Semantized geometric modeling method for solid fuel power device

Abstract

The invention provides a solid fuel power plant semantic geometric modeling method, which belongs to the technical field of geometric modeling, and comprises the following steps: constructing a semantic feature library containing feature kernels based on an open cascade technology and physical boundary semantics; establishing a structural solid geometric feature hierarchical framework based on feature traceability, injecting the structural solid geometric feature hierarchical framework into a parameterized network, and performing iteration by using a top-layer performance index to generate a parameterized model matrix; according to the parameterized model matrix, performing semantic entity segmentation on the physical domain by using a channel labeling technology based on a color system to obtain an enhanced three-dimensional entity model; compressing the complexity of the enhanced three-dimensional solid model by using a model symmetry solving algorithm based on a feature kernel to obtain a lightweight solid model, and completing semantic geometric modeling; according to the method, the problem that the digital and structured transmission of the design intention in the existing computer aided design software is difficult is solved.

Description

Technical Field

[0001] The present invention belongs to the technical field of geometric modeling, and in particular relates to a semantic geometric modeling method for a solid fuel power device. Background Art

[0002] Modern solid rocket motor research consistently strives for rapid and efficient solid rocket motor design. Digital modeling is an essential component of solid rocket motor design. However, in the integrated design process from digital modeling to digital simulation, 3D design software is typically purely "drawing" software. The digital 3D model retains only topological and dimensional information, leaving the design intent unstructured and undigitized. This disconnect not only requires designers to manually maintain the mapping between design intent and model parameters such as dimensions, materials, and boundaries, but also renders theoretically feasible digital operations extremely difficult in practice, such as automated structural meshing, automatic generation of full-time multi-physics simulation models, automated 3D model generation and bidirectional synchronous modification, enumeration of variable topology solutions without pre-set parameters, and semantic AI-assisted design. The synchronization issues between geometric models and simulation parameters primarily stem from the disconnect between design intent and digital implementation in current CAD software, which neglects the expression of design intent. Difficulties in meshing and pre-processing arise from the failure of existing software to structure the semantics of the model, making it impossible to automatically determine mesh partitioning and physical property assignments, forcing the software to rely on pre-set templates or manual interaction.

[0003] In existing design models, it's difficult to fully and systematically store and transmit design intent digitally. This problem not only complicates simulations and results in functional deficiencies, but also deprives modern design methods (such as artificial intelligence design) that rely on deeper digital information of a reliable foundation, hindering the further development of digital design methods for solid propulsion systems. Summary of the Invention

[0004] In response to the above-mentioned deficiencies in the prior art, the present invention provides a semantic geometric modeling method for a solid fuel power plant, which solves the problem of difficulty in digitalizing and structuring the transmission of design intent in existing computer-aided design software.

[0005] In order to achieve the above objectives, the technical solution adopted by the present invention is: a semantic geometric modeling method for a solid fuel power plant, comprising the following steps:

[0006] S1. Based on the open cascade technology, the solid fuel power plant is decoupled to obtain geometric features, and based on the physical boundary semantics, a semantic feature library containing feature cores is obtained;

[0007] S2. Based on the semantic feature library, using hierarchical Boolean operation rules and constructive solid geometry methods, a constructive solid geometric feature hierarchical framework is established, and the design parameters corresponding to each entity node are recorded to obtain geometric parameters;

[0008] S3. Injecting a parameterized network into the hierarchical framework of the constructed solid geometric features according to the geometric parameters, and iterating the lower-level geometric parameters of the hierarchical framework of the constructed solid geometric features using the top-level performance indicators to generate a parameterized model matrix;

[0009] S4. Import the parametric model matrix into the solid modeling kernel, use the color system-based channel annotation technology to perform semantic entity segmentation on the physical domain, and obtain an enhanced 3D solid model containing geometric topology and physical attribute labels;

[0010] S5. Use the feature kernel-based model symmetry solving algorithm to solve the symmetry of the enhanced three-dimensional solid model. According to the solution results, the complexity of the enhanced three-dimensional solid model is compressed to obtain a lightweight solid model and complete the semantic geometric modeling.

[0011] The beneficial effects of the present invention are as follows: the present invention constructs a semantic feature library, and based on feature tracing, performs dynamic modeling of a constructed solid geometric tree, and then uses a parameter-driven hierarchical modeling mechanism to generate semantically enhanced three-dimensional entities, and uses symmetry features to guide the lightweighting of the model to complete the semantic geometric modeling of the solid fuel power plant, making the digitalization and structured transmission of design intent in computer-aided design software simpler, and improving the efficiency of digitalization and structured transmission of design intent in computer-aided design software, thereby enhancing the freedom and flexibility of geometric modeling of solid power plants, eliminating unnecessary manual operations in geometric modeling and pre-processing processes, and laying the foundation for research on the optimization design of general solid power plants and the optimization of massive schemes.

[0012] Furthermore, the S1 includes the following steps:

[0013] S101. Decoupling multi-dimensional geometric features of the solid fuel power plant based on an open cascade technology to obtain decoupled geometric features;

[0014] S102. Based on the decoupled geometric features and the physical boundary semantics of the functional components in the solid fuel power plant, using the open source library in the open cascade technology, generate a structured feature class library containing topological constraints and assembly relationships;

[0015] S103. According to the structured feature library, a mapping rule library between geometric elements and engineering semantics is established to obtain a semantic feature library containing feature cores.

[0016] Furthermore, the S103 is specifically as follows:

[0017] Based on the structured feature library, a mapping rule library between geometric elements and engineering semantics is established. Based on the concepts of objectification and logical inheritance, the common features of solid fuel power plant component design are extracted. The feature core based on the basic geometric elements of the engine is obtained. By integrating the feature core, a semantic feature library containing the feature core is obtained.

[0018] The characteristic cores include: wing-shaped, cylindrical, frustum-shaped, ellipsoidal, butterfly-shaped and star-shaped.

[0019] The beneficial effects of the above further scheme are as follows: the present invention decomposes the geometric elements of the solid fuel power unit to form a feature core library with a multi-dimensional semantic labeling system, thereby realizing the deep integration of the physical concept and geometric form of the engine model; and proposes the concept of feature core to materialize the basic geometric elements that make up the engine structural parts, and through topological analysis and decomposition of the top-level structural parts, simplifies the modeling process of complex entities to the construction of feature cores and their combination relationships, thereby reducing the difficulty of open source modeling.

[0020] Furthermore, the S2 includes the following steps:

[0021] S201, according to the semantic feature library, based on the solid modeling method and using the modular concept, obtain the feature core module;

[0022] S202, combining the feature core module, using hierarchical Boolean operation rules and constructive solid geometry methods, constructing a solid geometry tree structure with the tree root as the component, the leaves as the feature core, and the intermediate nodes as the hierarchical Boolean operation operators, to obtain a constructive solid geometry feature hierarchical framework;

[0023] S203. Record the design parameters corresponding to each entity node to obtain geometric parameters.

[0024] The beneficial effects of the above-mentioned further scheme are as follows: the present invention is based on the dynamic modeling of the constructed solid geometry tree based on feature traceability, takes the feature class library as input, constructs a traceable constructed solid geometry process model through hierarchical Boolean operation rules, records the design parameters corresponding to each entity node, and realizes the two-way tracing of geometric operations and design intentions; and through the mapping rules of feature core semantics to geometry, realizes the explicit expression of design intentions, and solves the problem of easy loss of design intentions in traditional parametric modeling.

