A knowledge graph-based railway bridge BIM model auditing method

Abstract

The application discloses a kind of based on knowledge graph's railway bridge BIM model's auditing method, comprising the following steps: S1, railway bridge standard data is structured and constructs knowledge graph;S2, information in the BIM model to be audited is extracted and processed;S3, the standard rule base of auditing is constructed by Drools rule engine;S4, auditing is carried out.The method of the application is based on bridge engineering design specification, construction standard and experience knowledge in knowledge graph, formulates a set of comprehensive auditing rules, and these rules are constructed into rule base, utilizes pattern matching, calls rule base and audits each bridge component and its attribute in BIM model, to realize the automation, intelligent auditing of BIM model, significantly improve the auditing efficiency and accuracy, reduce the interference of human factor, ensure the compliance and reliability of final BIM model.

Description

Technical Field

[0001] This invention relates to the field of information technology and intelligent building engineering, and in particular to a method for reviewing railway bridge BIM models based on knowledge graphs. Background Technology

[0002] With the widespread application of Building Information Modeling (BIM) technology in railway bridge construction, BIM models have become an important tool for the entire lifecycle management of engineering projects, including design, construction, and operation and maintenance. BIM models not only integrate a large amount of geometric, physical, and attribute information, but also realize the visualization, parameterization, and collaboration of information, greatly improving the efficiency and accuracy of engineering management.

[0003] Traditional railway bridge BIM model review primarily relies on manual examination. Reviewers depend on design drawings, technical specifications, construction instructions, and other documents, combining their professional knowledge and experience to review each item of the BIM model. This method is not only time-consuming and labor-intensive but also prone to inconsistencies and errors due to human factors. Furthermore, with the continuous updating of design codes and construction standards, reviewers need to constantly learn and master new knowledge, which undoubtedly increases the difficulty and cost of the review work. Summary of the Invention

[0004] To address the problems in the background art, this invention proposes an efficient, accurate, and automated method for reviewing railway bridge BIM models based on knowledge graphs.

[0005] Therefore, the present invention adopts the following technical solution:

[0006] A knowledge graph-based method for reviewing railway bridge BIM models includes the following steps:

[0007] S1 involves structuring standard data for railway bridges and constructing a knowledge graph, including:

[0008] S11 structures the standard data of railway bridges to form a structured dataset suitable for knowledge graph construction. This standard data includes textual data from railway bridge design manuals, construction standards, engineering examples, and expert knowledge. The structured dataset includes the types and attributes of each bridge component in the railway bridge.

[0009] S12, construct the structured dataset obtained in S11 into a knowledge graph to fully present the attributes of each bridge component in the railway bridge, the relationships between bridge components, and the constraints.

[0010] S2, Extract and process information from the BIM model to be reviewed, including the following steps:

[0011] S21, Extract the actual dimensions, material properties, connection methods of component connection points, and stress information of bridge components of the railway bridge in the BIM model to be reviewed;

[0012] S22, Calculate the geometric matching degree G, which is used to determine the degree to which the actual size conforms to the knowledge graph;

[0013] S23, Calculate the material property matching degree M, wherein the material property matching degree M is used to determine the degree of consistency between the material property and the material property in the knowledge graph;

[0014] S24, calculate the rationality R of the component connection point relationship, wherein the rationality R of the component connection point relationship is used to determine the degree of conformity between the relationship of each component connection point and the provisions in the knowledge graph;

[0015] S25, Calculate the load distribution rationality L, wherein the load distribution rationality L is used to review the load distribution of the bridge structure in the BIM model;

[0016] S3, determine the review criteria for the BIM model to be reviewed based on the knowledge graph obtained in S1, and then construct a review criteria rule base using the Drools rule base and the review criteria. The review criteria rule base includes the review thresholds for G, M, R, and L.

[0017] S4 compares the results obtained in S2 with the review threshold in S3 to review the BIM model to be reviewed.

[0018] In S11: the types of bridge components include beams, columns, and piers;

[0019] The bridge component attributes include geometric attributes, stress information, standard and specification references, component connection point relationships, and text classification tags, among which:

[0020] The geometric properties include the standard dimensions of bridge components;

[0021] The material properties include the strength characteristics, durability, density, and corrosion resistance of bridge components;

[0022] The connection point relationship of the components is generated by using relationship extraction technology to identify the relationship between bridge components and thus generate the connection method between bridge components.

