, a processor 502, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the above-mentioned method for automatically generating semantically driven modeling and simulation code for aero-engine components.

[0082] Specifically, the computer device can be a computer terminal, a server, or a similar computing device.

[0083] In this embodiment, a computer-readable storage medium is provided, which stores a computer program that executes the above-described semantically driven method for automatically generating modeling and simulation code for aero-engine components.

[0084] Specifically, computer-readable storage media include both permanent and non-permanent, removable and non-removable media, which can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer-readable storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable storage media do not include transient media, such as modulated data signals and carrier waves.

[0085] Based on the same inventive concept, this invention also provides a semantically driven apparatus for automatically generating modeling and simulation code for aero-engine components, as described in the following embodiments. Since the principle behind the semantically driven apparatus for automatically generating modeling and simulation code for aero-engine components is similar to that of the method for automatically generating modeling and simulation code for aero-engine components, the implementation of the apparatus can refer to the implementation of the method for automatically generating modeling and simulation code for aero-engine components, and will not be repeated. As used below, the terms "unit" or "module" can refer to a combination of software and / or hardware that performs a predetermined function. Although the apparatus described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0086] Figure 6 This is a structural block diagram of a semantically driven device for automatically generating modeling and simulation code for aero-engine components according to an embodiment of the present invention, such as... Figure 6 As shown, it includes: a topology generation module 601, a cross-section function determination module 602, an equation construction module 603, and a code output module 604. The structure is described below.

[0087] The topology generation module 601 is used to identify the engine type and corresponding natural language description input by the user through semantic parsing, obtain the engine component type, airflow path relationship and mechanical connection relationship of rotating components, generate a structured flow path topology based on the engine component type and the airflow path relationship, and generate a rotor mechanical connection structure based on the mechanical connection relationship of rotating components. The section function determination module 602 is used to automatically assign section numbers to each aero-engine component based on the flow path topology and a preset section numbering system, generate a section number list for each aero-engine component, retrieve matching calculation functions from a preset component general calculation function library according to the component type in the flow path topology, and determine the input and output parameters and calling order of the calculation functions. The equation construction module 603 is used to generate a unified set of parameters required for the model based on the cross-section number list and the input and output parameters of each component, according to a preset parameter naming system; and to generate equilibrium equations and iterative variables for steady-state solution based on the flow path topology, the rotor mechanical connection structure, the cross-section number list and the unified parameter set, according to a preset engine modeling system, and to construct a nonlinear equation system. The code output module 604 is used to automatically generate a simulation solution program based on the flow path topology, the unified parameter set, the nonlinear equation system and the calling order, using a preset program template.

[0088] In one embodiment, the section function determination module includes: The cross-section numbering allocation unit is used to assign an inlet cross-section number, an outlet cross-section number, and a throat cross-section number to each aero-engine component according to the arrangement order and airflow path direction of each aero-engine component in the flow path topology, and generate a cross-section number list. The function retrieval unit is used to retrieve the calculation function corresponding to each aero-engine component from the component general calculation function library according to the engine component type in the flow path topology. The component general calculation function library includes calculation functions for the air intake, compression component, combustion chamber, turbine component, bypass duct, mixing chamber and nozzle. Each calculation function includes a function prototype, input and output parameters and characteristic structure information of the corresponding aero-engine component. The call order determination unit is used to determine the call order of the calculation functions for each of the aero-engine components based on the airflow direction of the flow path topology.

[0089] In one embodiment, the equation building module includes: The equilibrium equation construction unit is used to construct equilibrium equations based on the airflow path relationship in the flow path topology and the coaxial connection relationship in the rotor mechanical connection structure, wherein the equilibrium equations include the mass flow rate continuity equation, the power balance equation, and the static pressure balance equation. The iterative variable determination unit determines iterative variables from the unified parameter set according to a preset iterative variable selection method. The iterative variables include the relative physical rotational speed of the shaft, the pressure ratio of the compression component, the expansion ratio of the turbine component, and the ratio of the mass flow rate of the outer bypass duct to the inner bypass duct. The equation system construction unit is used to combine the equilibrium equation and the iteration variable to construct a nonlinear equation system with the iteration variable as the unknown.

