Building component attribute matching method and system based on BIM

Abstract

The invention discloses a BIM (Building Information Modeling)-based building component attribute matching method and a BIM-based building component attribute matching system, contour line segments of a target and a reference component are segmented, curvature change values are recorded, and the local feature expression capability is enhanced. Then, dynamic adjustment parameters are introduced to optimize eigenvector weighting processing, so that the matching process adapts to different shape complexity, and mismatching is avoided; and then, identifying a local similar region by comparing the corrected matching parameter set, and generating a registration mapping relation, thereby improving the accuracy of space consistency judgment. A matching priority matrix is constructed based on the mapping relation and the shape distribution difference, high-matching-degree candidate components are quickly screened, and the matching efficiency is improved. Finally, a feedback adjustment mechanism is adopted in the attribute verification stage, it is ensured that the matching result meets the preset precision requirement, and self-adaptive optimization is achieved. The method effectively solves the problems of low precision and poor stability in non-standard geometric component matching in the prior art, and has important practical value.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of architectural design, and particularly relates to a building component attribute matching method and system based on BIM. BACKGROUND

[0002] In the application of building information modeling (BIM), the attribute matching of building components is a key link to realize component identification, model updating and standardized management. The existing technology usually matches based on the basic geometric parameters (such as length, width and height) or simple shape features of the components, and relies on a preset rule library or fixed template for comparison. When facing components with complex shapes or non-standard geometric shapes, this method often fails to accurately extract features and establish effective matching relationships, resulting in large matching result errors, and even mis-matching.

[0003] Especially when dealing with components with curved surfaces, multi-segment line combinations or irregular boundary structures, the traditional method lacks effective means to depict local shape changes, and cannot fully reflect the spatial similarity and difference between components, thereby affecting the overall matching accuracy and application scope. SUMMARY

[0004] The purpose of the present application is to provide a building component attribute matching method and system based on BIM, which extracts component contour line segments and constructs feature sequences, optimizes matching parameters with a dynamic adjustment mechanism, significantly improves the matching accuracy of complex shape components, and solves the problems raised in the background technology.

[0005] To achieve the above purpose, the present application adopts the following technical solution: a building component attribute matching method based on BIM, comprising the following steps: Obtain the three-dimensional geometric data of the target component and the reference component, and extract the contour line segment set of the target component and the reference component, segment the contour line segment set, generate a feature sequence composed of multiple discrete points, and record the curvature change value of each line segment; Based on the feature sequence, calculate the shape distribution difference value of the target component and the reference component, and generate the corresponding feature vector, introduce a dynamic adjustment parameter, weight process the feature vector, and obtain the corrected matching parameter set; By comparing the corrected matching parameter set, determine the local similar area of the target component and the reference component, calculate the position offset of the target component and the reference component according to the local similar area, and generate a registration mapping relationship; Combine the registration mapping relationship and the shape distribution difference value to construct a matching priority matrix, and screen out high matching degree candidate components; The high matching degree candidate component is subjected to attribute verification, and if the matching accuracy does not meet the standard, the dynamic parameters are returned for adjustment until the preset matching threshold is met.

[0006] Preferably, the contour line segment set of the target component and the reference component is extracted, including: The triangular mesh representation of the target component and the reference component is extracted from the BIM model, and the spatial coordinates of each vertex are recorded; The edge detection is performed on the triangular mesh, and the edges with a normal angle difference between adjacent surfaces greater than a set threshold are identified as candidate boundary edges; The candidate boundary edges are connected in topological order to form continuous polyline, and the linear fitting is performed to obtain the parameters of each segment line; The change rate of the included angle between the segment lines is calculated, and the line segments with a change rate less than a set value are selected to form a contour line segment set.

[0007] Preferably, a feature sequence composed of multiple discrete points is generated, including: Each line segment in the contour line segment set is uniformly divided into multiple equal segments in the length direction, and a corresponding three-dimensional coordinate point sequence is generated at each equal division point; The direction change amount between adjacent three points in the coordinate point sequence is calculated in a sliding window manner, and the local curvature change value is determined according to the direction change amount; The curvature change values are arranged in order to form a feature sequence, and a threshold is set to remove points in low change areas and retain key points in high change areas to form a final feature sequence.