[0023] The stress information includes the standard stress conditions of bridge components;

[0024] The referenced standards and specifications include extracting design standard numbers and specification clauses from the text of railway bridge standard data;

[0025] The text classification tagging method uses keywords and rules to categorize different text fragments in railway bridge standard data into design standards, construction specifications, or safety regulations.

[0026] In S22, the geometric matching degree G is calculated using the following formula:

[0027]

[0028] Where i is the number of the bridge component; n is the total number of bridge components in the BIM model to be reviewed; v i It is the geometric importance weight of bridge component i, 0 <v i ≤1; L M (i) represents the actual dimensions of bridge component i in the BIM model to be reviewed; L K (i) represents the standard dimensions of bridge component i in the knowledge graph.

[0029] S23 specifically refers to:

[0030] The similarity u of material property k of bridge component i is calculated using the following formula. k (i):

[0031]

[0032] Where i is the number of the bridge component, Let k be the actual value of the k-th material property of bridge component i in the BIM model to be reviewed, where 1 ≤ k ≤ 4. Let be the standard value of the kth material property of bridge component i in the knowledge graph;

[0033] Then, the similarity M of the material property k of the railway bridge BIM model is calculated using the following formula. k :

[0034]

[0035] Where n is the total number of bridge components in the BIM model to be reviewed. The material strength weight of the k-th material property of bridge component i

[0036] The similarity scores for strength properties (M1), durability (M2), density (M3), and corrosion resistance (M4) were calculated.

[0037] Finally, the similarity scores of the four material properties are summed using the following formula to calculate the material property matching degree M:

[0038] M=w1.M1+w2.M2+w3.M3+w4.M4;

[0039] Where: w1, w2, w3, and w4 are the importance weights of the material strength properties, durability, density, and corrosion resistance of the railway bridge BIM model, respectively, 0. <w k ≤1.

[0040] S24 specifically includes the following steps:

[0041] S241, Define the connection matching degree function θ(C) M (j), C K (j)), the matching degree function θ(·) is used to quantify the actual component connection point relationship C. M (j) and the relationship between standard component connection points in the knowledge graph C K The similarity of (j), where j is the number of the component connection point, θ(·)∈[0,1];

[0042] S242, calculate the rationality R of the component connection point relationship in the BIM model to be reviewed using the following formula:

[0043]

[0044] Where m is the total number of component connection points in the BIM model to be reviewed, and z j The importance weight of component connection point j is determined based on actual working conditions, 0 <z j ≤1.

[0045] S25 specifically includes the following steps:

[0046] S251, the self-weight of bridge component i is calculated using the following formula:

[0047] W i =V i .ρ i .g;

[0048] Where i is the number of the bridge component, V i Let ρ be the volume of bridge component i. i Let be the material density of bridge component i, and g be the acceleration due to gravity.

[0049] Then calculate the distributed load q of bridge component i according to the following formula. i :

[0050] q i =A i .p;

[0051] Among them, A i Let p be the cross-sectional area of ​​component i, and p be the unit area load in the design file of the BIM model to be reviewed.

[0052] Importing W into Revit's public plugins iand q i Obtain the actual force value F of the component. M (i);

[0053] S252, calculate the load distribution rationality L of the BIM model to be reviewed using the following formula:

[0054]

[0055] Where n is the total number of bridge components in the BIM model to be reviewed, and s i The load importance weight of bridge component i is determined based on actual working conditions, 0 i ≤1, F M (i) represents the actual stress value of bridge component i in the BIM model to be reviewed, F K (i) represents the standard stress value of bridge component i in the knowledge graph.

[0056] Preferably, in S1: the connection method includes connection relationship, support relationship, dependency relationship, geometric constraint and physical constraint;

[0057] Wherein: the connection relationship is the connection method between bridge components; the support relationship is the mechanical dependence relationship between bridge components; the geometric constraint is the geometric relationship between bridge components, including relative position, angle, and alignment; and the physical constraint is the load-bearing limit and deformation constraint of bridge component materials that conform to physical properties and design requirements.

[0058] Preferably, in S12, the structured dataset is saved as a knowledge graph using the graph database Neo4j.

[0059] Compared with the prior art, the present invention has the following beneficial effects:

[0060] 1. This invention collects data related to railway bridge design, construction, and maintenance. Using a BiLSTM-CRF model, it structures engineering specifications, standards, and regulations into a knowledge graph, creating a comprehensive and systematic railway bridge engineering knowledge base, providing a foundation for subsequent BIM model review. By directly interacting with Revit using a custom API, it extracts BIM model information and establishes a mapping relationship between entities in the knowledge graph and information nodes in the BIM model. This achieves effective integration of BIM model data with rules and standards in the knowledge graph, enabling subsequent BIM model review.

