A method for modeling and topology checking of power cabin disassembly process based on multi-source constraint matrix

Abstract

This invention discloses a method for modeling and topology verification of power compartment disassembly and assembly processes based on a multi-source constraint matrix, relating to the field of special vehicle maintenance and assembly process modeling technology. The method includes: constructing a component connection graph model based on graph theory to describe the assembly structure of the power compartment system; extracting multi-source constraint information from the component connection graph model and fusing it into a multi-source process constraint matrix; generating a candidate set of disassembly and assembly sequences containing multiple potential candidate sequences by initially sorting the disassembly and assembly order of each component in the power compartment system; performing recursive topology verification and closed-loop constraint deadlock detection on the candidate set of disassembly and assembly sequences; and selecting the optimal candidate disassembly and assembly sequence from the detected candidate set. This invention solves the problems in existing technologies regarding the reliance on manual experience for disassembly and assembly sequence planning, lack of unified modeling of structural and spatial constraints, and insufficient handling of closed-loop dependencies and symmetrical components in power compartment disassembly and assembly operations.

Description

Technical Field

[0001] This invention relates to the field of special vehicle maintenance and assembly process modeling technology, and more specifically to a method for modeling and topology verification of power compartment disassembly and assembly processes based on a multi-source constraint matrix. Background Technology

[0002] The power compartment is a crucial component of the power system of special vehicles, typically integrating multiple key functional units such as the engine block, transmission components, fuel system, lubrication system, cooling system, electrical equipment, and auxiliary control devices. Due to the compact internal layout and highly integrated system structure of the power compartment, complex assembly and functional coupling relationships exist between various components, making the maintenance and disassembly process of the power compartment system highly complex. During the long-term operation of special vehicles, critical components within the power compartment require regular inspection, maintenance, or replacement. Maintenance tasks such as engine maintenance, fuel system inspection, lubrication system maintenance, and electrical equipment replacement often involve the hierarchical disassembly and reassembly of multiple components. Because of the large number of components, complex assembly levels, and the prevalent assembly dependencies, spatial obstruction relationships, and maintenance process constraints among these components, the disassembly and assembly sequence has strict logical dependencies. An unreasonable disassembly and assembly sequence can easily lead to conflicting disassembly and assembly paths, reduced operational efficiency, or even unexecutable maintenance steps, thereby affecting the maintenance efficiency and safety of special vehicles.

[0003] As the scale and integration of special vehicle systems continue to increase, traditional methods of planning disassembly and assembly operations, relying on maintenance manuals and the experience of maintenance personnel, are increasingly unable to meet the demands for efficient maintenance support. In recent years, with the development of digital maintenance, virtual maintenance simulation, and intelligent operation and maintenance technologies, special vehicle maintenance support is gradually moving towards digitalization and intelligence. In a digital maintenance system, by uniformly modeling vehicle structural information, assembly relationships, and maintenance process information, the maintenance process can be digitally expressed and automatically analyzed. This allows for the planning and verification of disassembly and assembly procedures before maintenance operations are implemented, improving maintenance efficiency and reducing the risks associated with human error. Therefore, how to utilize digital modeling methods to systematically describe the disassembly and assembly process of the power compartment of special vehicles and to automatically generate and verify the feasibility of disassembly and assembly sequences has become a crucial issue in maintainability engineering and digital maintenance research.

[0004] For modeling complex product assembly and disassembly processes, existing research has mainly developed the following three technical approaches:

[0005] The first type is a modeling method based on assembly relationship diagrams or product structure trees. This method constructs assembly dependencies between components and uses topological sorting algorithms to generate assembly or disassembly sequences. This type of method has an intuitive structural expression, but it usually only considers structural assembly relationships and is difficult to express spatial occlusion relationships and maintenance process constraints at the same time. The second category is process planning methods based on knowledge rules or expert experience. These methods generate disassembly and assembly sequences by pre-defining disassembly and assembly rules or maintenance procedures. This method has good engineering applicability in specific scenarios, but it relies on the construction of a manual rule base and is difficult to adapt to vehicle systems with complex structures and large configuration changes. The third category is maintenance planning methods that combine 3D models with spatial interference analysis. These methods determine the disassembly direction and assembly / disassembly sequence through geometric collision detection or spatial path analysis. While these methods emphasize spatial constraint analysis, they lack a unified modeling approach for structural dependencies and maintenance process constraints. Furthermore, when the system contains symmetrical components or cyclic dependencies, existing methods easily generate numerous redundant assembly / disassembly sequences, even forming closed-loop dependency structures, leading to reduced efficiency in assembly / disassembly sequence planning or deadlock problems.

[0006] Therefore, how to uniformly model multi-source information such as structural constraints, spatial constraints, and maintenance process constraints during the disassembly and assembly of the power compartment of special vehicles, and on this basis realize the automatic generation of disassembly and assembly sequences, topology feasibility verification, and closed-loop constraint deadlock detection, so as to improve the efficiency of disassembly and assembly process planning and ensure the executability of maintenance operations, has become an urgent problem to be solved in the maintainability design and digital maintenance technology of special vehicles. Summary of the Invention

[0007] In view of the above problems, the present invention proposes a method for modeling and topology verification of the power compartment disassembly and assembly process based on a multi-source constraint matrix to overcome or at least partially solve the above problems.

[0008] To achieve the above objectives, the present invention adopts the following technical solution: This invention provides a method for modeling and topology verification of the engine compartment disassembly and assembly process based on a multi-source constraint matrix, comprising the following steps: Acquire component structure information, assembly relationship information, and maintenance operation information of the power compartment system, and construct a component connection diagram model to describe the assembly structure of the power compartment system based on graph theory; Multi-source constraint information is extracted from the component connection diagram model, and a unified multi-source process constraint matrix is ​​constructed by fusing the multi-source constraint information. By initially sorting the disassembly and assembly sequence of each component in the power compartment system, a candidate set of disassembly and assembly sequences containing multiple potential candidate disassembly and assembly sequences is generated; where each candidate disassembly and assembly sequence represents the disassembly and assembly sequence of a component. For each candidate disassembly and assembly sequence in the candidate disassembly and assembly sequence set, a recursive topological test is performed. By traversing the candidate disassembly and assembly sequences and comparing them with the multi-source process constraint matrix, it is determined whether there is a constraint violation. Based on the number of violations, a violation index is constructed. Closed-loop constraint deadlock detection was performed on the power compartment system to obtain the system deadlock index; Based on the violation index and the system deadlock index, and combined with the disassembly and assembly path cost, the optimal candidate disassembly and assembly sequence is selected from the candidate set of disassembly and assembly sequences, and the corresponding disassembly and assembly process is generated.

[0009] Furthermore, the graph theory-based construction of the component connection graph model for describing the power compartment system assembly structure specifically includes: Based on the component structure information and assembly relationship information of the power compartment system, the structural parameters, spatial positional relationships and assembly connection relationships of each component are obtained; Based on the maintenance operation information of the power compartment system, the disassembly and assembly attributes of each assembly connection are determined; the disassembly and assembly attributes include the type of connecting parts, the constraint strength of the connection structure, and the disassembly and assembly difficulty coefficient. Each component is abstracted as a component node in a graph structure, and each node is assigned corresponding structural parameters and spatial positional relationships; the assembly and connection relationships between components are abstracted as edges in a graph structure; and the disassembly and assembly attributes are used as the weights of the corresponding edges. Based on the component nodes, edges, and weights, a component connection graph model of the power compartment system is constructed.

[0010] Furthermore, the multi-source constraint information includes structural constraint information, spatial constraint information, and maintenance process constraint information; The structural constraint information includes the assembly dependencies between the components; The spatial constraint information includes the assembly and disassembly sequence of each component due to spatial obstruction or spatial interference. The maintenance process constraint information includes the sequence of operations determined by maintenance safety specifications or maintenance procedure requirements.

[0011] Furthermore, the step of constructing a unified multi-source process constraint matrix by fusing the multi-source constraint information specifically includes: Based on the structural constraint information, spatial constraint information, and maintenance process constraint information, corresponding structural constraint submatrix, spatial constraint submatrix, and process constraint submatrix are constructed. The structural constraint submatrix, spatial constraint submatrix, and process constraint submatrix are fused to obtain a multi-source process constraint matrix.

[0012] Furthermore, each element in the structural constraint submatrix is ​​represented as follows:

[0013]

[0014] in, Representing component nodes i With component nodes The weighted structural constraint values ​​between them i Indicates the first i Each component node; Indicates the first Each component node; variables Indicates the first r Each connector is connected to the component node. i With component nodes The contribution value of the structural constraint relationship between them; r Representing component nodes i With component nodes The number of the connecting parts; variables Representing component nodes i With component nodes The total number of connecting parts that exist.

[0015] Furthermore, each element in the spatial constraint submatrix is ​​represented as follows:

[0016]

[0017] in, Representing component nodes i For component nodes Spatial constraints; Indicates the threshold for determining spatial occlusion; Representing component nodes i For component nodes Spatial occlusion degree, variable Representing component nodes i For component nodes The occlusion volume includes spatial overlap volume or projected occlusion volume; variables Representing component nodes The overall spatial volume.

[0018] Furthermore, the recursive topological test performed on each candidate disassembly / assembly sequence in the candidate disassembly / assembly sequence set, by traversing the candidate disassembly / assembly sequences and comparing them with the multi-source process constraint matrix to determine whether there is a constraint violation, and constructing a violation index based on the number of violations, specifically includes: The violation count for each candidate disassembly / assembly sequence in the candidate disassembly / assembly sequence set is calculated using a violation degree function; the violation degree function is expressed as:

[0019] in, This represents the sequence violation score calculated using a pairwise comparison method; Indicates the first in the candidate disassembly / assembly sequence One detachable component; Indicates the first in the sequence One detachable component; Represents the corresponding component in the multi-source process constraint matrix With components The constraints between them; N This indicates the total number of detachable parts in the candidate disassembly / assembly sequence; Count the number of constraint violations. And construct a violation index; the violation index is reflected by a violation penalty function, expressed as:

[0020]

[0021] in, ( ) indicates the candidate disassembly / assembly sequence S The penalty value for violations; To prevent extremely small positive numbers with a denominator of 0; pq This represents the constraint violation weighting coefficient, determined based on the number of constraint violations, and is used to measure the impact of different constraint violations on the system. Indicates candidate disassembly / assembly sequence S The corresponding indicator is whether the constraint has been violated.

