Three-layer association driven 4D construction network plan generation method and system thereof

Abstract

The application discloses a three-layer correlation driving 4D construction network plan generation method and system, belongs to the technical field of building construction informatization and BIM construction progress management, extracts component attributes from a BIM model, forms a full-dimensional attribute configuration containing monomer, floor, construction section, room, division, sub-division and type, and establishes a three-layer correlation relationship of BIM components, WBS nodes and construction processes; candidate WBS nodes and complete paths are generated according to a preset WBS generation strategy, binary projection inflation values and high-order evidence collapse values are obtained, and the loss degree of multi-dimensional combined constraints is comprehensively determined; the candidate WBS nodes are modified or reconstructed according to the loss degree; then, the construction network plan is generated from the modified or reconstructed WBS nodes, and is associated with the BIM components to form a 4D construction network plan; and the application can reduce false WBS nodes and incorrect processes, and improve the accuracy of the construction network plan and 4D simulation.

Description

Technical Field

[0001] This invention relates to the field of building construction informatization and BIM construction progress management technology, specifically to a three-layer association-driven 4D construction network plan generation method and system. Background Technology

[0002] With the deepening application of BIM technology in construction organization design and schedule management, generating a Work Breakdown Structure (WBS) based on the BIM model and further associating it with construction procedures, construction network plans, and 4D visualization models has become an important technical direction for improving construction planning efficiency. Existing technologies typically extract the professional and spatial attributes of components from the BIM model, such as sections, subsections, types, individual buildings, floors, construction sections, and rooms. These attributes are then sorted and combined according to a preset WBS generation strategy to form WBS nodes. These WBS nodes are then associated with construction procedures, schedule tasks, and BIM components, thereby achieving automated generation and 4D simulation of the construction plan.

[0003] However, in existing automatic WBS generation processes, systems often rely on binary key-value relationships between spatial discrete information and system discrete information for step-by-step filtering and node expansion. This involves deriving candidate WBS nodes through local "space-system" or "system-space" associations. While this approach improves WBS generation efficiency, it easily loses complete multi-dimensional combination constraints between individual buildings, floors, construction sections, rooms, sub-sections, and types. For example, the system might identify binary associations between a certain individual building, floor, and construction section with a certain professional system, leading to the incorrect generation of professional WBS nodes for that spatial combination. However, no actual BIM component may simultaneously satisfy this complete path, resulting in the generation of empty quantity nodes, fictitious construction tasks, or incorrect procedures.

[0004] The aforementioned problems further propagate to the construction network planning and 4D construction simulation stages, causing non-existent WBS nodes to be matched with construction procedures and incorporated into the network logic, resulting in distortions in schedule calculations, resource allocation, quantity statistics, and model schedule binding. This is particularly pronounced in large complexes, hospitals, airports, subway stations, and other projects with complex spaces and dense interdisciplinary collaborations, where local binary relationships hold true while complete multidimensional combinations do not. Summary of the Invention

[0005] The purpose of this invention is to provide a three-layer association-driven 4D construction network planning generation method and system to address the shortcomings in the prior art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a three-layer association-driven 4D construction network plan generation method, comprising: Extract component attributes from the BIM model to form a full-dimensional attribute configuration that includes individual units, floors, construction sections, rooms, divisions, sub-divisions, and types. Based on the full-dimensional attribute configuration, establish a three-layer relationship between BIM components, WBS nodes, and construction procedures. Candidate WBS nodes are generated according to the preset WBS generation strategy, and the complete path of the candidate nodes is obtained simultaneously. For the complete path, the binary projection expansion value, which represents the degree of binary key value expansion, and the high-order evidence collapse value, which represents the degree of component evidence collapse, are obtained respectively. The degree of loss of multidimensional combined constraints is determined by combining the binary projection expansion value and the higher-order evidence collapse value. When the degree of loss is lower than the preset threshold, the candidate WBS node is retained and its engineering quantity and lower-level process are corrected with the real component evidence set. When the degree of loss is not lower than the preset threshold, the generation of the candidate WBS node and its lower-level process is prohibited, and the process is rolled back to the upper-level valid node. The WBS subtree is reconstructed based on the real full-dimensional attribute configuration. The construction network plan is generated based on the revised or reconstructed WBS nodes and associated with BIM components to form a 4D construction network plan.

[0007] Preferably, the formation of the full-dimensional attribute configuration includes the following steps: The same component number is used for consistent identification of the same BIM component, and the original attributes, spatial outline inclusion results and sub-item identification results of the component are collected simultaneously. Field values ​​are written in a fixed field order of single unit, floor, construction section, room, section, sub-section and type, and source markers are configured for each field. For fields without valid values, null placeholders are written to make each BIM component form a full-dimensional attribute configuration with consistent field order and traceability.

[0008] Preferably, the establishment of a three-tiered relationship includes the following steps: Candidate WBS nodes are formed based on the actual values ​​of each field in the full-dimensional attribute configuration, and the component numbers that meet the node path value conditions are written into the component evidence set; the type field of the candidate WBS nodes is read and the corresponding construction procedures are matched; after verifying that the division, sub-division and type of each component in the component evidence set are consistent with the candidate WBS nodes, a corresponding record is established between the component number, the candidate WBS node number and the construction procedure number.