[0025] Furthermore, the S3 includes the following steps:

[0026] S301. Determine the main design geometric parameters of the solid fuel power plant based on the geometric parameters, and inject a parameterized network into the constructed three-dimensional geometric feature hierarchical framework;

[0027] S302: constructing a hierarchical framework of three-dimensional geometric features based on the injected parameterized network, using the main design geometric parameters as top-level skeleton model parameters, establishing a top-level skeleton model based on the feature kernel, parameterizing the feature kernel, associating the geometric parameters with the feature kernel size and component operations, and establishing a one-to-one correspondence between the geometric parameters and the feature kernel size;

[0028] S303, iterating the lower-level geometric parameters of the constructive solid geometric feature hierarchical framework based on the top-level performance indicators of the top-level skeleton model, changing the constraints between feature cores using size-driven methods, and combining the one-to-one correspondence between geometric parameters and feature core sizes to obtain a constructive solid geometric operation set;

[0029] S304: Based on the constructive solid geometry operation set, the feature kernel is processed using the constructive solid geometry operation to generate a parameterized model matrix that meets the preset size requirements.

[0030] Furthermore, the S304 specifically includes: dividing the feature library parameters into an object group and a tool group using an open cascade technology according to the constructed solid geometry operation set, performing a universal fusion operation on the tool group and the object group, and performing shape segmentation on the feature core according to the intersection result of the universal fusion operation to obtain a segmented feature core;

[0031] Based on the displacement and feature origin, the segmented feature core is processed by constructive solid geometry operation, and fuzzy Boolean operation is used to generate a parametric model matrix that meets the preset size requirements by setting the preset tolerance.

[0032] The beneficial effects of the above-mentioned further scheme are as follows: the present invention forms a formal description language and intelligent template engine for solid rocket engines by utilizing a parameter-driven hierarchical modeling mechanism. The top layer adopts a constraint-driven parametric skeleton template, which automatically derives the size constraints of each subsystem through key feature parameters and performance parameters, thereby realizing automatic two-way conversion between geometric models and design semantics; constructs a zero-code template extension framework to realize loose coupling between the feature core library and the three-dimensional modeling engine, greatly expanding the freedom of the component template library without changing the main program code; simplifies repetitive design work, while avoiding the entanglement of the underlying design theory and design technology of complex computer-aided design, saving a lot of learning costs for R&D personnel who lack experience and knowledge in related fields, and has a certain degree of friendliness.

[0033] Furthermore, the S4 includes the following steps:

[0034] S401, importing the parametric model matrix into the solid modeling kernel, using the solid modeling method to establish the internal relationship between the feature cores in the parametric model matrix to obtain the feature core relationship;

[0035] S402. Using a color system-based channel marking technology, mark the combustion surface, symmetry surface, and insulation surface, and use the marked preset color system value as a characteristic color;

[0036] S403: Perform semantic entity segmentation on the physical domain according to the characteristic color to obtain an enhanced three-dimensional entity model including geometric topology and physical attribute labels.

[0037] The beneficial effects of the above further scheme are: the present invention uses color information to mark the physical boundaries of the model, thereby avoiding the loss and conversion of the marking information format; and when the geometric model is imported into other computer-aided design platforms, the accompanying characteristic color information will follow, so that the computer-aided design platform can quickly determine the ownership of the surface by identifying the color information of the surface; by constructing a semantically enhanced structural entity operation unit at the bottom layer, each Boolean operation node carries manufacturing semantic information and introduces a certain tolerance to achieve stable processing of situations such as approximate consistency, slight interference or inability to align the model boundaries.

[0038] Furthermore, the S5 includes the following steps:

[0039] S501. Based on the symmetry of the model feature kernel, the enhanced three-dimensional solid model is simplified into a two-dimensional geometric figure consisting of lines extending from the center to the midpoint. Based on the two-dimensional geometric figure, the lines are considered as rays starting from the center point, and the starting rays are considered as repeated rays. The rays are sorted according to their angles with the zero-degree ray to form an angle series arranged from small to large. The difference sequence of the angle series is calculated to obtain an information string with angle data as elements.

[0040] S502: Consider the information string as a ring structure connected end to end, and determine the length of the information string;

[0041] If the length of the information string is an odd number, select the smallest element in the information string and calculate the differences between the elements on the left and right of the smallest element in turn. If the differences are all 0, the information string is defined as having an axisymmetric property. If the differences are not all 0, the symmetry number of the information string is defined as 1, and the process proceeds to step S505.

[0042] If the length of the information string is an even number, two adjacent elements with equal values in the information string are selected. If no adjacent elements with equal values can be found, the information string is directly determined to be non-axisymmetric. The left and right elements of the two selected elements are symmetrically subtracted, and the subtraction results are determined. If all subtraction results are 0, the information string is defined as having an axisymmetric property. The information string defined as having an axisymmetric property is split into multiple substrings, and the judgment process of S502 is repeated. If all subtraction results are not 0, the symmetry number of the information string is defined as 1, and the process proceeds to step S505.

[0043] S503, determine whether the current information string is axisymmetric and the maximum axisymmetric number is obtained, if so, proceed to step S505, otherwise proceed to step S504;

[0044] S504. Based on the information burst length, the period length of the information burst is traversed, and the period length of a valid period is retained to obtain multiple groups of period lengths. It is determined whether the information burst length is divisible by the period length. If so, corresponding elements in the multiple groups of information burst segments with lengths equal to the period length are compared to see whether they are equal. If they are equal, the maximum rotational symmetry number is obtained. If they are not equal, the symmetry number of the information burst is defined as 1.

[0045] S505, dividing the characteristic core into a strong symmetric characteristic core and a weak symmetric characteristic core according to the symmetry number, retaining the entire entity structure of the weak symmetric characteristic core with a symmetry number of 1, and dividing the strong symmetric characteristic core with multiple symmetric units into the minimum symmetric unit;

[0046] S506. Based on the weak symmetric feature core that retains the entire solid structure and the strong symmetric core that is divided into the smallest symmetric units, feature reduction is used to compress the complexity of the enhanced three-dimensional solid model to obtain a lightweight solid model and complete the semantic geometric modeling.

[0047] The beneficial effects of the above-mentioned further scheme are as follows: the present invention simplifies the three-dimensional model into a two-dimensional geometric figure by using a model symmetry solving algorithm based on a feature kernel, and then performs a symmetry analysis on the two-dimensional geometric model, thereby realizing model lightweighting guided by symmetric features; and automatically identifies redundant structures through feature symmetry analysis to realize model lightweight generation while maintaining the integrity of key semantic information. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 Flow chart of the method of the present invention.

[0049] Figure 2 This is a structural diagram of the method of this embodiment.

[0050] Figure 3 The standard form of the characteristic core of this embodiment is drawn.

[0051] Figure 4 This is a diagram of the structural three-dimensional geometric modeling of this embodiment.

[0052] Figure 5 This is a diagram of the method for tracing the construction of solid geometric features in this embodiment.

