CAD drawing real-time positioning method and device based on spatial index analysis

Abstract

The present invention discloses a method and device for real-time positioning of CAD drawings based on spatial index analysis, which relates to the technical field related to CAD drawing positioning. The method includes: performing multi-level splitting and identification on multi-version target CAD drawings; associating and indexing drawing version information, drawing layer information, and drawing component sets; performing semantic parsing and spatial association on each drawing text annotation; embedding and associating the drawing component spatial semantic tag set into a component-layer-drawing three-level association map; matching and indexing drawing semantic query requests based on the drawing spatial index mapping table; and performing real-time positioning and visual feedback based on the component query topology path. The method solves the technical problems existing in the prior art of difficulty in effectively processing complex spatial relationships and semantic information in drawings, low efficiency in retrieval of multi-version CAD drawings, insufficient immediacy, and inaccurate positioning, thereby achieving the technical effect of improving the retrieval efficiency, query immediacy, and positioning accuracy of multi-version drawings.

Description

Technical Field

[0001] The present application relates to the technical field related to CAD drawing positioning, and specifically to a method and device for real-time positioning of CAD drawings based on spatial index analysis. Background Art

[0002] CAD drawings often produce multiple versions, containing vast amounts of graphics, text, and annotation information, encompassing everything from macro-layouts to detailed design details for micro-components. However, in practice, it's difficult to quickly and accurately locate the required information within these multiple versions of CAD drawings. Existing CAD drawing retrieval and location methods, which mostly rely on simple keyword searches or manual browsing, are adequate for a small number of drawings, but are extremely inefficient when faced with a large number of CAD drawings across multiple versions. Furthermore, because the information in CAD drawings is highly spatially correlated and semantically complex, simple keyword matching cannot effectively understand the spatial relationships and semantic meaning of the components within the drawing. This results in significant deviations from actual requirements, making rapid retrieval and precise location difficult to achieve.

[0003] Therefore, in the current relevant technologies, there are technical problems such as difficulty in effectively processing complex spatial relationships and semantic information in drawings, low efficiency in retrieval of multi-version CAD drawings, lack of immediacy and inaccurate positioning. Summary of the Invention

[0004] This application solves the technical problems in the prior art of difficulty in effectively processing complex spatial relationships and semantic information in drawings, low efficiency in retrieval of multi-version CAD drawings, insufficient immediacy, and inaccurate positioning by providing a method and device for instant positioning of CAD drawings based on spatial index analysis. It achieves the technical effect of improving the efficiency of multi-version drawing retrieval, query immediacy, and positioning accuracy.

[0005] The present application provides a method for instant locating CAD drawings based on spatial index analysis, the method comprising: obtaining multi-version target CAD drawings, performing multi-level splitting and identification on the multi-version target CAD drawings, obtaining drawing version information, drawing layer information and a drawing component set; associating and indexing the drawing version information, drawing layer information and drawing component set, and building a component-layer-drawing three-level association map; performing semantic analysis and spatial association on each drawing text annotation in the multi-version target CAD drawings, and obtaining a drawing component spatial semantic tag set; embedding and associating the drawing component spatial semantic tag set into the component-layer-drawing three-level association map, and building a drawing spatial index mapping table; inputting a drawing semantic query request, matching and indexing the drawing semantic query request based on the drawing spatial index mapping table, and generating a component query topology path; performing instant locating and visual feedback in the multi-version target CAD drawings based on the component query topology path, and determining a drawing instant locating result.

[0006] In a possible implementation, the instant CAD drawing location method based on spatial index analysis also performs the following processing: obtaining drawing version identification rules, the identification content of the drawing version identification rules including project number, drawing type and timestamp; performing version classification and identification on the multi-version target CAD drawings according to the drawing version identification rules to obtain the drawing version information; performing layer classification on each version of the multi-version target CAD drawings according to preset layer classification rules to obtain the drawing layer information; performing component boundary identification and attribute type extraction on each version of the multi-version target CAD drawings to obtain the drawing component set.

[0007] In a possible implementation, the instant CAD drawing positioning method based on spatial index analysis further performs the following processing: building a drawing index coding system, the coding content of which includes drawing version, drawing layer and drawing component; indexing and coding the drawing version information, drawing layer information and drawing component set based on the drawing index coding system to obtain a drawing version identification code, a drawing layer identification code and a drawing component identification code; hierarchically associating the drawing version identification code, the drawing layer identification code and the drawing component identification code to obtain a drawing version-layer-component association identification code set; performing reverse tracing indexing based on the drawing version-layer-component association identification code set to build the component-layer-drawing three-level association map.

[0008] In a possible implementation, the instant CAD drawing positioning method based on spatial index analysis also performs the following processing: traverses and extracts the drawing annotation information of each drawing in the multi-version target CAD drawing to obtain a multi-version drawing annotation text set; performs text denoising and word segmentation processing on the multi-version drawing annotation text set according to the drawing application requirement standard to obtain a multi-version drawing annotation word set; mines and obtains the drawing professional field knowledge base of the multi-version target CAD drawing, performs graph cascade construction on the drawing professional field knowledge base, and obtains a drawing professional knowledge graph; performs semantic analysis and spatial association on the multi-version drawing annotation word set based on the drawing professional knowledge graph to obtain a drawing component spatial semantic label set.

[0009] In a possible implementation, the instant CAD drawing positioning method based on spatial index analysis also performs the following processing: semantic matching is performed on the multi-version drawing annotation word segmentation set based on the drawing professional knowledge graph to obtain a multi-version annotation matching semantic set; if the semantic matching fails, the word vector similarity of the multi-version drawing annotation word segmentation set and the drawing professional knowledge graph is calculated, and the multi-version annotation similar semantic set is determined based on the word vector similarity calculation result; the multi-version drawing annotation semantic set is obtained based on the multi-version annotation matching semantic set and the multi-version annotation similar semantic set; semantic label generation and component space association are performed on the multi-version drawing annotation semantic set to obtain a drawing component space semantic label set.

[0010] In a possible implementation, the CAD drawing instant positioning method based on spatial index analysis also performs the following processing: associating the multi-version drawing annotation semantic set and the drawing component set to obtain a multi-version drawing component semantic set; defining a drawing semantic label template, and using the drawing semantic label template to generate semantic labels for the multi-version drawing component semantic set to obtain a multi-version drawing component semantic label set; spatially positioning and identifying each component information in the drawing component set to obtain a multi-version drawing component spatial coordinate set; associating and mapping the multi-version drawing component semantic label set and the multi-version drawing component spatial coordinate set to obtain the drawing component spatial semantic label set.