[0009] Preferably, the generation of candidate WBS nodes and complete paths includes the following steps: parsing the preset WBS generation strategy into a strategy sequence containing field names and hierarchical order, and adding the type field to the end when the strategy sequence is missing; reading the field values ​​in the full-dimensional attribute configuration layer by layer according to the strategy sequence, and using the component evidence set of the node of the previous layer as the screening scope for the current layer from the second layer onwards; and continuously writing the field names, field values, hierarchical order and component evidence sets of each layer into the path record to obtain the complete path of the candidate WBS node.

[0010] Preferably, obtaining the binary projection inflation value includes the following steps: reading the non-empty spatial field values ​​and non-empty professional field values ​​in the complete path of the candidate WBS node to form a path attribute list; finding the binary key-value relationship between the spatial field values ​​and the professional field values ​​based on the path attribute list, and compiling a projection candidate list that can be formed by the binary key-value relationship; finding the real-existing full-dimensional attribute configurations with the complete path as the limiting condition, forming an empirical combination list, and determining the binary projection inflation value based on the degree of excess of the projection candidate list relative to the empirical combination list.

[0011] Preferably, the acquisition of the higher-order evidence collapse value includes the following steps: reading all non-empty field values ​​in the complete path of the candidate WBS node, and obtaining the component number set corresponding to each field value; calculating the common component number for adjacent fields and cross-level fields according to the hierarchical order of the complete path to obtain the number of second-order co-evidence; calculating the common component number for all non-empty field values ​​to obtain the total number of co-evidence, and determining the higher-order evidence collapse value based on the degree of reduction of the number of second-order co-evidence relative to the total number of co-evidence.

[0012] Preferably, the determination of the degree of loss of multidimensional combined constraints includes the following steps: limiting the binary projection expansion value and the higher-order evidence collapse value corresponding to the same candidate WBS node to the same value range; determining the participation ratio of the binary projection expansion value and the higher-order evidence collapse value according to the number of spatial fields, professional fields and component evidence sets in the complete path; merging the binary projection expansion value and the higher-order evidence collapse value according to the participation ratio, and performing reverse verification in combination with the number of co-evidence of all orders to obtain the degree of loss of multidimensional combined constraints.

[0013] Preferably, the correction or reconstruction of candidate WBS nodes includes the following steps: when the degree of loss of multidimensional combination constraints is lower than a preset threshold, the real component evidence set is re-extracted with the complete path as a limiting condition, and the engineering quantity and lower-level construction procedures of the candidate WBS node are corrected with the real component evidence set; when the degree of loss of multidimensional combination constraints is not lower than the preset threshold, the candidate WBS node and its lower-level construction procedures are deleted, and the process is backtracked along the complete path to the upper-level valid node where the component evidence set is not empty and the degree of loss of multidimensional combination constraints is lower than the preset threshold; based on the complete path of the upper-level valid node, the actual subsequent field value combinations are extracted from the full-dimensional attribute configuration, and the WBS subtree is reconstructed.

[0014] Preferably, the formation of the 4D construction network plan includes the following steps: using the modified or reconstructed WBS nodes as formal WBS nodes, forming construction tasks based on the real component evidence set, quantity, type fields, and lower-level construction procedures of the formal WBS nodes; determining the duration according to the quantity and unit time completion of the construction tasks, and establishing the predecessor and successor relationships between construction tasks based on the spatial and professional fields in the complete path; writing the time information of the construction tasks into the construction time attribute of the corresponding BIM component, so that the BIM component forms a 4D association of completed state, construction state, and non-constructed state with the construction network plan.

[0015] This invention also provides a three-layer association-driven 4D construction network planning generation system, comprising: The three-layer association establishment module extracts component attributes from the BIM model to form a full-dimensional attribute configuration that includes individual units, floors, construction sections, rooms, sub-sections, and types. Based on the full-dimensional attribute configuration, a three-layer association relationship is established between BIM components, WBS nodes, and construction procedures. The candidate WBS node generation module generates candidate WBS nodes according to a preset WBS generation strategy and simultaneously obtains the complete path of the candidate nodes. For the complete path, it obtains the binary projection expansion value, which represents the degree of binary key value expansion, and the high-order evidence collapse value, which represents the degree of component evidence collapse. The node handling module comprehensively determines the degree of loss of multidimensional combination constraints based on the binary projection expansion value and the high-order evidence collapse value. When the degree of loss is lower than the preset threshold, the candidate WBS node is retained and its engineering quantity and lower-level process are corrected with the real component evidence set. When the degree of loss is not lower than the preset threshold, the generation of the candidate WBS node and its lower-level process is prohibited, and the process is rolled back to the upper-level valid node. The WBS subtree is reconstructed based on the real full-dimensional attribute configuration. The construction network plan generation module generates a construction network plan based on the revised or reconstructed WBS nodes and associates it with BIM components to form a 4D construction network plan.

