Knowledge-guided visual simulation method for railway bridge construction process

Abstract

The invention discloses a knowledge-guided visual simulation method for a railway bridge construction process, belongs to the field of railway bridge construction, and solves the problems that the railway bridge construction process in the prior art relates to many elements, is difficult in relation capture, is not clear in behavior state description, is continuously iterated and updated along with the promotion of the construction process, is complex in environment change, and is poor in reliability. And the incidence relation between the bridge main body model and the visual simulation environment is difficult to establish. The method comprises the following steps: acquiring railway bridge construction data including GIS data, field data and construction data; constructing an entity-process-behavior three-domain associated railroad bridge construction scene knowledge graph based on the obtained railroad bridge construction data; and performing knowledge-guided dynamic visual simulation of the railway bridge construction process based on the railway bridge construction scene knowledge graph. The method is used for visual simulation of the railway bridge construction process.

Description

Technical Field

[0001] The invention discloses a knowledge-guided visual simulation method for railway bridge construction process, which is used for visual simulation of railway bridge construction process and belongs to the field of railway bridge construction. Background Art

[0002] Railway bridges are large, complex, and sophisticated dynamic system projects with long spans, many risk factors, and fast-changing construction processes. The construction process of railway bridges is complex and fluid, and the relationship between the bridge body and the surrounding environment is dynamic and changeable, which brings huge safety risks. Therefore, accurate control of the construction process is crucial. Through virtual simulation technology, risks can be predicted and controlled in advance by rehearsing the operation process in a virtual environment to ensure the safety of the construction process and improve the efficiency and quality of engineering management.

[0003] At present, the simulation method of railway bridge construction process usually relies on parametric modeling software to design bridge component models according to construction requirements, and then the models are moved and assembled according to the actual operation conditions to simulate the construction process. Since different construction processes require different component models and there is no unified standard for different operation operations, it is difficult to organize and schedule the model and the simulation process has poor adaptability. In addition, this method usually considers the fluidity of the model in isolation, ignoring the temporal and spatial correlation between the construction process and the surrounding geographical environment. In the actual construction process, the scene objects and environment often change dynamically over time. Therefore, it is necessary to improve and optimize this simulation method to achieve an accurate description of the behavior state of the scene entities and establish an efficient mapping relationship between the main engineering model and the geographical environment to support the rapid visualization of various types of railway bridge construction processes.

[0004] Knowledge graph is an advanced knowledge management technology, whose core components are triples of "entity-relationship-entity" or "entity-relationship-attribute". These triples are interconnected through the attributes of entities and the correlation between entities, forming a comprehensive and interconnected knowledge network, and providing a basis for the bridge construction simulation process. However, the existing knowledge graph technology has the problems of weak perception of domain objects and difficulty in describing behavioral states in the complex bridge construction process, which makes it difficult to fully cope with the changing situations in the construction process in practical applications. Therefore, it is urgent to establish a model that can accurately describe the behavioral states of scene entities and enhance the ability of knowledge graphs in describing dynamically changing spatiotemporal relationships, so as to better support the scene modeling of railway engineering and the efficient reuse of domain knowledge.

[0005] In summary, the following technical problems exist in the visualization simulation of railway bridge construction process:

[0006] 1. The railway bridge construction process involves many elements, difficult to capture relationships, and unclear descriptions of behavior states. As the construction process progresses, the bridge structure is constantly iterated and updated, and the environment changes are complex, making it difficult to establish the relationship between the main bridge model and the visual simulation environment;

[0007] 2. The simulation process of railway bridge construction involves a large amount of scene information and is difficult to organize and schedule. In addition, the process and methods of bridge construction are complex and diverse, and there is a lack of adaptable and flexible simulation methods. Summary of the invention

[0008] The purpose of the present invention is to provide a knowledge-guided visual simulation method for the railway bridge construction process, so as to solve the problems that the railway bridge construction process in the prior art involves many elements, it is difficult to capture relationships, and the description of behavior states is unclear. As the construction process progresses, the bridge structure is constantly iterated and updated, and the environment changes are complex, making it difficult to establish the relationship between the main bridge model and the visual simulation environment.