[0022] Furthermore, deadlock detection of closed-loop constraints is performed using a deadlock detection function; the deadlock detection function is expressed as:

[0023]

[0024] in, Indicates the deadlock index of the engine compartment disassembly and assembly system; ( ) indicates an indicator function; Represents the first in the system A closed-loop constraint structure; This represents the set of constraint edges contained in the closed-loop structure; The first in the multi-source process constraint matrix Line 1 The elements of the column represent component nodes. With component nodes The overall assembly and disassembly sequence constraints between them.

[0025] Furthermore, it also includes: identifying symmetrical components in the power compartment system: if the spatial distance between two components is less than a preset similarity threshold, then the two components are identified as symmetrical components. Set component nodes With component nodes The spatial position vectors are respectively and Then their symmetry relationship can be determined by the following function:

[0026] And define the symmetry function:

[0027] in, Representing component nodes With component nodes The symmetrical relationship between them; Representing component nodes With component nodes Spatial distance between; vector Representing component nodes Spatial position coordinates; vector Representing component nodes Spatial position coordinates; symbols Represents Euclidean distance; parameters This represents the symmetry determination threshold; if the spatial distance between two components is less than the preset similarity threshold, then the two components are determined to be symmetrical.

[0028] Furthermore, the optimal candidate disassembly / assembly sequence is selected from the candidate disassembly / assembly sequence set using a candidate disassembly / assembly sequence optimization objective function; the candidate disassembly / assembly sequence optimization objective function is expressed as:

[0029]

[0030] in, ( ) represents the objective function for optimizing the candidate disassembly / assembly sequence; ( ) represents the penalty value for violations in the candidate disassembly / assembly sequence; 1 represents ( The weighting coefficients of ) Indicates the deadlock index of the engine compartment disassembly and assembly system; 2 indicates Weighting coefficients; ( ) represents the disassembly / assembly path cost function; 3 indicates ( The weighting coefficients of ) Indicates detachable components in the candidate disassembly / assembly sequence. Disassembly costs; N This indicates the total number of detachable components in the candidate disassembly / assembly sequence.

[0031] As can be seen from the above technical solution, compared with the prior art, the present invention discloses a method for modeling and topology verification of the power compartment disassembly and assembly process based on a multi-source constraint matrix, which has the following beneficial effects: (1) Enhanced Multi-Source Constraint Unified Modeling and Disassembly / Assembly Logic Expression Capabilities: Traditional special vehicle power compartment disassembly / assembly process planning often relies on maintenance technical manuals or maintenance personnel experience, making it difficult to systematically describe the structural dependencies, spatial occlusion relationships, and maintenance process specifications of the internal components of the power compartment. This results in the disassembly / assembly logic being difficult to express digitally and to be automated for analysis and verification. This invention establishes a power compartment component connection graph model, abstracting each structural component inside the power compartment into graph structure nodes and the assembly connection relationships between components into graph structure edges, thereby forming a unified structural relationship expression model. On this basis, structural constraints, spatial constraints, and maintenance process constraints are further extracted, and a multi-source process constraint matrix is ​​constructed through a unified matrix expression form. This achieves a unified expression of constraints from different sources in the same mathematical model, enabling the assembly dependencies, spatial interference relationships, and maintenance operation specifications in the power compartment disassembly / assembly process to be systematically described and quantified, thereby significantly improving the digital modeling capability of the special vehicle power compartment disassembly / assembly logic.

[0032] (2) Topology verification and deadlock detection mechanisms improve the feasibility and planning reliability of candidate disassembly and assembly sequences: In the power compartment system of special vehicles, due to the possible multi-level dependencies and cyclic constraint structures between components, traditional methods based on simple topology sorting are prone to generating candidate disassembly and assembly sequences that do not meet the constraints, and even the problem of cyclic dependencies causing disassembly and assembly paths to be unexecutable. Based on generating a candidate set of disassembly and assembly sequences, this invention introduces a recursive topology verification algorithm. By traversing the candidate disassembly and assembly sequences step by step and combining the multi-source process constraint matrix, the feasibility of each step is judged. At the same time, a violation evaluation index is constructed to quantitatively evaluate the disassembly and assembly sequence that violates the constraints. In addition, this invention further introduces a symmetrical component structure identification and closed-loop constraint deadlock detection mechanism. By analyzing the spatial structure characteristics and cyclic paths in the component connection diagram, potential symmetrical structures and closed-loop dependencies in the system are identified. Deadlock constraints are removed through sequence adjustment or auxiliary operations, thereby effectively avoiding logical conflicts in the disassembly and assembly process and improving the feasibility and stability of candidate disassembly and assembly sequence planning.

[0033] (3) Optimization of candidate disassembly and assembly sequences and improvement of maintenance operation efficiency: In traditional disassembly and assembly planning, the disassembly and assembly sequence often lacks a systematic evaluation mechanism, making it difficult to quantitatively compare different disassembly and assembly paths and select the optimal solution. This invention establishes a comprehensive evaluation model for candidate disassembly and assembly sequences, unifying and integrating factors such as violation index, deadlock detection results, and disassembly and assembly path costs, constructing a multi-objective evaluation function, and systematically evaluating and optimizing candidate disassembly and assembly sequences to obtain candidate disassembly and assembly sequences that meet all constraints and have the optimal operation cost. At the same time, after obtaining the optimal candidate disassembly and assembly sequence, it is converted into a standardized maintenance operation procedure, and supplemented with operation instructions, tool requirements, and safety tips in conjunction with maintenance technical documents, so that the generated disassembly and assembly process not only meets the logical sequence requirements but can also directly guide actual maintenance operations, thereby improving maintenance operation efficiency and reducing operational risks.

[0034] (4) Enhanced Digitalization and Intelligence in Special Vehicle Maintenance Processes: This invention achieves a fully digital description of the entire process, from structural information acquisition and constraint relationship modeling to candidate disassembly / assembly sequence generation and sequence optimization, by constructing a power compartment component connection diagram model, a multi-source process constraint matrix, and a topology verification and optimization mechanism. Compared to traditional disassembly / assembly planning methods that rely on manual experience, this invention can automatically analyze the structural relationships and disassembly / assembly logic within the power compartment and generate executable disassembly / assembly process schemes that meet various constraints. This not only improves the efficiency of candidate disassembly / assembly sequence planning but also enhances the standardization and repeatability of the maintenance process. Furthermore, the method of this invention can be implemented without additional modifications to existing vehicle structures or monitoring systems, has good engineering applicability, and can be widely applied to the disassembly / assembly operation planning, maintenance simulation, and digital maintenance support systems of special vehicle power systems.

[0035] In summary, this invention forms a complete technology chain encompassing "structural information parsing—multi-source constraint modeling—candidate disassembly / assembly sequence generation—topology verification and deadlock detection—sequence optimization—standardized process generation." Compared to existing methods, this invention has significant advantages in disassembly / assembly logic expression capabilities, candidate disassembly / assembly sequence planning efficiency, and maintenance operation executability. It can effectively improve the planning efficiency and reliability of power compartment maintenance operations for special vehicles, providing efficient and reliable technical support for maintainability design and digital maintenance assurance of special vehicles. Attached Figure Description

[0036] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0037] Figure 1 This is a schematic diagram of the process for modeling and topology verification of the power compartment disassembly and assembly procedure based on a multi-source constraint matrix provided in an embodiment of the present invention. Figure 2 This is a schematic diagram of the connection model of the power compartment components provided in the embodiment of the present invention; Figure 3 This is a schematic diagram illustrating the construction of multi-source process constraint matrix and the fusion of constraint relationships provided in an embodiment of the present invention; Figure 4 This is a schematic diagram of the sequence screening and statistical results during the candidate disassembly and assembly sequence planning process provided in this embodiment of the invention; Figure 5 This is a schematic diagram illustrating the proportion of multi-source constraints provided in an embodiment of the present invention. Detailed Implementation

[0038] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0039] like Figures 1 to 3 As shown, the method for modeling and topology verification of the power compartment disassembly and assembly process based on a multi-source constraint matrix provided by this invention mainly includes the following steps: S1. Obtain component structure information, assembly relationship information and maintenance operation information of the power compartment system, and construct a component connection diagram model to describe the assembly structure of the power compartment system based on graph theory; S2. Extract multi-source constraint information from the component connection diagram model, and construct a unified multi-source process constraint matrix by fusing the multi-source constraint information; S3. By initially sorting the disassembly and assembly sequence of each component in the power compartment system, a candidate set of disassembly and assembly sequences containing multiple potential candidate disassembly and assembly sequences is generated; where each candidate disassembly and assembly sequence represents the disassembly and assembly sequence of a component. S4. Perform a recursive topological test on each candidate disassembly and assembly sequence in the candidate disassembly and assembly sequence set. By traversing the candidate disassembly and assembly sequences and comparing them with the multi-source process constraint matrix, determine whether there is a violation of the constraint. Construct a violation index based on the number of violations. S5. Perform closed-loop constraint deadlock detection on the power compartment system and obtain the system deadlock index; S6. Based on the violation index and system deadlock index, and combined with the disassembly and assembly path cost, select the optimal candidate disassembly and assembly sequence from the candidate disassembly and assembly sequence set, and generate the corresponding disassembly and assembly process.

[0040] The power compartment disassembly and assembly process modeling and topology verification method based on a multi-source constraint matrix provided in this invention addresses the problems of complex structural relationships, strong process dependencies, and difficulty in digitally representing disassembly and assembly logic during the maintenance and disassembly of power compartments in special vehicles. By establishing a component connection diagram model, a multi-source process constraint matrix, and a recursive topology verification algorithm, a unified modeling of disassembly and assembly logic relationships is achieved. Combined with deadlock detection and sequence optimization methods, an executable disassembly and assembly process plan is generated, thereby realizing the systematic modeling and optimization of the power compartment disassembly and assembly process for special vehicles, improving the efficiency and accuracy of special vehicle maintenance plan planning.

[0041] It should be noted that the labels S1-S6 above are only for ease of explanation and do not limit the execution order of the steps. Next, each of the above steps will be explained in detail.