[0053] Figure 6 This is a diagram of the parametric modeling mechanism of this embodiment.

[0054] Figure 7 This is a flowchart of the facet parent-child relationship of this embodiment.

[0055] Figure 8 This is an example of entity color restoration in this embodiment.

[0056] Figure 9 This is for symmetry sake to simplify the description of this embodiment.

[0057] Figure 10 This is the symmetry judgment criterion diagram of this embodiment.

[0058] Figure 11 This is a diagram demonstrating the algorithm of the wing-column type grain of this embodiment.

[0059] Figure 12 A sketch of the engine of this embodiment.

[0060] Figure 13 A diagram showing the composition of the feature core library designed for the solid propulsion system components of this embodiment.

[0061] Figure 14 This is a conceptual diagram of the feature core combination of this embodiment.

[0062] Figure 15 This is a sketch of the engine grain of this embodiment. Here, X1 represents the front aperture diameter, X2 represents the front aperture length, X3 represents the front step length, X4 represents the mid-section aperture, X5 represents the wing front distance, X6 represents the grain design length, X7 represents the first wing height, X8 represents the first wing angle, X9 represents the first wing fillet, X10 represents the rear step length, X11 represents the second wing angle, X12 represents the second wing fillet, X13 represents the third wing angle, X14 represents the rear step length, X15 represents the rear step length, X16 represents the second wing height, X17 represents the wing thickness, X18 represents the third wing fillet, and X19 represents the minimum rotation angle.

[0063] Figure 16 This is a sketch of the engine combustion chamber of this embodiment. Among them, X1, the short half axis of the head, X2, the height of the front head joint, X3, the thickness of the insulation layer of the front head thickened area, X4, the width of the front head joint, X5, the length of the front head joint cone neck, X6, the cone angle of the front head joint cone neck, X7, the eccentricity of the front head joint, X8, the radius of the front opening of the grain, X9, the thickness of the barrel section insulation layer, X10, the thickness of the front skirt, X11, the width of the front skirt joint, X12, the height of the front skirt joint, X13, the thickness of the combustion chamber shell, X14, the length of the front skirt, X15, the width of the connection between the barrel section and the head, X16, length of the connection joint between the barrel section and the head, X17, length of the thread of the connection joint between the barrel section and the head, X18, height of the thread of the connection joint between the barrel section and the head, X19, radius of the opening of the rear head, X20, height difference between the rear head and the insulation layer, X21, height difference between the insulation layer and the rear opening of the charge column, X22, thickness of the insulation layer in the thickened area of the rear head, X23, cone angle of the cone neck of the rear head joint, X24, height of the rear head joint, X25, eccentricity of the rear head joint, X26, width of the rear head joint, X27, length of the cone neck of the rear head joint.

[0064] Figure 17 This is a sketch of the engine nozzle of this embodiment. Among them, X1, inlet radius, X2, height of the convergent section ablation layer step, X3, height of the convergent section shell step, X4, length of the convergent section ablation layer step, X5, length of the first convergent section insulation layer step, X6, length of the second convergent section insulation layer step, X7, width of the first convergent section shell step, X8, width of the second convergent section shell step, X9, shell outer convergence angle, X10, length of the first back wall step, X11, length of the back wall step, X12, height of the first back wall step, X1 3. Minimum shell thickness, X14, convergence angle, X15, height of the second back wall step, X16, back wall thickness, X17, throat liner thickness, X18, height of the first throat liner step, X19, critical section length of the throat liner, X20, throat radius, X21, length of the second back wall step, X22, height of the second throat liner step, X23, exposed length, X24, small end thickness of the insulation layer of the diffusion section, X25, small end thickness of the ablation layer of the diffusion section, X26, expansion angle, X27, outlet radius.

[0065] Figure 18 This is an example diagram of the physical boundary marking of this embodiment.

[0066] Figure 19 1 is an example diagram of RGB values of a set of characteristic colors in this embodiment.

[0067] Figure 20 An example diagram of modeling the grain and combustion chamber casing for a certain type of engine in this embodiment.

[0068] Figure 21 An example diagram of modeling the insulation layer, nozzle, and entire engine of a certain type of engine in this embodiment. DETAILED DESCRIPTION

[0069] The specific embodiments of the present invention are described below to facilitate understanding of the present invention by those skilled in the art. However, it should be clear that the present invention is not limited to the scope of the specific embodiments. For those skilled in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the appended claims, these changes are obvious, and all inventions and creations utilizing the concepts of the present invention are protected.

[0070] Before describing this embodiment, the following terms are explained:

[0071] OCCT: Open Cascade Technology;

[0072] STEP: A format for transferring 3D model data between different CAD software;

[0073] TCL: Tool Command Language;

[0074] RGB: color system;

[0075] XDE: Extended Data Exchange Module in Open Cascade Technology;

[0076] BREP: Boundary Representation in Open Cascade Technology;

[0077] XCAF: Data organization framework in open cascade technology, used to store assembly information and other attribute data;

[0078] CAD: Computer-aided design;

[0079] Creo: CAD design software package;

[0080] UG: Interactive CAD.

[0081] Example 1

[0082] like Figure 1 As shown, the present invention provides a semantic geometric modeling method for a solid fuel power plant, and its implementation method is as follows:

[0083] S1. Based on the open cascade technology, the solid fuel power plant is decoupled to obtain geometric features. Based on the physical boundary semantics, a semantic feature library containing feature cores is obtained. The specific steps are as follows:

[0084] S101. Decoupling multi-dimensional geometric features of the solid fuel power plant based on an open cascade technology to obtain decoupled geometric features;

[0085] S102. Based on the decoupled geometric features and the physical boundary semantics of the functional components in the solid fuel power plant, using the open source library in the open cascade technology, generate a structured feature class library containing topological constraints and assembly relationships;

[0086] S103. Based on the structured feature library, a mapping rule library between geometric elements and engineering semantics is established to obtain a semantic feature library containing feature cores, specifically:

[0087] According to the structured feature library, by establishing a mapping rule library between geometric elements and engineering semantics, based on the ideas of objectification and logical inheritance, the common features of the design of solid fuel power plant components are extracted to obtain the feature core based on the basic geometric elements of the engine. By integrating the feature cores, a semantic feature library containing the feature cores is obtained.

[0088] In this embodiment, Figure 2As shown in the figure, a semantic geometric modeling method for solid fuel power plants is proposed, which includes a feature library, a method for tracing the geometric features of solid structures based on feature decomposition, a top-down parametric modeling mechanism, a three-dimensional model generation method with semantic information, and a minimization model method considering symmetric features. The advantage of this geometric modeling method is that it realizes the synchronous structured preservation of the geometric modeling and design intent information of the solid fuel power plant, and greatly expands the freedom of the component template library without changing the main program code.

[0089] In this embodiment, a semantic feature library is constructed and geometric element parsing is performed. Based on open cascade technology, multi-dimensional geometric feature decoupling is performed on the solid fuel power plant to obtain decoupled geometric features. Based on the physical boundary semantics of functional components such as the combustion chamber, nozzle, and grain, combined with the open source C++ class library in open cascade technology, a structured feature class library containing topological constraints and assembly relationships is generated, and a mapping rule library between geometric elements and engineering semantics is established.