[0008] Preferably, a corresponding feature vector is generated, including: The feature sequence of the target component and the feature sequence of the reference component are respectively subjected to normalization processing to obtain point sequences with consistent lengths; The spatial distance between corresponding points in the point sequences is calculated one by one to form a distance set, and the overall shape distribution difference value is calculated according to the distance set; The overall shape distribution difference value and the direction change information of each point are combined to form a multi-dimensional feature vector containing spatial position and shape change information.

[0009] Preferably, the feature vector is subjected to weighted processing to obtain a corrected matching parameter set, including: An initial weight factor is set, and the same weight is given to each item in the feature vector to form an initial weight sequence; According to the current matching level and the preset maximum level, a dynamic decay factor is calculated, and some elements in the weight sequence are updated; Multiply the updated weight sequence with the feature vector corresponding to obtain a weighted parameter group, normalize the weighted parameter group to obtain a corrected matching parameter set.

[0010] Preferably, determining the local similar region of the target component and the reference component comprises: Calculating the difference between the corrected matching parameter set of the target component and the matching parameter set of the reference component to obtain a difference sequence; Setting a similarity threshold value, screening out parameter pairs that meet the conditions, and recording the corresponding index position to form a candidate matching index group; Identifying continuous index segments in the candidate matching index group, each segment having a length not less than a set minimum length, as potential local similar segments; Verifying the spatial consistency of the local similar segments, and if the index interval between adjacent segments is less than a set value, merging them to form the final local similar region.

[0011] Preferably, generating a registration mapping relationship comprises: Extracting the three-dimensional coordinates of each matching point in the local similar region in the target component and the reference component to form two point sets respectively; Calculating the average center point deviation between the point sets to obtain a deviation vector, and performing spatial translation correction on the target component based on the deviation vector; Establishing a one-to-one correspondence between the corrected points and the corresponding reference points to form a registration mapping table.

[0012] Preferably, constructing a matching priority matrix to screen out high matching degree candidate components comprises: According to the registration mapping relationship, extracting the spatial error between each matching point pair to form an error sequence; Fusing the error sequence with the corresponding shape distribution difference value to calculate the comprehensive deviation of each matching group; Taking the component number as the row and the matching position as the column, constructing a matching priority matrix, and the matrix elements reflect the matching strength of each position; Selecting the positions corresponding to the two smallest values in each row in the matching priority matrix, and marking the associated components as a high matching degree candidate component set.

[0013] Preferably, verifying the attributes of the high matching degree candidate components comprises: Extracting the geometric attributes and design attributes of the high matching degree candidate components, and comparing them with the corresponding attributes of the target component item by item to generate an attribute deviation set, and calculating the comprehensive error rate according to the attribute deviation set; Setting a matching accuracy threshold value, if the comprehensive error rate exceeds the threshold value, triggering a parameter adjustment mechanism to increase the weight concentration; The weighting processing procedure is re-executed using the updated weights, and the matching verification procedure is entered again until the matching accuracy requirement is met.

[0014] In another aspect, the present application provides a BIM-based building component attribute matching system, comprising the following steps: A sequence construction module is configured to obtain three-dimensional geometric data of a target component and a reference component, extract a contour line segment set of the target component and the reference component, perform segmentation processing on the contour line segment set, generate a feature sequence composed of a plurality of discrete points, and record a curvature change value of each line segment; A parameter optimization module is configured to calculate a shape distribution difference value of the target component and the reference component based on the feature sequence, generate a corresponding feature vector, introduce a dynamic adjustment parameter, perform weighting processing on the feature vector, and obtain a corrected matching parameter set; A space mapping analysis module is configured to determine a local similar region of the target component and the reference component by comparing the corrected matching parameter set, calculate a position offset of the target component and the reference component according to the local similar region, and generate a registration mapping relationship; A candidate screening module is configured to construct a matching priority matrix in combination with the registration mapping relationship and the shape distribution difference value, and screen out a high-matching-degree candidate component; A parameter feedback adjustment module is configured to perform attribute verification on the high-matching-degree candidate component, return to adjust the dynamic parameter if the matching accuracy does not meet the standard, and until a preset matching threshold is met.