[0090] In one embodiment, the balance equation construction unit is further configured to construct a mass flow continuity equation between two adjacent aero-engine components that can independently calculate their outlet flow rates, based on the airflow path relationship in the flow path topology; if there is additional mass injection or extraction, the mass flow continuity equation is explicitly modified; wherein the aero-engine components that can independently calculate their outlet flow rates include the compressor component, the turbine component, and the nozzle; construct a power balance equation between the compressor component and the turbine component on the same rotating shaft based on the coaxial connection relationship in the rotor mechanical connection structure; if there is extraction of power, the extraction term is explicitly subtracted from the power balance equation; and construct a static pressure balance equation at the inlet of the mixing chamber when a mixing structure between the outer bypass duct and the core duct, implemented by the mixing chamber, is identified based on the airflow path relationship in the flow path topology.

[0091] In one embodiment, the code output module includes: The flow path calculation function generation unit is used to generate flow path calculation functions that sequentially call each of the aero-engine components along the airflow direction according to the calling order. The flow path calculation function includes the input and output parameters passing statements between the unified parameter set and the local variables of the flow path calculation function. The flow path calculation function is used to take the current value of the iteration variable as input and output each section parameter. The residual calculation function construction unit is used to construct a residual calculation function based on the equilibrium equations in the nonlinear equation system, wherein the residual calculation function is used to calculate the residual value of each equilibrium equation under the current iteration variable value; An iterative solver interface generation unit is used to generate an iterative solver interface based on the iterative variables in the nonlinear equation system, wherein the iterative solver interface is used to call the Newton-Raphson iteration method or the fixed-point iteration method to solve the nonlinear equation system. The program assembly unit is used to assemble the flow path calculation function, the residual calculation function, and the iterative solver interface according to a preset program template to generate a simulation solution program. The program template includes a main simulation loop framework, parameter initialization code, and result output code.

[0092] In one embodiment, the above-described apparatus further includes a consistency check module.

[0093] In one embodiment, the consistency check module includes: The logical consistency check unit is used to perform logical consistency checks on the results of semantic parsing based on the physical mechanism of aero-engines and calculate the confidence level evaluation. The logical consistency check includes detecting whether there are isolated components, whether the rotor power chain is closed, and whether the flow path is continuous. The visualization generation and display unit is used to generate a visualized flow path topology diagram if the confidence level evaluation is lower than a preset threshold, and to display the aero-engine components or connection relationships that fail the logical consistency check in the graphical interface of the visualized flow path topology diagram. A graphical correction interface unit is used by users to modify the flow path topology and / or the rotor mechanical connection structure through a graphical correction interface, and to update the flow path topology and the rotor mechanical connection structure. The correction record storage and feedback unit is used to save the user's correction records to the system library and to use them for iterative optimization of the semantic parsing.

[0094] The logical consistency check unit is also used to check whether there are isolated components in the flow path topology according to the aerodynamic and thermodynamic principles of aero-engines. If isolated components are found, the logical consistency is deemed unqualified. An isolated component is defined as a component with no upstream airflow inflow and no downstream airflow outflow. Based on the rotor power chain closure principle in the physical mechanism of aero-engines, the unit checks whether the power chain in the rotor mechanical connection structure is closed. If an unclosed power chain is found, the logical consistency is deemed unqualified. A closed power chain means that there is at least one compression component and one turbine component on each shaft, and the power transmission path is complete. Based on the flow path continuity principle in the aerodynamic and thermodynamic principles of aero-engines, the unit checks whether the airflow path in the flow path topology is continuous. If a discontinuous flow path is found, the logical consistency is deemed unqualified. Flow path continuity means that the inlet and outlet sections of adjacent components on each airflow branch from the inlet to the nozzle are sequentially connected without any breaks. Based on the results of each logical consistency check, a confidence evaluation corresponding to the semantic parsing result is generated.

[0095] The embodiments of the present invention achieve the following technical effects: By employing semantic parsing technology (preferably a model based on a large language), the system automatically recognizes the user-input natural language description of the engine mechanism, generating a structured flow path topology and rotor mechanical connection structure without requiring manual analysis of component types, airflow paths, and mechanical connections. Based on this, it automatically completes the allocation of section numbers, retrieval and determination of component calculation functions' calling order, generation of a unified parameter set, construction of equilibrium equations and iterative variables, and generation of simulation solution code. The entire modeling process eliminates the need for manual coding or drag-and-drop module manipulation, completely transforming the inefficient and error-prone nature of traditional manual programming or graphical modeling methods. This reduces the modeling cycle from weeks to minutes, making it particularly suitable for the rapid evaluation of numerous candidate solutions during the overall conceptual design phase of aero-engines.