[0009] In order to achieve the above object, the technical solution adopted by the present invention is:

[0010] A knowledge-guided visual simulation method for railway bridge construction process includes the following steps:

[0011] Step 1: Obtain railway bridge construction data, including GIS data, field data, and construction data;

[0012] Step 2: construct a railway bridge construction scenario knowledge graph that associates the three domains of entity, process, and behavior based on the acquired railway bridge construction data;

[0013] Step 3: Perform knowledge-guided dynamic visualization simulation of the railway bridge construction process based on the railway bridge construction scenario knowledge graph.

[0014] Further, the specific steps of step 2 are:

[0015] Step 2.1, based on the construction data, the behavior states of each construction model in the railway bridge construction scenario at different construction stages are integrated, and an entity-process-behavior three-domain association expression model is given, wherein the construction data includes the bridge structure and the construction process;

[0016] Step 2.2: Construct a railway bridge construction scenario knowledge graph based on the acquired data and the given entity-process-behavior three-domain association expression model. The specific steps are as follows:

[0017] Step 2.21, based on the given entity-process-behavior three-domain association expression model, the three-domain concepts of the domain data in the railway bridge construction scenario are extracted, and the domain ontology of each concept in the three-domain concepts is extracted, that is, the construction of the knowledge graph pattern layer is completed, wherein the domain data includes design data, model data and monitoring data, the three-domain concepts include entity domain, process domain and behavior domain, and the domain ontology refers to the pattern layer, including the three concept nodes of entity domain, process domain and behavior domain and their association relationships, and the association relationships include hierarchical relationship, semantic association relationship and spatiotemporal coupling relationship;

[0018] The extraction steps of the entity domain are:

[0019] Establish a bounding box of the entire railway bridge, and select the bottommost component model of the railway bridge in the bounding box of the entire railway bridge as the starting point;

[0020] Establish a bounding box of the component model based on the starting point, and send out rays based on the six faces of the bounding box of the component model, obtain the relationship and category between the component models intersecting with the ray, and put the adjacent relationship into the same column, and put the same category into the same group, where the column and group are used as the storage method of the railway bridge entity domain;

[0021] The component models where the rays intersect are respectively used as starting points to send out rays, and the railway bridge entity domain is continuously constructed until the construction is completed;

[0022] The extraction steps of process domain and behavior domain are:

[0023] First, at the macro scale, based on the acquired domain data, construction data, and extracted entity domains, the construction plan, design data, and construction schedule of the railway bridge are manually analyzed to divide the construction phases into different stages;

[0024] Secondly, at the meso-scale, each construction stage is divided into specific construction tasks, and the construction tasks are associated and bound with the entity objects in the entity domain to ensure that each task can be further mapped to specific operational behaviors;

[0025] Then, at the microscopic scale, after each task is refined into specific operation behavior steps, the construction process flow and specific operation behaviors involving each entity object in the entity domain are extracted, and the construction process flow and specific operation behaviors are sorted and combined to form a process sequence and operation sequence for all entity objects in the entity domain; based on the process sequence, the time arrangement and correlation relationship of the construction tasks corresponding to all entity objects in the entity domain are sorted out, and a complete construction process framework is established to reflect the execution order of the construction tasks, that is, the process domain is constructed;

[0026] Based on the operation sequence, the specific behavior and interaction process of each entity object in the entity domain is recorded to form a behavior network that reflects the actions and interactions of each entity object, that is, the behavior domain is constructed;

[0027] Step 2.22: After the knowledge graph model layer is constructed, the concept nodes and association relationships in the domain ontology are mapped to the entity nodes and association relationships in the data layer to realize the organization and storage of data, that is, the construction of the data layer is realized;

[0028] Step 2.23: Use the open source graph database Neo4j to store and visualize the triple information in the data layer, thus obtaining an instantiated railway bridge construction scenario knowledge graph.

[0029] Further, the specific steps of step 3 are:

[0030] Step 3.1: Based on the railway bridge construction scenario knowledge graph, event-driven railway bridge construction scenario is constructed on demand. The specific steps are as follows:

[0031] Step 3.11, using railway bridge construction events as the driving force, based on construction data and design data, artificially analyze the component models, geographical environment, and construction processes associated with railway bridge construction events, identify the characteristic entities, process flows, and behavior elements required in each railway bridge construction event, and map them to the railway bridge construction scenario knowledge graph, and identify the entity nodes in the railway bridge construction event and their interrelated relationships;