[0015] The present application has the following advantages compared with the prior art: The present application significantly improves the accuracy and stability of complex shape component matching through fine geometric feature extraction and dynamic parameter adjustment. First, the contour line segments of the target and reference components are segmented and the curvature change value is recorded, enhancing the local feature expression capability. Then, dynamic adjustment parameters are introduced to optimize the feature vector weighting processing, making the matching process adapt to different shape complexities and avoiding mis-matching. Subsequently, the local similar region is identified by comparing the corrected matching parameter set, and the registration mapping relationship is generated, improving the accuracy of spatial consistency judgment. Based on this mapping relationship and shape distribution difference, a matching priority matrix is constructed to quickly screen high-matching-degree candidate components, improving the matching efficiency. Finally, a feedback adjustment mechanism is adopted in the attribute verification stage to ensure that the matching result meets the preset accuracy requirement, realizing adaptive optimization. This method effectively solves the problems of low accuracy and poor stability in non-standard geometric component matching in the prior art, and has important practical value. BRIEF DESCRIPTION OF DRAWINGS

[0016] Figure 1 A structural diagram of the present application; Figure 2 A structural diagram of the present application. DETAILED DESCRIPTION

[0017] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. The specific embodiments described here are only used to explain the present application, and are not used to limit the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work belong to the scope of protection of the present application.

[0018] The present application provides a BIM-based building component attribute matching method as shown in Figure 1 The present application provides a BIM-based building component attribute matching method as shown in Step one: obtaining the three-dimensional geometric data of the target component and the reference component, and extracting the contour line segment set of the target component and the reference component; specifically including the following steps: Extract the triangular mesh representation of the target component and the reference component from the BIM model, and record the spatial coordinates (x, y, z) of each vertex as the basic data source for subsequent geometric analysis.

[0019] Edge detection is performed on the triangular mesh, and edges with a normal angle difference between adjacent faces greater than a set threshold θ are identified as candidate boundary edges; by calculating the normal vector angle of adjacent triangular facets, it is determined whether the two faces form a significant turning. Let the normal vectors of the two faces be and , then the angle When θ> θ (set threshold), the edge is considered as a boundary edge.

[0020] The candidate boundary edges are connected in topological order to form continuous polylines, and the parameters of each segment are obtained by straight line fitting, wherein any segment AB is determined by two points ; the unordered boundary edges are converted into a structured line segment sequence, which establishes a unified description basis for subsequent feature extraction and matching.

[0021] The angle change between each segment is calculated, and the line segments with a change rate less than a set value ε are selected to form a contour line segment set.

[0022] Step two: segmenting the contour line segment set to generate a feature sequence composed of multiple discrete points, and recording the curvature change value of each segment; specifically including the following steps: Divide each line segment in the contour line segment set into n equal segments along the length direction, and generate a corresponding three-dimensional coordinate point sequence at each equal point , i=1 to p; The sliding window method is used to calculate the direction change between three adjacent points in the coordinate point sequence. The window length is set to 3, and any three consecutive points are selected. , calculate its in-plane rotation angle: , use the vector dot product formula to calculate the direction angle of the line connecting two adjacent points, reflecting the local path turning situation. Among them, and The direction vectors of the front and back segments are divided by their modulus to obtain the unit vector, and the dot product result is used to solve the angle .

[0023] According to the direction change Determine the local curvature change value ,in for and The length of the chord between the three adjacent points; the angle between the three adjacent points Divide by the straight line distance between the first and last points (i.e. the chord length ) to represent the intensity of the direction change within a unit length, thereby defining the local curvature change value at the point .

[0024] Arrange the curvature change values ​​k_i in order to form a feature sequence, and set the threshold ε to eliminate For points with a value smaller than ε, the key points in the high-variation area are retained to form the final feature sequence, which effectively compresses the data size and highlights the key parts of the shape, improves the efficiency and accuracy of subsequent matching, and is particularly suitable for the rapid identification and comparison of non-standard components.