[0090] Based on the ideas of objectification and logical inheritance, common problems in the design of solid propulsion system components were studied, and a series of feature cores based on the basic geometric elements of the engine were formed, forming a feature core template library with semantic information, and the feature cores are universal in the model; the feature cores include wing shapes, cylinders, truncated cones, ellipsoids, disk shapes and star shapes, and the feature cores are lightweight, portable and scalable.

[0091] In this embodiment, to avoid dependence on commercial software, open cascade technology is introduced as the core engine of the geometric modeling method. Open cascade technology provides a highly mature open source C++ class library and a supporting TCL script-based modeling environment. By absorbing the core concept of solid modeling, the components of the solid rocket engine are decomposed into a series of basic geometric elements to form a feature core template library. Dynamic scripting technology is used to achieve the storage, transmission, and expansion of design information, avoiding the complexity of the source code development model and the monotony of the script development model.

[0092] The concept of feature core proposed in this paper is explained as follows: Feature core can be regarded as the basic geometric elements that make up the engine structure. By performing topological analysis and decomposition on the top-level structural parts, the modeling process of complex entities is simplified to the construction of feature cores and their combination relationships, which reduces the difficulty of open source modeling.

[0093] like Figure 3As shown in the figure, the standard format of the feature core includes the supporting modeling script feature.tcl, the preview files feature.step and feature.png, and the model definition file feature-core.json. The feature-core.json file has the function of storing model semantic information, including object identification, attribute description and functional operation.

[0094] In this embodiment, from the perspective of engine modeling, in the three-dimensional solid model, the model semantic information includes the following:

[0095] Object identification: represented by a unique ID, used to characterize a specific entity;

[0096] Attribute description: The attributes of objects or components in the model, such as size, shape, material, etc., give them specific physical and visual characteristics. Geometric information is summarized as private parameters of the feature kernel, and material information is summarized as the characteristics of the physical block;

[0097] Functional operations: Associate objects, components, or surfaces in the model to describe their functions, operations, and usage methods, including color-based physical surface annotation, assembly implementation, and meshing information.

[0098] S2. Based on the semantic feature library, using hierarchical Boolean operation rules and constructive solid geometry methods, a constructive solid geometric feature hierarchical framework is established, and the design parameters corresponding to each entity node are recorded to obtain geometric parameters. The specific steps are as follows:

[0099] S201, according to the semantic feature library, based on the solid modeling method and using the modular concept, obtain the feature core module;

[0100] S202, combining the feature core module, using hierarchical Boolean operation rules and constructive solid geometry methods, constructing a solid geometry tree structure with the tree root as the component, the leaves as the feature core, and the intermediate nodes as the hierarchical Boolean operation operators, to obtain a constructive solid geometry feature hierarchical framework;

[0101] S203. Record the design parameters corresponding to each entity node to obtain geometric parameters.

[0102] In this embodiment, dynamic modeling of a constructive solid geometry tree based on feature traceability is performed: based on a semantic feature library, a traceable constructive solid geometry process model is constructed through hierarchical Boolean operation rules, the design parameters corresponding to each entity node are recorded, and bidirectional tracing of geometric operations and design intent is achieved to obtain geometric parameters;

[0103] Since each feature core is constructed in a discrete manner and lacks inherent correlation, it is necessary to establish a hierarchical framework for constructive solid geometry features and perform constructive solid geometry operations based on displacement and feature origin. By performing transformation operations such as translation, rotation, and scaling on the feature cores, as well as Boolean operations such as union, intersection, and difference, the solid power device can be constructed.

[0104] like Figure 4 As shown in the figure, the constructive solid geometry tree structure can record the basic composition and operation logic of the solid fuel power plant, where the tree root represents the defined components, the leaves represent the feature core, and the middle nodes represent the operators of Boolean operations, thus obtaining a constructive solid geometry feature hierarchy framework;

[0105] The modular concept is incorporated into the solid modeling method, using less-variable features to address changing modeling requirements. Standardization of the feature core module adapts to the individualization and diversification of the design environment. The feature core module manages the computational logic and assembly sequence of the feature core. The feature core can be considered the basic geometric elements that make up the engine structure. By performing topological analysis and decomposition on the top-level structural parts, the modeling process of complex entities is simplified to the construction of feature cores and their combination relationships, reducing the difficulty of open source modeling.

[0106] We further obtain a method for tracing the origin of three-dimensional features based on feature cores to achieve structured storage of model semantics. The fusion logic between multiple feature cores is as follows: Figure 5 shown.

[0107] S3. Based on the geometric parameters, a parameterized network is injected into the hierarchical framework of the constructed solid geometric features. The lower-level geometric parameters of the hierarchical framework of the constructed solid geometric features are iterated using the top-level performance indicators to generate a parameterized model matrix. The specific steps are as follows:

[0108] S301. Determine the main design geometric parameters of the solid fuel power plant based on the geometric parameters, and inject a parameterized network into the constructed three-dimensional geometric feature hierarchical framework;

[0109] S302: constructing a hierarchical framework of three-dimensional geometric features based on the injected parameterized network, using the main design geometric parameters as top-level skeleton model parameters, establishing a top-level skeleton model based on the feature kernel, parameterizing the feature kernel, associating the geometric parameters with the feature kernel size and component operations, and establishing a one-to-one correspondence between the geometric parameters and the feature kernel size;

[0110] S303, iterating the lower-level geometric parameters of the constructive solid geometric feature hierarchical framework based on the top-level performance indicators of the top-level skeleton model, changing the constraints between feature cores using size-driven methods, and combining the one-to-one correspondence between geometric parameters and feature core sizes to obtain a constructive solid geometric operation set;

[0111] S304: Based on the constructive solid geometry operation set, the feature core is processed using constructive solid geometry operations to generate a parametric model matrix that meets preset size requirements. Specifically, based on the constructive solid geometry operation set, the feature library parameters are divided into object groups and tool groups using open cascade technology, a universal fusion operation is performed on the tool group and the object group, and shape segmentation is performed on the feature core based on the intersection result of the universal fusion operation to obtain a segmented feature core.

[0112] Based on the displacement and feature origin, the segmented feature core is processed by constructive solid geometry operation, and fuzzy Boolean operation is used to generate a parametric model matrix that meets the preset size requirements by setting the preset tolerance.

[0113] In this embodiment, a parameter-driven hierarchical modeling mechanism is used: a parametric constraint network is injected into the hierarchical structure of the constructed solid geometry. The top-level performance indicators drive the iteration of the lower-level geometric parameters to generate a parametric model matrix that meets multidisciplinary constraints, completing the automatic transfer from system indicators to part dimensions.

[0114] A dimension-driven approach is used to change the feature core formed by geometric constraints to construct a set of three-dimensional geometric operations. A one-to-one correspondence is established between the parameter sequence and the dimension sequence of the feature core. When the dimension sequence as a parameter is assigned different values, the size of the corresponding feature core will also change accordingly. At the same time, the feature core is driven to generate engine parts that meet the dimension requirements through constructing three-dimensional geometric operations.

[0115] According to the overall technical requirements, the main design geometric parameters of the solid device are determined, and the main design geometric parameters are used as the key parameters of the top skeleton model. The top skeleton model based on the characteristic core is established, and the parameters are associated with the characteristic core size and component operations.