[0016] The technical effects and advantages provided by the present invention in the above technical solution are as follows: The core technical approach of this invention, which distinguishes it from existing technologies, lies in its departure from relying solely on spatial-system binary key-value relationships to directly generate WBS nodes. Instead, it first establishes a full-dimensional attribute configuration at the BIM component level, encompassing individual units, floors, construction sections, rooms, sub-sections, and types. Furthermore, it simultaneously preserves the complete path and component evidence set during candidate WBS node generation. Consequently, each WBS node can be traced back to the actual BIM component and its complete attribute combination, preventing existing technologies from erroneously constructing non-existent WBS nodes due to valid local binary associations. This improves the consistency between WBS nodes, quantities, and construction procedures from the source.

[0017] This invention further assesses the degree of loss of multidimensional combined constraints from two directions—binary relation expansion and component evidence contraction—using binary projection expansion values ​​and higher-order evidence collapse values, respectively. Based on the assessment results, candidate WBS nodes are corrected or reconstructed. For nodes with minor loss, the quantities and lower-level processes are corrected using the actual component evidence set; for nodes with severe loss, the generation of WBS subtrees is prohibited, and reconstruction is rolled back. This method can prevent false nodes from entering the construction network plan, reduce erroneous processes, schedule errors, and incorrect 4D bindings, and improve the accuracy and traceability of the 4D construction network plan. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0019] Figure 1 This is a flowchart of the method of the present invention.

[0020] Figure 2 This is a flowchart of the system modules of the present invention. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] Example 1, please refer to Figure 1 As shown in this embodiment, the three-layer association-driven 4D construction network plan generation method includes: Component attributes are extracted from the BIM model to form a full-dimensional attribute configuration that includes individual units, floors, construction sections, rooms, divisions, sub-divisions, and types. Based on the full-dimensional attribute configuration, a three-layer relationship between BIM components, WBS nodes, and construction procedures is established.

[0023] In one implementation, each building information component is uniquely read to obtain its component number, component name, component family, component material, component geometric boundary, component location point, floor identifier, and component professional identifier as original attributes. For the same building information component, the same component number is used throughout subsequent processing, ensuring that the original attributes, spatial contour inclusion results, and component identification results all trace back to the same data source. Spatial contour inclusion results are obtained through the inclusion relationship between the component's geometric boundary and the boundaries of individual buildings, floors, construction sections, and rooms. When the component's geometric boundary completely falls within a spatial contour, the name of that spatial contour is written into the corresponding field. When the component's geometric boundary crosses multiple spatial contours, the intersection volume between the component's geometric boundary and each spatial contour is used as the classification basis, and the name of the spatial contour with the highest intersection volume value is recorded. Simultaneously, a list of intersecting spaces is maintained to prevent loss of component classification.

[0024] The identification results of building information components are determined. The component name, component family, component professional identifier, and component material are sequentially matched with a pre-set component identification table to obtain the section, sub-section, and type. The matching order is as follows: first, determine the section based on the component professional identifier; then, determine the sub-section based on the component family; and finally, determine the type based on the component name and component material. If a unique result cannot be obtained for a previous item, the next item is used to limit it until a unique section, sub-section, and type are obtained. If multiple results still exist, the result consistent with the component family is retained, and its source is marked as a family-limited result. This yields the original attributes, spatial outline inclusion results, and component identification results corresponding to the same component number.

[0025] The initial attribute configuration is organized according to a fixed field order of unit, floor, construction section, room, division, sub-division, and type. Each field has a value field and a source field. The value field contains the corresponding attribute name, and the source field contains the attribute acquisition method, including original attribute acquisition, spatial outline inclusion acquisition, sub-item identification acquisition, and manual revision acquisition. If a field does not have a valid value, a null placeholder is written in the value field, and a null value source description is written in the source field. The null placeholder does not participate in subsequent construction procedure matching but is retained in the full-dimensional attribute configuration to maintain the integrity of the field order. After organization, each building information component forms a traceable full-dimensional attribute configuration, which includes at least the component number, 7 field values, and 7 field sources.

[0026] Candidate work breakdown structures (WBSs) are formed based on the actual values ​​of each field in the full-dimensional attribute configuration. During node formation, the unit field is read first, followed by the floor field, construction section field, and room field to form a spatial path; then, the sub-section field, sub-sub-section field, and type field are read to form a professional path. The spatial and professional paths are connected according to the preset work breakdown order to form a complete node path. For fields with null placeholder values, the field position is not deleted but reserved in the node path to ensure consistent path lengths for nodes of the same type. For each candidate WBS node formed, the component numbers that meet the node's path value conditions are written into the candidate WBS node's component evidence set. If the component evidence set is empty, the candidate WBS node does not enter the construction procedure matching stage; if the component evidence set is not empty, it proceeds to subsequent association.

[0027] Construction procedures are matched based on the type field of candidate work breakdown nodes and component evidence sets. First, the type field of the candidate work breakdown node is read to obtain the set of construction procedure names corresponding to the type field. Then, the full-dimensional attribute configuration corresponding to each component number in the component evidence set is read, and its division, sub-division, and type are checked to ensure consistency with the candidate work breakdown node. If consistent, the construction procedure name is written to the candidate work breakdown node, and a one-to-one correspondence record is established between the component number, the candidate work breakdown node number, and the construction procedure number. Through this process, building information components provide the true attribute source, candidate work breakdown nodes carry component evidence sets, and construction procedures are obtained by relying on the type field and component evidence sets, thus forming a closed-loop association between building information components, work breakdown nodes, and construction procedures.