[0061] ​2. The method of this invention, based on bridge engineering design specifications, construction standards, and experiential knowledge from a knowledge graph, formulates a comprehensive set of review rules. These rules are then constructed into a rule base. Using pattern matching, the rule base is invoked to review each bridge component and its attributes in the BIM model, thereby achieving automated and intelligent review of the BIM model. This significantly improves review efficiency and accuracy, reduces interference from human factors, and ensures the compliance and reliability of the final BIM model. Simultaneously, it provides strong data support for project management and decision-making. Attached Figure Description

[0062] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0063] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments.

[0064] like Figure 1 As shown, the knowledge graph-based railway bridge BIM model review method of the present invention includes the following steps:

[0065] S1 involves structuring the standard data for railway bridges and constructing a knowledge graph, including the following steps:

[0066] S11 structures the standard data for railway bridges, forming a structured dataset suitable for knowledge graph construction, which can then be used as a comparison basis in BIM model review. The standard data for railway bridges includes textual data from railway bridge design manuals, construction standards, engineering examples, and expert knowledge. Specifically:

[0067] A combined model of Bidirectional Long Short-Term Memory (BiLSTM) network and Conditional Random Field (CRF) is used to achieve deep analysis and semantic information capture of standard railway bridge data, structuring key information within it. The structured dataset includes bridge component types and bridge component attributes;

[0068] Bridge components include beams, columns, and piers;

[0069] Bridge component attributes include geometric attributes, stress information, standard and specification references, component connection point relationships, and text classification tags, among which:

[0070] 1) Geometric properties include the standard dimensions of bridge components, including length, width, and height;

[0071] 2) Material properties include the strength, durability, density, and corrosion resistance of bridge components;

[0072] 3) The component connection point relationship is generated by using relationship extraction technology to identify the relationships between bridge components. The connection methods include connection relationship, support relationship, dependency relationship, geometric constraint and physical constraint. Among them: connection relationship is the connection method between bridge components; support relationship is the mechanical dependency relationship between bridge components; geometric constraint is the geometric relationship between bridge components, including relative position, angle and alignment; physical constraint is the load-bearing limit and deformation constraint of bridge component materials that meet the physical properties and design requirements.

[0073] 4) Stress information includes the standard stress conditions of bridge components;

[0074] 5) Standard and specification references include extracting design standard numbers and specification clauses from the text of railway bridge standard data;

[0075] 6) Text classification tagging uses keywords and rules to categorize different text fragments in railway bridge standard data into “design standards”, “construction specifications” or “safety regulations” so that entries can be distinguished by category in the knowledge graph.

[0076] S12, Constructing a knowledge graph for railway bridge engineering:

[0077] The structured dataset obtained from S11 is constructed into a knowledge graph to provide a basis for comparison in BIM model review. The specific process is as follows:

[0078] First, the bridge component types in S11 are used as nodes in the knowledge graph, the bridge component attributes are used as child nodes in the knowledge graph, and the connection points of the bridge components are used as edges in the knowledge graph.

[0079] Then, the structured dataset will be saved as a knowledge graph using the graph database Neo4j.

[0080] Knowledge graphs efficiently manage bridge component types and attributes. Each node and edge carries specific labels and attribute information for accurate queries and rapid matching. Neo4j's query optimization capabilities are used to build an index structure, ensuring rapid access to standard component and relationship data during model review.

[0081] The knowledge graph obtained in step S1 is a comprehensive standard data model of railway bridges, which can fully present the attributes of bridge components, the relationships between bridge components, and the constraints.

[0082] S2, Extract and process information from the BIM model to be reviewed:

[0083] By directly interacting with Revit using a custom API, information from the BIM model to be reviewed can be extracted, including the geometric, material, and relational information of bridge components in the model. The specific steps are as follows:

[0084] S21, the API interface used to extract BIM model data to be reviewed, is as follows:

[0085] Installation and project setup: Install Visual Studio and Revit API SDK, and extract the actual dimensions, material properties, connection methods of component connection points, and stress information of bridge components from the BIM model to be reviewed.