[0011] In a possible implementation, the instant CAD drawing positioning method based on spatial index analysis also performs the following processing: embedding the drawing component spatial semantic tag set into the component-layer-drawing three-level association map for cross-layer association to obtain an initial drawing index mapping table; performing conflict detection on the initial drawing index mapping table to obtain a conflict index component; and correcting and updating the initial drawing index mapping table based on the conflict index component to obtain a drawing space index mapping table.

[0012] In a possible implementation, the instant CAD drawing positioning method based on spatial index analysis also performs the following processing: classifying and identifying the conflicting index components to obtain the component index conflict type; constructing an index conflict type correction rule, using the index conflict type correction rule to match and correct the component index conflict type to determine the component index correction information; updating the initial drawing index mapping table based on the component index correction information to obtain the drawing space index mapping table.

[0013] In a possible implementation, the instant positioning method of CAD drawings based on spatial index analysis also performs the following processing: obtaining the index conflict correction type, which includes spatial overlap conflict, semantic label conflict and cross-layer reference conflict; based on the index conflict correction type, performing correction strategy association and correction strategy integration respectively, and constructing the index conflict type correction rule.

[0014] The present application also provides a CAD drawing instant positioning device based on spatial index analysis, the device comprising: a drawing multi-level splitting unit, used to obtain multi-version target CAD drawings, perform multi-level splitting identification on the multi-version target CAD drawings, and obtain drawing version information, drawing layer information and drawing component set; an association map building unit, used to associate and index the drawing version information, drawing layer information and drawing component set, and build a component-layer-drawing three-level association map; a semantic parsing unit, used to perform semantic parsing and spatial association on each drawing text annotation in the multi-version target CAD drawings, and obtain a drawing component space semantic label set; an index mapping table building unit, used to embed and associate the drawing component space semantic label set into the component-layer-drawing three-level association map, and build a drawing space index mapping table; a query topology path generating unit, used to input a drawing semantic query request, match and index the drawing semantic query request based on the drawing space index mapping table, and generate a component query topology path; an instant positioning result determining unit, used to perform instant positioning and visual feedback in the multi-version target CAD drawings based on the component query topology path, and determine the drawing instant positioning result.

[0015] The proposed method and device for real-time CAD drawing location based on spatial index analysis in this application is intended to perform multi-level split identification on multi-version target CAD drawings; associate and index the drawing version information, drawing layer information, and drawing component set; perform semantic analysis and spatial association on each drawing text annotation; embed the drawing component spatial semantic tag set into the component-layer-drawing three-level association map; match and index drawing semantic query requests based on the drawing spatial index mapping table; and perform real-time location and visual feedback based on the component query topology path to determine the real-time location result of the drawing. This solves the technical problems existing in the prior art of the difficulty in effectively handling complex spatial relationships and semantic information in drawings, low efficiency in retrieval of multi-version CAD drawings, insufficient immediacy, and inaccurate location, achieving the technical effect of improving the efficiency of multi-version drawing retrieval, query immediacy, and location accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] To more clearly illustrate the technical solutions of the embodiments of the present disclosure, the accompanying drawings of the embodiments of the present disclosure are briefly introduced below. Flowcharts are used in this application to illustrate the operations performed by the apparatus according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed in precise order. Instead, various steps may be processed in reverse order or simultaneously as needed. Furthermore, other operations may be added to these processes, or one or more operations may be removed from these processes.

[0017] Figure 1 A flow chart of a method for instant locating CAD drawings based on spatial index analysis provided in an embodiment of the present application.

[0018] Figure 2 This is a schematic diagram of the structure of a CAD drawing instant positioning device based on spatial index analysis provided in an embodiment of the present application.

[0019] Explanation of the accompanying symbols: drawing multi-level splitting unit 10, association graph building unit 20, semantic parsing unit 30, index mapping table building unit 40, query topology path generation unit 50, instant positioning result determination unit 60. DETAILED DESCRIPTION

[0020] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below.

[0021] In order to make the purpose, technical solutions and advantages of this application clearer, the application will be further described in detail below with reference to the accompanying drawings. The described embodiments should not be regarded as limiting this application. All other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0022] In the following description, reference is made to “some embodiments”, which describes a subset of all possible embodiments, but it will be understood that “some embodiments” may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict, and the terms “first\second” involved are merely to distinguish similar objects and do not represent a specific ordering of the objects. The terms “including” and “having” and any variations are intended to cover non-exclusive inclusions, for example, a process, method, device, product or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this application belongs. The terms used herein are for the purpose of describing the embodiments of this application only.

[0023] The embodiment of the present application provides a method for real-time positioning of CAD drawings based on spatial index analysis, such as Figure 1 As shown, the method includes:

[0024] Step S100: Acquire a multi-version target CAD drawing, perform multi-level splitting and identification on the multi-version target CAD drawing, and obtain drawing version information, drawing layer information, and a drawing component set.

[0025] Preferably, different versions of CAD drawings in different design stages of the project, i.e., multi-version target CAD drawings, are collected. There may be differences in the overall layout, detailed design, and dimensioning of these drawings. Then, the multi-version target CAD drawings are split and identified at multiple levels, i.e., the complex CAD drawing information is refined and classified to obtain the drawing version information, drawing layer information, and drawing component set. Specifically, each version of the CAD drawing includes a specific version number, version name, and version update time, etc. By extracting and identifying the drawing version information, the differences between different versions of the drawings can be clarified. For example, the version number may increase according to a rule, such as from 1.0, 1.1, 1.2, etc., and each version number corresponds to different design modification content; the version name describes the main modification content or applicable stage of the version, such as "preliminary design version", "detailed design version", "construction version", etc.; the version update time records the specific time when the version was created or last modified.