[0028] Candidate WBS nodes are generated according to the preset WBS generation strategy, and the complete path of the candidate nodes is obtained simultaneously. For the complete path, the binary projection expansion value, which represents the degree of binary key value expansion, and the high-order evidence collapse value, which represents the degree of component evidence collapse, are obtained respectively.

[0029] In one implementation, a preset work breakdown structure generation strategy is first parsed into a strategy sequence. The strategy sequence consists of several strategy items arranged hierarchically. Each strategy item includes a field name and a hierarchical number. The field names are derived from unit, floor, construction section, room, division, sub-division, and type. During parsing, the first strategy item is used as the root field, and the strategy items from the second to the last are used as expansion fields layer by layer. A unique hierarchical number is assigned to each strategy item. If duplicate fields exist in the strategy sequence, only the first field is retained; if a type field is missing, it is added to the last position to ensure that candidate work breakdown nodes are associated with construction procedures.

[0030] After completing the strategy sequence, the corresponding field values ​​in the full-dimensional attribute configuration are read layer by layer according to the strategy sequence. The first layer uses the set of component numbers corresponding to all full-dimensional attribute configurations as the filtering scope, reads the non-empty values ​​under the first-layer fields, merges them according to the same value, forming first-layer candidate work decomposition nodes, and writes the corresponding component number into the component evidence set of that node. From the second layer onwards, each layer uses the component evidence set of the previous layer's candidate work decomposition nodes as the filtering scope for this layer, reading the current field value only within this scope. Therefore, the source of the current layer's candidate work decomposition nodes is limited to the range of actual components in the previous layer, avoiding the mixing of field values ​​between different parent nodes.

[0031] Each time a candidate work decomposition node is formed, the component evidence set for that layer is updated synchronously. The update method is as follows: in the component evidence set of the previous layer, the component number whose current field value is equal to the value of the node in this layer is selected, and the selection result is used as the component evidence set for this layer; if the current field is a null placeholder, the field position is retained, and only the component evidence set of the previous layer is passed to this layer as is. The complete path of the candidate work decomposition node is obtained by writing the field name, field value, and layer number of each layer consecutively. The record format of the complete path is "layer number, field name, field value, component evidence set". Through this record format, any candidate work decomposition node can trace back its parent node, field source, and component evidence set change process.

[0032] After obtaining the complete path of the candidate work decomposition nodes, the binary projection inflation value is further obtained. First, the non-empty spatial and professional field values ​​in the complete path are read. Spatial fields include unit, floor, construction section, and room; professional fields include division, sub-division, and type, thus forming a path attribute list. Then, the binary key-value relationship is checked item by item to see if there are any associated records between spatial and professional field values. For the same complete path, its parent path remains unchanged, and the current node's layer and its lower layers are used as the expansion scope to collect all field value strings allowed by the binary key-value relationship, resulting in a projection candidate list. Each item in the projection candidate list is a complete field value string derived from the binary key-value relationship and contains at least one spatial field value and one professional field value.

[0033] Using the complete path as a constraint, records that simultaneously satisfy all non-empty field values ​​of the complete path are searched in the full-dimensional attribute configuration to obtain the empirical combination list. Each item in the empirical combination list originates from the full-dimensional attribute configuration corresponding to the actual component number. The projection candidate list is compared with the empirical combination list. Any field value string in the projection candidate list that does not exist in the empirical combination list is counted as an excess quantity. The binary projection inflation value is obtained by dividing the excess quantity by the number of projection candidate lists; when the number of projection candidate lists is 0, the binary projection inflation value is recorded as 0. A value of 0 indicates that no excess items were generated in the projection candidate list; a value of 1 indicates that none of the projection candidate lists are supported by the empirical combination list.

[0034] For example, if a candidate list for projection of a complete path contains 10 field value strings, and only 4 of them can be found in the empirical combination list, then the excess is 6. The binary projection inflation value is 6 divided by 10, resulting in 0.6. This calculation result indicates that 60% of the candidate content derived based on binary key-value relationships cannot be verified by full-dimensional attribute configuration.

[0035] The higher-order evidence collapse value is obtained based on the complete path. First, all non-empty field values ​​in the complete path are extracted, and the component number set corresponding to each field value is obtained. The component number set corresponding to the field value is obtained by reverse retrieval of the full-dimensional attribute configuration, that is, when the corresponding field value of any component is equal to that field value, the component number is added to the component number set of that field value.

[0036] According to the hierarchical order of the complete path, the common component numbers are calculated for adjacent fields and cross-level fields. Adjacent fields are two fields with consecutive hierarchical numbers, and cross-level fields are two fields with non-consecutive hierarchical numbers but belonging to the same complete path. The number of common component numbers for each pair of field values ​​is equal to the number of intersections of the sets of component numbers corresponding to the two field values. The number of intersections of all adjacent and cross-level fields is added together and divided by the number of field pairs involved in the calculation to obtain the second-order co-certification number; when the number of field pairs involved in the calculation is 0, the second-order co-certification number is recorded as 0.