[0042] In step S1 above, the component structure information, assembly relationship information and maintenance operation information of the power compartment system are obtained, and a component connection diagram model for describing the assembly structure of the power compartment system is constructed based on graph theory. This invention focuses on the power compartment system of a special vehicle, characterized by its compact integration of multiple components and complex spatial layout. Preferably, a power compartment for a certain type of engine is used as an example to analyze the structure of the core power unit and its auxiliary units. The power compartment mainly includes the engine body, intake manifold, exhaust guide device, oil pump assembly, fuel pump assembly, hydraulic pump assembly, starter generator, accessory transmission housing, cooling ducts, cable supports, sensor supports, fuel lines, oil lines, cable bundle clamps, fireproof and heat-insulating covers, and mounting beams. These components, located within the power compartment, are characterized by limited installation space, complex assembly levels, diverse connection methods, and high requirements for maintenance accessibility. Therefore, they are suitable as the research object for modeling and topology verification of the disassembly and assembly process in this invention.

[0043] Specifically, the system acquires component structural information, assembly relationship information, and maintenance operation information for the power compartment system. Component structural information includes a 3D assembly model of the power compartment, providing information on component geometry, dimensional boundaries, spatial location, and assembly hierarchy. Assembly relationship information includes a product structure tree and parts list, providing component codes and hierarchical relationships between assemblies, subassemblies, and parts. Maintenance operation information includes maintenance process cards, assembly process documents, and maintenance technical data, providing information on component detachment directions, work sequences, and operational constraints. Based on the component structure and assembly relationship information of the power compartment system, the structural parameters, spatial relationships, and assembly connection relationships of each component are obtained. The structural parameters include the geometric parameters, installation positions, and connection methods of key components such as engine assemblies, transmission assemblies, cooling assemblies, and auxiliary equipment assemblies. Specifically, this includes component name, component number, associated assembly, geometric envelope dimensions, installation position coordinates, attitude direction, detachable direction, connection method, number of connecting parts, and information on adjacent components. Assembly connection relationships include assembly forms such as bolted connections, flange connections, pin connections, and snap-fit ​​connections. Preferably, key connection areas such as the connection between the engine body and accessory transmission housing, the accessory transmission housing and the lubricating pump assembly, the lubricating pump assembly and the bracket, the fuel pump assembly and the mounting base, the cooling duct and the bulkhead, and the cable bracket and the mounting beam are analyzed in detail to accurately identify the assembly dependencies between components. For flexible or semi-flexible connectors such as fuel lines, lubricating oil lines, cable bundles, and corrugated hoses, it is also preferable to record their connection port positions, clamping points, and loosening sequence information to avoid ignoring the impact of flexible connections on the disassembly and assembly sequence during modeling.

[0044] Based on the maintenance operation information of the power compartment system, the disassembly and assembly attributes of each assembly connection are determined; these disassembly and assembly attributes include the type of connecting parts, the constraint strength of the connection structure, and the disassembly and assembly difficulty coefficient. To enhance the engineering practicality of the model, this embodiment preferably classifies the power compartment components, including: core functional components (engine body, accessory transmission housing, starter generator, oil pump, fuel pump, etc.); piping and cable components (fuel hoses, oil hoses, cable bundles, connectors, etc.); installation and support components (mounting beams, mounting seats, brackets, and connecting plates, etc.); and protection and airflow guiding components (fireproof covers, heat insulation panels, air ducts, and maintenance covers, etc.). This classification method allows for the assignment of different node and edge types based on the functional attributes of the components during subsequent graph modeling, enabling the graph model to not only express component connection relationships but also reflect connection methods and assembly / disassembly dependencies.

[0045] After extracting the structural data, the raw data is standardized. Since naming conventions and hierarchical granularity may differ in 3D models, process documents, and technical data, this embodiment preferably uses unified encoding for component names, connector names, and mounting point numbers. For example, different descriptions such as "oil pump assembly" and "oil pump component" are uniformly mapped to the same component identifier, and "mounting bolts" and "flange screws" are merged into the connector type field. Simultaneously, location labels are added for components with spatial orientation characteristics. This standardization process ensures consistency between nodes and edges in the subsequent graphical model, avoiding redundant modeling or constraint omissions.

[0046] After obtaining the above information, the power compartment structure is analyzed, and each component is abstracted as a node in a graph structure. Each node is assigned corresponding structural parameters and spatial positional relationships. The assembly and connection relationships between components are abstracted as edges in a graph structure. The disassembly and assembly attributes are used as the weights of the corresponding edges. Based on the above nodes, edges and weights, a component connection graph model of the power compartment system is constructed.

[0047] The component connection diagram model is represented as follows:

[0048] in:

[0049] in, This is a component connection diagram model used to represent the overall assembly relationship model of the power compartment system; Represents a set of component nodes; i Indicates the first i Each component node; Indicates the first Each component node; n Indicates the total number of removable components in the power compartment system; n Represents a set of component nodes The corresponding number inn Each component node; The set of edges representing the connection relationships between components; Representing component nodes With component nodes There are assembly connection relationships between them, including physical assembly relationships, functional connection relationships, or maintenance dependency relationships; W This represents a set of edge attributes used to record disassembly and assembly attributes such as connector type, constraint strength, and disassembly / assembly difficulty; each element in this set of disassembly and assembly attributes... Represented as:

[0050] in, Representing component nodes With component nodes The assembly / disassembly attributes between them; variables Representing component nodes With component nodes The type of connecting parts is used to distinguish different assembly methods such as bolt connections, pin connections, or snap-fit ​​connections; variables Representing component nodes With component nodes The strength coefficient of the connection structure between them is used to describe the constraint strength of the connection structure; variables Representing component nodes With component nodes The disassembly and assembly difficulty coefficient is used to describe the complexity of the operation required to disassemble the connector.

[0051] In actual modeling, connection edges include not only rigid assembly connections but also constraints such as pipe clamping, wiring harness fixing, and protective component installation. For example, the fuel pump assembly and mounting base can be modeled as a bolt connection edge, the fuel rigid pipe and pump body port as a flange connection edge, and the cable harness and cable bracket as a clamping constraint edge. For situations where pipes, wiring harnesses, or protective components must be removed before the main component can be disassembled, auxiliary edges can be added to the drawing to reflect the prerequisite dependencies in the disassembly and assembly process.

[0052] To more accurately reflect the disassembly and assembly characteristics of the power compartment, this embodiment also preferably introduces node-level and edge-level information. Core equipment nodes are located at the upper level, while auxiliary components such as brackets, clamps, and covers are located at the lower level; the connection relationships that determine the main disassembly and assembly sequence are defined as first-level constraint edges, and the connection relationships that only affect local operations are defined as second-level constraint edges.

[0053] In addition, spatial attributes are attached to nodes and edges for subsequent spatial constraint identification. Each node records the component's bounding box, centroid coordinates, and detachable direction; each edge records the connection point location and disassembly operation direction. For example, the starter generator node records its installation coordinates, the fireproof and heat-insulating cover node records its envelope range and shielding relationship, and the pipeline connection edge records the port location and loosening direction.

[0054] After establishing the component connection diagram, the diagram model needs to be checked for consistency. This involves checking for missing nodes using the product structure tree, checking for missing edges using the assembly model's adjacency relationships, and verifying that the connection relationships conform to the actual disassembly and assembly logic using maintenance process documents. For power compartment structures with different configurations, configuration identifiers and version numbers can also be recorded to support the modeling needs of different structural forms.

[0055] In summary, this step achieves standardized extraction of structural information and assembly relationships of power compartment components by analyzing the 3D assembly model of the power compartment, the product structure tree, and maintenance technical data. Furthermore, by constructing a component connection diagram model, it provides a unified and reliable data foundation for subsequent multi-source constraint modeling, candidate disassembly and assembly sequence generation, and topology verification.

[0056] In step S2 above, after completing the acquisition of power compartment component structural information and component connection diagram modeling in step S1, this embodiment further extracts multi-source constraint information from the component connection diagram model and constructs a unified multi-source process constraint matrix by fusing the multi-source constraint information. The main purpose of this step is to uniformly model the various constraint relationships existing in the power compartment disassembly and assembly process based on the component connection diagram model, and to convert constraint information from different sources into a computable matrix form, thereby providing a unified constraint expression for subsequent candidate disassembly and assembly sequence generation and topology verification.

[0057] After completing the component connection diagram modeling, the constraint relationships that may arise during the disassembly and assembly process are analyzed to obtain multi-source constraint information. Since disassembly and assembly operations in special vehicle power compartment systems are typically affected by structural constraints, spatial constraints, and process constraints simultaneously, this embodiment first extracts the above three types of constraint information and constructs corresponding constraint matrices for each.

[0058] First, structural constraint information is extracted, and a structural constraint submatrix is ​​established. .

[0059] Structural constraint information primarily stems from the assembly dependencies between components. For example, the oil pump assembly is bolted to the accessory drive housing, the fuel pump assembly is connected to the housing via a mounting bracket, and the fuel line is connected to the fuel pump via a flange joint. During disassembly, associated connections or dependent components must be disconnected before the target component can be disassembled. Therefore, the assembly dependencies between components can be identified based on the connection relationships in the component connection diagram, and a structural constraint sub-matrix can be established. The structural constraint submatrix The elements in the table are represented as follows:

[0060] in, Represents the structural constraint submatrix; N × N The matrix dimension represents the structural constraint submatrix, and this dimension is consistent with the total number of detachable parts in the system; The first submatrix of the structural constraint matrix Line 1 The elements of the column represent component nodes. With component nodes Weighted structural constraint values ​​between them; Indicates the component nodes during the assembly and disassembly process. Must precede component nodes Disassembled; when Must precede During disassembly, the matrix element takes a value of 1; when there is no structural dependency between the two, the value takes a value of 0; when... Must precede The value is -1 when disassembling.

[0061] Furthermore, in actual power compartment structures, a component is often connected to other components via multiple connectors; for example, a fuel pump assembly may be fixed to a bracket by multiple bolts. Therefore, structural constraints can also be weighted and represented by the number of connectors.

[0062] in, Representing component nodes i With component nodes Weighted structural constraint values ​​between variables; Indicates the first r Each connector is connected to the component node. i With component nodes The contribution value of the structural constraint relationship between them; r Representing component nodes i With component nodes The number of the connecting parts; variables Representing component nodes i With component nodes The total number of connecting parts that exist.

[0063] Spatial constraint information is then extracted, and a spatial constraint submatrix is ​​established. .

[0064] Spatial constraint information includes the assembly / disassembly sequence of components due to spatial obstruction or interference. When one component blocks the disassembly path of another, the obstructing component must be removed first. Specifically, due to the compact internal structure of the power compartment, some components may be obstructed by other components during disassembly. For example, the starter-generator assembly may be located on the side of the accessory drive housing, and its outer side may be obstructed by a fireproof heat shield or cooling duct. Therefore, the obstructing component must be removed before disassembly can proceed. To identify such spatial relationships, this embodiment utilizes component envelope information in the three-dimensional assembly model to analyze the spatial positional relationships between components.