[0025] Step 3: Based on the feature sequence, calculate the shape distribution difference between the target component and the reference component, and generate a corresponding feature vector; specifically, the following steps are included: Normalize the feature sequence of the target component and the feature sequence of the reference component respectively to obtain a point sequence with the same length and ; Point sequence and Calculate the Euclidean distance of the corresponding points one by one , forming a distance set D; using the Euclidean distance formula between two points in three-dimensional space to measure the geometric distance between corresponding points, this distance reflects the degree of morphological difference between the two sequences at the corresponding position. )and Represent the spatial coordinates of the i-th point in the target sequence and reference sequence respectively.

[0026] Calculate the overall shape distribution difference value according to the distance set D ; sum the Euclidean distances between all corresponding point pairs and take the average as the overall shape distribution difference indicator E, the smaller the value, the closer the shapes of the two components. The parameter p represents the number of points in the sequence, is the cumulative sum of the distances of all point pairs.

[0027] Combine the difference value E with the directional change amount of each point to form a multi-dimensional feature vector containing spatial position and shape change information . Combine the overall shape difference value E and the local point pair distance Construct a multi-dimensional feature vector V that not only considers the consistency of macroscopic shape, but also takes into account the matching of microscopic detail features. This combination enhances the feature description capability, so that subtle differences between different components can also be identified.

[0028] Step four: introduce a dynamic adjustment parameter to weight the feature vector and obtain a corrected matching parameter set; specifically including the following steps: Set the initial weight factor w, assign the same weight to each item in the feature vector V according to the position, and construct the initial weight sequence W = [w, w, …, w]; at the beginning of the matching process, all feature dimensions (such as overall shape difference, point distance, etc.) are considered to be of equal importance, so an initial weight factor w is uniformly set, and each element of the feature vector V is assigned the same weight value, and the initial weight sequence W is constructed.

[0029] According to the current matching level l and the preset maximum level L, calculate the dynamic attenuation factor , and update the elements in the second half of the weight sequence to ; the dynamic attenuation factor decreases with the increase of the level l, which is used to adjust the value of the second half of the weight sequence. In the formula, L is the preset maximum matching level.

[0030] Multiply the updated weight sequence W and the feature vector V to obtain the weighted parameter group ; multiply the weight value corresponding to each feature dimension with the feature value in that dimension to form a new weighted parameter group A, which is an importance adjustment based on the weight of the original feature vector.

[0031] Normalize the weighted parameter group to obtain the corrected matching parameter set , where S is the sum of all elements in A.

[0032] Step five: determine the local similar area of the target component and the reference component by comparing the corrected matching parameter set; specifically including the following steps: Compare the corrected matching parameter set of the target component a set of matching parameters of the reference component performing the difference calculation item by item to obtain a difference sequence ; wherein and respectively represent the matching parameters of the target component and the reference component at the i-th position.

[0033] set a similarity threshold δ, filter out the parameter pairs satisfying <δ, and record the corresponding index positions to form a candidate matching index group I; effectively filter out the non-matching regions with large differences, and focus on the potential matching regions.

[0034] identify continuous index segments in the candidate matching index group I, each segment having a length not less than a set minimum length l, as potential local similar segments; eliminate scattered and isolated matching points, and extract continuous segments with actual structural meaning. Perform spatial consistency verification on the local similar segments, and if the index interval between adjacent segments is less than g, merge them to form the final local similar region.

[0035] Step six: according to the local similar region, calculate the position offset of the target component and the reference component, and generate a registration mapping relationship; specifically including the following steps: extract the three-dimensional coordinates of each matching point in the local similar region in the target component and the reference component to form point sets and ; extract the corresponding geometric point positions in the target component and the reference component, and each point is represented by a three-dimensional spatial coordinate (x, y, z) to form two point sets T (target) and R (reference) for spatial alignment analysis.

[0036] calculate the average center point deviation between the point sets T and R to obtain the offset vector , i=1 to q; this formula obtains an offset vector ΔP representing the overall translation trend by summing the coordinate differences of the corresponding points in the reference point set R and the target point set T and then taking the average, where q is the total number of matching points, represents the position difference of the i-th point pair in the three-dimensional space.