[0037] Next, calculate the common component number for all non-empty field values ​​in the complete path to obtain the total number of co-evidences at all levels. During calculation, starting from the set of component numbers corresponding to the first non-empty field value, find the intersection with the set of component numbers corresponding to subsequent non-empty field values. The number of component numbers in the final intersection is the total number of co-evidences at all levels. The higher-order evidence collapse value is obtained by dividing the difference between the second-order co-evidence count and the total number of co-evidences at all levels by the second-order co-evidence count. When the second-order co-evidence count is 0, the higher-order evidence collapse value is recorded as 0; when the total number of co-evidences at all levels is greater than the second-order co-evidence count, the higher-order evidence collapse value is recorded as 0.

[0038] For example, if a complete path contains 6 non-empty field values, the number of field pairs involved in the calculation is 15, and the sum of the common component numbers for each field pair is 45, then the number of second-order co-evidence is 45 divided by 15, resulting in 3. If the number of common component numbers for all 6 non-empty field values ​​is 1, then the higher-order evidence collapse value is the number of second-order co-evidence (3) minus the number of all-order co-evidence (1), divided by the number of second-order co-evidence (3), resulting in 0.67. This result indicates that the candidate work decomposition node has component evidence at the pairwise field level, but only retains 33% co-evidence support at the complete path level, posing a risk of forming candidate nodes through local association splicing.

[0039] Through the above processing, the formation of candidate work decomposition nodes no longer relies solely on the layer-by-layer expansion of discrete field values, but instead simultaneously preserves the complete path and component evidence sets during node formation. The binary projection inflation value measures the difference between the binary key value derivation results and the empirical results of the full-dimensional attribute configuration, while the higher-order evidence collapse value measures the degree to which field-pair evidence shrinks into complete path evidence. These two values ​​correspond to two different directions: excessive expansion of binary relations and higher-order collapse of component evidence, respectively, providing a calculable basis for determining the degree of loss of multidimensional combination constraints in subsequent processes.

[0040] The degree of loss of multidimensional combined constraints is determined by combining the binary projection expansion value and the higher-order evidence collapse value. When the degree of loss is lower than the preset threshold, the candidate WBS node is retained and its engineering quantity and lower-level process are corrected with the real component evidence set. When the degree of loss is not lower than the preset threshold, the generation of the candidate WBS node and its lower-level process is prohibited, and the process is rolled back to the upper-level valid node. The WBS subtree is reconstructed based on the real full-dimensional attribute configuration.

[0041] In one implementation, the degree of loss of multidimensional combination constraints is jointly determined by the binary projection inflation value, the higher-order evidence collapse value, the composition of the complete path field, and the number of component evidence sets. First, the binary projection inflation value and the higher-order evidence collapse value corresponding to the same candidate work decomposition node are read and limited to a range of 0 to 1. When either value is below 0, it is treated as 0; when either value is above 1, it is treated as 1, thus obtaining the normalized binary projection inflation value and the normalized higher-order evidence collapse value. This process ensures that both are within the same value range, avoiding deviations in subsequent merging results due to different value scales.

[0042] The participation weight is determined based on the field composition and the number of component evidence sets within the complete path. The number of non-empty spatial fields in the complete path is denoted as the number of spatial fields, the number of non-empty professional fields is denoted as the number of professional fields, and the number of component evidence sets corresponding to candidate work decomposition nodes is denoted as the number of evidence sets. First, the spatial basis weight is calculated: Spatial basis weight = Number of spatial fields ÷ (Number of spatial fields + Number of professional fields); then, the professional basis weight is calculated: Professional basis weight = Number of professional fields ÷ (Number of spatial fields + Number of professional fields). When the number of evidence sets is 0, the evidence correction value is 1; when the number of evidence sets is not 0, the evidence correction value is 1 ÷ (Number of evidence sets + 1). The participation weight of the binary projection inflation value = Spatial basis weight × (1 - Evidence correction value); the participation weight of the higher-order evidence collapse value = Professional basis weight + Evidence correction value. To ensure the sum of the two participation weights equals 1, the participation weight of the binary projection inflation value is divided by the sum of the two participation weights to obtain the binary correction weight; the participation weight of the higher-order evidence collapse value is divided by the sum of the two participation weights to obtain the higher-order correction weight. This processing allows the effect of binary key expansion to be enhanced when the spatial field hierarchy occupies more space, and the effect of component evidence collapse to be enhanced when the amount of evidence is less.

[0043] The initial loss factor is then calculated. Initial loss factor = Normalized binary projection dilation value × Binary correction weight + Normalized higher-order evidence collapse value × Higher-order correction weight. For example, if the normalized binary projection dilation value is 0.60, the normalized higher-order evidence collapse value is 0.70, the binary correction weight is 0.40, and the higher-order correction weight is 0.60, then the initial loss factor = 0.60 × 0.40 + 0.70 × 0.60 = 0.66.

[0044] After obtaining the initial loss level, reverse verification is performed using the total number of co-evidences at all levels corresponding to the complete path. The total number of co-evidences at all levels is the number of component numbers corresponding to all non-empty field values ​​of the complete path. First, the total support rate is calculated: Total Support Rate = Total Number of Co-evidences at All Levels ÷ Number of Evidences; when the number of evidences is 0, the total support rate is 0; when the total support rate is higher than 1, it is treated as 1. Then, the reverse lack of evidence level is calculated: Reverse lack of evidence level = 1 - Total Support Rate. The multi-dimensional combined constraint loss level = Initial Loss Level × 0.70 + Reverse lack of evidence level × 0.30. Through this calculation, the initial loss level can reflect the combined impact of binary expansion and evidence collapse, and the reverse lack of evidence level can verify whether the candidate work decomposition node has real component support at the complete path level.