[0086] S22, Calculate the geometric matching degree G:

[0087] The geometric consistency of the BIM model to be reviewed is calculated using equation (1). G is used to verify whether the actual dimensions of the bridge components in the BIM model to be reviewed meet the design standards, ensuring the accuracy of the precision and position of each bridge component.

[0088]

[0089] Where i is the bridge component number, n is the total number of bridge components in the BIM model to be reviewed, and v i The geometric importance weight of bridge component i (0) <v i ≤1), L M (i) represents the actual dimensions of bridge component i in the BIM model to be reviewed, L K (i) represents the standard dimensions of bridge component i in the knowledge graph.

[0090] The closer the geometric matching degree G is to 1, the higher the geometric accuracy and the more the geometric structure of the BIM model to be reviewed conforms to the standard.

[0091] S23, Calculate the material property matching degree M:

[0092] Calculate the material matching degree of the BIM model to be reviewed. M is used to check whether the actual material properties of the components in the railway bridge BIM model match the standard material properties specified in the knowledge graph. Material properties include strength characteristics, durability, density, and corrosion resistance. The steps are as follows:

[0093] First, the properties of each material are quantified, including:

[0094] The strength characteristics data are the compressive strength values ​​of the component material (unit: MPa);

[0095] Durability data is quantified using standard durability ratings or expected service life.

[0096] The density data refers to the density value of the component material (unit: kg / m3);

[0097] The corrosion resistance data is quantified by classifying the corrosion resistance grades of materials as A and B into numerical values: A = 1, B = 2.

[0098] Then, a material similarity function is defined to obtain the material similarity u of material property k of bridge component i. k (i), as in equation (2):

[0099]

[0100] in, Let k be the actual material property value of bridge component i in the BIM model to be reviewed, where 1 ≤ k ≤ 4. The specific material properties are strength, durability, density, and corrosion resistance. Let u be the k-th standard material property of bridge component i in the knowledge graph. The smaller the difference, the better. k (i) The closer to 1, the higher the similarity of the material property k of bridge component i.

[0101] Calculate the k-similarity M of material properties in a railway bridge BIM model. k As shown in equation (3):

[0102]

[0103] in, The material strength weight of the k-th material property of bridge component i

[0104] The similarity scores for strength characteristics (M1), durability (M2), density (M3), and corrosion resistance (M4) were calculated.

[0105] The similarity scores of the four material properties are summed to calculate the material property matching degree M, as shown in equation (4):

[0106] M=w1.M1+w2.M2+w3.M3+w4.M4 (4)

[0107] Where: w1, w2, w3, and w4 are the importance weights of the material strength characteristics, durability, density, and corrosion resistance of the railway bridge BIM model, respectively (0, 1, 2, 3, and 4). <w k ≤1). The closer the M value is to 1, the higher the material property matching degree.

[0108] S24, Calculate the rationality R of the connection point relationship of the components:

[0109] Analyze the connection node information between components in the BIM model to be reviewed, check the connection method, and compare the actual connection method with the design specification requirements.

[0110] To determine the degree to which the connection points of bridge components conform to the specifications in the knowledge graph, and to ensure the reliability of the connections between components, the following steps are taken:

[0111] S241, Define the connection matching degree function θ(C) M (j), C K (j)), the matching degree function is used to quantify the actual component connection point relationship C. M (j) and the relationship between standard component connection points in the knowledge graph C K The similarity of (j), where j is the number of the component connection point, and θ(·)∈[0,1].

[0112] Specifically, if the actual connection method is completely consistent with the standard connection method, then θ(C) M (j), C K (j))=1; If the actual connection method is not exactly the same as the standard connection method, but can be considered an acceptable substitute in some scenarios, θ can be set to a specific value between 0 and 1 according to the degree of acceptability to reflect "partial matching"; If the actual connection method is completely different from the standard connection method and does not conform to the design specifications, then set θ(C) to 1. M (j), C K (j))=0.

[0113] To facilitate calculations, a component connection method quantification table is established, containing possible connection method combinations and their corresponding matching degree values. See the table below:

[0114] Table 1. Measurement of Component Connection Methods in Railway Bridge BIM Model

[0115]

[0116]

[0117] S242, calculate the rationality R of the connection point relationship of the BIM model components to be reviewed, as shown in equation (5):

[0118]

[0119] Where m is the total number of component connection points in the BIM model to be reviewed, and z j The importance weight of component connection point j is determined based on the actual working conditions, (0 <z j ≤1). Among them, the closer the value of R is to 1, the more the connection point relationship of the components conforms to the standard.