[0032] Step 3.12: Based on the GIS data and the entity nodes and their associations obtained in step 3.11, the entity objects represented by the entity nodes are modeled respectively, and the geographical environment model is established by integrating multi-source data. A virtual railway bridge construction scene is constructed based on the geographical environment model and the requirements of the railway bridge construction event, wherein the entity object is the category of the component model in the entity domain, including the main railway bridge project and the geographical environment. For the main railway bridge project, modeling is performed by assembling step by step, and the association relationship of the main project is encapsulated. The GIS data includes terrain data and image data, and the multi-source data includes elevation data DEM, document data DOM, vector data, and point cloud data;

[0033] Step 3.2: Visualize and simulate the railway bridge construction process based on the constructed railway bridge construction scenario, specifically:

[0034] Construct a railway bridge construction event model that can capture and express the dynamic changes of time and space. The railway bridge construction event model introduces the time axis to associate the changes of objects in three-dimensional space with time during the railway bridge construction process, forming a four-dimensional space-time framework to describe the state changes and task progress of various entity objects during the railway bridge construction process. The railway bridge construction event model is defined as:

[0035] M=[T,t,P,E]

[0036] in,

[0037] Wherein, T represents the time of occurrence of a railway bridge construction event, which is used to express the time node of a railway bridge construction event, t represents the duration of the railway bridge construction event, which is used to represent the duration of each railway bridge construction event, P represents the position and posture information of the component model, which is used to determine the position and posture information of each component model involved in the railway bridge construction event, E represents the type of railway bridge construction event, which is used to describe the type of operation event in the railway bridge construction process, R represents the rotation matrix of the component model, and p represents the three-dimensional coordinate information of the component model;

[0038] In the osgEarth three-dimensional engine, the parameters in the railway bridge construction event model in the event sequence are semantically mapped into animation parameters in the key frame to obtain the animation parameter file, which is then parsed by the tinyxml2 library. After parsing, the osg::Animationpath interface is used to obtain the frame animation of the component model changing over time, and the frame animation is loaded into the constructed railway bridge construction scene to achieve a visual expression of the construction process.

[0039] Compared with the prior art, the advantages of the present invention are:

[0040] 1. The present invention designs an entity-process-behavior three-domain association expression model, and proposes a method for automatically traversing and extracting bridge entities based on bounding boxes to extract entities and relationships to establish a bridge construction scene domain ontology, and accordingly establish a knowledge graph to achieve flexible storage and dynamic expression of knowledge. That is, through the method of bounding box and ray intersection, it can automatically identify and extract component models in railway bridges, and accurately capture the spatial adjacent relationship between components. This method not only improves the efficiency and accuracy of railway bridge entity domain extraction, but also provides reliable basic data support for subsequent knowledge graph construction, solves the problems of difficulty in entity domain extraction, inaccurate classification, and fuzzy hierarchical relationships in the prior art, and achieves the technical effect of automatic, precise and structured extraction of railway bridge entity domains. At the same time, through the extraction of process domains and behavior domains, it can achieve clear expression of construction process architecture and dynamic tracking of entity object behavior, effectively solves the problem of difficulty in clearly expressing the complex association between construction process and object behavior in the prior art, improves the accuracy of construction process simulation, and ensures the overall controllability of the construction process;

[0041] Second, the present invention proposes a knowledge-guided dynamic visualization simulation method for the railway bridge construction process. With events as the driving force, a construction virtual scene is constructed through three-domain analysis, the characteristics of scene elements are clarified, and an event model that takes into account the changes in time and space is defined to realize the visualization simulation of the railway bridge construction process.

[0042] 3. The present invention regards the entire railway bridge construction process as an event tree through a four-dimensional space-time railway bridge event model, simulates the construction events of each tree node by setting different parameters in the event model, and instantiates the parameters to obtain the key frame parameters of each railway bridge construction event to achieve simulation;

[0043] 4. The present invention realizes the visual simulation of the railway bridge construction process in the osgEarth three-dimensional engine by using animation for spatiotemporal narrative description, intuitively presenting each construction step in a three-dimensional scene in chronological order, clearly expressing the timing and dynamic changes of the construction process, enhancing the authenticity and immersion of the simulation, and improving the flexibility and practicality of the simulation. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0045] Figure 1 It is the overall framework diagram of the present invention;

[0046] Figure 2 Schematic diagram of the entity-process-behavior three-domain association expression model in the present invention;

[0047] Figure 3 It is a schematic diagram of the entity domain extraction process in the bridge construction scene of the present invention;

[0048] Figure 4 The process of constructing the knowledge graph of the railway bridge construction scenario in the present invention;