[0045] It should be noted that the preset threshold is preferably 0.60. When the number of non-empty fields in the complete path is no less than 6 and the number of pieces of evidence does not exceed 2, the threshold is 0.50; when the number of non-empty fields in the complete path is no less than 6 and the number of pieces of evidence is no less than 10, and the number of co-evidence at all levels is no less than 3, the threshold is 0.70; in other cases, the preferred value of 0.60 is used. The above rules impose stricter restrictions on candidate work decomposition nodes with many field levels but insufficient evidence, and also prevent candidate work decomposition nodes with stable complete path evidence from being mistakenly deleted.

[0046] When the degree of loss of multidimensional combination constraints is less than a preset threshold, candidate work decomposition nodes are retained. First, using the complete path as a constraint, component numbers that simultaneously satisfy all non-empty field values ​​are re-extracted from the full-dimensional attribute configuration to form a real component evidence set. Then, the original component evidence set is replaced with the real component evidence set, and the quantity of the candidate work decomposition node is re-summarized based on the quantity attributes of each component in the real component evidence set. The quantity summarization uses a component-by-component accumulation method; the node quantity equals the sum of the corresponding quantity values ​​of each component in the real component evidence set. Subsequently, the type field is read, and lower-level operations are matched only when the real component evidence set is not empty; the range of components corresponding to lower-level operations must not exceed the range of the real component evidence set. Thus, candidate work decomposition nodes with minor losses are retained, but their quantities and lower-level operations are limited to the range of real component evidence.

[0047] When the degree of loss of multidimensional combination constraints is not lower than a preset threshold, the candidate work decomposition node enters the process of prohibiting generation. First, the candidate work decomposition node is deleted from the work decomposition structure to be output, and its subordinate process matching is stopped; the subordinate processes temporarily obtained based on the node are also deleted. Then, backtracking upwards along the complete path, the node with the most recent component evidence set not empty and the degree of loss of multidimensional combination constraints is lower than the preset threshold is found, which is taken as the superior valid node. Using the complete path of the superior valid node as a constraint, the actual subsequent field value combinations are extracted from the full-dimensional attribute configuration, and the work decomposition subtree under the superior valid node is reconstructed according to the original preset work decomposition generation strategy. During reconstruction, each new node is simultaneously written with the complete path, the actual component evidence set, and the quantity of work, so that the reconstructed work decomposition subtree comes only from the actual full-dimensional attribute configuration, thereby preventing false nodes formed by splicing binary key-value relationships from continuing to enter the construction network plan.

[0048] The construction network plan is generated based on the revised or reconstructed WBS nodes and associated with BIM components to form a 4D construction network plan.

[0049] In one implementation, the revised or restructured Work Breakdown Structure (WBS) nodes are first used as formal WBS nodes. Each formal WBS node contains a complete path, a set of actual component evidence, quantities, a type field, and lower-level construction procedures. The set of actual component evidence for the formal WBS node is read to obtain the BIM component number corresponding to that node; the type field and lower-level construction procedures are read to form the construction tasks corresponding to that formal WBS node. When a formal WBS node corresponds to multiple lower-level construction procedures, multiple construction tasks are formed according to the order of the construction procedures, and the same complete path and corresponding BIM component number are written into each construction task, enabling the construction tasks to be traced back to the formal WBS node and BIM components.

[0050] The duration of a construction task is determined based on the quantities and construction procedures of the formal WBS nodes. The duration is calculated by dividing the node quantity by the quantity completed per unit time; the quantity completed per unit time is determined by the labor, machinery, or work team capacity corresponding to the construction procedure. For example, if the duct installation quantity for a formal WBS node is 120 meters, and the corresponding work team's daily completion rate is 30 meters, then the duration of this construction task is 120 divided by 30, resulting in 4 days. If the same construction task includes multiple BIM components, the quantities of each BIM component in the actual component evidence set are first accumulated, and then the duration is calculated. If the calculated result is not an integer number of days, the planned duration is determined by rounding up to ensure that the construction task period covers all quantities.

[0051] The construction network logical relationship is generated based on the complete path and construction sequence of the formal WBS nodes. For multiple construction procedures within the same formal WBS node, a sequential relationship is formed according to the order of procedures, such as setting out, installing supports and hangers, installing pipelines, pressure testing, and insulation being connected in sequence. For the relationship between different formal WBS nodes, the spatial sequence is determined according to the spatial fields in the complete path: within the same building, underground floor tasks precede above-ground floor tasks; within the same floor, construction sections are connected according to the preset construction section sequence; within the same construction section, the main structure, secondary structure, electromechanical installation, and decoration and finishing corresponding to the professional fields are connected according to the preset professional sequence. For formal WBS nodes with the same parent path but different professional fields, a sequential relationship is established only when the professional sequence requirement exists; when there is no professional sequence requirement, parallel construction tasks are allowed.