[0026] Preferably, the drawing layer information refers to the multiple layers contained in the CAD drawing, each layer has its specific purpose and function, which may include but is not limited to the outline layer, centerline layer, annotation layer, fill layer, etc. By splitting and identifying the drawing layer information, the name of each layer, the visibility of the layer, the color of the layer, the line type and other attributes of the layer are determined. For example, the outline layer uses a solid line to represent the shape of the object, and the color is black; the centerline layer uses a dotted line to represent the axis of symmetry of the object, and the color is red; at the same time, the hierarchical relationship and organizational structure between the layers are obtained, and the interrelated layers and independent layers are clearly defined. Drawing components refer to the basic elements that constitute the graphics in the CAD drawing, such as points, lines, surfaces, circles, rectangles, polygons, and complex graphics composed of basic elements, such as mechanical parts, building components, etc. Splitting and identifying the drawing component set is to identify the type, position, size, shape and other geometric information of each component, as well as the component's attribute information, such as name, material, number, etc., and then form a drawing component set.

[0027] Preferably, step S100 also includes step S110, obtaining a drawing version identification rule, wherein the identification content of the drawing version identification rule includes a project number, a drawing type, and a timestamp; step S120, performing version classification identification on the multi-version target CAD drawings according to the drawing version identification rule to obtain the drawing version information; step S130, performing layer classification on each version of the multi-version target CAD drawings according to a preset layer classification rule to obtain the drawing layer information; step S140, performing component boundary identification and attribute type extraction on each version of the multi-version target CAD drawings to obtain the drawing component set.

[0028] Preferably, different versions of drawings in a project are obtained, and their drawing version identification rules are identified. The identification content includes the project number, drawing type, and timestamp. The project number is a number unique to each project. The project number can be used to quickly locate a specific project and distinguish drawings of different projects. The drawing type refers to various types of CAD drawings, such as architectural drawings, structural drawings, electrical drawings, etc. For example, "ARCH" represents architectural drawings and "STRUCT" represents structural drawings. The timestamp records the creation or modification time of the drawing version. Multi-version target CAD drawings are classified and identified according to the drawing identification rules, that is, each drawing is assigned a version identification that complies with the rules, so that the project, type, and version time of each drawing can be clearly understood, and the drawing version information can be obtained.

[0029] Preferably, the layers of each version of the multi-version target CAD drawing are classified using preset layer classification rules, wherein the preset layer classification rules are set to classify the layers in the drawing, such as stipulating that all graphic elements representing walls are placed in the "wall layer" and all annotation information is placed in the "annotation layer", thereby clarifying the content and purpose of each layer in each drawing, thereby obtaining detailed drawing layer information, including layer name, element type contained in the layer, etc. Computer graphics processing is then used to identify component boundaries and extract attribute types for each version of the drawing in the multi-version target CAD drawing, specifically including identifying the boundaries of each component in the drawing and extracting attribute information of each component, such as the component name, material, size, etc. Finally, the identified and extracted component information is combined into a drawing component set.

[0030] Step S200 , the drawing version information, the drawing layer information and the drawing component set are associated and indexed to build a three-level association map of components, layers and drawings.

[0031] Preferably, the drawing version information, drawing layer information and drawing component set are associated and indexed, that is, a correspondence between the drawing version information, drawing layer information and drawing component set is established, including quickly determining elements in other related data sets through one element, similar to establishing an association table. Specifically, each component in a CAD drawing is located on a specific layer. By recording the layer information to which each component belongs, a mapping relationship from component to layer is established in the atlas, that is, an association between component and layer is established; each CAD drawing contains multiple layers, and different versions of drawings may have different layer structures. By determining which drawing each layer belongs to and the version information of the drawing, a mapping relationship from layer to drawing is established in the atlas, that is, an association between layer and drawing is established; then, the mapping associations between component and layer, layer and drawing are integrated to establish a three-level association map of component-layer-drawing, that is, based on component, with layer as the intermediate bridge, connected to drawing version information, each component can be traced back to its layer and drawing version. Conversely, the layers and components contained therein can also be gradually located from the drawing version information. Through the three-level association map, you can quickly locate the layer and drawing version where the component is located, improving the efficiency and accuracy of information query.

[0032] Furthermore, step S200 also includes step S210, building a drawing index coding system, the coding content of the drawing index coding system includes drawing version, drawing layer and drawing component; step S220, indexing and coding the drawing version information, drawing layer information and drawing component set based on the drawing index coding system to obtain a drawing version identification code, a drawing layer identification code and a drawing component identification code; step S230, hierarchically associating the drawing version identification code, the drawing layer identification code and the drawing component identification code to obtain a drawing version-layer-component association identification code set; step S240, performing reverse tracing indexing based on the drawing version-layer-component association identification code set to build the component-layer-drawing three-level association map.

[0033] Preferably, a drawing index coding system is established. The coding content of the drawing index coding system includes the drawing version, drawing layer and drawing component, that is, the drawing version, drawing layer and drawing component information are encoded. Each information corresponds to a coding rule. Specifically, the project number, design date, number of revisions, etc. are integrated to index and identify each version of the drawing to obtain a drawing version identification code. For example, the drawing version code of a construction project is "PROJ001-20250427-V3", where "PROJ001" is the project number, "20250427" is the date, and "V3" means the third time. Modified version; Combined with the layer name, function, purpose, etc., the drawing layer information is indexed and encoded to obtain the drawing layer identification code, such as "WALL-LAYER-01" represents the first wall layer, and "WINDOW-LAYER-02" represents the second window layer; Combined with the component type, size, position and other attribute information, the drawing component set is indexed and encoded to obtain the drawing component identification code, such as the code of a circular column may be "COLUMN-CIRCLE-D500-POS(10, 20)", which means a circular column with a diameter of 500, located at the coordinate (10, 20).

[0034] Preferably, the drawing version identification code, the drawing layer identification code and the drawing component identification code are associated according to a hierarchical relationship to form a drawing version-layer-component association identification code set to reflect the hierarchical structure of the drawing information, that is, the component exists in a specific layer, and the layer belongs to a specific version of the drawing, so that the layer and drawing version to which each component belongs can be clearly understood; finally, the drawing version-layer-component association identification code set is used for reverse traceability indexing, that is, starting from the component identification code, the layer and drawing version to which the component belongs are reversely searched through the hierarchical relationship in the association identification code set, thereby building the component-layer-drawing three-level association map. Specifically, according to the component identification code, the layer identification code to which it is located can be found through the association relationship, and then the drawing version identification code to which the layer belongs can be determined. The association relationship between all components, layers and drawing versions is displayed in the form of a map, presenting the hierarchical structure and mutual relationship of information in the CAD drawing, so that users can quickly locate the required component, layer or drawing version, thereby improving the management and query efficiency of drawing information.