[0049] Figure 5 It is a schematic diagram of three-domain analysis of the event-driven bridge scenario in the present invention;

[0050] Figure 6 This is a schematic diagram of the scene modeling process in the present invention;

[0051] Figure 7 A schematic diagram of a virtual scene for building a railway bridge in the present invention;

[0052] Figure 8 It is a schematic diagram of the simulation process of the railway bridge construction event model in the present invention;

[0053] Fig. 9 It is a schematic diagram of the instantiation process of the event model of the bridge construction process in the present invention;

[0054] Fig.10 It is a flowchart of the process domain and behavior domain extraction in the present invention. DETAILED DESCRIPTION

[0055] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0056] like Figure 1 As shown, it mainly includes obtaining railway bridge construction data, constructing a railway bridge construction scenario knowledge graph that associates the three domains of entity, process and behavior, and dynamic visual simulation of the railway bridge construction process guided by knowledge.

[0057] First, obtain railway bridge construction data, including GIS data, field data, and construction data;

[0058] Based on the acquired railway bridge construction data, a railway bridge construction scene knowledge graph with entity-process-behavior three-domain association is constructed. That is, the railway bridge construction scene knowledge graph includes a pattern layer and a data layer. For the pattern layer of the knowledge graph, an ontology library is usually used for management. The ontology determines the concept nodes and association relationships in the knowledge graph, which is an important basis for constructing the knowledge graph. According to the detailed classification of concepts such as physical entities, construction processes, and behavioral states in the bridge scene in the three-domain association expression model, the domain ontology containing the upper and lower relationships, semantic association relationships, and spatiotemporal coupling relationships between the concepts can be extracted, that is, the construction of the knowledge graph pattern layer is completed; the data layer mainly includes the data and relationships refined on the basis of the pattern layer, and the concept node-association relationship in the ontology structure is mapped to the entity node-association relationship in the data layer, thereby realizing the organization and storage of data, that is, realizing the construction of the data layer. Subsequently, the present invention uses the open source graph database Neo4j to store and visualize the triple information in the data layer. This knowledge graph instantiated with the railway bridge construction scene domain ontology can realize a dynamic construction knowledge management process, and thereby ensure the accurate conduct of the railway bridge construction simulation. The specific steps are:

[0059] Traditional knowledge graphs often face defects such as unclear information organization, difficulty in capturing dynamic changes, and unclear behavior descriptions when dealing with complex engineering projects. In response to the above problems, the present invention fully considers the bridge structure and construction process, integrates the behavior status of each object (component model) at different construction stages, and designs an entity-process-behavior three-domain association expression model by extracting railway bridge scene entities and analyzing the construction process flow, realizing efficient management of the full-scene information of railway bridge construction, that is, integrating the behavior status of each construction model in the railway bridge construction scene at different construction stages based on construction data, and giving an entity-process-behavior three-domain association expression model, where the construction data includes bridge structure and construction process;

[0060] Based on the acquired data and the given entity-process-behavior three-domain association expression model, a railway bridge construction scenario knowledge graph is constructed. The specific steps are as follows:

[0061] Based on the given entity-process-behavior three-domain association expression model, the three-domain concepts of the domain data in the railway bridge construction scenario are extracted, and the domain ontology of each concept in the three-domain concept is extracted, that is, the construction of the knowledge graph pattern layer is completed. Among them, the domain data includes design data, model data and monitoring data, the three-domain concepts include entity domain, process domain and behavior domain, and the domain ontology refers to the pattern layer, including three concept nodes of entity domain, process domain and behavior domain and their association relationships. The association relationships include hierarchical relationships, semantic association relationships and spatiotemporal coupling relationships.

[0062] In this three-domain association expression model, the entity domain covers all physical and abstract objects in the bridge construction process, including various components of the railway bridge, construction equipment, environmental information, etc. Each entity object has specific attributes and states. Therefore, in order to ensure the complete expression of the entity object in the model, the bounding box is used to traverse the bridge scene model for extraction. The specific process is as follows: Figure 3 , after the extraction is completed, multiple entity domain lists and groups are obtained. The list represents the connection relationship between bridge entities, and the group represents the same entity type. The extraction steps of the entity domain are:

[0063] Establish a bounding box of the entire railway bridge, and select the bottommost component model of the railway bridge in the bounding box of the entire railway bridge as the starting point;