[0052] Perform network time calculations for construction tasks. The start time of a construction task without predecessor tasks is set as the project start time. The earliest start time of any construction task is equal to the maximum of the earliest finish times of all its predecessor tasks, and the earliest finish time is equal to the earliest start time plus the duration. After all construction tasks have completed the forward calculation, reverse calculations are performed using the project's planned finish time as the starting point. The latest finish time of any construction task is equal to the minimum of the latest start times of all its successor tasks, and the latest start time is equal to the latest finish time minus the duration. Total float is equal to the latest start time minus the earliest start time; construction tasks with a total float of 0 are recorded as critical construction tasks. This yields a construction network plan that includes construction tasks, durations, predecessor tasks, successor tasks, earliest start time, earliest finish time, latest start time, latest finish time, and total float.

[0053] The construction network plan is time-linked with BIM components to form a 4D construction network plan. For each construction task, its corresponding BIM component number, earliest start time, and earliest finish time are read and written into the construction time attribute of the corresponding BIM component. When the same BIM component corresponds to multiple construction tasks, multiple time intervals are written for the BIM component according to the construction sequence. During 4D display, the date is used as the time index to read BIM components that have been completed, are under construction, or have not yet started before that date, and assign them a completed, under construction, or not-under-construction status, respectively. The completed status corresponds to a task whose earliest finish time is earlier than the current date, the under construction status corresponds to a current date between the earliest start time and the earliest finish time of the task, and the not-under-construction status corresponds to a task whose earliest start time is later than the current date. Through the above processing, the corrected or reconstructed formal WBS nodes, construction network plan tasks, and BIM components are consistent under the same complete path and real component evidence set, thus forming a traceable and verifiable 4D construction network plan.

[0054] Example 2, please refer to Figure 2 As shown, the three-layer association-driven 4D construction network planning generation system described in this embodiment includes: The three-layer association establishment module extracts component attributes from the BIM model to form a full-dimensional attribute configuration that includes individual units, floors, construction sections, rooms, sub-sections, and types. Based on the full-dimensional attribute configuration, a three-layer association relationship is established between BIM components, WBS nodes, and construction procedures. The candidate WBS node generation module generates candidate WBS nodes according to a preset WBS generation strategy and simultaneously obtains the complete path of the candidate nodes. For the complete path, it obtains the binary projection expansion value, which represents the degree of binary key value expansion, and the high-order evidence collapse value, which represents the degree of component evidence collapse. The node handling module comprehensively determines the degree of loss of multidimensional combination constraints based on the binary projection expansion value and the high-order evidence collapse value. When the degree of loss is lower than the preset threshold, the candidate WBS node is retained and its engineering quantity and lower-level process are corrected with the real component evidence set. When the degree of loss is not lower than the preset threshold, the generation of the candidate WBS node and its lower-level process is prohibited, and the process is rolled back to the upper-level valid node. The WBS subtree is reconstructed based on the real full-dimensional attribute configuration. The construction network plan generation module generates a construction network plan based on the revised or reconstructed WBS nodes and associates it with BIM components to form a 4D construction network plan.

[0055] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A three-layer association-driven 4D construction network plan generation method, characterized in that, include: Extract component attributes from the BIM model to form a full-dimensional attribute configuration that includes individual units, floors, construction sections, rooms, divisions, sub-divisions, and types. Based on the full-dimensional attribute configuration, establish a three-layer relationship between BIM components, WBS nodes, and construction procedures. Establishing a three-tier relationship involves the following steps: Candidate WBS nodes are formed based on the actual values ​​of each field in the full-dimensional attribute configuration, and the component numbers that meet the node path value conditions are written into the component evidence set; the type field of the candidate WBS nodes is read and matched with the corresponding construction procedures; after verifying that the division, sub-division and type of each component in the component evidence set are consistent with the candidate WBS nodes, a corresponding record is established between the component number, the candidate WBS node number and the construction procedure number. Candidate WBS nodes are generated according to the preset WBS generation strategy, and the complete path of the candidate nodes is obtained simultaneously. For the complete path, the binary projection expansion value, which represents the degree of binary key value expansion, and the high-order evidence collapse value, which represents the degree of component evidence collapse, are obtained respectively. Obtaining the binary projection inflation value involves the following steps: reading the non-empty spatial field values ​​and non-empty professional field values ​​in the complete path of the candidate WBS node to form a path attribute list; finding the binary key-value relationship between the spatial field values ​​and the professional field values ​​based on the path attribute list, and compiling a projection candidate list that can be formed by the binary key-value relationship; Using the complete path as a constraint, we search for real-existing full-dimensional attribute configurations to form an empirical combination list, and determine the binary projection inflation value based on the degree to which the projection candidate list exceeds the empirical combination list. The acquisition of the higher-order evidence collapse value includes the following steps: reading all non-empty field values ​​in the complete path of the candidate WBS node, and obtaining the component number set corresponding to each field value; calculating the common component number for adjacent fields and cross-level fields according to the hierarchical order of the complete path to obtain the number of second-order co-evidence; calculating the common component number for all non-empty field values ​​to obtain the total number of co-evidence, and determining the higher-order evidence collapse value based on the degree of reduction of the total number of co-evidence relative to the number of second-order co-evidence; The degree of loss of multidimensional combined constraints is determined by combining the binary projection expansion value and the higher-order evidence collapse value. When the degree of loss is lower than the preset threshold, candidate WBS nodes are retained, and their engineering quantity and lower-level procedures are corrected with the real component evidence set. When the loss level is not lower than the preset threshold, the generation of the candidate WBS node and its subordinate processes is prohibited, and the process is rolled back to the upper-level valid node. The WBS subtree is reconstructed based on the actual full-dimensional attribute configuration. The determination of the degree of loss of multidimensional combination constraints includes the following steps: limiting the binary projection expansion value and the higher-order evidence collapse value corresponding to the same candidate WBS node to the same value range; determining the participation ratio of the binary projection expansion value and the higher-order evidence collapse value according to the number of spatial fields, professional fields and component evidence sets in the complete path; The degree of loss of multidimensional combined constraints is obtained by combining the binary projection expansion value and the high-order evidence collapse value according to the participation ratio, and by combining the number of co-evidence of all orders for reverse verification. The construction network plan is generated based on the revised or reconstructed WBS nodes and associated with BIM components to form a 4D construction network plan.