[0065] The degree of spatial occlusion can be represented by an occlusion degree function:

[0066] Among them, variables Representing component nodes i For component nodes Spatial occlusion degree, variable Representing component nodes i For component nodes The occlusion volume includes spatial overlap volume or projected occlusion volume; variables Representing component nodes The overall spatial volume; A spatial constraint submatrix can be established based on the degree of occlusion. The elements in this spatial constraint submatrix are represented as follows:

[0067] in, The first spatial constraint submatrix Line 1 The elements of the column represent component nodes. i For component nodes Spatial constraints; This indicates the threshold for determining spatial occlusion; when the degree of spatial occlusion... Greater than the spatial occlusion determination threshold When, it indicates a component node. i The spatial obstruction of component nodes The disassembly path requires first disassembling the component nodes. i .

[0068] Finally, process constraint information is extracted, and a process constraint submatrix is ​​formed. .

[0069] Maintenance process constraints are derived from maintenance operation specifications and safety requirements, including the operational sequence determined by maintenance safety specifications or maintenance procedures. For example, power must be disconnected or pressure released before disassembling electrical equipment; fuel pressure must be released and valves closed before disassembling fuel system components; and lubricating oil must be drained or oil lines shut off before disassembling lubricating oil system components. These operational requirements are usually clearly specified in maintenance technical documents, and therefore need to be converted into disassembly / assembly sequence constraints. This process constraint sub-matrix The elements in the table are represented as follows:

[0070] in, For process constraint submatrix The Middle Line 1 The elements of the column represent component nodes. i With component nodes Process constraints between them; Obtaining the structural constraint submatrix Spatial constraint submatrix and process constraint submatrix Next, the constraint information from different sources needs to be unified and integrated to construct a multi-source process constraint matrix for the power compartment disassembly and assembly operations. This step constructs the multi-source process constraint matrix in a unified matrix form, enabling the constraint relationships from different sources to be expressed in the same mathematical model. This multi-source process constraint matrix is ​​represented as follows:

[0071] in, This represents a multi-source process constraint matrix for the engine compartment disassembly and assembly operations, used to uniformly describe the process constraint relationships from different sources. This represents the structural constraint submatrix, used to reflect the assembly structural dependencies between various components of the power compartment; This represents a spatial constraint submatrix, used to describe the assembly / disassembly sequence constraints between components caused by spatial occlusion or spatial interference. This represents a process constraint submatrix, used to describe the operational sequence constraints arising from maintenance safety specifications or maintenance process requirements. (Coefficients) , , These represent the weight coefficients of the structural constraint submatrix, spatial constraint submatrix, and process constraint submatrix, respectively, used to adjust the influence of different constraint sources in the comprehensive constraint matrix; The multi-source process constraint matrix The elements in the text are represented as follows:

[0072] in, The first in the multi-source process constraint matrix Line 1 The elements of the column represent component nodes. i With component nodes The overall assembly and disassembly sequence constraints between them; For structural constraint submatrices The Middle Line 1 The elements of the column represent component nodes. i With component nodes Weighted structural constraint values ​​between them; For spatially constrained submatrices The Middle Line 1 The elements of the column represent component nodes. i For component nodes Spatial constraints; For process constraint submatrix The Middle Line 1 The elements of the column represent component nodes. i With component nodes Process constraints between them.

[0073] Using the above method, disassembly and assembly constraints from different sources can be uniformly mapped into the same matrix structure, thus forming a complete multi-source process constraint model. This model can simultaneously describe the assembly dependencies, spatial occlusion relationships, and maintenance operation specifications in the power compartment system, providing basic constraints for subsequent candidate disassembly and assembly sequence generation, topology verification, and deadlock detection.

[0074] In step S3 above, the assembly and disassembly sequences of each component in the power compartment system are initially sorted to form multiple possible sequences. Sequences that clearly do not meet structural constraints are then preliminarily screened using a multi-source process constraint matrix. To improve generation efficiency, a graph search or topological sorting method can be used to generate a candidate set of assembly and disassembly sequences; each candidate sequence represents the assembly and disassembly sequence of a component. Specifically: Assuming the power compartment system has a total of n For detachable components, due to structural, spatial, and technological constraints, not all permutations and combinations can serve as valid candidate disassembly / assembly sequences. Therefore, this embodiment preferably employs a topological sorting method based on constraint matrices to generate an initial candidate set of disassembly / assembly sequences.

[0075] First, based on the multi-source process constraint matrix Construct a component dependency graph. When the matrix elements satisfy:

[0076] Then the component node is considered Must precede component nodes Disassemble and create directed edges in the dependency graph:

[0077] This allows us to form a directed graph representing the assembly / disassembly sequence constraints. :

[0078] in: A set of component nodes; For the constraint matrix The derived set of order constraint edges.

[0079] Based on this, a topological sorting algorithm is used to generate candidate disassembly / assembly sequences that satisfy the basic constraints. Specifically, for any component node... Its in-degree is defined as:

[0080] in: Representing component nodes in-degree; ( The function is an indicator function that takes the value 1 if the condition is true and 0 otherwise. Representing component nodes Must precede component nodes i Disassembly.

[0081] When the in-degree of a node is 0, it means that there are no unmet pre-constraints for that component, so it can be used as a candidate component for the current disassembly / assembly step.

[0082] When generating candidate disassembly / assembly sequences, this embodiment uses a search tree generation method to traverse all possible disassembly / assembly paths. Specifically, the current set of detachable components is denoted as:

[0083] in: Indicates inclusion A collection of components with detachable parts; This represents the set of remaining components that have not yet been added to the sequence prefix; This indicates that the component currently has no unmet prior constraints; This represents the set of currently disassembled parts. When there are multiple disassembleable parts, multiple disassembly / reassembly paths can be generated by expanding the search tree nodes. For example, when the set... When there are multiple nodes, any one of them can be selected as the next disassembly component, thus forming different candidate disassembly and assembly sequence branches.

[0084] To avoid generating a large number of invalid or duplicate sequences, this embodiment further introduces pruning rules to constrain the search process. When a portion of the sequence violates the conditions in the constraint matrix, the expansion of that branch stops. Specifically, for the current set of disassembled parts... If there are components satisfy:

[0085] This indicates the component node. There are still unmet preconditions. In this case, they are not allowed to be added to the current candidate assembly / disassembly sequence, so as to avoid generating a sequence that violates the constraints.

[0086] Through the above topological sorting and search tree expansion process, a candidate set of disassembly and assembly sequences that satisfy the basic constraints can be generated:

[0087] Where: Ω represents the candidate set of disassembly / assembly sequences; This indicates the first candidate disassembly / assembly sequence; M This indicates the number of candidate sequences generated.

[0088] In practical applications, the number of candidate sequences may increase rapidly with the increase of the number of components. Therefore, this embodiment also performs preliminary screening of candidate sequences based on factors such as component importance, constraint weight, or disassembly / assembly path length, so as to retain the most representative set of disassembly / assembly paths.

[0089] Through the above steps, a candidate set of disassembly and assembly sequences for the power compartment system can be generated based on the multi-source constraints, providing a candidate solution space for subsequent recursive topology verification, violation calculation, and optimization of candidate disassembly and assembly sequences.

[0090] In step S4 above, after generating the candidate set of disassembly and assembly sequences in step S3, this embodiment further performs a recursive topological test on each candidate disassembly and assembly sequence in the candidate set. This involves traversing the candidate disassembly and assembly sequences and comparing them with the multi-source process constraint matrix to determine if there are any constraint violations, and constructing a violation index based on the number of violations. Specifically, this includes: Let any candidate disassembly / assembly sequence generated in step S3 be represented as:

[0091] in: Indicates the candidate disassembly / assembly sequence; N Indicates the first in the candidate disassembly / assembly sequence N One component; N This indicates the total number of detachable components in the candidate disassembly / assembly sequence.

[0092] For any component in the candidate disassembly / assembly sequence s It is necessary to determine whether there are still unmet pre-conditions when the component is selected to the current assembly / disassembly position. If a component node exists... satisfy If the component has not been removed before the current sequence position, then it indicates that the current component... s The timing of disassembly did not meet the constraints. Among them, s Representing component nodes i With components s The overall assembly and disassembly sequence constraints between them. Representing component nodes Must precede components s Disassembly.

[0093] Therefore, this embodiment establishes a recursive topological verification function. For candidate disassembly / assembly sequences... S The first in p There are detachable components, and the set of their preceding detached components is defined as follows:

[0094] in, Indicates detachable parts In the prelude One disassembled component; Detachable parts The feasibility determination function can be expressed as:

[0095] in, Indicates detachable components in the candidate disassembly / assembly sequence. The feasibility determination result is whether all preceding constraints are satisfied. Indicates detachable parts Corresponding component nodes ; Representing component nodes i With component nodes The overall assembly and disassembly sequence constraints between them; Representing component nodes Must precede component nodes Disassembly; This indicates a collection of disassembled parts.

[0096] Based on this, a recursive topological verification rule is established for the entire candidate assembly / disassembly sequence. Starting from the first component of the sequence, each component is checked sequentially to see if it meets the constraints; if the current component meets the constraints, the next component is checked; if the constraints are not met, a violation is recorded, and the subsequent components are checked. This yields the sequence. Global topology check function:

[0097] in, Indicates candidate disassembly / assembly sequence The number of recursive topology violations; Indicates detachable components in the candidate disassembly / assembly sequence. The feasibility assessment results. When When the candidate disassembly / assembly sequence satisfies all constraints, it is a topologically feasible sequence; when If this occurs, it indicates that there is a component disassembly sequence in the candidate sequence that violates the constraints; N This indicates the total number of detachable components in the candidate disassembly / assembly sequence.

[0098] Furthermore, to more accurately reflect the deviation between the candidate sequence and the constraint matrix, this embodiment uses a pairwise comparison method to establish a violation function, the expression of which is:

[0099] in, This represents the sequence violation score calculated using a pairwise comparison method; Indicates the first in the candidate disassembly / assembly sequence One detachable component; Indicates the first in the sequence One detachable component; Represents the corresponding component in the multi-source process constraint matrix With components The constraints between them; if This indicates that a component located later should have been placed earlier, thus constituting a violation. This formula can be used to count all violations of the order constraint in the entire candidate sequence.