[0064] Establish a bounding box of the component model based on the starting point, and send out rays based on the six faces of the bounding box of the component model, obtain the relationship and category between the component models intersecting with the ray, and put the adjacent relationship into the same column, and put the same category into the same group, where the column and group are used as the storage method of the railway bridge entity domain;

[0065] The component models where the rays intersect are respectively used as starting points to send out rays, and the railway bridge entity domain is continuously constructed until the construction is completed;

[0066] The process domain describes the various stages and steps of bridge construction, as well as the logical relationships between these stages and steps. It reflects the temporal and spatial changes of entities and relationships during the construction process and is the core of the progress of bridge construction. The behavior domain describes the specific operations and behaviors performed by each entity during the construction process. It is the key execution layer in the bridge construction process, ensuring that various construction tasks are successfully completed. The information of the process domain and the behavior domain is mainly obtained through prior knowledge, design data, research reports, etc.

[0067] In the above-mentioned integrated expression model of entity-process-behavior domain association, the behavior domain promotes the progress of the construction process through specific operations, the process domain steps are reflected in the state changes of the entity objects, and the changes in the entity domain in turn feedback and affect the subsequent construction behavior.

[0068] The extraction steps of process domain and behavior domain are:

[0069] First, at the macro scale, based on the acquired domain data, construction data, and extracted entity domains, the construction plan, design data, and construction schedule of the railway bridge are manually analyzed to divide the construction phases into different stages;

[0070] Secondly, at the meso-scale, each construction stage is divided into specific construction tasks, and the construction tasks are associated and bound with the entity objects in the entity domain to ensure that each task can be further mapped to specific operational behaviors;

[0071] Then, at the microscopic scale, after each task is refined into specific operation behavior steps, the construction process flow and specific operation behaviors involving each entity object in the entity domain are extracted, and the construction process flow and specific operation behaviors are sorted and combined to form a process sequence and operation sequence for all entity objects in the entity domain; based on the process sequence, the time arrangement and correlation relationship of the construction tasks corresponding to all entity objects in the entity domain are sorted out, and a complete construction process framework is established to reflect the execution order of the construction tasks, that is, the process domain is constructed;

[0072] Based on the operation sequence, the specific behavior and interaction process of each entity object in the entity domain is recorded to form a behavior network that reflects the actions and interactions of each entity object, that is, the behavior domain is constructed.

[0073] After the knowledge graph model layer is constructed, the concept nodes and association relationships in the domain ontology are mapped to the entity nodes and association relationships in the data layer to realize the organization and storage of data, that is, the construction of the data layer is realized;

[0074] The open source graph database Neo4j is then used to store and visualize the triple information in the data layer, thus obtaining an instantiated railway bridge construction scenario knowledge graph.

[0075] Dynamic visualization simulation of railway bridge construction process based on knowledge graph of railway bridge construction scenario.

[0076] In order to simulate the construction process of railway bridges, it is first necessary to build models of the main bridge project and the relevant geographical environment to generate a virtual scene. However, the construction machinery model and geographical environment involved in the railway construction process have complex semantic information and associations. How to integrate and manage the models in the railway long-term project and apply them has become a key issue. For this reason, this case uses construction geographic events as the driving force, analyzes the characteristics of bridge construction scenes from three dimensions: entity, process, and behavior, and identifies the key objects in the corresponding scenes and the relationships between them by mapping event features to the knowledge graph.

[0077] After the scene entity-relationships obtained through the above process, these entities are modeled separately, and the poses of the models are restricted by the multi-level semantic relationships analyzed to construct a virtual scene of the construction process. The modeling process is as follows: Figure 6 As shown in the figure. The smallest unit of bridge main engineering modeling is the primitive model. A primitive model library that can adapt to scheduling needs is established, and semantic information is encapsulated to support spatial analysis of model objects. The main engineering model is formed by step-by-step assembly. The combination and assembly of primitive models form component models, and the assembly of component models constitutes the main model. The geographic environment model is generated by integrating multi-source spatiotemporal data, such as DEM, DOM, vector data, etc.