[0096] By employing a pre-defined cross-section numbering system, parameter naming system, and a standardized universal calculation function library for components, it is ensured that models built by different configurations and engineers are completely consistent in cross-section numbering, variable naming, and parameter organization. This standardized system enables models to be shared, reused, and maintained seamlessly among teams, significantly reducing learning costs, collaboration costs, and the difficulty of subsequent upgrades, and overcoming the shortcomings of traditional modeling, such as chaotic model structures and difficulty in inheritance.

[0097] Based on the airflow path relationships in the flow path topology and the coaxial connection relationships in the rotor mechanical connection structure, the mass flow continuity equation, power balance equation, and static pressure balance equation are automatically constructed. Iterative variables such as the relative physical speed of the rotor shaft, the pressure ratio of the compression components, the expansion ratio of the turbine components, and the bypass ratio are automatically selected to form a complete set of steady-state nonlinear equations. This process transforms the traditional equation construction work, which relies on the experience of domain experts, into a rule-driven automated process. It is not only applicable to dual-rotor turbofan engines but can also be extended to complex configurations such as three-rotor, variable-cycle, and adaptive-cycle engines, significantly reducing the modeling threshold.

[0098] When generating the simulation program, extrapolation logic is automatically embedded: when the iterative variable of a component exceeds the data boundary of the preset characteristic map during the solution process, the system constructs an extrapolation model based on similarity criteria and physical limits, calculates the extrapolated value as the output value of that component at the current operating point, and generates a warning message output interface to remind the user. This technical solution enables the model to continue operating outside the coverage of the characteristic map, solving the problem that traditional component-level models cannot be calculated due to insufficient characteristic map data under transitional or extreme operating conditions. This significantly broadens the applicability of the model, especially suitable for scenarios with incomplete characteristic data in the exploration of new engine schemes.

[0099] The flow path topology and rotor mechanical connection structure generated by semantic parsing undergo logical consistency checks based on engine design rules (including isolated component detection, rotor power chain closure detection, and flow path continuity detection), outputting a confidence score. When the confidence score indicates an anomaly, a visual flow path topology diagram is automatically generated, and a graphical correction interface is provided. Users can quickly correct parsing errors by dragging, adding, deleting, or modifying parameters. Correction records are stored in a rule base for iterative optimization in subsequent parsing. This mechanism combines the automation advantages of artificial intelligence with the experience of human experts, ensuring efficiency in large-scale modeling while providing necessary error correction and knowledge accumulation methods, overcoming the lack of credibility in pure black-box AI modeling in engineering applications.

[0100] Based on a preset program template, the system automatically generates a complete simulation solution program, including a flow path calculation function that sequentially calls the calculation functions of each component along the airflow direction, a residual calculation function that calculates the residuals of the equilibrium equation, and an iterative solver interface that calls the Newton-Raphson or fixed-point iterative method. It also automatically assembles the main simulation loop framework, parameter initialization code, and result output code. The generated code has a clear structure and high readability, and can be directly used for simulation calculations without manual writing. This significantly reduces code development costs and ensures consistency in code style and logical structure across different engineers, facilitating subsequent integration, debugging, and maintenance.

[0101] Through semantic parsing and rule-driven approaches, it can support complex engine configurations such as dual-rotor, triple-rotor, variable-cycle, and adaptive-cycle engines, as well as complex structures such as parallel flow paths, multiple shafts, and pump-injection systems. During the overall conceptual design phase, designers only need to modify the natural language description to quickly generate models and simulation code for different configurations, enabling parallel evaluation and rapid iteration of multiple schemes. This accelerates engine product development and reduces development risks and testing costs.

[0102] Obviously, those skilled in the art should understand that the modules or steps of the above-described embodiments of the present invention can be implemented using general-purpose computing devices. They can be centralized on a single computing device or distributed across a network of multiple computing devices. Optionally, they can be implemented using computer-executable program code, thereby storing them in a storage device for execution by a computing device. In some cases, the steps shown or described can be performed in a different order than those presented here, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. Thus, the embodiments of the present invention are not limited to any particular hardware and software combination.