[0037] perform spatial translation correction on the target component based on the offset vector ΔP, so that ; perform a uniform translation operation on all matching points of the target component to make it approach the reference component in the three-dimensional space, and complete the preliminary spatial alignment.

[0038] establish a one-to-one correspondence between the corrected point and the corresponding to form a registration mapping table . After completing the spatial translation correction, establish a one-to-one correspondence between each corrected point The original point corresponding to it in the reference component Pairing, constructing a structured mapping table to record the precise spatial correspondence between the target and the reference component.

[0039] Step seven: Combine the registration mapping relationship and the shape distribution difference value to construct a matching priority matrix, and screen out high-matching-degree candidate components; specifically including the following steps: According to the registration mapping relationship, the spatial distance error between each matching point pair is extracted , to form an error sequence ; The error sequence E and the corresponding shape distribution difference value D are normalized and fused to calculate the comprehensive deviation of each matching group , where λ is the balance factor; this step introduces the balance factor λ to linearly combine the error sequence E reflecting the spatial position difference and the value D describing the shape feature difference, to obtain a deviation index considering both factors , which is used to more comprehensively evaluate the quality of matching. The balance factor λ is used to adjust the relative importance of the two indicators in the final score.

[0040] Construct a matching priority matrix P with component number as row and matching position as column, and the matrix element corresponds to the matching strength of the jth candidate component; the best matching position of each component can be quickly identified by comparing the numerical values in the matrix.

[0041] Select the positions corresponding to the two smallest values in each row in the matching priority matrix P, and mark the associated components as the high-matching-degree candidate component set C, effectively reducing the possible matching range and focusing on the components most likely to successfully pair.

[0042] Step eight: Perform attribute verification on the high-matching-degree candidate components, and if the matching accuracy does not meet the standard, return to adjust the dynamic parameters until the preset matching threshold is met; specifically including the following steps: Extract the geometric properties and design properties of the high-matching-degree candidate components, and compare them one by one with the corresponding properties of the target component to generate an attribute deviation set ; this step systematically extracts the key attributes of the candidate components (such as length, width, height, material, purpose, etc.) and compares them one by one with the corresponding attributes of the target component, calculates the difference between each attribute, and forms a deviation set ΔA that describes the overall attribute difference.

[0043] Calculate the comprehensive error rate according to the attribute deviation set ΔA; this formula takes the average of the sum of all attribute deviations to obtain a comprehensive error rate ρ that reflects the overall attribute difference between the candidate component and the target component, where a represents the total number of attributes involved in the comparison represents the sum of each attribute deviation.

[0044] A matching accuracy threshold τ is set, if ρ>τ, the parameter adjustment mechanism is triggered, the weight concentration in the dynamic adjustment parameter is increased, and the wherein α is an adjustment proportion factor; the weighting process is re-executed using the updated weight sequence, and the matching verification process is entered again, and the above operations are repeated until ρ≤τ. A closed-loop matching process is constructed to ensure that the final output matching result always meets the accuracy standard required by engineering applications, and is suitable for intelligent decision-making scenarios of component identification and replacement in large-scale BIM models.

[0045] On the other hand, the present application provides a BIM-based building component attribute matching system, as shown in Figure 2 The system comprises the following steps: A sequence construction module is configured to obtain three-dimensional geometric data of a target component and a reference component, extract a contour line segment set of the target component and the reference component, perform segmentation processing on the contour line segment set, generate a feature sequence composed of a plurality of discrete points, and record the curvature change value of each line segment. A parameter optimization module is configured to calculate a shape distribution difference value of the target component and the reference component based on the feature sequence, generate a corresponding feature vector, introduce a dynamic adjustment parameter, and perform weighting processing on the feature vector to obtain a corrected matching parameter set. A spatial mapping analysis module is configured to determine a local similar area of the target component and the reference component by comparing the corrected matching parameter set, calculate a position offset of the target component and the reference component according to the local similar area, and generate a registration mapping relationship. A candidate screening module is configured to construct a matching priority matrix in combination with the registration mapping relationship and the shape distribution difference value, and screen out a high-matching-degree candidate component. A parameter feedback adjustment module is configured to perform attribute verification on the high-matching-degree candidate component, and return to adjust the dynamic parameter until the preset matching threshold is met if the matching accuracy does not meet the standard.