[0120] S25, Calculate the rationality of load distribution L:

[0121] The stress information of each component in the overall bridge structure extracted from S21 is compared with the design load distribution requirements. The deviation between the actual stress of each component and the design standard stress is calculated, and the load distribution of the bridge structure in the BIM model is reviewed to ensure the stability and safety of the overall structure. The steps are as follows:

[0122] S251, based on geometric data and material properties, calculate the actual load data and calculate the self-weight of bridge component i, as shown in equation (6):

[0123] W i =V i .ρ i .g (6)

[0124] Among them, V i Let ρ be the volume of bridge component i. i Let be the material density of bridge component i, and g be the acceleration due to gravity.

[0125] Calculate the distributed load q of bridge component i i As shown in equation (7):

[0126] q i =A i .p (7)

[0127] Among them, A i Let be the cross-sectional area of ​​component i, and p be the unit area load in the design document.

[0128] Using the Revit plugin Autodesk Robot Structural Analysis, the self-weight W of bridge component i was calculated. i and distributed load q i Import,

[0129] The model was then subjected to stress simulation based on the structural conditions. Through mechanical analysis, the actual stress values ​​F of the components were obtained. M (i). S252, calculate the load distribution rationality L of the BIM model to be reviewed, as shown in equation (8):

[0130]

[0131] Among them, s i The load importance weight of bridge component i is determined based on actual working conditions, 0 i ≤1, F M (i) represents the actual stress value of bridge component i in the BIM model to be reviewed, F K (i) represents the standard stress value of bridge component i in the knowledge graph.

[0132] ​The closer the value of load distribution rationality L is to 0, the smaller the deviation between the actual stress result and the standard stress value, indicating that the load distribution is more reasonable.

[0133] S3 builds a standard rule base for auditing using the Drools rule engine:

[0134] First, the review criteria for the BIM model to be reviewed are determined by using the knowledge graph obtained from the railway bridge standard data in S1.

[0135] The audit criteria are logically formalized into a computer-understandable format (logical expression, rule template, or scripting language) and imported into the Drools rule engine to build a complete audit criteria rule library. This ensures that the rules are expressed accurately and efficiently, so that the Drools rule engine can automatically manage, store, and call the audit criteria in the audit criteria rule library.

[0136] The audit standard rule base includes audit thresholds for G, M, R, and L.

[0137] S4, proceed with the review:

[0138] The results obtained from S2 are compared with the review threshold of S3 to review the BIM model to be reviewed.

Claims

1. A knowledge graph-based auditing method for a railway bridge BIM model, characterized in that, Comprise the following steps: S1, the railway bridge standard data is structured and the knowledge graph is built, comprising: S11, the railway bridge standard data is structured, forming a structured data set suitable for knowledge graph construction, the railway bridge standard data comprising text data in railway bridge design manual, construction standard, engineering example and expert knowledge; the structured data set comprises the type and attribute of each bridge component in the railway bridge: S12, the structured data set obtained in S11 is constructed into the form of a knowledge graph, for complete presentation of the attributes of each bridge component in the railway bridge, the relationship between bridge components and the constraint conditions; S2, extracting and processing the information in the BIM model to be audited, comprising the following steps: S21, extracting the actual size, material attribute and connection mode of the component connection point of each component of the railway bridge in the BIM model to be audited, and the stress information of the bridge component; S22, calculating the geometric matching degree G, which is used to judge the degree of conformity of the actual size with the provisions in the knowledge graph; S23, calculating the material attribute matching degree M, which is used to judge the degree of conformity of the material attribute with the material attribute in the knowledge graph; S24, calculating the component connection point relationship rationality R, which is used to judge the degree of conformity of the relationship of each component connection point with the provisions in the knowledge graph; S25, calculating the load distribution rationality L, which is used to review the load distribution of the bridge structure in the BIM model; S3, determining the audit standard of the BIM model to be audited through the knowledge graph obtained in S1, and constructing an audit standard rule base through the Drools rule base and the audit standard, the audit standard rule base comprising the audit threshold values of G, M, R and L; S4, comparing the results obtained in S2 with the audit threshold values in S3 to audit the BIM model to be audited.