[0078] Considering the difficulty in management caused by the large number of elements in the construction scene model, this case is based on the osgEarth three-dimensional earth simulation engine, and uses PagedLOD algorithm, multi-threading, multi-level model-terrain-image fusion and other technologies to complete the model management and scene generation. In osgEarth, all layer objects are organized in the scene tree structure, and the functions of moving, scaling, rotating and so on can be realized by operating the nodes in the scene. In the scene, all the main bridge engineering models are managed by the osg::Group container node, and the organization and scheduling of the model are realized by operating its nodes; the osgEarth::MapNode node is used to manage geographic environment information, including images, elevations, vectors, mapping data and other information; then the PagedLOD algorithm is used to divide the entire scene into blocks, and a hierarchical model from low detail to high detail is created for each block. Finally, only the blocks within the viewing cone are rendered, and the generated railway three-dimensional information scene is as follows: Figure 7 shown.

[0079] That is, based on the railway bridge construction scene knowledge graph, event-driven railway bridge construction scene is constructed on demand. The specific steps are:

[0080] Taking railway bridge construction events as the driving force, we manually analyze the component models, geographical environment and construction processes associated with railway bridge construction events based on construction data and design data, identify the characteristic entities, process flows and behavioral elements required in each railway bridge construction event, and map them to the railway bridge construction scenario knowledge graph, identifying the entity nodes in the railway bridge construction event and their interrelated relationships;

[0081] Based on GIS data and the obtained entity nodes and association relationships, the entity objects represented by the entity nodes are modeled respectively, and the geographical environment model is established by integrating multi-source data. Based on the geographical environment model and the requirements of the railway bridge construction event, a virtual railway bridge construction scene is constructed, wherein the entity object is the category of the component model in the entity domain, including the main railway bridge project and the geographical environment. For the main railway bridge project, modeling is performed by assembling step by step, and the association relationship of the main project is encapsulated. GIS data includes terrain data and image data, and multi-source data includes elevation data DEM, document data DOM, vector data and point cloud data;

[0082] Railway bridge construction is not only a display of engineering technology, but also a complex process of dynamic changes in space-time information. During the construction of railway bridges, the positions and states of various objects and elements are constantly changing. These changes are continuous and interrelated in time and space. In order to accurately manage and simulate each link of railway bridge construction, this case established an event model that can capture and express these dynamic changes in time and space. By introducing a timeline, the model associates the changes of objects in three-dimensional space during the construction process with time to form a four-dimensional space-time framework. It is used to describe the state changes and task progress of various objects during the construction process. The construction process is visualized by semantically mapping them to animation parameters in key frames and parsing frame animations. The specific process is as follows: Figure 8 As shown. That is, the railway bridge construction process is visualized and simulated based on the constructed railway bridge construction scene, specifically:

[0083] Based on the constructed railway bridge construction scene, a railway bridge construction event model is constructed that can capture and express the dynamic changes in time and space. The railway bridge construction event model introduces the time axis to associate the changes of objects in three-dimensional space with time during the railway bridge construction process, forming a four-dimensional space-time framework to describe the state changes and task progress of various entity objects during the railway bridge construction process. The railway bridge construction event model is defined as:

[0084] M=[T,t,P,E]

[0085] in,

[0086] Wherein, T represents the time of occurrence of a railway bridge construction event, which is used to express the time node of a railway bridge construction event, t represents the duration of the railway bridge construction event, which is used to represent the duration of each railway bridge construction event, P represents the position and posture information of the component model, which is used to determine the position and posture information of each component model involved in the railway bridge construction event, R represents the type of railway bridge construction event, which is used to describe the type of operation event in the railway bridge construction process, R represents the rotation matrix of the component model, and p represents the three-dimensional coordinate information of the component model;

[0087] In the above-mentioned railway bridge construction event model, the spatial displacement or transformation process of the railway bridge construction event model can be regarded as one or more events. When a certain time node is reached, the event will be triggered, and the events involved in the entity in a period of time are completely constrained by time and space. The railway bridge construction event model is extended to the entire railway bridge construction process to achieve accurate simulation and management of the entire bridge construction process. Specifically, the entire railway bridge construction event can be regarded as an event directory tree first, and each event sequence process operation is regarded as each sub-node in the directory tree. The operation parameters covered by each sub-node element are obtained from the domain and construction knowledge graph, and are encapsulated and stored through xml files. Through this operation, the parameter file after the event model is instantiated can be obtained, and then the parameter files of each sub-node are mapped to key frames arranged in time sequence. By integrating these key frames, the animation parameter file of the bridge construction process can be obtained. The overall process is as follows: Fig. 9 As shown. The animation parameter file is parsed through the animation interface provided by osgEarth, and applied to the relevant models involved in the railway bridge construction event to obtain the frame animation. Loading the frame animation into the constructed railway bridge construction scene can realize the correlation between the construction process and the surrounding geographical environment, thereby realizing the dynamic simulation of the railway bridge construction process. That is, in the osgEarth three-dimensional engine, each parameter in each railway bridge construction event model in the event sequence is semantically mapped to the animation parameter in the key frame to obtain the animation parameter file, and then the animation parameter file is parsed through the tinyxml2 library. After parsing, the osg::Animationpath interface is used to obtain the frame animation of the component model changing over time, and the frame animation is loaded into the constructed railway bridge construction scene to realize the visual expression of the construction process.