[0116] The basic principle of parametric modeling mechanism is as follows Figure 6 As shown in the figure, based on the skeleton template, each feature core is parameterized, which can greatly simplify repetitive design work and avoid the entanglement of complex CAD underlying design theories and design technologies. For R&D personnel who lack experience and relevant field knowledge, it saves a lot of learning costs and is more user-friendly.

[0117] From the perspective of the data chain, data flow follows a top-down sharing mechanism. The geometric modeling module captures and integrates the design parameters from the global project, and passes the data step by step to each model, physical block and feature core unit. From the perspective of geometric modeling construction, model generation follows a bottom-up logical order. The topological structure of the model relies on Boolean operations. The basic geometric elements are generated by the feature core TCL script, and then the dynamic loading and inter-group calculation of the geometric elements are implemented based on the C++ library. The operational attributes of the feature core include "addition", "removal" and "symmetry". The first feature core is displayed as "addition" attribute by default. The component relationship is established according to the assembly logic, and the engine parts and the overall structure are finally formed;

[0118] Open Cascade technology divides parameters into two groups: objects and tools. Based on a common fusion operation, all input shapes are segmented according to the intersection results. Furthermore, for cells that require multiple operations, fuzzy Boolean operations are introduced. By setting a certain tolerance, stable processing is achieved for cases where the model boundaries are approximately consistent, slightly interfered with, or cannot be aligned.

[0119] S4. Import the parametric model matrix into the solid modeling kernel and use the color-based channel annotation technology to perform semantic entity segmentation on the physical domain to obtain an enhanced 3D solid model containing geometric topology and physical attribute labels. The specific steps are as follows:

[0120] S401, importing the parametric model matrix into the solid modeling kernel, using the solid modeling method to establish the internal relationship between the feature cores in the parametric model matrix to obtain the feature core relationship;

[0121] S402. Using a color system-based channel marking technology, mark the combustion surface, symmetry surface, and insulation surface, and use the marked preset color system value as a characteristic color;

[0122] S403: Perform semantic entity segmentation on the physical domain according to the characteristic color to obtain an enhanced three-dimensional entity model including geometric topology and physical attribute labels.

[0123] In this embodiment, semantically enhanced 3D solid generation: the parametric model is imported into the solid modeling kernel, and RGB multi-channel annotation technology is used to perform semantic entity segmentation on physical domains such as combustion field and structural boundaries, and an enhanced 3D solid model (enhanced STEP model) containing both geometric topology and physical attribute labels is output;

[0124] Based on the solid modeling method, the intrinsic relationship between feature cores is established. In view of the need for boundary recognition in the simulation of solid device performance, color information is used to realize the information annotation of physical surfaces. The surface feature colors in the CAD model are used to avoid the loss and conversion of the annotation information format. The geometric model is drawn in CAD, and the combustion surface, symmetry surface and insulation surface are respectively annotated with the defined RGB values as their feature colors. When the model is imported into other CAD platforms, the feature color information attached to the model will be included, so other CAD can determine the ownership of the surface by identifying the color information of the surface.

[0125] Importing and exporting geometric models between CAD software inevitably involves file format conversion, a process that makes it difficult to ensure the smooth transfer of additional information. Surface numbers, names, and feature colors are important additional information for geometric entities. Because different CAD platforms use different surface numbering and naming conventions, it is impossible to use surface numbers and names to identify geometric boundaries. Using surface feature colors in CAD models can avoid the loss and conversion of annotation information formats. Furthermore, the 8-bit RGB color system provided by CAD offers a variety of options for identification, effectively avoiding the problem of surface information confusion.

[0126] Solid modeling primarily records the process of generating a solid device model based on the construction of solid geometry operations using feature kernels, rather than its final shape or geometry. This method uses a constructive solid geometry tree structure to describe the relationship between feature kernels and complex shapes. The root of the tree represents the defined shape, the leaves represent the feature kernels, and the intermediate nodes represent transformations or Boolean operators.

[0127] In this example, while a reasonable topology can be constructed after performing Boolean operations within the Open Cascade technology framework, the newly generated entities fail to effectively inherit the color information of the original geometric surfaces because the 3D model follows the isolation strategy of geometry and additional behavior. To address this issue, the Extended Data Exchange module of Open Cascade technology is applied. XDE can encapsulate metadata such as color, layer information, detailed descriptions of components, and validation properties, along with the basic boundary representation BREP geometry data, into an XCAF document.

[0128] Given that Boolean operations make it difficult to directly obtain the face inheritance relationship between the newly generated entity and the original entity, and that tracing the parent-child relationship by performing pairwise operations on each face is computationally intensive, an optimization strategy is proposed. The faces of the newly generated entity always come from the base model and the tool model, and have an either-or mutually exclusive relationship. The entity with fewer face elements in the two is considered the marked entity, and the closed surface obtained by fuzzy Boolean operations on the set of marked entity faces is obtained. Then, Boolean subtraction operations are performed on each face of the newly generated entity and the closed surface one by one. If the result is empty, the entity to which it belongs is a marked unit, otherwise it is an unmarked unit.

[0129] like Figure 7 As shown, the specific coloring process is as follows:

[0130] Read the base model and tool model as Object entity and Tool entity respectively, and expose all facets;

[0131] Calculate the number of facets of Object and Tool, and select the one with fewer facets as the marked entity, otherwise it is the unmarked entity. The facets of the marked entity are recorded as marked surfaces;

[0132] Perform Boolean operations between Object and Tool, generate the result sub-entity, and obtain all its facets;

[0133] Perform a Boolean subtraction operation on each face of the result entity and the marked surface, and assign the result entity the color attribute of its parent face. If the number of face elements in the operation result is 0, the face comes from the marked entity, otherwise it comes from the unmarked entity;

[0134] After the program is finished running, export the STEP file with color information;

[0135] The above method is used to color the surface, and the colors of all sub-surfaces are successfully restored, such as Figure 8 As shown, (a) represents before restoration, and (b) represents after restoration.

[0136] S5. Use the feature kernel-based model symmetry solution algorithm to solve the symmetry of the enhanced 3D solid model. Based on the solution results, compress the complexity of the enhanced 3D solid model to obtain a lightweight solid model and complete the semantic geometric modeling. The specific steps are as follows:

[0137] S501. Based on the symmetry of the model feature kernel, the enhanced three-dimensional solid model is simplified into a two-dimensional geometric figure consisting of lines extending from the center to the midpoint. Based on the two-dimensional geometric figure, the lines are considered as rays starting from the center point, and the starting rays are considered as repeated rays. The rays are sorted according to their angles with the zero-degree ray to form an angle series arranged from small to large. The difference sequence of the angle series is calculated to obtain an information string with angle data as elements.

[0138] S502: Consider the information string as a ring structure connected end to end, and determine the length of the information string;

[0139] If the length of the information string is an odd number, select the smallest element in the information string and calculate the differences between the elements on the left and right of the smallest element in turn. If the differences are all 0, the information string is defined as having an axisymmetric property. If the differences are not all 0, the symmetry number of the information string is defined as 1, and the process proceeds to step S505.