[0046] In addition, each of the above modules is further configured to implement other steps of the above-mentioned BIM-based building component attribute matching method when executed, as follows: Taking the matching of a prefabricated beam component in a certain fabricated building project as an example, the specific operation process and calculation process of the BIM-based building component attribute matching method provided by the present application in actual engineering are demonstrated. I. Introduction In a certain prefabricated building project, the designer needs to quickly find a similar component to the current target beam component from the component library for replacement. Due to the complex shape of the component (such as with variable cross-section, hole, etc.), the traditional matching method based on size or type label is easy to misjudge. Therefore, the method described in the application is used for high-precision automatic matching.

[0047] II. Data preparation and input Target component: No. L01, located in the current model, its geometric information is extracted through IFC format.

[0048] Reference component library: contains 5 candidate beam components (L11-L15).

[0049] All components are imported as triangular mesh, and the vertex coordinates have been extracted.

[0050] III. Step execution and calculation example Step 1: Obtain three-dimensional geometric data and extract contour line segment set Extract the triangular mesh representation of the target component L01 and each reference component, and record the spatial coordinates (x, y, z) of each vertex; Edge detection is performed on the mesh, and the normal angle difference threshold θ=30° is set to identify all boundary edges; Connect the boundary edges into continuous polylines and fit them into multiple straight lines; Calculate the angle change rate between each line segment , filter <ε (ε=10°) line segment to form the final contour line segment set.

[0051] Example: Assume that L01 extracts 8 main contour line segments, of which 4 meet <10°, forming the final contour line segment set.

[0052] Step 2: Generate feature sequence and record curvature change value Divide the above 4 line segments into n=10 segments in length direction to get p=40 discrete points; Use a sliding window (length 3) to calculate the in-plane rotation angle between adjacent three points : ; Calculate the local curvature change value , where is the chord length between and ; Set the curvature threshold ε=0.1, remove <ε points, and keep the key points to form the final feature sequence.

[0053] Example result: L01 finally retains a feature sequence composed of 15 key points.

[0054] Step three: Calculate shape distribution difference value, generate feature vector Normalize target component L01 and reference component L11 respectively to obtain point sequences of consistent length And ; Calculate the Euclidean distance between corresponding points : ; Calculate the overall shape distribution difference value ; Assuming that for L01 and L11, E=0.12 is calculated; Construct a feature vector .

[0055] Example feature vector (only the first 5 items are listed): V=[0.12, 0.11, 0.13, 0.10, 0.14].

[0056] Step four: Introduce dynamic adjustment parameters, weighted processing to generate a corrected parameter set Set the initial weight w=0.5, construct the initial weight sequence W=[0.5, 0.5,..., 0.5]; The current matching level l=1, the maximum level L=3, calculate the dynamic decay factor: =1-(1 / 3)=0.6667; Update the weight of the second half: =0.5*0.6667=0.333; After weighting , and after normalization, the corrected matching parameter set M is obtained.

[0057] Example corrected parameter set (first 5 items): M=[0.12*0.5=0.06, 0.11*0.5=0.055, 0.13*0.5=0.065, 0.10*0.333=0.033, 0.14*0.333=0.047].

[0058] Step five: Determine the local similar region Compare the corrected parameter set of L01 and L11 And , calculate the difference ; Set the similarity threshold δ=0.05, filter Points with a difference less than δ form the candidate index group I; Identify consecutive index segment length ≥ l = 3 as potential local similar segments; If the interval between adjacent segments < g = 2, merge to form the final local similar region.

[0059] Example result: two consecutive matching segments are found, with lengths of 5 and 4 respectively, and after merging, a continuous region is formed.

[0060] Step six: generate registration mapping relationship Extract the local similar region And q = 9; Calculate the offset vector ; Assume ΔP = (1.1, -0.3, 0.0); Correct the target point: ; Build a registration mapping table .

[0061] Step seven: build a matching priority matrix and screen high matching components Extract the spatial error of the registration point pair , form the error sequence E; Combine the shape distribution difference D = E (reuse) with the balance factor λ = 0.5 to calculate the comprehensive deviation: .