[0035] Step S300 : performing semantic analysis and spatial association on each drawing text annotation in the multi-version target CAD drawings to obtain a drawing component spatial semantic tag set.

[0036] Preferably, the target CAD drawings contain various text annotations for conveying design intent, explaining component attributes and technical requirements. For example, architectural drawings may have annotations of wall thickness, door and window sizes, and room functions; mechanical drawings may have annotations of part tolerances, surface roughness, and material types; semantic parsing of the text annotations in each drawing refers to converting the text annotations into semantic information that can be understood and processed by computers. Specifically, the vocabulary in the text annotations is identified and classified. For example, in the annotation "column: C30 concrete, side length 500mm", it is identified that "column" is the component name, "C30 concrete" is the material attribute, and "side length 500mm" is the size attribute; semantic parsing of the vocabulary in the text Understanding, for example, determines from the annotations that "C30 concrete" and "side length 500mm" are attribute information describing the component "column"; then spatially associate the text annotations of each drawing, that is, spatially associate the text annotations after semantic analysis with the components in the drawing, and determine the specific component corresponding to each text annotation. For example, according to the position of the text annotation in the drawing, find the components nearby, determine whether the annotation is related to it, and analyze the geometric relationship between the text annotation and the component, such as inclusion, adjacency, alignment, etc.; finally, give each drawing component a label that can describe its semantic information and spatial position, and then form a drawing component spatial semantic label set, which is conducive to accurate drawing space index mapping and fast query positioning.

[0037] Furthermore, step S300 also includes step S310, traversing and extracting the drawing annotation information of each drawing in the multi-version target CAD drawing to obtain a multi-version drawing annotation text set; step S320, performing text denoising and word segmentation processing on the multi-version drawing annotation text set according to the drawing application requirement standard to obtain a multi-version drawing annotation word set; step S330, mining and obtaining the drawing professional field knowledge base of the multi-version target CAD drawing, performing graph cascade construction on the drawing professional field knowledge base to obtain a drawing professional knowledge graph; step S340, performing semantic analysis and spatial association on the multi-version drawing annotation word set based on the drawing professional knowledge graph to obtain a drawing component spatial semantic label set.

[0038] Preferably, the text annotation information used to illustrate the meaning, size, technical requirements, etc. of the graphics of each drawing is traversed and extracted from the multi-version target CAD drawings to form a multi-version drawing annotation text set, and then the multi-version drawing annotation text set is subjected to text denoising and word segmentation processing according to the drawing application requirement standards. Specifically, different drawing application scenarios have different requirement standards for annotation texts. The multi-version drawing annotation text set is subjected to text denoising, that is, irrelevant symbols, repeated content, incorrect spellings and other noise information are removed to make the text cleaner and more accurate. The multi-version drawing annotation text set is then subjected to word segmentation processing, that is, the denoised text is divided according to the boundaries of words, and continuous text strings are divided into separate words. For example, "building structure design description" is segmented into "architecture", "structure", "design" and "description" to facilitate more convenient semantic analysis and processing of the text.

[0039] Preferably, drawing-related knowledge is mined based on relevant professional literature, standard specifications, design manuals and other materials in the drawing professional field, and a drawing professional field knowledge base of multiple versions of target CAD drawings is obtained, which includes professional terms, concepts, relationships, specifications, etc., and then a graph cascade is constructed for the drawing professional field knowledge base. Specifically, professional terms, concepts, component names, etc. are used as graph nodes, and node relationships (such as "belong to", "include", "associated", etc.) are used as graph edges. The knowledge in the knowledge base is analyzed and associated to form a drawing professional knowledge graph.

[0040] Preferably, the knowledge and relationships in the drawing professional knowledge map are used to understand the information in the annotation word set of multiple versions of drawings. Specifically, each word in the annotation word set is matched and associated with the node in the knowledge map to determine its accurate meaning in the professional field, including the semantic role and attributes in the drawing and the combination relationship between words; the annotation information is associated with the spatial position in the drawing, that is, by analyzing the position of the annotation in the drawing and the spatial relationship with other graphic elements, the position of the specific component corresponding to each annotation in the drawing is determined; finally, a drawing component spatial semantic label set is constructed, which includes the component name, attributes (such as material, size, etc.) and its spatial position information in the drawing, etc., which helps to achieve more efficient drawing information retrieval and positioning.

[0041] Furthermore, step S340 also includes step S341, performing semantic matching on the multi-version drawing annotation word segmentation set based on the drawing professional knowledge graph to obtain a multi-version annotation matching semantic set; step S342, if the semantic matching fails, calculating the word vector similarity of the multi-version drawing annotation word segmentation set and the drawing professional knowledge graph, and determining the multi-version annotation similar semantic set based on the word vector similarity calculation result; step S343, obtaining a multi-version drawing annotation semantic set based on the multi-version annotation matching semantic set and the multi-version annotation similar semantic set; step S344, performing semantic label generation and component space association on the multi-version drawing annotation semantic set to obtain a drawing component space semantic label set.

[0042] Preferably, semantic matching refers to comparing each word or phrase in the multi-version drawing annotation word set with the nodes and relationships in the drawing professional knowledge graph, and finding whether the annotation word matches the node name or relationship description in the knowledge graph. If a certain annotation word can find an exact matching node or relationship in the knowledge graph, its corresponding semantic information is extracted, and finally a multi-version annotation matching semantic set is obtained; when a certain annotation word does not have a directly matching semantic in the knowledge graph, its word vector similarity with each node in the knowledge graph is calculated, where the word vector is used to convert the word into a numerical representation in the vector space, and the similarity between the words is measured by calculating the distance between the vectors (such as cosine similarity), and then a threshold is set according to the calculation result of the word vector similarity (such as similarity greater than 0.8), and the annotation words whose similarity with the nodes in the knowledge graph exceeds the threshold and their similar semantic information are formed into a multi-version annotation similar semantic set.

[0043] Preferably, the multi-version annotation matching semantic set and the multi-version annotation similar semantic set are then merged to obtain a multi-version drawing annotation semantic set, which includes various semantic contents in the multi-version drawing annotations; finally, the information in the multi-version drawing annotation semantic set is processed to generate a semantic label for each annotation, summarizing the core semantic features of the component or information represented by the annotation; the generated semantic label is then associated with the spatial position information of the specific component in the drawing, including analyzing the position and coordinates of the annotation in the drawing, and the relative relationship with other graphic elements, etc., to determine the specific position and range of each component in the drawing space; finally, a set containing the semantic labels of each component in the drawing and its position information in space is obtained, that is, the drawing component space semantic label set, which helps to achieve more intelligent and efficient drawing information management.