[0140] If the length of the information string is an even number, two adjacent elements with equal values in the information string are selected. If no adjacent elements with equal values can be found, the information string is directly determined to be non-axisymmetric. The left and right elements of the two selected elements are symmetrically subtracted, and the subtraction results are determined. If all subtraction results are 0, the information string is defined as having an axisymmetric property. The information string defined as having an axisymmetric property is split into multiple substrings, and the judgment process of S502 is repeated. If all subtraction results are not 0, the symmetry number of the information string is defined as 1, and the process proceeds to step S505.

[0141] S503, determine whether the current information string is axisymmetric and the maximum axisymmetric number is obtained, if so, proceed to step S505, otherwise proceed to step S504;

[0142] S504. Based on the information burst length, the period length of the information burst is traversed, and the period length of a valid period is retained to obtain multiple groups of period lengths. It is determined whether the information burst length is divisible by the period length. If so, corresponding elements in the multiple groups of information burst segments with lengths equal to the period length are compared to see whether they are equal. If they are equal, the maximum rotational symmetry number is obtained. If they are not equal, the symmetry number of the information burst is defined as 1.

[0143] S505, dividing the characteristic core into a strong symmetric characteristic core and a weak symmetric characteristic core according to the symmetry number, retaining the entire entity structure of the weak symmetric characteristic core with a symmetry number of 1, and dividing the strong symmetric characteristic core with multiple symmetric units into the minimum symmetric unit;

[0144] S506. Based on the weak symmetric feature core that retains the entire solid structure and the strong symmetric core that is divided into the smallest symmetric units, feature reduction is used to compress the complexity of the enhanced three-dimensional solid model to obtain a lightweight solid model and complete the semantic geometric modeling.

[0145] In this embodiment, symmetry-guided model lightweighting: Analyze and enhance the geometric or physical symmetry in the model, compress the model complexity through feature reduction, and ultimately output a lightweight solid model with complete semantics and optimal expression efficiency;

[0146] Engine components generally exhibit rotationally symmetrical structural characteristics. Therefore, the overall structure can be represented by a simplified shape of 1 / n, and the symmetry of the engine components can be judged based on the symmetry of the characteristic core set of the components. The characteristic cores are divided into two categories: strong symmetry and weak symmetry. When a strong symmetric characteristic core is introduced into the model, it will not change the symmetry, while the introduction of a weak symmetric characteristic core may lead to adjustments and changes in the symmetry of the model. By simplifying the weak symmetric characteristic core into multiple lines extending from the center to the midpoint, the model is converted into a two-dimensional geometric figure composed of multiple center lines. The problem of symmetry analysis of the model is transformed into the problem of identifying and matching repeated / symmetric sequences in an ordered array, and then the model is minimized by the established symmetry judgment criteria.

[0147] Feature kernels can be divided into two categories: strong symmetry and weak symmetry. Strong symmetry feature kernels, such as rotational bodies such as cylinders and truncated cones, exhibit equivalence in the circumferential direction, and the cutting ratio does not affect their ability to describe shape features. In contrast, weak symmetry feature kernels, such as wing-shaped structures, have array numbers that are the core indicator affecting the central symmetry characteristics. Therefore, when a strong symmetry feature kernel is introduced into a model, it will not bring about a change in symmetry, while the introduction of a weak symmetry feature kernel may lead to adjustments and changes in the model's symmetry.

[0148] Given that each weakly symmetric characteristic core, such as a wing-shaped structure, can be simplified to a line extending from the center to the midpoint in the axial direction, such as Figure 9 As shown, assuming that the model contains several such characteristic cores, the starting deflection angle α of each characteristic core and its array number N are extracted. On this basis, the geometric simplified model of any characteristic core can be described by an arithmetic progression. The specific expression is:

[0149] α=α+360n / N;

[0150] Where n = 1, 2, ..., N represents the nth array, and α represents the deflection angle of the center line of each feature core relative to the zero-degree reference line. After the above simplification, the enhanced three-dimensional solid model is transformed into a two-dimensional geometric figure composed of multiple center lines. The problem of symmetry analysis of the model is transformed into the problem of identifying and matching repeated or symmetric sequences in an ordered array, which reduces the complexity of automatically determining the symmetry of components.

[0151] In this embodiment, a model symmetry solution algorithm based on characteristic kernel is proposed, such as Figure 10 As shown, the core steps are as follows:

[0152] A1. Encoding: The model is simplified into a two-dimensional geometric figure. The model is abstracted into multiple rays starting from the center point. To achieve a geometric closed loop, the starting ray is treated as a repeated ray. The angles of these rays relative to the 0-degree ray are arranged in ascending order to form an angle sequence. The difference sequence of this sequence is calculated to obtain the information string A. Information string A contains the geometric information of the angles between adjacent rays in the clockwise direction starting from the minimum starting angle of the simplified model. Each element represents this angle data.

[0153] A2. Axisymmetry determination: Figure 10 As shown in (a), consider the information string A as a ring structure connected end to end. For an information string of odd length, find the element with the smallest value and calculate the difference between the elements on its left and right sides. If all the differences are 0, the information string is considered to be axially symmetric; otherwise, the symmetry number is 1, and the process proceeds to step A4.

[0154] For an even-length information string, two adjacent elements with equal values are searched. If no adjacent elements with equal values are found in the even-length information string, the information string is directly determined to be non-axisymmetric. A symmetrical subtraction operation is performed on the left and right sides of the two adjacent elements with equal values. If all subtraction results are 0, the information string is determined to have an axisymmetric property. For the information string determined to have an axisymmetric property, it is split into multiple substrings and the judgment process of step A2 is repeated. If the information string is determined to be axisymmetric and the maximum axisymmetric number is obtained, step A4 is performed; otherwise, step A3 is performed.

[0155] A3. Rotational symmetry determination: To determine rotational symmetry, when the information string length is L, traverse all possible period lengths i, i∈[1,L-1], and check whether each length is a valid period. For each possible period length i, check whether the information string length L is divisible by the period length i. If it is divisible, further compare the corresponding elements in multiple groups of information string segments of length i to see if they are equal. If they are equal, the maximum rotational symmetry number is obtained; otherwise, the symmetry number is 1.

[0156] A4. Model segmentation: For models with a symmetry number of 1, their entire physical structure is retained; for other models with multiple symmetry units, the models with the largest axial symmetry number and the largest rotational symmetry number are segmented into the smallest symmetry units.