[0100] To differentiate the impact of different types of violations on the disassembly and assembly operations, this embodiment further introduces a violation penalty function to count the number of times constraints are violated. Furthermore, a violation index for candidate disassembly / assembly sequences is constructed. This violation index reflects the degree of deviation between the candidate disassembly / assembly sequences and the constraints, and serves as the evaluation basis for subsequent sequence optimization. The violation index is embodied by a violation penalty function, expressed as:

[0101]

[0102]

[0103] To clarify the correspondence between sequence positions and component nodes, the following definition is made: ;in, ( ) indicates the candidate disassembly / assembly sequence S The penalty value for violations; Indicates the number of times constraints were violated; To prevent extremely small positive numbers with a denominator of 0; pq This represents the constraint violation weighting coefficient, determined based on the number of constraint violations, and is used to measure the impact of different constraint violations on the system. Indicates candidate disassembly / assembly sequence S The indicator quantity indicating whether the corresponding constraint has been violated; Representing component nodes i With component nodes The overall assembly and disassembly sequence constraints between them; Representing component nodes i In the candidate disassembly / assembly sequence S Position number in; Representing component nodes In the candidate disassembly / assembly sequence S The position number in the matrix. When the multi-source process constraint matrix specifies the component node. i Must precede component nodes If the components are disassembled but the actual sequence is reversed, then a constraint violation is considered to have occurred.

[0104] In actual engine compartment disassembly and assembly, the impact of different types of violations varies. For example, mistakenly starting the generator before removing the fireproof and heat-insulating cover is usually a violation related to spatial obstruction, while disconnecting fuel lines without releasing fuel pressure is a process-related violation with a higher risk. Therefore, this embodiment preferably sets penalty weights based on the source of the constraint. Violations caused by process constraints can be assigned higher weights, violations caused by structural dependencies can be assigned medium weights, and spatial constraints that only affect local operations can be assigned lower weights, making the violation evaluation more consistent with the characteristics of actual maintenance operations.

[0105] Furthermore, during the recursive topology check, the current set of detachable parts can also be updated synchronously. Let the first step be... After the initial inspection, the remaining undisassembled parts are as follows:

[0106] The next set of optional components is:

[0107] This set represents the detachable objects among the remaining components that satisfy all prior constraints after the current partial components have been removed. By dynamically maintaining this set during the recursive verification process, the candidate sequences can be progressively verified, providing a basis for subsequent search backtracking and local sequence correction.

[0108] In summary, this step establishes a recursive topology verification mechanism based on the comprehensive constraint matrix to verify each candidate disassembly / assembly sequence generated in step three, and quantifies the feasibility of the sequence using a violation count function and a violation penalty function. This method can identify disassembly / assembly sequence relationships that violate structural, spatial, and process constraints, and provides evaluation indicators for subsequent symmetrical component identification, closed-loop constraint deadlock detection, and optimal candidate disassembly / assembly sequence generation.

[0109] In step S5 above, after completing the recursive topology check and violation calculation in step S4, this embodiment further performs closed-loop constraint deadlock detection on the power compartment system to obtain the system deadlock index. The main purpose of this step is to identify disassembly and assembly conflicts in the power compartment structure caused by symmetrical installation or cyclic dependency, thereby avoiding unexecutable operation paths in the candidate disassembly and assembly sequence planning process.

[0110] In the power compartment system, some components are typically installed symmetrically, such as symmetrically arranged fuel line supports, dual-sided cooling duct mounting brackets, and paired cable bundle clamps. These components are highly similar in spatial location and connection relationship. If they are not identified during the candidate assembly / disassembly sequence generation process, it can easily lead to duplicate calculations or redundant sequence branches. Therefore, this embodiment first identifies the symmetrical components in the power compartment structure.

[0111] Specifically, the spatial coordinate information and geometric attributes of the components obtained in step S1 are used to compare the spatial position vectors of the component nodes. Let the component nodes... With component nodes The spatial position vectors are respectively and Then their symmetry relationship can be determined by the following function:

[0112] And define the symmetry function:

[0113] in, Representing component nodes With component nodes The symmetrical relationship between them; Representing component nodes With component nodes Spatial distance between; vector Representing component nodes Spatial position coordinates; vector Representing component nodes Spatial position coordinates; symbols Represents Euclidean distance; parameters This represents the symmetry threshold. If the spatial distance between two components is less than the preset similarity threshold, the two components are considered symmetrical. Once symmetrical components are identified, they can be grouped into the same symmetrical group, and this group of components can be treated as equivalent nodes during the candidate assembly / disassembly sequence generation process, thereby reducing the number of candidate sequences and improving subsequent computational efficiency.

[0114] After identifying symmetrical components, it is also necessary to detect any closed-loop constraints that may exist in the powertrain system. Closed-loop constraints typically arise from cyclic dependencies between multiple components, such as components... Disassembly requires removing the components first. , and components Disassembly requires removing the components first. Meanwhile, components It also depends on components In this case, the following cyclic constraint will be formed:

[0115] This type of circular dependency can lead to deadlock during the assembly / disassembly process, meaning there is no assembly / disassembly sequence that satisfies all constraints. Therefore, it is necessary to perform loop closure checks on the constraint graph.

[0116] In this embodiment, the comprehensive constraint matrix constructed in step two is used... Constructing a constraint graph If the matrix elements satisfy If the value is greater than 0, then a directed edge is constructed in the graph. This method yields a directed graph representing the assembly / disassembly dependencies:

[0117] in For a set of component nodes, Let be the set of directed edges formed by constraints.

[0118] To detect the presence of closed-loop structures in the diagram, this embodiment introduces a deadlock detection function to perform the aforementioned closed-loop constraint deadlock detection. This deadlock detection function is used to identify closed-loop constraint structures formed by cyclic dependencies in the power compartment component connection diagram, thereby determining whether the candidate disassembly / assembly sequence contains an unexecutable deadlock state; its expression is:

[0119]

[0120] in, Indicates the deadlock index of the engine compartment disassembly and assembly system. This indicates that there are no circular dependencies in the constraint graph, and the candidate assembly / disassembly sequence can theoretically be obtained through topological sorting; when If this happens, it indicates that there is a closed-loop constraint in the system, and the disassembly and assembly logic needs further adjustment. ( The ) indicates an indicator function that takes a value of 1 when a loop structure is detected, and a value of 0 otherwise. Represents the first in the system h A closed-loop constraint structure; h This represents the set of constraint edges contained in the closed-loop structure; The first in the multi-source process constraint matrix Line 1 The elements of the column represent component nodes. With component nodes The overall assembly and disassembly sequence constraints between them.

[0121] When a closed-loop constraint structure exists in a system, it indicates that the components are interdependent, making the disassembly and assembly process impossible. In this case, the source of the conflict can be identified by analyzing the critical constraint edges in the loop path. For example, in some structures, a cooling duct may obstruct the fuel pump assembly, which in turn is assembled with the duct support, thus forming a local loop constraint. By identifying the critical components in this loop path, the disassembly and assembly sequence can be readjusted or auxiliary steps can be introduced, such as first removing the support fixtures or releasing the piping connections, thereby eliminating the loop dependency.

[0122] Through the above-mentioned process of symmetric component identification and closed-loop constraint detection, potential disassembly deadlock problems can be identified before further optimization of candidate disassembly sequences, and redundant sequences caused by symmetric structures can be reduced, thereby improving the efficiency and reliability of candidate disassembly sequence planning.

[0123] In summary, this step, by identifying symmetrical components in the power compartment structure and detecting cyclic dependencies in the integrated constraint diagram, enables the early identification and handling of potential disassembly / assembly deadlock issues, providing a foundation for subsequent optimization of candidate disassembly / assembly sequences and generation of executable procedures.

[0124] In step S6 above, based on the violation index and system deadlock index, and combined with the disassembly and assembly path cost, the optimal candidate disassembly and assembly sequence is selected from the candidate disassembly and assembly sequence set, and the corresponding disassembly and assembly procedure is generated. The main purpose of this step is to select candidate disassembly and assembly sequences that meet all constraints and have high execution efficiency from the candidate disassembly and assembly sequence set Ω, and convert them into standardized disassembly and assembly procedures that can directly guide maintenance operations.

[0125] Specifically, the optimal candidate disassembly / assembly sequence is selected from the candidate disassembly / assembly sequence set using a candidate disassembly / assembly sequence optimization objective function. This objective function comprehensively considers the violation penalty, deadlock risk, and disassembly / assembly path cost of the candidate disassembly / assembly sequence to perform multi-objective evaluation and optimization screening, thereby obtaining the candidate disassembly / assembly sequence that satisfies all constraints and has the optimal operation cost. The candidate disassembly / assembly sequence optimization objective function is expressed as follows:

[0126]

[0127] in, ( ) represents the objective function for optimizing the candidate disassembly and assembly sequence, which is used to find the optimal candidate disassembly and assembly sequence under the constraints. ( ) represents the penalty value for violations in the candidate disassembly / assembly sequence; 1 represents ( The weighting coefficients of ) Indicates the deadlock index of the engine compartment disassembly and assembly system; 2 indicates Weighting coefficients; ( ) represents the disassembly / assembly path cost function; 3 indicates ( The weighting coefficients of ) Indicates detachable components in the candidate disassembly / assembly sequence. The disassembly cost. For example, the operating cost can be set higher for parts that require special tools to disassemble; and relatively lower for simple disassembly parts, such as protective covers or bracket bolts. N This indicates the total number of detachable components in the candidate disassembly / assembly sequence.

[0128] After obtaining the comprehensive evaluation function, the optimal candidate disassembly / assembly sequence with the smallest evaluation value can be obtained by calculating all sequences in the candidate set Ω of disassembly / assembly sequences:

[0129] in, Ω represents the optimal candidate disassembly / assembly sequence; Ω represents the candidate set of disassembly / assembly sequences. Represents a sequence The comprehensive evaluation function is as follows. This sequence, while satisfying structural, spatial, and technological constraints, can complete the engine compartment disassembly and assembly task with relatively low operating costs.

[0130] After determining the optimal candidate disassembly and assembly sequence, this embodiment further generates executable maintenance procedures based on the component order in the sequence. Specifically, the disassembly operation of each component in the sequence is converted into standardized procedure steps, and supplemented with operation instructions based on maintenance technical data. For example, when a step in the sequence is to disassemble the oil pump assembly, the corresponding procedure can be generated: first, loosen the fixing bolts; second, disconnect the oil pipeline connection; third, remove the relevant cable interface; and finally, remove the oil pump assembly from the accessory transmission housing.