[0044] Furthermore, step S344 also includes step A1, associating the multi-version drawing annotation semantic set with the drawing component set to obtain a multi-version drawing component semantic set; step A2, defining a drawing semantic label template, using the drawing semantic label template to generate semantic labels for the multi-version drawing component semantic set, and obtaining a multi-version drawing component semantic label set; step A3, spatially positioning and identifying each component information in the drawing component set to obtain a multi-version drawing component spatial coordinate set; step A4, associating and mapping the multi-version drawing component semantic label set and the multi-version drawing component spatial coordinate set to obtain the drawing component spatial semantic label set.

[0045] Preferably, the multi-version drawing annotation semantic set and the drawing component set are associated, that is, the semantic information in the annotation semantic set is matched and connected with the specific components in the component set according to the correspondence between the component name or number and the component mentioned in the annotation, so as to obtain the multi-version drawing component semantic set; the drawing semantic label template is pre-defined, including the format, content composition and value range of the semantic label, etc., and the defined drawing semantic label template is used to process the semantic information of each component in the multi-version drawing component semantic set, extract and organize relevant information according to the requirements of the template, generate corresponding semantic labels, and combine them to form a multi-version drawing component semantic label set.

[0046] Preferably, the spatial position of each component in the drawing component set is determined and identified using CAD drawing coordinates, including obtaining the coordinate position of the component on the drawing plane (such as X, Y coordinates), and the position information in three-dimensional space (if it is a three-dimensional drawing), and then combining the spatial positioning identification information of all components to generate a multi-version drawing component spatial coordinate set, which records the spatial position coordinates of each component in the multi-version drawings; finally, the multi-version drawing component semantic label set and the multi-version drawing component spatial coordinate set are associated and mapped, that is, the semantic label of each component in the semantic label set is associated with its corresponding spatial position information in the spatial coordinate set, and the drawing component spatial semantic label set is obtained, which integrates the semantic information and spatial position information of the component to facilitate more efficient query, analysis, positioning and management of the target CAD version drawing.

[0047] Step S400: embed and associate the drawing component space semantic tag set into the component-layer-drawing three-level association map to construct a drawing space index mapping table.

[0048] Preferably, the drawing component spatial semantic label set is embedded and associated with the component-layer-drawing three-level association map, that is, the information in the drawing component spatial semantic label set is connected and integrated with the corresponding elements in the component-layer-drawing three-level association map. Specifically, the spatial semantic label of each component is associated with the component node in the map, so that the components in the map not only have their hierarchical relationship information in the drawing, but also contain detailed spatial and semantic description information. For example, find the component node representing a window in the map, and embed the spatial semantic label of the window (such as position coordinates, size, function, etc.) into the node, so that the components in the map can more comprehensively reflect the actual situation in the drawing.

[0049] Preferably, a drawing space index mapping table is then constructed based on the enriched three-level association map, which is used to quickly locate relevant components, layers or drawing information in the drawing according to different query conditions. For example, the name of the component, an attribute in the semantic tag (such as function, material, etc.), or the spatial position information of the component (coordinate range, etc.) can be used as an index key to quickly find the position of the corresponding component in the map, and then obtain the layer where the component is located, the drawing to which it belongs, and other related detailed information, thereby improving the efficiency of drawing management, query and analysis, and facilitating users to quickly and accurately obtain the required drawing space information.

[0050] Furthermore, step S400 also includes step S410, embedding the drawing component space semantic tag set into the component-layer-drawing three-level association map for cross-layer association to obtain an initial drawing index mapping table; step S420, performing conflict detection on the initial drawing index mapping table to obtain a conflict index component; step S430, correcting and updating the initial drawing index mapping table based on the conflict index component to obtain a drawing space index mapping table.

[0051] Preferably, in actual CAD drawings, components in different layers may have spatial or semantic connections. The drawing component space semantic label set containing component spatial information (such as position, size, etc.) and semantic information (such as component type, function, etc.) is cross-layer associated with the component nodes in the component-layer-drawing three-level association map to establish associations between components in different layers. Based on the enriched three-level association map, an initial drawing index mapping table is constructed, that is, certain characteristics of the component (such as component name, attributes in the semantic label, spatial position, etc.) are used as index keys, and a mapping relationship is established between them and the component's position in the map and related layers and drawing information. The index key can be used to quickly locate the component in the map and obtain its related information.

[0052] Preferably, the initial drawing index mapping table is subjected to conflict detection, i.e., checking whether there is a situation where the index key is inconsistent with the actual component information or association relationship, or there are situations where different index keys point to the same component but the information is contradictory, etc. For example, there may be two different semantic labels pointing to the same component, but the component functions or attributes described by the two labels conflict with each other; or the component found according to the spatial position index key does not match the component actually at that position, etc.; and then the component corresponding to the index where the conflict is detected is determined as the conflicting index component; then, for the conflicting index component, the cause of the conflict is analyzed, and the relevant information in the initial drawing index mapping table is modified, adjusted or supplemented, for example, if the conflict is caused by an error in the semantic label, the semantic label is corrected; if the spatial position index is inaccurate, the correct spatial position information of the component is re-determined, etc.; finally, a corrected and improved drawing index mapping table, i.e., a drawing space index mapping table, is obtained, which can more accurately reflect the spatial position, semantic information and association relationship of the components in the drawing, so as to facilitate efficient query, management and analysis of CAD drawings.

[0053] Furthermore, step S430 also includes step S431, classifying and identifying the conflict index components to obtain the component index conflict type; step S432, constructing an index conflict type correction rule, using the index conflict type correction rule to match and correct the component index conflict type, and determining the component index correction information; step S433, updating the initial drawing index mapping table based on the component index correction information to obtain the drawing space index mapping table.