[0157] In this embodiment, the wing column grain is taken as an example to demonstrate the above symmetry solving process. The characteristic cores of the grain mainly include cylinders, frustums, ellipsoids and wings, among which cylinders, frustums and ellipsoids are all rotational bodies and are considered as strong symmetry characteristic cores, which have no significant effect on the overall symmetry; however, the wings are weak symmetry characteristic cores in the grain components, and their array number and wing type have a greater impact on the overall symmetry; the types of wings are mainly divided into front wings and rear wings, and the array number is the number of uniform arrangements of each wing at the corresponding latitude. For a single-wing engine, the number of rotational symmetries of the grain is equal to the number of rotational symmetries of the wings; but in front and rear wing engines, the situation will be complicated, and it is necessary to comprehensively consider the symmetry numbers of the front and rear wings to determine the overall symmetry of the grain;

[0158] Specifically, the number of arrays of the front wings in the charge model is set to 3, the starting deflection angle is set to 60°, and the number of arrays of the rear wings is set to 6, and the starting deflection angle is set to 0°; first, the charge model is simplified and coded, and the angle sequence formed by the front wings is [60°, 180°, 300°, 420°], and the angle sequence formed by the rear wings is [0°, 60°, 120°, 180°, 240°, 300°, 360°]; the angle sequence of the front wings and the angle sequence of the rear wings are merged to obtain the angle sequence of the charge model [0°, 60°, 60°, 120°, 180°, 240°, 300°, 360°]. 0°, 180°, 240°, 300°, 300°, 360°, 420°]; and calculate the difference sequence of the sequence to get the sequence [60, 0, 60, 60, 0, 60, 60, 0, 60]; and apply the symmetry judgment algorithm. Since there are three repeated periodic units [60, 0, 60] in the difference sequence, the symmetry number of the model is determined to be 3; at this time, the minimum symmetry unit is the geometric entity clamped by rotating clockwise from the minimum starting angle 0° to 120°; in actual design, the wing-column type charge generally takes a single wing or a geometric configuration with overlapping front and rear wings, such as Figure 11 The algorithm is demonstrated as shown.

[0159] Example 2

[0160] In this embodiment, Figure 12 As shown in the figure, the design sketch of a certain type of solid fuel power plant adopts an internal combustion single-wing type grain, casting process, ordinary single nozzle and single thrust configuration engine structure according to the structural form selected by the overall design department. The engine head is designed to be ellipsoidal, and the convergent and divergent sections of the nozzle are both conical structures. The specific parameters are as follows: Figure 13 As shown;

[0161] Based on the various technical implementation schemes described above, the present invention provides a semantic geometric modeling method for a solid fuel power plant, which specifically includes the following steps:

[0162] Establishing a feature core library: Based on the core concepts of modularization and object-orientedness, a feature core library for the design of solid fuel power plant components is summarized, and the feature core is universal in the model, such as Figure 13 By flexibly combining feature cores, the entities of each engine component can be constructed, such as Figure 14 The feature kernel combination shown.

[0163] Establish a parametric modeling template based on the traceability method of constructive solid geometric features: Aiming at the structural design of a certain type of solid fuel power plant, a top-down design approach is adopted to plan the overall design process. Given the overall parameters, based on empirical formulas and structural geometric constraints, parameter associations are established between the overall machine and subcomponents, as well as between subcomponents and feature cores, ultimately forming a three-dimensional solid model of the solid power plant. The subcomponent construction is specifically as follows:

[0164] For the construction of the grain parts, the "subtraction" modeling idea is adopted, that is, the inner cavity parts are cut out from the complete matrix to obtain the required shape; since there is no closed cavity inside the grain, each grain contains at most one inner cavity part, thus simplifying the grain modeling problem into the modeling of two independent parts: the matrix and the inner cavity; the wing column type charge is a wall-cast grain, and its matrix fits tightly to the inner surface of the shell / insulation layer and is formed accordingly; the grain core mold presents the inner hole shape of a stepped single wing. The outer contour entity consists of two ellipsoidal feature cores and one cylindrical feature core; the grain core mold presents the inner hole shape of a stepped single wing. When considering the possible influence of the insulation layer on the shape contour of the grain, the cylindrical feature core can be used for cutting at both ends. By performing the Boolean difference operation of the matrix and the grain core mold entity, the grain solid model is obtained. When considering the possible influence of the insulation layer on the shape contour of the grain, the cylindrical feature core can be used for cutting at both ends; its parameterized template is as follows Figure 15 As shown;

[0165] For the construction of combustion chamber parts, the design work usually covers two core aspects: shell structure design and thermal protection design, specifically involving the design of front and rear heads, barrel structure design, connection structure design, and insulation layer design. The ellipsoidal, threaded barrel joint and head joint feature core design, as well as the insulation layer feature core, are used. During the modeling process, the cylindrical feature core and the ellipsoidal feature core are effectively reused; its parametric templates are as follows: Figure 16 As shown;

[0166] The nozzle assembly is complex and its assembly relationships are complicated. Multi-level model data duplication and sharing can easily lead to the complication of model dependencies, which makes subsequent design modifications difficult. Therefore, a common nozzle is selected as the basic structural form and the model is simplified. Since the nozzle parts are only solved for mass and do not involve performance calculations, rounded corners are not used to ensure model stability. Its parameterized template is as follows: Figure 17 As shown;

[0167] Color-based physical boundary annotation: In view of the need for boundary identification in solid charge performance simulation, color information is used to implement physical surface information annotation. In order to avoid conflicts with the defined colors in the model, a set of unconventional RGB value examples are listed, and their corresponding relationships with different boundaries are as follows: Figure 18 As shown, Figure 18 (a) is the pre-processing charge for interior ballistics simulation. Figure 18 (b) Pre-processing the charge for structural integrity simulation; taking the charge model as an example, marking the physical boundary between the charge structural integrity and the interior ballistics calculation, such as Figure 19 As shown;

[0168] Minimization model calculation: The grain model is a single-wing, with an array of six wings and a starting deflection angle of 0°. Simplified encoding of the grain model yields the angle sequence [0°, 60°, 120°, 180°, 240°, 300°, 360°]. The difference sequence of this angle sequence is calculated as [60°, 60°, 60°, 60°, 60°, 60°]. Using a symmetry algorithm, since this difference sequence contains six repeating periodic units [60, 60, 60, 60, 60], the model's symmetry number is determined to be 6.

[0169] In this embodiment, the results are analyzed, and the modeled charge, nozzle, insulation layer, combustion chamber entity and the assembled whole machine are as follows: Figure 20 and Figure 21 As shown, Figure 20 (a) is the drug column, Figure 20 (b) is the combustion chamber, Figure 21 (c) is a thermal insulation layer, Figure 21 (d) is the nozzle, Figure 21 (e) For the complete machine, the STEP file generated based on open cascade technology shows the consistency of the solid model in the mainstream software Creo and UG; in particular, the core mold of the grain is generally required for simulation calculation, minimization and color annotation;

[0170] At the same time, the present invention was used to model a solid fuel power plant, generating geometric files of the grain, nozzle, insulation layer, and combustion chamber. The files were imported into commercial software for model review, and the results were compared and analyzed. The following conclusions were drawn:

[0171] Compared with common CAD modeling methods, the present invention realizes the physical boundary annotation of the model through color information;

[0172] The three-dimensional model constructed by the present invention through the open source platform still maintains the integrity and stability of the structure in commercial software.