[0131] Furthermore, to improve the safety and standardization of maintenance operations, operational prompts can be added simultaneously when generating disassembly and assembly procedures, such as tool type, operating direction, and safety precautions. For example, a prompt to release fuel system pressure before disconnecting fuel lines, and a prompt to disconnect power before disconnecting electrical equipment. In this way, the generated disassembly and assembly procedures not only meet logical sequence requirements but also conform to actual maintenance operation standards.

[0132] Finally, the generated disassembly and assembly procedures are output in a standardized format to form a complete engine compartment maintenance and disassembly / assembly plan. This plan can be used to guide on-site maintenance operations and also as input data for maintenance analysis and simulation.

[0133] In summary, this step optimizes and filters candidate disassembly and assembly sequences by establishing a comprehensive evaluation model, and converts the optimal candidate disassembly and assembly sequence into an executable maintenance procedure, thereby completing the modeling and inspection process of the power compartment disassembly and assembly procedure and realizing the automated planning and standardized generation of the power compartment disassembly and assembly operation process.

[0134] Through the above steps, this invention establishes a complete modeling and inspection process for the engine compartment disassembly and assembly, achieving a fully digital description of the entire process from structural information acquisition, constraint relationship modeling, candidate disassembly and assembly sequence generation to sequence optimization. Compared with traditional methods that rely on manual experience to formulate disassembly and assembly procedures, this invention enables structured modeling and automated analysis of the engine compartment disassembly and assembly process, improving the efficiency and reliability of candidate disassembly and assembly sequence planning, while reducing operational risks caused by unreasonable disassembly and assembly sequences. This provides an effective technical means for engine compartment maintenance operation planning and digital maintenance applications.

[0135] To verify the effectiveness, accuracy, efficiency, and practicality of the proposed multi-source constraint matrix-based modeling and topology verification method for power compartment disassembly and assembly in engineering applications, and to address the industry pain points of complex structural relationships, strong process dependencies, and difficulty in digitally expressing disassembly and assembly logic in the maintenance and disassembly of complex equipment power compartments, this verification takes the power compartment of a certain type of large special vehicle as the research object and conducts full-process, multi-dimensional experimental verification. This type of power compartment is a typical compact integrated complex equipment compartment with multiple components. It integrates core functional components such as the engine body, accessory transmission housing, lubricating oil pump assembly, fuel pump assembly, and starter generator, as well as auxiliary components such as cooling ducts, fireproof heat insulation covers, fuel / lubricating oil pipelines, various cable brackets, and mounting beams. It has significant characteristics such as limited installation space, complex assembly levels, diverse connection forms, high maintenance accessibility requirements, and numerous disassembly and assembly constraints. Its maintenance and disassembly process is subject to strict structural, spatial, and process constraints, making it a typical applicable scenario for the method of this invention. The verification results have strong engineering reference value and generalizability.

[0136] This verification strictly followed the six core steps proposed in this invention: "structural information acquisition and connection graph modeling → multi-source constraint extraction and matrix construction → generation of candidate disassembly and assembly sequences → recursive topology verification and violation calculation → symmetric component identification and closed-loop constraint deadlock detection → candidate disassembly and assembly sequence optimization and executable process generation." Through quantitative statistical modeling results, constraint information extraction effects, sequence screening and optimization data, and multi-dimensional performance comparisons with traditional manual experience planning methods and simple topology sorting methods, the technical superiority and engineering practicality of the method proposed in this invention in digital modeling, sequence planning, and logical verification of the power compartment disassembly and assembly process were comprehensively verified.

[0137] I. Experimental Foundation and Preliminary Preparations: (I) Basic characteristics of the experimental subjects: The power compartment of a certain type of large special vehicle being verified this time, as the core unit of the vehicle's power system, has its internal components laid out according to the principle of "functional integration and compact space," encompassing four main categories of core components: First, core assembly functional components, including the engine block, accessory transmission housing, starter generator, lubricating oil pump assembly, fuel pump assembly, and hydraulic pump assembly; second, pipeline and cable components, including fuel hoses / rigid hoses, lubricating oil lines, cable bundles, connectors, and fixing clamps; third, installation and support components, including compartment mounting beams, component mounting seats, various brackets, connecting plates, and vibration isolators; and fourth, protection and airflow guiding components, including fireproof and heat insulation covers, heat insulation plates, cooling ducts, air ducts, and maintenance covers. The components are connected by various methods such as bolts, flanges, pins, clips, and clamps, resulting in close assembly dependencies. Furthermore, some components are constrained by space obstructions and pre-process requirements, making it easy for manual planning of candidate disassembly and assembly sequences to lead to logical errors, deadlock conflicts, and low operational efficiency.

[0138] (II) Acquisition of Experimental Data Sources: To ensure the accuracy of modeling and constraint extraction, this verification employed a multi-source information fusion approach to comprehensively acquire information related to the power compartment structure, assembly, and maintenance. The information sources included: 3D assembly model: Provides the geometric shape, dimensional boundaries, spatial coordinates, assembly level, and spatial interference relationship between components; Product structure tree: Provides a hierarchical structure of component codes, affiliations, assemblies-subassemblies-parts, clearly defining the subordinate relationships of each component; Parts list: Provides basic attributes such as part name, specifications, quantity, and material, and provides a basis for unified part coding; Maintenance process cards and assembly process documents: provide process information such as the direction in which parts can be disassembled, standard operating sequence, prerequisite operating conditions, and tool usage requirements; Maintenance Technical Manual: Provides core requirements such as safety restrictions for component disassembly and assembly, process specifications (such as pressure release, power disconnection, etc.), and disassembly and assembly guidelines for fault repair.

[0139] By standardizing and integrating the above multi-source information, the basic information of 24 key detachable components was extracted, laying the data foundation for subsequent component connection diagram modeling and multi-source constraint matrix construction.

[0140] (III) Experimental hardware and software environment: The hardware environment used in this verification was: Intel Core i7-12700H processor, 32GB DDR5 memory, and 512GB NVMe SSD; the software environment was: Python 3.9 (with NetworkX graph theory library and NumPy numerical computation library), UGNX 12.0 (3D model analysis), and Matlab R2022b (matrix calculation and algorithm simulation). All algorithms in the experiment were implemented based on the mathematical model and process proposed in this invention, ensuring the independence of the verification process and the authenticity of the results.

[0141] II. Implementation and Result Analysis of the Entire Experiment: This verification was strictly implemented according to the six-step core process of this invention. Quantitative data statistics and analysis were conducted at each stage to comprehensively verify the effectiveness of the method. The implementation process and results of each stage are as follows: (I) Phase 1: Acquisition of structural information of engine compartment components and modeling of component connection diagrams: Based on the results of multi-source information analysis, standardized data processing was performed on 24 key components, including unified coding of component names and numbers (eliminating naming differences between different information sources), calibration of geometric envelope dimensions and installation position coordinates, and sorting out connection methods and adjacent component lists. At the same time, special attention was paid to information such as connection ports, clamping fulcrums, and loosening sequences of flexible / semi-flexible connectors such as fuel lines and lubricating oil lines, to avoid ignoring the impact of flexible connections on the disassembly and assembly sequence during modeling.

[0142] Based on this, a graph theory modeling method is adopted to abstract 24 key components into a graph node set V, and 41 assembly connection relationships between components (including 7 categories such as bolts, flanges, clamps, and functional accessories) into a set of connection edges with attributes E. Each edge is assigned attributes such as connection type, constraint strength, disassembly capability, and disassembly direction restriction, constructing a component connection graph model G=(V,E,W) (W is the set of connection attributes). At the same time, node hierarchy and edge hierarchy information are introduced into the model: core equipment nodes are the upper layer, and auxiliary component nodes such as brackets and clamps are the lower layer; strong constraint edges that determine the main assembly and disassembly sequence are first-level constraint edges, and weak constraint edges that only affect local operations are second-level constraint edges. Spatial attributes (bounding box, centroid coordinates, connection point position, etc.) are added to all nodes and edges to provide support for subsequent spatial constraint identification.

[0143] Finally, the component connection diagram model was checked for consistency using the product structure tree, 3D assembly model, and maintenance process cards to ensure there were no missing nodes, omitted edges, or logically incorrect connections. Version identifiers for the component assembly configurations were also recorded to guarantee the model's adaptability. This stage successfully constructed a power compartment component connection diagram model that conforms to engineering realities and has complete information, providing a unified structural foundation for subsequent multi-source constraint extraction.

[0144] (II) Phase Two: Extraction of Multi-Source Constraint Information and Construction of Multi-Source Process Constraint Matrix Based on the component connection diagram model, the system extracts three types of core constraint information during the disassembly and assembly of the power compartment: structural constraints, spatial constraints, and process constraints. Corresponding constraint matrices are constructed for each, and finally, they are integrated into a unified multi-source process constraint matrix through weight coefficients, thereby realizing the mathematical and unified expression of multiple constraints.

[0145] Structural constraint extraction: Based on the edge relationships of the component connection graph, the assembly dependency relationships between components are identified, and a structural constraint matrix Cs is constructed. A total of 36 structural constraints are extracted, accounting for 53.7% of the total constraints. These constraints mainly include "the component can only be disassembled after the connecting parts are removed" and "the dependent component must be disassembled before the main component". Spatial constraint extraction: Utilizing the 3D envelope information of the component, through the occlusion function... Calculate the degree of spatial occlusion between components, set the occlusion judgment threshold θ=0.2, and construct the spatial constraint matrix. A total of 19 spatial constraints were extracted, accounting for 28.4% of the total constraints. These mainly include constraints such as "the occluding component must be disassembled before the occluded component" and "avoid spatial interference in the disassembly path". Process constraint extraction: Analyze maintenance technical manuals and process documents, convert safety operating procedures into disassembly and assembly sequence constraints, and construct a process constraint matrix. C t A total of 12 process constraints were extracted, accounting for 17.9% of the total constraints. These constraints mainly include "pressure must be released before disassembling fuel system components" and "power must be disconnected before disassembling electrical equipment".

[0146] Subsequently, weighting coefficients were set as α=0.5 (structural constraint), β=0.3 (spatial constraint), and γ=0.2 (process constraint), and the formula was used... By integrating the three types of constraint matrices, a multi-source process comprehensive constraint matrix C is formed, resulting in 67 valid disassembly and assembly sequence constraint relationships. This achieves a unified mathematical expression for constraints from different sources and of different types, providing core constraint basis for subsequent candidate disassembly and assembly sequence generation and topology verification.