[0088] Therefore, the present invention realizes the on-demand construction of railway construction scene process scenes and takes into account the accurate expression of the behavior status of scene entities. Compared with the prior art, the advantages and features of this case are as follows: a three-domain association expression model of entity-process-behavior is designed, from which entities and relationships are extracted to establish the domain ontology of bridge construction scenes, and a knowledge graph is established based on this to achieve flexible storage and dynamic expression of knowledge; a knowledge-guided dynamic visualization simulation method for railway bridge construction process is proposed, which takes events as the driving force, clarifies the characteristics of scene elements, and defines an event model that takes into account the changes in time and space. A frame animation of the railway bridge construction process is created in the osgEarth three-dimensional earth simulation engine to achieve visual simulation of the construction process. Since this method integrates the knowledge graph of the railway bridge construction scene and takes into account the four-dimensional space-time state of the specific scene objects in the construction process, it can flexibly adapt to each link of the bridge construction project, so it has strong universality and is suitable for the simulation of various railway bridge construction processes.

[0089] In view of the above technical difficulties, this case analyzes the characteristics of bridge construction scenes, designs entity-process-behavior three-domain association expression model, extracts scene entities and relationships, constructs railway construction scene knowledge map, and clarifies the behavior state of objects involved in railway bridge construction events that change with time and space. At the same time, with events as the driving force, the entities, processes, and behaviors of the construction process are clarified, and a virtual simulation platform is established based on the osgEarth three-dimensional engine. The dynamic evolution process of railway bridge construction events is described in detail in the three-dimensional virtual scene through the event model and mapped as animation key frames. The virtual simulation of the railway bridge construction process is completed by parsing the key frames, solving the problems of unclear element information organization, difficulty in capturing dynamic changes in the environment, and unclear description of object behavior in the existing railway bridge construction process simulation methods.

Claims

1. A knowledge-guided visual simulation method for railway bridge construction process, characterized in that: The steps include: Step 1: Obtain railway bridge construction data, including GIS data, field data, and construction data; Step 2: construct a railway bridge construction scenario knowledge graph that associates the three domains of entity, process, and behavior based on the acquired railway bridge construction data; Step 3: Perform knowledge-guided dynamic visualization simulation of the railway bridge construction process based on the railway bridge construction scenario knowledge graph.