[0062] Build a matching priority matrix P, with row numbers representing component numbers and column numbers representing matching positions; Select the two components corresponding to the smallest value in each row, and mark them as the high matching candidate set C.

[0063] Example result: C = {L11, L13}.

[0064] Step eight: attribute verification and parameter feedback adjustment Extract the geometric attributes (length, width, height, material) and design attributes (purpose, load level) of L11 and L13; Compare with L01 item by item to generate attribute deviation set ; Calculate the comprehensive error rate ; Assume ρ = 0.15; Pre-set matching accuracy threshold τ = 0.1, ρ > τ, trigger parameter adjustment mechanism: , α = 0.2; Original = 0.5, new = 0.5 + 0.2 * (1 - 0.5) = 0.6; Back to step four re-weighting, matching, verification, until p < t.

[0065] After the second iteration p = 0.09, meet the conditions, output L11 is the best matching component.

[0066] This embodiment fully demonstrates how to use the method described in the application, from geometry extraction, feature construction, parameter optimization, spatial registration, priority sorting to attribute verification, finally successfully select the component L11 from the component library that best meets the requirements of the target component, the whole process is highly automated, high matching accuracy, suitable for large-scale BIM environment intelligent component retrieval and replacement task.

[0067] Finally, it should be noted that: the above only for the preferred embodiments of the present application, and not for the purpose of limiting the present application, although the foregoing embodiments of the present application are described in detail, for those skilled in the art, it still can be modified, or part of the technical features of the equivalent replacement, within the spirit and principles of the present application, any modification, equivalent replacement, improvement, etc., should be included within the scope of the present application.

Claims

1. A building component attribute matching method based on BIM, characterized in that: The following steps are involved: Acquiring three-dimensional geometric data of a target component and a reference component, extracting a set of contour line segments of the target component and the reference component, segmenting the set of contour line segments to generate a feature sequence consisting of multiple discrete points, and recording a curvature change value of each line segment; Based on the feature sequence, the shape distribution difference between the target component and the reference component is calculated, and a corresponding feature vector is generated. Dynamic adjustment parameters are introduced to perform weighted processing on the feature vector to obtain a revised matching parameter set; By comparing the corrected matching parameter set, the local similarity region between the target component and the reference component is determined, and based on the local similarity region, the position offset between the target component and the reference component is calculated, and a registration mapping relationship is generated; Combining the registration mapping relationship with the shape distribution difference value, a matching priority matrix is ​​constructed to screen out candidate components with high matching degree; The attributes of the high-matching candidate components are verified. If the matching accuracy does not meet the requirements, the dynamic parameters are adjusted until the preset matching threshold is met.

2. A BIM-based building component attribute matching method according to claim 1, characterized in that: Extracting a set of contour line segments of the target component and the reference component, including: Extract the triangular mesh representation of the target component and the reference component from the BIM model and record the spatial coordinates of each vertex; Performing edge detection on the triangular mesh to identify edges whose normal angle difference between adjacent faces is greater than a set threshold as candidate boundary edges; Connect the candidate boundary edges into a continuous polyline in topological order, and obtain the parameters of each segment by straight line fitting; The change rate of the angle between the line segments is calculated, and the line segments with a change rate less than a set value are screened out to form a contour line segment set.

3. The BIM-based building component attribute matching method according to claim 1, characterized in that: Generate a feature sequence consisting of multiple discrete points, including: Each line segment in the contour line segment set is evenly divided into a plurality of equal segments along the length direction, and a corresponding three-dimensional coordinate point sequence is generated at each equal segment point; Calculating the direction change between three adjacent points in the coordinate point sequence using a sliding window method, and determining the local curvature change value according to the direction change; The curvature change values ​​are arranged in order to form a feature sequence, and a threshold is set to eliminate points in low-change areas, and key points in high-change areas are retained to form the final feature sequence.

4. The BIM-based building component attribute matching method according to claim 1, characterized in that: Generate the corresponding feature vector, including: Normalize the feature sequence of the target component and the feature sequence of the reference component respectively to obtain point sequences with the same length; Calculating spatial distances for corresponding points in the point sequence one by one to form a distance set, and calculating an overall shape distribution difference value based on the distance set; The overall shape distribution difference value is combined with the direction change information of each point to form a multi-dimensional feature vector containing spatial position and shape change information.