[0054] Preferably, the conflicting index components are classified and identified to obtain the component index conflict type. For example, there may be semantic conflicts, that is, different semantic labels contradict each other in describing the same component; spatial conflicts, the spatial position of the index does not match the actual spatial position of the component; or association conflicts, errors in the association relationship between the component and the layer or drawing, etc. Based on the actual application requirements and industry standards of CAD drawings, corresponding correction rules are formulated according to different component index conflict types. For example, for semantic conflicts, the accurate semantics of the component is determined according to its geometric characteristics, location and relationship with other components; for spatial conflicts, the correct spatial position of the component is recalculated and determined based on the scale of the drawing, the coordinate system and the position of other reference components; for association conflicts , refer to the drawing design logic and layer management specifications to repair the association relationship between components and layers, drawings; then match the conflict type of each conflict index component with the constructed index conflict type correction rule, and correct the conflict according to the corresponding rule, so as to obtain the component index correction information of each conflict index component, including the correct semantic label, spatial position information and accurate association relationship, etc.; finally, use the component index correction information to update the information of the corresponding conflict index component in the initial drawing index mapping table, that is, replace the error or conflict information in the initial table with the corrected component information, and obtain the drawing space index mapping table, which can more accurately reflect the actual situation of the components in the CAD drawing, and facilitate users to quickly and accurately obtain drawing information data through indexing.

[0055] Furthermore, step S432 also includes step B1, obtaining the index conflict correction type, which includes spatial overlap conflict, semantic label conflict and cross-layer reference conflict; step B2, performing correction strategy association and correction strategy integration based on the index conflict correction type, and constructing the index conflict type correction rule.

[0056] Preferably, different types of index conflicts in the drawing space index mapping table are clarified, including spatial overlap conflicts, semantic label conflicts and cross-layer reference conflicts. Among them, spatial overlap conflicts refer to the situation in which the spatial position indexes of different components overlap in the drawing space due to coordinate marking errors, scale setting problems or inaccurate component space definitions, that is, according to the index information, two or more components seem to occupy the same spatial position; semantic label conflicts refer to the situation in which the semantic labels assigned to components are contradictory or inaccurate due to manual marking errors, inconsistent semantic specifications of different versions of drawings, etc.; cross-layer reference conflicts refer to problems in the associated references of components between different layers due to chaotic layer management, failure to correctly handle cross-layer relationships during drawing updates, etc.

[0057] Preferably, correction strategies are associated based on the index conflict correction type. Specifically, for spatial overlapping conflicts, the coordinate information of the components is rechecked and calibrated, and the positions of overlapping components are adjusted according to the scale of the drawing and other accurately marked components to ensure that each component has a unique and reasonable position in space; for semantic label conflicts, the geometric shape of the component, the connection relationship with other components and the relevant instructions of the drawing are used to determine its accurate function and type, and then the semantic label is updated; for cross-layer reference conflicts, the link settings between layers are checked to confirm the existence and correctness of the referenced component in the target layer, and the cross-layer reference relationship is repaired or re-established according to the actual design intent; finally, the corresponding correction strategies for each index conflict correction type are integrated to form index conflict type correction rules, which can provide specific and feasible solutions for different types of index conflicts, thereby improving the accuracy and reliability of the drawing space index mapping table.

[0058] Step S500: input a drawing semantic query request, perform matching indexing on the drawing semantic query request based on the drawing space index mapping table, and generate a component query topology path.

[0059] Preferably, the user inputs a drawing semantic query request according to needs, that is, a query instruction generated based on the semantic information of the drawing (component type, function, location, etc.). The drawing semantic query request input by the user is then compared and matched with the index in the drawing space index mapping table, including searching and filtering each field in the table. The components that meet the requirements are determined based on the semantic conditions in the query request, and the logical path from the query request to finding the components that meet the conditions is determined, that is, a logical path based on the semantic structure and index relationship of the drawing. For example, the component set of type "window" is found from the "semantic label" field of the index mapping table, and then the components located on the first floor are further filtered out from this set through the "spatial coordinates" or "location description" fields. The graph is represented as a directed graph from the query condition node (such as "first floor window") to the specific component node, where the edges represent the logical steps and relationships of the query, and ultimately a logical path from the query request to the target component is formed, that is, the component query topology path, so as to facilitate the determination of the corresponding component according to the query request.

[0060] Step S600: performing real-time positioning and visual feedback in the multi-version target CAD drawings based on the component query topology path, and determining the real-time positioning result of the drawings.

[0061] Preferably, a target component that meets the query criteria is quickly found in multiple versions of CAD drawings based on a component query topology path. Specifically, a search and location is performed in each version of the CAD drawing according to the steps indicated by the topology path. If the topology path indicates that a component with a specific semantic tag should be found first, the corresponding tag is searched in the relevant data structure of the CAD drawing, and then the search is continued according to subsequent steps until the specific location of the target component in the current version of the CAD drawing is found. After the target component in the multiple versions of the CAD drawing is located, the location result is presented to the user in a visual manner on the graphical interface of the CAD drawing. For example, the target component is distinguished from other non-target components using a specific color, line style, or transparency, so that the user can intuitively see the location of the query result in the drawing. Detailed properties of the component (name, size, material, etc.) are also provided to facilitate a more comprehensive understanding of the target component. Finally, the instant location result of the drawing is determined, including the accurate location information of the target component found in the multiple versions of the CAD drawing and the related visual display status, which helps to quickly and accurately determine the specific location and related information of the queried component in the CAD drawing.

[0062] In the above, refer to Figure 1 The instant locating method of CAD drawings based on spatial index analysis according to an embodiment of the present invention is described in detail. Figure 2A device for real-time positioning of CAD drawings based on spatial index analysis according to an embodiment of the present invention is described.

[0063] The CAD drawing instant location device based on spatial index analysis according to the embodiment of the present invention is used to solve the technical problems existing in the prior art, such as the difficulty in effectively processing complex spatial relationships and semantic information in drawings, low efficiency in retrieval of multi-version CAD drawings, insufficient immediacy, and inaccurate location, thereby achieving the technical effect of improving the efficiency of multi-version drawing retrieval, query immediacy, and location accuracy. Figure 2 As shown, the CAD drawing instant positioning device based on spatial index analysis includes: a drawing multi-level splitting unit 10, an association graph building unit 20, a semantic parsing unit 30, an index mapping table construction unit 40, a query topology path generation unit 50, and an instant positioning result determination unit 60.