2. The knowledge graph based auditing method of railway bridge BIM model according to claim 1, characterized in that, In S11: The bridge component type comprises beam, column and pier; The bridge component attribute comprises geometric attribute, stress information, standard specification reference, component connection point relationship and text classification mark, wherein: The geometric attribute comprises the standard size of the bridge component; The material attribute comprises the strength characteristic, durability, density and corrosion resistance of the bridge component; The component connection point relationship is to generate the connection mode between bridge components by identifying the relationship between bridge components using relationship extraction technology; The stress information comprises the standard stress condition of the bridge component; The standard specification reference comprises extracting the design standard number and specification clause from the text of the railway bridge standard data; The text classification mark is to classify different text fragments in the railway bridge standard data into design standards, construction specifications or safety regulations through keyword and rule methods.

3. The method for auditing a knowledge graph-based railway bridge BIM model according to claim 2, characterized in that, In S22, the geometric matching degree G is calculated by the following formula: wherein i is the number of the bridge member; n is the total number of bridge members in the BIM model to be audited; v i is the geometric importance weight of the bridge member i, 0 < v i ≤ 1; L M (i) is the actual size of the bridge member i in the BIM model to be audited; L K (i) is the standard size corresponding to the bridge member i in the knowledge graph.

4. The method for auditing the railway bridge BIM model based on the knowledge graph according to claim 2, characterized in that, S23 is specifically: The similarity u of the material properties k of the bridge member i is calculated by the following formula k (i): wherein i is the number of the bridge member, is the actual value of the kth material property of the bridge member i in the BIM model to be audited, 1≤k≤4, is the standard value of the kth material property of the bridge member i in the knowledge graph; The similarity M of the material property k of the railway bridge BIM model is calculated again by the following formula k : Wherein, n is the total number of bridge components in the BIM model to be audited, material strength importance weight of the kth material attribute of the bridge component i The strength characteristic similarity M1, the durability similarity M2, the density similarity M3 and the corrosion resistance similarity M4 are calculated; Finally, the four material attribute similarities are added together to calculate the material attribute matching degree M by the following formula: M = w1.M1 + w2.M2 + w3.M3 + w4.M4; Wherein: w1, w2, w3, w4 are the importance weights of the material strength characteristics, durability, density and corrosion resistance of the railway bridge BIM model respectively, 0 k ≤1.

5. The method for auditing a knowledge graph based railway bridge BIM model as claimed in claim 2, wherein, S24 specifically comprises the following steps: S241, defining a connection mode matching degree function θ(C M (j), C K (j)), the matching degree function θ(·) being used to quantify the similarity of the actual component connection point relationship C M (j) and the standard component connection point relationship C K (j) in the knowledge graph, wherein j is the number of the component connection point, and θ(·) ∈ [0, 1]. S242, calculate the component connection point relationship rationality R of the BIM model to be audited by the following formula: Wherein, m is the total number of component connection points in the BIM model to be audited, z j is the importance weight of the component connection point j, determined according to the actual working condition, 0<z j ≤1.

6. The method for auditing a knowledge graph based railway bridge BIM model as claimed in claim 2, wherein, S25 specifically comprises the following steps: S251, calculate the self-weight of the bridge component i by the following formula: W i = V i . p i . g; where i is the number of the bridge element, V i is the volume of the bridge element i, p i is the material density of the bridge element i, g is the gravitational acceleration; The distributed load q of the bridge member i is calculated again according to the following formula i : q i = A i .p; where A i is the cross-sectional area of the member i, and p is the load per unit area in the design file of the BIM model under review; Importing W into the public plug-in of Revit i and q i , obtaining the actual force value F of the component M (i); S252, calculate the load distribution rationality L of the BIM model to be audited by the following formula: Wherein, n is the total number of bridge components in the BIM model to be audited, s i is the load importance weight of the bridge component i, determined according to the actual working condition, 0 < s i ≤ 1, F M (i) is the actual stress value of the bridge component i in the BIM model to be audited, F K (i) is the standard stress value of the bridge component i in the knowledge graph.

7. The method for auditing a knowledge graph-based railway bridge BIM model according to claim 2, characterized in that: In S1: the connection mode includes connection relationship, support relationship, dependency relationship, geometric constraint and physical constraint; Wherein: the connection relationship is the connection mode between the bridge components; the support relationship is the mechanical dependency relationship between the bridge components; the geometric constraint is the geometric relationship between the bridge components, including relative position, angle, alignment; the physical constraint is the load bearing limit and deformation constraint of the bridge component material conforming to the physical properties and design requirements.

8. The method for auditing a knowledge graph-based railway bridge BIM model according to claim 2, characterized in that: In S12, the structured data set is saved as a knowledge graph by using a graph database Neo4j.

Images