[0103] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, various modifications and variations can be made to the embodiments of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A semantically driven method for automatic generation of modeling and simulation code for aero-engine components, characterized in that, include: By semantic parsing, the engine type and corresponding natural language description input by the user are identified, and the engine component type, airflow path relationship and mechanical connection relationship of rotating components are obtained. A structured flow path topology is generated based on the engine component type and the airflow path relationship, and a rotor mechanical connection structure is generated based on the mechanical connection relationship of rotating components. Based on the flow path topology, a cross section number is automatically assigned to each aero-engine component according to a preset cross section numbering system, generating a cross section number list for each aero-engine component. Based on the component type in the flow path topology, a matching calculation function is retrieved from a preset component general calculation function library, and the input and output parameters and calling order of the calculation function are determined. Based on the cross-section number list and input / output parameters of each component, a unified parameter set required for the model is generated according to a preset parameter naming system. Based on the flow path topology, the rotor mechanical connection structure, the cross-section number list and the unified parameter set, a balance equation and iterative variables for steady-state solution are generated according to a preset engine modeling system, and a nonlinear equation system is constructed. Based on the flow path topology, the unified parameter set, the nonlinear equation system, and the calling order, a simulation solution program is automatically generated using a preset program template, including: According to the calling order, flow path calculation functions are generated that sequentially call each of the aero-engine components along the airflow direction. Each flow path calculation function includes statements for passing input and output parameters between the unified parameter set and the local variables of the flow path calculation function. The flow path calculation function is used to take the current value of the iteration variable as input and output the parameters of each section. Based on the equilibrium equations in the nonlinear equation system, a residual calculation function is constructed. This residual calculation function is used to calculate the residual value of each equilibrium equation under the current value of the iteration variable. Based on the iteration variables in the nonlinear equation system, an iterative solver interface is generated. This iterative solver interface is used to call the Newton-Raphson iteration method or the fixed-point iteration method to solve the nonlinear equation system. The flow path calculation function, the residual calculation function, and the iterative solver interface are assembled according to a preset program template to generate a simulation solution program. This program template includes a main simulation loop framework, parameter initialization code, and result output code.

2. The method for automatic generation of semantic-driven modeling and simulation code for aero-engine components as described in claim 1, characterized in that, Based on the flow path topology, a cross-section number is automatically assigned to each aero-engine component according to a preset cross-section numbering system, generating a cross-section number list for each aero-engine component. According to the component type in the flow path topology, a matching calculation function is retrieved from a preset general-purpose component calculation function library, and the input / output parameters and calling order of the calculation function are determined, including: Based on the arrangement order and airflow path direction of each aero-engine component in the flow path topology, an inlet section number, an outlet section number, and a throat section number are assigned to each aero-engine component according to the preset section numbering system, and a section number list is generated. Based on the engine component type in the flow path topology, a calculation function corresponding to each aero-engine component is retrieved from the component general calculation function library. The component general calculation function library includes calculation functions for the air intake, compression component, combustion chamber, turbine component, bypass duct, mixing chamber, and nozzle. Each calculation function includes a function prototype, input and output parameters, and characteristic structure information of the corresponding aero-engine component. Based on the airflow direction of the flow path topology, the order in which the calculation functions of each aero-engine component are called is determined.

3. The method for automatic generation of semantic-driven modeling and simulation code for aero-engine components as described in claim 1, characterized in that, Based on the flow path topology, the rotor mechanical connection structure, the cross-section number list, and the unified parameter set, equilibrium equations and iterative variables for steady-state solution are generated according to the preset engine modeling system, and a nonlinear equation system is constructed, including: Based on the airflow path relationship in the flow path topology and the coaxial connection relationship in the rotor mechanical connection structure, a balance equation is constructed, wherein the balance equation includes a mass flow rate continuity equation, a power balance equation, and a static pressure balance equation. According to the preset method for selecting iterative variables, iterative variables are determined from the unified parameter set, wherein the iterative variables include the relative physical rotational speed of the shaft, the pressure ratio of the compression component, the expansion ratio of the turbine component, and the ratio of the mass flow rate of the outer bypass duct to the inner bypass duct. The equilibrium equation and the iteration variable are combined to construct a system of nonlinear equations with the iteration variable as the unknown.

4. The method for automatic generation of semantic-driven modeling and simulation code for aero-engine components as described in claim 3, characterized in that, Based on the airflow path relationships in the described flow path topology and the coaxial connection relationships in the described rotor mechanical connection structure, an equilibrium equation is constructed, including: Based on the airflow path relationship in the flow path topology, a mass flow continuity equation is constructed between two adjacent aero-engine components that can independently calculate their outlet flow. If there is additional mass injection or extraction, the mass flow continuity equation is explicitly modified. The aero-engine components that can independently calculate their outlet flow include the compressor, turbine, and nozzle. Based on the coaxial connection relationship in the rotor mechanical connection structure, a power balance equation is constructed between the compression component and the turbine component on the same shaft. If there is power extraction, the power extraction term is explicitly deducted from the power balance equation. Based on the airflow path relationship in the flow path topology, when a mixing structure between the outer bypass duct and the core duct, implemented by the mixing chamber, is identified, a static pressure balance equation is constructed at the inlet of the mixing chamber.