[0064] The drawing multi-level splitting unit 10 is used to obtain multi-version target CAD drawings, perform multi-level splitting identification on the multi-version target CAD drawings, and obtain drawing version information, drawing layer information and drawing component set; the association map building unit 20 is used to associate and index the drawing version information, drawing layer information and drawing component set, and build a three-level component-layer-drawing association map; the semantic parsing unit 30 is used to perform semantic parsing and spatial association on each drawing text annotation in the multi-version target CAD drawings to obtain a drawing component space semantic tag set; the index mapping table building unit 40 is used to embed and associate the drawing component space semantic tag set into the three-level component-layer-drawing association map, and build a drawing space index mapping table; the query topology path generation unit 50 is used to input a drawing semantic query request, match and index the drawing semantic query request based on the drawing space index mapping table, and generate a component query topology path; the instant positioning result determination unit 60 is used to perform instant positioning and visual feedback in the multi-version target CAD drawings based on the component query topology path, and determine the drawing instant positioning result.

[0065] The specific configuration of the multi-level drawing splitting unit 10 will be described in detail below. The multi-level drawing splitting unit 10 further includes: obtaining a drawing version identification rule, wherein the identification content of the drawing version identification rule includes a project number, a drawing type, and a timestamp; performing version classification and identification on the multi-version target CAD drawings according to the drawing version identification rule to obtain the drawing version information; performing layer classification on each version of the multi-version target CAD drawings according to a preset layer classification rule to obtain the drawing layer information; and performing component boundary identification and attribute type extraction on each version of the multi-version target CAD drawings to obtain the drawing component set.

[0066] The specific configuration of the association map construction unit 20 will be described in detail below. The association map construction unit 20 further includes: building a drawing index coding system, wherein the coding content of the drawing index coding system includes drawing version, drawing layer, and drawing component; indexing and coding the drawing version information, drawing layer information, and drawing component set based on the drawing index coding system to obtain a drawing version identification code, a drawing layer identification code, and a drawing component identification code; hierarchically associating the drawing version identification code, the drawing layer identification code, and the drawing component identification code to obtain a drawing version-layer-component association identification code set; and performing reverse tracing indexing based on the drawing version-layer-component association identification code set to build the component-layer-drawing three-level association map.

[0067] The specific configuration of the semantic parsing unit 30 will be described in detail below. The semantic parsing unit 30 further includes: traversing and extracting the annotation information of each drawing in the multi-version target CAD drawing to obtain a multi-version drawing annotation text set; performing text denoising and word segmentation processing on the multi-version drawing annotation text set according to the drawing application requirement standard to obtain a multi-version drawing annotation word set; mining and obtaining the drawing professional domain knowledge base of the multi-version target CAD drawing, performing graph cascade construction on the drawing professional domain knowledge base to obtain a drawing professional knowledge graph; performing semantic parsing and spatial association on the multi-version drawing annotation word set based on the drawing professional knowledge graph to obtain a drawing component spatial semantic label set.

[0068] The specific configuration of the semantic parsing unit 30 will be described in detail below. The semantic parsing unit 30 further includes: performing semantic matching on the multi-version drawing annotation word segmentation set based on the drawing professional knowledge graph to obtain a multi-version annotation matching semantic set; if the semantic matching fails, calculating the word vector similarity between the multi-version drawing annotation word segmentation set and the drawing professional knowledge graph, and determining a multi-version annotation similar semantic set based on the word vector similarity calculation result; obtaining a multi-version drawing annotation semantic set based on the multi-version annotation matching semantic set and the multi-version annotation similar semantic set; performing semantic label generation and component space association on the multi-version drawing annotation semantic set to obtain a drawing component space semantic label set.

[0069] The specific configuration of the semantic parsing unit 30 will be described in detail below. The semantic parsing unit 30 further includes: associating the multi-version drawing annotation semantic set with the drawing component set to obtain a multi-version drawing component semantic set; defining a drawing semantic label template, using the drawing semantic label template to generate semantic labels for the multi-version drawing component semantic set to obtain a multi-version drawing component semantic label set; spatially locating and identifying each component information in the drawing component set to obtain a multi-version drawing component spatial coordinate set; and associating and mapping the multi-version drawing component semantic label set with the multi-version drawing component spatial coordinate set to obtain the drawing component spatial semantic label set.

[0070] The specific configuration of the index mapping table construction unit 40 will be described in detail below. The index mapping table construction unit 40 further includes: embedding the drawing component space semantic tag set into the component-layer-drawing three-level association map for cross-layer association to obtain an initial drawing index mapping table; performing conflict detection on the initial drawing index mapping table to obtain a conflict index component; and correcting and updating the initial drawing index mapping table based on the conflict index component to obtain a drawing space index mapping table.

[0071] The specific configuration of the index mapping table construction unit 40 will be described in detail below. The index mapping table construction unit 40 further includes: classifying and identifying the conflicting index components to obtain component index conflict types; constructing index conflict type correction rules, using the index conflict type correction rules to match and correct the component index conflict types to determine component index correction information; and updating the initial drawing index mapping table based on the component index correction information to obtain the drawing space index mapping table.

[0072] The specific configuration of the index mapping table construction unit 40 will be described in detail below. The index mapping table construction unit 40 further includes: obtaining index conflict correction types, wherein the index conflict correction types include spatial overlap conflicts, semantic label conflicts, and cross-layer reference conflicts; and performing correction strategy association and correction strategy integration based on the index conflict correction types to construct the index conflict type correction rules.

[0073] The CAD drawing real-time positioning device based on spatial index analysis provided by an embodiment of the present invention can execute the CAD drawing real-time positioning method based on spatial index analysis provided by any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0074] Although the present application makes various references to certain modules in the apparatus according to the embodiments of the present application, any number of different modules may be used and run on the user terminal and / or server, and the various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other and are not used to limit the scope of protection of the present invention.

[0075] The above specific embodiments do not constitute a limitation on the scope of protection of this application. Those skilled in the art should understand that various modifications, combinations, and substitutions may be made based on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this application shall be included within the scope of protection of this application.

Claims

1. A CAD drawing instant positioning method based on spatial index analysis, characterized in that: The method comprises: Acquire a multi-version target CAD drawing, perform multi-level splitting and identification on the multi-version target CAD drawing, and obtain drawing version information, drawing layer information, and a drawing component set; Associating and indexing the drawing version information, drawing layer information, and drawing component set to build a three-level component-layer-drawing association map; Performing semantic parsing and spatial association on the text annotations of each drawing in the multi-version target CAD drawings to obtain a drawing component spatial semantic tag set; Embed and associate the drawing component space semantic tag set into the component-layer-drawing three-level association map to construct a drawing space index mapping table; Input a drawing semantic query request, perform matching indexing on the drawing semantic query request based on the drawing space index mapping table, and generate a component query topology path; Based on the component query topology path, real-time positioning and visual feedback are performed in the multi-version target CAD drawings to determine the real-time positioning result of the drawings.