2. The three-layer association-driven 4D construction network planning generation method according to claim 1, characterized in that, The formation of a full-dimensional attribute configuration includes the following steps: The same component number is used for consistent identification of the same BIM component, and the original attributes, spatial outline inclusion results and sub-item identification results of the component are collected simultaneously. Field values ​​are written in a fixed field order of single unit, floor, construction section, room, section, sub-section and type, and source markers are configured for each field. For fields without valid values, null placeholders are written to make each BIM component form a full-dimensional attribute configuration with consistent field order and traceability.

3. The three-layer association-driven 4D construction network planning generation method according to claim 1, characterized in that, The generation of candidate WBS nodes and complete paths includes the following steps: parsing the preset WBS generation strategy into a strategy sequence containing field names and hierarchical order, and adding the type field to the end when the strategy sequence is missing; reading the field values ​​in the full-dimensional attribute configuration layer by layer according to the strategy sequence, and using the component evidence set of the previous layer node as the filtering scope for the current layer from the second layer onwards; and continuously writing the field names, field values, hierarchical order and component evidence sets of each layer into the path record to obtain the complete path of the candidate WBS node.

4. The three-layer association-driven 4D construction network planning generation method according to claim 1, characterized in that, The correction or reconstruction of candidate WBS nodes includes the following steps: When the degree of loss of multidimensional combination constraints is lower than a preset threshold, the real component evidence set is re-extracted with the complete path as the limiting condition, and the engineering quantity and lower-level construction procedures of the candidate WBS node are corrected with the real component evidence set; when the degree of loss of multidimensional combination constraints is not lower than the preset threshold, the candidate WBS node and its lower-level construction procedures are deleted, and the process is backtracked along the complete path to the upper-level valid node where the component evidence set is not empty and the degree of loss of multidimensional combination constraints is lower than the preset threshold; based on the complete path of the upper-level valid node, the actual subsequent field value combinations are extracted from the full-dimensional attribute configuration, and the WBS subtree is reconstructed.

5. The three-layer association-driven 4D construction network planning generation method according to claim 1, characterized in that, The formation of a 4D construction network plan includes the following steps: First, the modified or reconstructed WBS nodes are used as formal WBS nodes. Construction tasks are then formed based on the actual component evidence set, quantities, type fields, and lower-level construction procedures of the formal WBS nodes. Second, the duration of each construction task is determined according to its quantity and completion rate per unit time. Third, the preceding and following relationships between construction tasks are established based on the spatial and professional fields in the complete path. Finally, the time information of the construction tasks is written into the construction time attribute of the corresponding BIM component, enabling the BIM component to form a 4D association of completed, under-construction, and unconstructed states along with the construction network plan.

6. A three-layer association-driven 4D construction network plan generation system, used to implement the three-layer association-driven 4D construction network plan generation method according to any one of claims 1-5, characterized in that, include: The three-layer association establishment module extracts component attributes from the BIM model to form a full-dimensional attribute configuration that includes individual units, floors, construction sections, rooms, sub-sections, and types. Based on the full-dimensional attribute configuration, a three-layer association relationship is established between BIM components, WBS nodes, and construction procedures. The candidate WBS node generation module generates candidate WBS nodes according to a preset WBS generation strategy and simultaneously obtains the complete path of the candidate nodes. For the complete path, it obtains the binary projection expansion value, which represents the degree of binary key value expansion, and the high-order evidence collapse value, which represents the degree of component evidence collapse. The node handling module comprehensively determines the degree of loss of multidimensional combined constraints based on the binary projection expansion value and the high-order evidence collapse value. When the degree of loss is lower than the preset threshold, the candidate WBS node is retained, and its engineering quantity and subsequent processes are corrected with the real component evidence set. When the loss level is not lower than the preset threshold, the generation of the candidate WBS node and its subordinate processes is prohibited, and the process is rolled back to the upper-level valid node. The WBS subtree is reconstructed based on the actual full-dimensional attribute configuration. The construction network plan generation module generates a construction network plan based on the revised or reconstructed WBS nodes and associates it with BIM components to form a 4D construction network plan.

Images