5. The BIM-based building component attribute matching method according to claim 1, characterized in that: The feature vectors are weighted to obtain a modified matching parameter set, including: Setting an initial weight factor, assigning the same weight to each item in the feature vector, and forming an initial weight sequence; Calculate the dynamic attenuation factor based on the current matching level and the preset maximum level, and update some elements in the weight sequence; The updated weight sequence is multiplied by the corresponding feature vector to obtain a weighted parameter group, and the weighted parameter group is normalized to obtain a revised matching parameter set.

6. The BIM-based building component attribute matching method according to claim 1, characterized in that: Determine the local similarity between the target component and the reference component, including: Calculate the difference between the corrected matching parameter set of the target component and the matching parameter set of the reference component item by item to obtain a difference sequence; Set a similarity threshold, filter out parameter pairs that meet the conditions, and record the corresponding index positions to form a candidate matching index group; Identifying consecutive index segments in the candidate matching index group, each segment having a length not less than a set minimum length, as potential locally similar segments; The spatial consistency of the local similar segments is verified. If the index interval between adjacent segments is less than a set value, they are merged to form a final local similar region.

7. The BIM-based building component attribute matching method according to claim 1, characterized in that: Generate registration mapping relationships, including: Extracting the three-dimensional coordinates of each matching point in the local similarity area in the target component and the reference component to form two groups of point sets respectively; Calculating the average center point deviation between the point sets to obtain an offset vector, and performing spatial translation correction on the target component based on the offset vector; A one-to-one correspondence is established between the corrected points and the corresponding reference points to form a registration mapping table.

8. The BIM-based building component attribute matching method according to claim 4, characterized in that: Build a matching priority matrix to screen out highly matching candidate components, including: According to the registration mapping relationship, the spatial error between each matching point pair is extracted to form an error sequence; Fusing the error sequence with the corresponding shape distribution difference value to calculate the comprehensive deviation of each set of matches; A matching priority matrix is ​​constructed with component numbers as rows and matching positions as columns. The matrix elements reflect the matching strength of each position. The positions corresponding to the two smallest values ​​in each row of the matching priority matrix are selected, and the associated components are marked as a set of high-matching candidate components.

9. The BIM-based building component attribute matching method according to claim 1, characterized in that: Performing attribute verification on the candidate component with a high matching degree includes: Extracting the geometric properties and design properties of the candidate components with high matching scores, comparing them item by item with the corresponding properties of the target components, generating a property deviation set, and calculating a comprehensive error rate based on the property deviation set; Setting a matching accuracy threshold. If the comprehensive error rate exceeds the threshold, a parameter adjustment mechanism is triggered to increase the weight concentration. The weighted processing process is re-executed using the updated weights, and the matching verification process is entered again until the matching accuracy requirements are met.

10. A BIM-based building component attribute matching system for implementing the method according to any one of claims 1 to 9, characterized in that: The following steps are involved: A sequence construction module is used to obtain the three-dimensional geometric data of the target component and the reference component, extract the contour line segment set of the target component and the reference component, segment the contour line segment set, generate a feature sequence composed of multiple discrete points, and record the curvature change value of each segment; a parameter optimization module for calculating the shape distribution difference between the target component and the reference component based on the feature sequence, generating a corresponding feature vector, introducing a dynamic adjustment parameter, and performing weighted processing on the feature vector to obtain a revised matching parameter set; A spatial mapping analysis module is used to determine the local similarity area between the target component and the reference component by comparing the corrected matching parameter set, calculate the position offset between the target component and the reference component based on the local similarity area, and generate a registration mapping relationship; A candidate screening module is used to combine the registration mapping relationship and the shape distribution difference value to construct a matching priority matrix and screen out candidate components with high matching degree; The parameter feedback adjustment module is used to verify the attributes of the high-matching candidate components. If the matching accuracy does not meet the requirements, the module returns to adjust the dynamic parameters until the preset matching threshold is met.

Images