2. The CAD drawing instant positioning method based on spatial index analysis according to claim 1, characterized in that: The obtaining of drawing version information, drawing layer information and drawing component set includes: Obtaining a drawing version identification rule, wherein the identification content of the drawing version identification rule includes a project number, a drawing type, and a timestamp; Performing version classification identification on the target CAD drawings with multiple versions according to the drawing version identification rule to obtain the drawing version information; Classifying each version of the target CAD drawing in multiple versions by using a preset layer classification rule to obtain the layer information of the drawing; Component boundary identification and attribute type extraction are performed on each version of the multi-version target CAD drawing to obtain the drawing component set.

3. The method for real-time positioning of CAD drawings based on spatial index analysis according to claim 1, characterized in that: The component-layer-drawing three-level association map includes: Establishing a drawing index coding system, wherein the coding content of the drawing index coding system includes drawing version, drawing layer and drawing component; Indexing and coding the drawing version information, drawing layer information, and drawing component set based on the drawing index coding system to obtain a drawing version identification code, a drawing layer identification code, and a drawing component identification code; hierarchically associating the drawing version identification code, the drawing layer identification code, and the drawing component identification code to obtain a drawing version-layer-component association identification code set; A reverse traceability index is performed based on the drawing version-layer-component association identification code set to build the component-layer-drawing three-level association map.

4. The method for real-time positioning of CAD drawings based on spatial index analysis according to claim 1, wherein: Obtaining the drawing component spatial semantic label set includes: Traversing and extracting the annotation information of each drawing in the multi-version target CAD drawings to obtain a multi-version drawing annotation text set; Performing text denoising and word segmentation processing on the multi-version drawing annotation text set according to the drawing application requirement standard to obtain a multi-version drawing annotation word segmentation set; Mining and obtaining a drawing professional domain knowledge base of the multi-version target CAD drawings, and performing graph cascade construction on the drawing professional domain knowledge base to obtain a drawing professional knowledge graph; Based on the drawing professional knowledge graph, semantic parsing and spatial association are performed on the multi-version drawing annotation word set to obtain a drawing component spatial semantic label set.

5. The method for real-time positioning of CAD drawings based on spatial index analysis according to claim 4, characterized in that: The obtaining of the drawing component space semantic tag set includes: Performing semantic matching on the multi-version drawing annotation word sets based on the drawing professional knowledge graph to obtain a multi-version annotation matching semantic set; If the semantic matching fails, the word vector similarity between the multi-version drawing annotation word set and the drawing professional knowledge graph is calculated, and the multi-version annotation similar semantic set is determined based on the word vector similarity calculation result; Obtaining a multi-version drawing annotation semantic set according to the multi-version annotation matching semantic set and the multi-version annotation similar semantic set; Semantic tags are generated and component space association is performed on the multi-version drawing annotation semantic sets to obtain a drawing component space semantic tag set.

6. The method for real-time positioning of CAD drawings based on spatial index analysis according to claim 5, characterized in that: Generating semantic tags and associating component spaces with the semantic sets of the multi-version drawing annotations to obtain a semantic tag set of the drawing component spaces includes: Associating the multi-version drawing annotation semantic set with the drawing component set to obtain a multi-version drawing component semantic set; Defining a drawing semantic label template, and using the drawing semantic label template to generate semantic labels for the multi-version drawing component semantic set to obtain the multi-version drawing component semantic label set; Performing spatial positioning identification on each component information in the drawing component set to obtain a multi-version drawing component spatial coordinate set; The multi-version drawing component semantic tag set and the multi-version drawing component space coordinate set are associated and mapped to obtain the drawing component space semantic tag set.

7. The method for real-time positioning of CAD drawings based on spatial index analysis according to claim 1, characterized in that: The construction of the drawing space index mapping table includes: Embed the drawing component space semantic tag set into the component-layer-drawing three-level association map to perform cross-layer association to obtain an initial drawing index mapping table; Performing conflict detection on the initial drawing index mapping table to obtain a conflict index component; The initial drawing index mapping table is corrected and updated based on the conflict index component to obtain a drawing space index mapping table.

8. The method for real-time positioning of CAD drawings based on spatial index analysis according to claim 7, characterized in that: The obtaining of the paper space index mapping table includes: Classifying and identifying the conflicting index components to obtain component index conflict types; Constructing an index conflict type correction rule, using the index conflict type correction rule to match and correct the component index conflict type, and determining component index correction information; The initial drawing index mapping table is updated based on the component index correction information to obtain the drawing space index mapping table.

9. The method for real-time positioning of CAD drawings based on spatial index analysis according to claim 8, characterized in that: The index conflict type correction rule is constructed, including: Obtaining index conflict correction types, including spatial overlap conflicts, semantic label conflicts, and cross-layer reference conflicts; Based on the index conflict correction type, correction strategy association and correction strategy integration are performed respectively to construct the index conflict type correction rule.

10. A CAD drawing real-time positioning device based on spatial index analysis, characterized in that: The device is used to implement the CAD drawing instant positioning method based on spatial index analysis according to any one of claims 1 to 9, and the device comprises: A drawing multi-level splitting unit is used to obtain a multi-version target CAD drawing, perform multi-level splitting identification on the multi-version target CAD drawing, and obtain drawing version information, drawing layer information and drawing component set; A correlation map building unit is used to associate and index the drawing version information, drawing layer information and drawing component set, and build a three-level component-layer-drawing correlation map; A semantic parsing unit, configured to perform semantic parsing and spatial association on the text annotations of each drawing in the multi-version target CAD drawing to obtain a set of spatial semantic labels for the drawing components; An index mapping table construction unit is used to embed and associate the drawing component space semantic tag set into the component-layer-drawing three-level association map to construct a drawing space index mapping table; A query topology path generating unit is configured to input a drawing semantic query request, match and index the drawing semantic query request based on the drawing space index mapping table, and generate a component query topology path; The instant positioning result determining unit is used to perform instant positioning and visual feedback in the multi-version target CAD drawings based on the component query topology path, and determine the instant positioning result of the drawings.

Images