Claims

1. A semantic geometric modeling method for a solid fuel power plant, characterized by: The following steps are involved: S1. Based on the open cascade technology, the solid fuel power plant is decoupled to obtain geometric features, and based on the physical boundary semantics, a semantic feature library containing feature cores is obtained; S2. Based on the semantic feature library, using hierarchical Boolean operation rules and constructive solid geometry methods, a constructive solid geometric feature hierarchical framework is established, and the design parameters corresponding to each entity node are recorded to obtain geometric parameters; S3. Injecting a parameterized network into the hierarchical framework of the constructed solid geometric features according to the geometric parameters, and iterating the lower-level geometric parameters of the hierarchical framework of the constructed solid geometric features using the top-level performance indicators to generate a parameterized model matrix; S4. Import the parametric model matrix into the solid modeling kernel, use the color system-based channel annotation technology to perform semantic entity segmentation on the physical domain, and obtain an enhanced 3D solid model containing geometric topology and physical attribute labels; S5. Use the feature kernel-based model symmetry solving algorithm to solve the symmetry of the enhanced three-dimensional solid model. According to the solution results, the complexity of the enhanced three-dimensional solid model is compressed to obtain a lightweight solid model and complete the semantic geometric modeling.

2. The semantic geometric modeling method for a solid fuel power plant according to claim 1, characterized in that: Said S1 comprises the following steps: S101. Decoupling multi-dimensional geometric features of the solid fuel power plant based on an open cascade technology to obtain decoupled geometric features; S102. Based on the decoupled geometric features and the physical boundary semantics of the functional components in the solid fuel power plant, using the open source library in the open cascade technology, generate a structured feature class library containing topological constraints and assembly relationships; S103. According to the structured feature library, a mapping rule library between geometric elements and engineering semantics is established to obtain a semantic feature library containing feature cores.

3. The semantic geometric modeling method for a solid fuel power plant according to claim 1, characterized in that: The S103 is specifically as follows: Based on the structured feature library, a mapping rule library between geometric elements and engineering semantics is established. Based on the concepts of objectification and logical inheritance, the common features of solid fuel power plant component design are extracted. The feature core based on the basic geometric elements of the engine is obtained. By integrating the feature core, a semantic feature library containing the feature core is obtained. The characteristic cores include: wing-shaped, cylindrical, frustum-shaped, ellipsoidal, butterfly-shaped and star-shaped.

4. The semantic geometric modeling method for a solid fuel power plant according to claim 1, characterized in that: The S2 comprises the following steps: S201, according to the semantic feature library, based on the solid modeling method and using the modular concept, obtain the feature core module; S202, combining the feature core module, using hierarchical Boolean operation rules and constructive solid geometry methods, constructing a solid geometry tree structure with the tree root as the component, the leaves as the feature core, and the intermediate nodes as the hierarchical Boolean operation operators, to obtain a constructive solid geometry feature hierarchical framework; S203. Record the design parameters corresponding to each entity node to obtain geometric parameters.

5. The semantic geometric modeling method for a solid fuel power plant according to claim 1, characterized in that: The S3 includes the following steps: S301. Determine the main design geometric parameters of the solid fuel power plant based on the geometric parameters, and inject a parameterized network into the constructed three-dimensional geometric feature hierarchical framework; S302: constructing a hierarchical framework of three-dimensional geometric features based on the injected parameterized network, using the main design geometric parameters as top-level skeleton model parameters, establishing a top-level skeleton model based on the feature kernel, parameterizing the feature kernel, associating the geometric parameters with the feature kernel size and component operations, and establishing a one-to-one correspondence between the geometric parameters and the feature kernel size; S303, iterating the lower-level geometric parameters of the constructive solid geometric feature hierarchical framework based on the top-level performance indicators of the top-level skeleton model, changing the constraints between feature cores using size-driven methods, and combining the one-to-one correspondence between geometric parameters and feature core sizes to obtain a constructive solid geometric operation set; S304: Based on the constructive solid geometry operation set, the feature kernel is processed using the constructive solid geometry operation to generate a parameterized model matrix that meets the preset size requirements.

6. The semantic geometric modeling method for a solid fuel power plant according to claim 5, characterized in that: Specifically, S304 includes: dividing the feature library parameters into an object group and a tool group using an open cascade technology according to a set of constructed solid geometry operations, performing a universal fusion operation on the tool group and the object group, and performing shape segmentation on the feature core based on an intersection result of the universal fusion operation to obtain a segmented feature core; Based on the displacement and feature origin, the segmented feature core is processed by constructive solid geometry operation, and fuzzy Boolean operation is used to generate a parametric model matrix that meets the preset size requirements by setting the preset tolerance.

7. The semantic geometric modeling method for a solid fuel power plant according to claim 1, characterized in that: The S4 comprises the following steps: S401, importing the parametric model matrix into the solid modeling kernel, using the solid modeling method to establish the internal relationship between the feature cores in the parametric model matrix to obtain the feature core relationship; S402. Using a color system-based channel marking technology, mark the combustion surface, symmetry surface, and insulation surface, and use the marked preset color system value as a characteristic color; S403: Perform semantic entity segmentation on the physical domain according to the characteristic color to obtain an enhanced three-dimensional entity model including geometric topology and physical attribute labels.

8. The semantic geometric modeling method for a solid fuel power plant according to claim 1, characterized in that: The S5 comprises the following steps: S501. Based on the symmetry of the model feature kernel, the enhanced three-dimensional solid model is simplified into a two-dimensional geometric figure consisting of lines extending from the center to the midpoint. Based on the two-dimensional geometric figure, the lines are considered as rays starting from the center point, and the starting rays are considered as repeated rays. The rays are sorted according to their angles with the zero-degree ray to form an angle series arranged from small to large. The difference sequence of the angle series is calculated to obtain an information string with angle data as elements. S502: Consider the information string as a ring structure connected end to end, and determine the length of the information string; If the length of the information string is an odd number, select the smallest element in the information string and calculate the differences between the elements on the left and right of the smallest element in turn. If the differences are all 0, the information string is defined as having an axisymmetric property. If the differences are not all 0, the symmetry number of the information string is defined as 1, and the process proceeds to step S505. If the length of the information string is an even number, two adjacent elements with equal values in the information string are selected. If no adjacent elements with equal values can be found, the information string is directly determined to be non-axisymmetric. The left and right elements of the two selected elements are symmetrically subtracted, and the subtraction results are determined. If all subtraction results are 0, the information string is defined as having an axisymmetric property. The information string defined as having an axisymmetric property is split into multiple substrings, and the judgment process of S502 is repeated. If all subtraction results are not 0, the symmetry number of the information string is defined as 1, and the process proceeds to step S505. S503, determine whether the current information string is axisymmetric and the maximum axisymmetric number is obtained, if so, proceed to step S505, otherwise proceed to step S504; S504. Based on the information burst length, the period length of the information burst is traversed, and the period length of a valid period is retained to obtain multiple groups of period lengths. It is determined whether the information burst length is divisible by the period length. If so, corresponding elements in the multiple groups of information burst segments with lengths equal to the period length are compared to see whether they are equal. If they are equal, the maximum rotational symmetry number is obtained. If they are not equal, the symmetry number of the information burst is defined as 1. S505, dividing the characteristic core into a strong symmetric characteristic core and a weak symmetric characteristic core according to the symmetry number, retaining the entire entity structure of the weak symmetric characteristic core with a symmetry number of 1, and dividing the strong symmetric characteristic core with multiple symmetric units into the minimum symmetric unit; S506. Based on the weak symmetric feature core that retains the entire solid structure and the strong symmetric core that is divided into the smallest symmetric units, feature reduction is used to compress the complexity of the enhanced three-dimensional solid model to obtain a lightweight solid model and complete the semantic geometric modeling.

Images