2. A knowledge-guided railway bridge construction process visualization simulation method according to claim 1, characterized in that: The specific steps of step 2 are: Step 2.1, based on the construction data, the behavior states of each construction model in the railway bridge construction scenario at different construction stages are integrated, and an entity-process-behavior three-domain association expression model is given, wherein the construction data includes the bridge structure and the construction process; Step 2.2: Construct a railway bridge construction scenario knowledge graph based on the acquired data and the given entity-process-behavior three-domain association expression model. The specific steps are as follows: Step 2.21, based on the given entity-process-behavior three-domain association expression model, the three-domain concepts of the domain data in the railway bridge construction scenario are extracted, and the domain ontology of each concept in the three-domain concepts is extracted, that is, the construction of the knowledge graph pattern layer is completed, wherein the domain data includes design data, model data and monitoring data, the three-domain concepts include entity domain, process domain and behavior domain, and the domain ontology refers to the pattern layer, including the three concept nodes of entity domain, process domain and behavior domain and their association relationships, and the association relationships include hierarchical relationship, semantic association relationship and spatiotemporal coupling relationship; The extraction steps of the entity domain are: Establish a bounding box of the entire railway bridge, and select the bottommost component model of the railway bridge in the bounding box of the entire railway bridge as the starting point; Establish a bounding box of the component model based on the starting point, and send out rays based on the six faces of the bounding box of the component model, obtain the relationship and category between the component models intersecting with the ray, and put the adjacent relationship into the same column, and put the same category into the same group, where the column and group are used as the storage method of the railway bridge entity domain; The component models where the rays intersect are respectively used as starting points to send out rays, and the railway bridge entity domain is continuously constructed until the construction is completed; The extraction steps of process domain and behavior domain are: First, at the macro scale, based on the acquired domain data, construction data, and extracted entity domains, the construction plan, design data, and construction schedule of the railway bridge are manually analyzed to divide the construction phases into different stages; Secondly, at the meso-scale, each construction stage is divided into specific construction tasks, and the construction tasks are associated and bound with the entity objects in the entity domain to ensure that each task can be further mapped to specific operational behaviors; Then, at the microscopic scale, after each task is refined into specific operation behavior steps, the construction process flow and specific operation behaviors involving each entity object in the entity domain are extracted, and the construction process flow and specific operation behaviors are sorted and combined to form a process sequence and operation sequence for all entity objects in the entity domain; based on the process sequence, the time arrangement and correlation relationship of the construction tasks corresponding to all entity objects in the entity domain are sorted out, and a complete construction process framework is established to reflect the execution order of the construction tasks, that is, the process domain is constructed; Based on the operation sequence, the specific behavior and interaction process of each entity object in the entity domain is recorded to form a behavior network that reflects the actions and interactions of each entity object, that is, the behavior domain is constructed; Step 2.22: After the knowledge graph model layer is constructed, the concept nodes and association relationships in the domain ontology are mapped to the entity nodes and association relationships in the data layer to realize the organization and storage of data, that is, the construction of the data layer is realized; Step 2.23: Use the open source graph database Neo4j to store and visualize the triple information in the data layer, thus obtaining an instantiated railway bridge construction scenario knowledge graph.

3. A knowledge-guided railway bridge construction process visualization simulation method according to claim 2, characterized in that: The specific steps of step 3 are: Step 3.1: Based on the railway bridge construction scenario knowledge graph, event-driven railway bridge construction scenario is constructed on demand. The specific steps are as follows: Step 3.11, using railway bridge construction events as the driving force, based on construction data and design data, artificially analyze the component models, geographical environment, and construction processes associated with railway bridge construction events, identify the characteristic entities, process flows, and behavior elements required in each railway bridge construction event, and map them to the railway bridge construction scenario knowledge graph, and identify the entity nodes in the railway bridge construction event and their interrelated relationships; Step 3.12: Based on the GIS data and the entity nodes and their associations obtained in step 3.11, the entity objects represented by the entity nodes are modeled respectively, and the geographical environment model is established by integrating multi-source data. A virtual railway bridge construction scene is constructed based on the geographical environment model and the requirements of the railway bridge construction event. The entity object is the category of the component model in the entity domain, including the main railway bridge project and the geographical environment. For the main railway bridge project, modeling is performed by assembling step by step, and the association relationship of the main project is encapsulated. GIS data includes terrain data and image data, and multi-source data includes elevation data DEM, document data DOM, vector data and point cloud data: Step 3.2: Visualize and simulate the railway bridge construction process based on the constructed railway bridge construction scenario, specifically: Construct a railway bridge construction event model that can capture and express the dynamic changes of time and space. The railway bridge construction event model introduces the time axis to associate the changes of objects in three-dimensional space with time during the railway bridge construction process, forming a four-dimensional space-time framework to describe the state changes and task progress of various entity objects during the railway bridge construction process. The railway bridge construction event model is defined as: M=[T,t,P,E] in, Among them, T represents the time of occurrence of the railway bridge construction event, which is used to express the time node of a railway bridge construction event. t represents the duration of the railway bridge construction event, which is used to represent the duration of each railway bridge construction event. P represents the position and posture of the component model, which is used to determine the position and posture information of each component model involved in the railway bridge construction event. E represents the type of railway bridge construction event, which is used to describe the type of operation event in the railway bridge construction process. R represents the rotation matrix of the component model, and p represents the three-dimensional coordinate information of the component model: In the osgEarth three-dimensional engine, the parameters in the railway bridge construction event model in the event sequence are semantically mapped to animation parameters in the key frame to obtain the animation parameter file, which is then parsed by the tinyxml2 library. After parsing, the osg::Animationpath interface is used to obtain the frame animation of the component model changing over time, and the frame animation is loaded into the constructed railway bridge construction scene to achieve a visual expression of the construction process.

Images