5. The method for automatic generation of semantic-driven modeling and simulation code for aero-engine components as described in any one of claims 1 to 4, characterized in that, Also includes: Based on the physical mechanism of aero-engines, the results of semantic parsing are subjected to logical consistency checks to generate confidence evaluations. The logical consistency checks include detecting whether there are isolated components, whether the rotor power chain is closed, and whether the flow path is continuous. If the confidence level evaluation is lower than a preset threshold, a visual flow path topology diagram is generated, and the aircraft engine components or connection relationships that fail the logical consistency check are displayed in the graphical interface of the visual flow path topology diagram. Users can modify the flow path topology and / or the rotor mechanical connection structure through a graphical modification interface, and update the flow path topology and the rotor mechanical connection structure. The user's correction records are saved to the system library and used for iterative optimization of the semantic parsing.

6. The method for automatic generation of semantic-driven modeling and simulation code for aero-engine components as described in claim 5, characterized in that, Based on the physical mechanisms of aero-engines, the results of semantic parsing are subjected to logical consistency checks to generate confidence assessments, including: Based on the aerodynamic and thermodynamic principles of aero engines, check whether there are isolated components in the flow path topology. If there are isolated components, the logic consistency is deemed unqualified. The isolated component is a component with no upstream airflow inflow and no downstream airflow outflow. According to the rotor power chain closure principle in the physical mechanism of aero-engines, check whether the power chain in the rotor mechanical connection structure is closed. If there is an unclosed power chain, the logic consistency is deemed unqualified. The power chain closure means that there is at least one compression component and one turbine component on each shaft and the power transmission path is complete. Based on the principle of flow path continuity in the aerodynamic thermodynamics of aero engines, check whether the airflow path in the flow path topology is continuous. If there is a discontinuous flow path, the logic consistency is deemed unqualified. The flow path continuity means that the inlet and outlet sections of adjacent components on each airflow branch from the intake to the nozzle are connected in sequence without any breaks. The confidence level evaluation corresponding to the semantic parsing result is generated based on the results of each logical consistency check.

7. A semantically driven device for automatic generation of modeling and simulation code for aero-engine components, characterized in that, include: The topology generation module is used to identify the engine type and corresponding natural language description input by the user through semantic parsing, obtain the engine component type, airflow path relationship and mechanical connection relationship of rotating components, generate a structured flow path topology based on the engine component type and the airflow path relationship, and generate a rotor mechanical connection structure based on the mechanical connection relationship of rotating components. The section function determination module is used to automatically assign section numbers to each aero-engine component based on the flow path topology and a preset section numbering system, generate a section number list for each aero-engine component, retrieve matching calculation functions from a preset component general calculation function library according to the component type in the flow path topology, and determine the input and output parameters and calling order of the calculation functions; The equation construction module is used to generate a unified set of parameters required for the model based on the cross-section number list and the input and output parameters of each component, according to a preset parameter naming system. Based on the flow path topology, the rotor mechanical connection structure, the cross-section number list and the unified parameter set, it generates equilibrium equations and iterative variables for steady-state solution according to a preset engine modeling system, and constructs a nonlinear equation system. The code output module is used to automatically generate a simulation solution program based on the flow path topology, the unified parameter set, the nonlinear equation system and the calling order, using a preset program template. The code output module includes: The flow path calculation function generation unit is used to generate flow path calculation functions that sequentially call each of the aero-engine components along the airflow direction according to the calling order. The flow path calculation function includes the input and output parameters passing statements between the unified parameter set and the local variables of the flow path calculation function. The flow path calculation function is used to take the current value of the iteration variable as input and output each section parameter. The residual calculation function construction unit is used to construct a residual calculation function based on the equilibrium equations in the nonlinear equation system, wherein the residual calculation function is used to calculate the residual value of each equilibrium equation under the current iteration variable value; An iterative solver interface generation unit is used to generate an iterative solver interface based on the iterative variables in the nonlinear equation system, wherein the iterative solver interface is used to call the Newton-Raphson iteration method or the fixed-point iteration method to solve the nonlinear equation system. The program assembly unit is used to assemble the flow path calculation function, the residual calculation function, and the iterative solver interface according to a preset program template to generate a simulation solution program. The program template includes a main simulation loop framework, parameter initialization code, and result output code.

8. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the semantic-driven method for automatically generating modeling and simulation code for aero-engine components as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that performs the semantically driven method for automatically generating modeling and simulation code for aero-engine components according to any one of claims 1 to 6.

Images