[0147] (III) Stage Three: Generation of the Candidate Set of Disassembly and Assembly Sequences Based on the multi-source process comprehensive constraint matrix C, a directed graph of component dependency relationships is constructed. (The set of sequential constraint edges derived from the constraint matrix) is used to generate a candidate set of disassembly and assembly sequences that satisfy the basic disassembly and assembly logic by combining a topological sorting algorithm based on the constraint matrix with search tree expansion and pruning rules.

[0148] First, the set of detachable components for each step is determined by calculating the in-degree of nodes (an in-degree of 0 indicates no unmet pre-order constraints). A search tree expansion method is then used to traverse all possible disassembly / assembly paths. Simultaneously, a strict pruning rule is introduced: if a portion of the sequence violates the constraint matrix conditions, expansion of that branch is stopped to avoid generating invalid or duplicate sequences. Experimental results show that for the power compartment system with 24 key components, a total of 326 candidate disassembly / assembly sequences are generated in this stage. This is a significant reduction in the number of permutations and combinations compared to the unconstrained case, significantly reducing the search space for subsequent sequence verification and optimization, and improving the algorithm's execution efficiency.

[0149] (iv) Phase Four: Recursive Topology Check and Violation Calculation A full-process recursive topology check was performed on 326 candidate disassembly and assembly sequences. By traversing the component order in the sequence and combining the multi-source process comprehensive constraint matrix C, the feasibility judgment function was constructed to determine whether the component disassembly timing meets the preceding constraint requirements step by step. The violation calculation function is used to quantitatively evaluate the topological feasibility of each sequence.

[0150] During the inspection process, if any component still has unmet prerequisite constraints during disassembly, a violation is recorded and the violation count is determined using a violation count function. Count the total number of violations and use the violation penalty function. The overall violation penalty value was calculated (penalty weights were set according to constraint type, with process constraints having the highest penalty, followed by structural constraints, and spatial constraints having the lowest penalty, reflecting the risk characteristics of actual engineering). Experimental results show that among the 326 candidate sequences, 188 sequences have varying degrees of constraint conflicts (violation count ≥ 1), and only 138 sequences satisfy all constraints or have extremely low violation rates, making them feasible disassembly / assembly paths. This stage successfully achieved accurate screening of candidate sequences, providing a high-quality candidate solution space for subsequent deadlock detection and sequence optimization.

[0151] (V) Phase 5: Symmetrical component structure identification and closed-loop constraint deadlock detection: To address the assembly characteristics of symmetrical components and potential closed-loop constraint deadlock issues in the power compartment structure, this phase focuses on symmetrical component identification, closed-loop constraint detection and removal, further optimizing candidate sequences, and eliminating logical conflicts during disassembly and assembly.

[0152] Symmetrical component structure identification: Utilizing the spatial coordinate information of components, a symmetry determination function is used. (ε is the symmetry judgment threshold, which is set to 50mm in this experiment). Three sets of symmetrically installed components were successfully identified in the power compartment structure, including left and right fuel line brackets, double-sided cable clamps, and double-sided cooling duct fixing components. Each set of symmetrical components was divided into the same structural group and regarded as equivalent nodes, which effectively reduced the repeated calculations in subsequent sequence optimization and improved the algorithm efficiency. Closed-loop constraint deadlock detection and resolution: Based on the constraint directed graph Gc constructed from the multi-source process comprehensive constraint matrix C, a closed-loop detection function is used to resolve the deadlock. The experiment successfully identified one closed-loop deadlock in the detection graph, caused by a circular dependency formed by the cooling duct, fuel pump assembly, and their mounting bracket (the cooling duct obstructs the fuel pump assembly, there is an assembly dependency between the fuel pump assembly and the mounting bracket, and the mounting bracket is spatially constrained with the cooling duct). To address this closed-loop constraint, this invention analyzes the critical constraint edges of the circular path and introduces auxiliary disassembly / assembly steps. By prioritizing the removal of the fixed connection of the mounting bracket, the circular dependency is successfully resolved, avoiding deadlock issues during disassembly / assembly and ensuring the practical operability of all feasible sequences.

[0153] (vi) Phase Six: Optimization of Candidate Disassembly / Assembly Sequences and Generation of Executable Procedures For the 138 feasible candidate disassembly and assembly sequences after topology verification and deadlock detection, a multi-index comprehensive evaluation model was established to screen the optimal candidate disassembly and assembly sequence and convert the optimal sequence into a standardized disassembly and assembly procedure that can directly guide on-site maintenance.

[0154] Comprehensive evaluation model construction: Using the violation penalty value P(S), deadlock index D(G_c), and disassembly / assembly path cost L(S) as core evaluation indicators, a comprehensive evaluation function is constructed. Among them, the cost of disassembly and assembly path To calculate the operational costs of component disassembly (including time, tools, and safety risks), weighting coefficients ω1=0.4, ω2=0.3, and ω3=0.3 are set to achieve a weighted comprehensive evaluation of multiple indicators. Optimal sequence selection: The 138 feasible sequences were quantitatively calculated using a comprehensive evaluation function. With the minimum evaluation value as the optimization objective, one optimal candidate disassembly and assembly sequence was selected. This sequence has the shortest disassembly and assembly path, the lowest operation complexity, and the lowest maintenance cost under the premise of meeting all structural, spatial, and process constraints, which meets the actual operation requirements of the project. Executable process generation: Each component disassembly operation in the optimal candidate disassembly and assembly sequence is converted into a standardized maintenance procedure. The specific operation content, tool type, and operation direction are supplemented with the maintenance technical manual, and necessary safety precautions (such as pressure release, power disconnection, collision prevention, etc.) are added to form a complete power compartment disassembly and assembly procedure plan that can directly guide on-site operations. This realizes the transformation from digital sequence to engineering procedure and completes the full-process implementation of the method of this invention.

[0155] III. Statistical Analysis of Experimental Quantitative Results: The core quantization results at each stage of this validation are presented in tabular form to intuitively reflect the modeling effectiveness, constraint extraction effectiveness, algorithm efficiency, and sequence planning quality of the method of this invention. The specific statistical results are as follows: Table 1. Quantitative Statistics of Power Cabin System Structural Modeling and Multi-Source Constraint Extraction

[0156] Table 2 Constraint Matrix of Multi-Source Processes for Typical 8 Core Components

[0157] Note: Matrix element C ij =1 indicates that component i must be disassembled before component j (there is a disassembly / assembly sequence constraint), C ij =0 indicates no direct constraint relationship; rows represent pre-disassembly components and columns represent post-disassembly components, intuitively reflecting the constraint relationship between core components.

[0158] Table 3. Time consumption statistics for each calculation stage of the method of the present invention.

[0159] Table 4. Multi-dimensional performance comparison between the method of the present invention and the traditional candidate assembly / disassembly sequence planning method.

[0160] To more intuitively demonstrate the effectiveness of the method of this invention in candidate disassembly / assembly sequence screening and constraint modeling, the experimental data were visualized and analyzed, and statistical charts of candidate disassembly / assembly sequence screening and proportions of multi-source constraints were drawn, such as... Figure 4 and Figure 5 As shown.

[0161] like Figure 4 As shown, the method of the present invention obtains a total of 326 candidate disassembly and assembly sequences in the candidate sequence generation stage. After recursive topology verification and violation evaluation, 138 feasible candidate disassembly and assembly sequences are selected, and finally one optimal candidate disassembly and assembly sequence is determined. This shows that the method of the present invention can effectively reduce the search space and improve the efficiency of sequence planning.

[0162] like Figure 5As shown, structural constraints, spatial constraints, and process constraints account for 53.7%, 28.4%, and 17.9% of the multi-source process constraint matrix, respectively, indicating that the power compartment assembly and disassembly process is simultaneously affected by multiple types of constraints. This invention achieves the fusion expression of constraint information from different sources through a unified constraint matrix model, thereby improving the accuracy and reliability of candidate assembly and disassembly sequence planning.

[0163] IV. Key experimental conclusions and analysis of methodological advantages: Through full-process experimental verification of the power compartment of a certain type of large special vehicle, and by combining the quantitative results of each stage with multi-dimensional comparisons with traditional methods, the following key conclusions can be drawn, fully demonstrating the technical superiority and engineering practicality of the method of this invention: (i) Digital and unified modeling of the complex engine compartment disassembly and assembly logic has been achieved: This invention transforms the complex structural assembly relationships, spatial occlusion relationships, and maintenance process constraints of the power compartment into a standardized graph theory model and mathematical matrix through a component connection diagram model and a multi-source process constraint matrix. This solves the industry pain points of traditional methods, such as the difficulty in digitally representing disassembly and assembly logic and the inability to uniformly utilize scattered constraint information. The modeling process fully considers practical engineering factors such as rigid / flexible components, strong / weak constraints, and spatial attributes, ensuring the completeness and accuracy of the model and laying a core foundation for the digital planning of complex equipment disassembly and assembly processes.

[0164] (ii) Significantly improved the efficiency and quality of candidate assembly / disassembly sequence planning: This invention utilizes a sequence generation method combining "topological sorting + search tree expansion + pruning rules" to effectively control the size of candidate sequences, reducing the number of candidate sequences from 812 in traditional topological sorting to 326, a reduction of 59.8%. Simultaneously, through recursive topological verification, symmetric component identification, and deadlock detection and resolution, it achieves precise screening and optimization of candidate sequences, increasing the number of feasible sequences by 14.0%. Furthermore, all feasible sequences achieve zero violations and no deadlocks, with sequence quality far exceeding traditional methods. The entire planning process takes only 4.3 seconds, a 62.9% reduction compared to traditional topological sorting and a significant 95.5% reduction compared to manual experience-based planning, substantially improving the efficiency of candidate assembly / disassembly sequence planning.

[0165] (iii) It completely solves the deadlock and conflict problems that are prone to occur in traditional methods: To address the common closed-loop constraint deadlock problem in the assembly and disassembly of complex equipment, this invention proposes a specialized deadlock detection function and constraint resolution strategy. By analyzing the critical constraint edges of the loop path and introducing auxiliary assembly and disassembly steps, it successfully resolved one closed-loop constraint deadlock discovered in the experiment, achieving 100% elimination of deadlock risk. Traditional manual experience-based planning methods cannot identify deadlocks in advance, and traditional topology sorting methods can only detect deadlocks but cannot effectively resolve them. This invention fundamentally solves this technical problem, ensuring the practical feasibility of the assembly and disassembly process.

[0166] (iv) It has achieved end-to-end implementation from digital sequence to engineering process: This invention does not simply generate digital candidate disassembly and assembly sequences, but after selecting the optimal sequence, it converts it into standardized and engineered executable disassembly and assembly procedures. It supplements the core information required for on-site operations, such as specific operation content, tool requirements, and safety precautions. The generated procedure plan can directly guide on-site maintenance operations, realizing end-to-end implementation from algorithm model to engineering application. It solves the problem of traditional digital methods being "heavy on theory and light on application" and has strong engineering practicality.

[0167] (v) Possesses good versatility and scalability: The method of this invention is based on graph theory models and matrix calculations. It does not depend on specific power compartment models. By adjusting parameters such as component nodes, constraint relationships, and weight coefficients, it can be adapted to power compartments of different types and scales of complex equipment. It can even be extended to the planning of complex compartment / equipment disassembly and assembly processes in aviation, aerospace, and shipbuilding fields. It has good versatility and scalability and has broad industry application prospects.

[0168] V. Summary of Experimental Verification: This study, using the power compartment of a certain type of large special vehicle as the research object, comprehensively verified the entire process and multiple dimensions of the six core steps of the power compartment disassembly and assembly process modeling and topology verification method based on multi-source constraint matrix of the present invention. Through quantitative statistics, algorithm simulation and method comparison, the effectiveness, accuracy and efficiency of the method of the present invention in complex structure digital modeling, unified expression of multi-source constraints, efficient planning of candidate disassembly and assembly sequences, accurate detection and resolution of logical conflicts, and generation of engineering processes were fully verified.

[0169] Experimental results show that the method of the present invention effectively solves the core problems of complex structural relationships, strong process dependence, and difficulty in digitally expressing disassembly and assembly logic in the maintenance and disassembly of power compartments of special vehicles. It significantly improves the efficiency and quality of disassembly and assembly process planning, completely eliminates the risk of deadlock and conflict, and the generated standardized disassembly and assembly process can directly guide on-site operations, providing reliable technical support for digital maintenance assurance, maintainability design, and intelligent maintenance of power compartments of complex equipment.

[0170] The method of this invention combines theoretical innovation and engineering practicality, breaking through the technical bottleneck of traditional candidate disassembly and assembly sequence planning methods. It has important application value in the fields of high-end equipment manufacturing and maintenance such as special vehicles, aviation, aerospace, and ships, and can provide new technical ideas and methodological references for the digital, automated, and intelligent planning of complex equipment disassembly and assembly processes in the industry.

[0171] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0172] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for modeling and topology checking of power cabin disassembly process based on multi-source constraint matrix, characterized in that, Includes the following steps: Acquire component structure information, assembly relationship information, and maintenance operation information of the power compartment system, and construct a component connection diagram model to describe the assembly structure of the power compartment system based on graph theory; Multi-source constraint information is extracted from the component connection diagram model, and a unified multi-source process constraint matrix is ​​constructed by fusing the multi-source constraint information. By initially sorting the disassembly and assembly sequence of each component in the power compartment system, a candidate set of disassembly and assembly sequences containing multiple potential candidate disassembly and assembly sequences is generated; where each candidate disassembly and assembly sequence represents the disassembly and assembly sequence of a component. For each candidate disassembly and assembly sequence in the candidate disassembly and assembly sequence set, a recursive topological test is performed. By traversing the candidate disassembly and assembly sequences and comparing them with the multi-source process constraint matrix, it is determined whether there is a constraint violation. Based on the number of violations, a violation index is constructed. Closed-loop constraint deadlock detection was performed on the power compartment system to obtain the system deadlock index; Based on the violation index and the system deadlock index, and combined with the disassembly and assembly path cost, the optimal candidate disassembly and assembly sequence is selected from the candidate set of disassembly and assembly sequences, and the corresponding disassembly and assembly process is generated.

2. The method for modeling and topology verification of the power compartment disassembly and assembly process based on a multi-source constraint matrix as described in claim 1, characterized in that, The graph theory-based model for constructing component connection diagrams to describe the assembly structure of the power compartment system specifically includes: Based on the component structure information and assembly relationship information of the power compartment system, the structural parameters, spatial positional relationships and assembly connection relationships of each component are obtained; Based on the maintenance operation information of the power compartment system, the disassembly and assembly attributes of each assembly connection are determined; the disassembly and assembly attributes include the type of connecting parts, the constraint strength of the connection structure, and the disassembly and assembly difficulty coefficient. Each component is abstracted as a component node in a graph structure, and each node is assigned corresponding structural parameters and spatial positional relationships; the assembly and connection relationships between components are abstracted as edges in a graph structure; and the disassembly and assembly attributes are used as the weights of the corresponding edges. Based on the component nodes, edges, and weights, a component connection graph model of the power compartment system is constructed.

3. The method for modeling and topology verification of the power compartment disassembly and assembly process based on a multi-source constraint matrix as described in claim 1, characterized in that, The multi-source constraint information includes structural constraint information, spatial constraint information, and maintenance process constraint information; The structural constraint information includes the assembly dependencies between the components; The spatial constraint information includes the assembly and disassembly sequence of each component due to spatial obstruction or spatial interference. The maintenance process constraint information includes the sequence of operations determined by maintenance safety specifications or maintenance procedure requirements.

4. The method for modeling and topology verification of the power compartment disassembly and assembly process based on a multi-source constraint matrix as described in claim 3, characterized in that, The construction of a unified multi-source process constraint matrix by fusing the multi-source constraint information specifically includes: Based on the structural constraint information, spatial constraint information, and maintenance process constraint information, corresponding structural constraint submatrix, spatial constraint submatrix, and process constraint submatrix are constructed. The structural constraint submatrix, spatial constraint submatrix, and process constraint submatrix are fused to obtain a multi-source process constraint matrix.

5. The method for modeling and topology verification of the power compartment disassembly and assembly process based on a multi-source constraint matrix as described in claim 4, characterized in that, Each element in the structural constraint submatrix is ​​represented as follows: in, Representing component nodes i With component nodes The weighted structural constraint values ​​between them i Indicates the first i Each component node; Indicates the first Each component node; variables Indicates the first r Each connector is connected to the component node. i With component nodes The contribution value of the structural constraint relationship between them; r Representing component nodes i With component nodes The number of the connecting parts; variables Representing component nodes i With component nodes The total number of connecting parts that exist.

6. The method for modeling and topology verification of the power compartment disassembly and assembly process based on a multi-source constraint matrix as described in claim 4, characterized in that, Each element in the spatial constraint submatrix is ​​represented as follows: in, Representing component nodes i For component nodes Spatial constraints; Indicates the threshold for determining spatial occlusion; Representing component nodes i For component nodes Spatial occlusion degree, variable Representing component nodes i For component nodes The occlusion volume includes spatial overlap volume or projected occlusion volume; variables Representing component nodes The overall spatial volume.

7. The method for modeling and topology verification of the power compartment disassembly and assembly process based on a multi-source constraint matrix as described in claim 1, characterized in that, The recursive topological test is performed on each candidate disassembly / assembly sequence in the candidate disassembly / assembly sequence set. This involves traversing the candidate disassembly / assembly sequences and comparing them with the multi-source process constraint matrix to determine whether there are any constraints violated. A violation index is then constructed based on the number of violations. Specifically, this includes: The violation count for each candidate disassembly / assembly sequence in the candidate disassembly / assembly sequence set is calculated using a violation degree function; the violation degree function is expressed as: in, This represents the sequence violation score calculated using a pairwise comparison method; Indicates the first in the candidate disassembly / assembly sequence One detachable component; Indicates the first in the sequence One detachable component; Represents the corresponding component in the multi-source process constraint matrix With components The constraints between them; N This indicates the total number of detachable parts in the candidate disassembly / assembly sequence; Count the number of constraint violations. And construct a violation index; the violation index is reflected by a violation penalty function, expressed as: in, ( ) indicates the candidate disassembly / assembly sequence S The penalty value for violations; To prevent extremely small positive numbers with a denominator of 0; pq This represents the constraint violation weighting coefficient, determined based on the number of constraint violations, and is used to measure the impact of different constraint violations on the system. Indicates candidate disassembly / assembly sequence S The corresponding indicator is whether the constraint has been violated.

8. The method for modeling and topology verification of the power compartment disassembly and assembly process based on a multi-source constraint matrix as described in claim 1, characterized in that, Deadlock detection in closed-loop constraints is performed using a deadlock detection function; the deadlock detection function is expressed as follows: in, Indicates the deadlock index of the engine compartment disassembly and assembly system; ( ) indicates an indicator function; Represents the first in the system A closed-loop constraint structure; This represents the set of constraint edges contained in the closed-loop structure; The first in the multi-source process constraint matrix Line 1 The elements of the column represent component nodes. With component nodes The overall assembly and disassembly sequence constraints between them.

9. The method for modeling and topology verification of the power compartment disassembly and assembly process based on a multi-source constraint matrix as described in claim 1, characterized in that, Also includes: Symmetrical component identification of parts in the power compartment system: Set component nodes With component nodes The spatial position vectors are respectively and Then their symmetry relationship can be determined by the following function: And define the symmetry function: in, Representing component nodes With component nodes The symmetrical relationship between them; Representing component nodes With component nodes Spatial distance between; vector Representing component nodes Spatial position coordinates; vector Representing component nodes Spatial position coordinates; symbols Represents Euclidean distance; parameters This represents the symmetry determination threshold; if the spatial distance between two components is less than the preset similarity threshold, then the two components are determined to be symmetrical.

10. The method for modeling and topology verification of the power compartment disassembly and assembly process based on a multi-source constraint matrix as described in claim 1, characterized in that, The optimal candidate disassembly / assembly sequence is selected from the candidate disassembly / assembly sequence set by using an optimization objective function; the optimization objective function for the candidate disassembly / assembly sequence is expressed as: in, ( ) represents the objective function for optimizing the candidate disassembly / assembly sequence; ( ) represents the penalty value for violations in the candidate disassembly / assembly sequence; 1 represents ( The weighting coefficients of ) Indicates the deadlock index of the engine compartment disassembly and assembly system; 2 indicates Weighting coefficients; ( ) represents the disassembly / assembly path cost function; 3 indicates ( The weighting coefficients of ) Indicates detachable components in the candidate disassembly / assembly sequence. Disassembly costs; N This indicates the total number of detachable components in the candidate disassembly / assembly sequence.

Images