A prefabricated fence construction process digital management method and system

Abstract

The present application relates to the technical field of fence construction, in particular to a prefabricated fence construction process digital management method and system, comprising the following steps: obtaining component identification, positioning, hoisting posture and installation image data, constructing a site digital twin model, correlating components with designed installation positions, forming a state sequence according to node time, screening hoisting or installation nodes, comparing positioning data with designed positions to obtain installation offset, combining posture, image and offset to determine abnormality level and generate instructions, in the present application, through the component identity, spatial position, hoisting posture, installation image and node state, a traceable chain is formed for production, transportation, approach, stacking, hoisting, installation and acceptance, the site model synchronously presents component state and offset risk, abnormality level drives review, adjustment, archiving and closed loop, reduces record breakpoints and experience judgment deviation, improves timeliness of deviation identification, accuracy of quality traceability and consistency of acceptance basis.

Description

Technical Field

[0001] This invention relates to the field of wall construction technology, and in particular to a digital management method and system for the construction process of prefabricated walls. Background Technology

[0002] The field of fence construction technology involves the collection, storage, and processing of construction objects, construction stages, process records, and management data during fence construction. Specifically, digital management of the precast fence construction process refers to the process where on-site terminals collect records of precast fence component arrival, positioning, installation, connection, and acceptance, upload the construction process data to a server, and the server writes the data into a database according to project number, component number, process node, and acceptance field. The management terminal then reads the corresponding records from the database and generates a construction process ledger.

[0003] Current precast wall construction management relies heavily on manual ledgers, on-site photos, and scattered records to confirm component transfer. Information on key milestones such as production, transportation, on-site stacking, hoisting, installation, and acceptance is difficult to connect continuously. The correspondence between component identity and design location is easily affected by record lag. Hoisting posture and installation deviations mainly rely on on-site experience for judgment. Anomaly detection depends on post-event review. There is a lack of synchronous mapping between the spatial state of the construction site and the state of the components. It is difficult for managers to grasp the risk of deviation and the closed-loop status of the process in a timely manner, resulting in a longer rework and investigation cycle, scattered quality acceptance criteria, and insufficient accuracy in the whole process traceability of components. Summary of the Invention

[0004] The purpose of this invention is to address the shortcomings of existing technologies by proposing a digital management method for the construction process of prefabricated walls.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: a digital management method for the construction process of prefabricated walls, comprising the following steps: Acquire component identification data, component positioning data, hoisting posture data, and installation image data generated at the production, transportation, arrival, stacking, hoisting, installation, and acceptance stages of prefabricated wall components; and construct a digital twin model of the construction site based on the component positioning data, the hoisting posture data, and the installation image data. The prefabricated wall components are associated with the designed installation location using the component identification data. The component status sequence of the prefabricated wall components is determined according to the chronological order of the occurrence of production nodes, transportation nodes, arrival nodes, stacking nodes, hoisting nodes, installation nodes, and acceptance nodes. The component status sequence is then marked in the construction site digital twin model. Node states of type hoisting node or installation node are selected from the component state sequence, and the prefabricated wall components corresponding to the node states are determined. The positioning data of the corresponding components are compared with the designed installation position in the construction site digital twin model to determine the installation offset. Based on the hoisting posture data, the installation image data, and the installation offset, the construction anomaly level is determined, and a component review instruction, hoisting adjustment instruction, or acceptance archiving instruction is generated according to the construction anomaly level.

[0006] Furthermore, the steps for constructing the digital twin model of the construction site include: Read the electronic identification carrier on the prefabricated wall component, extract the component's unique code, production batch code, and design component number, match the component's unique code with the design component number one-to-one, write the production batch code into the corresponding component's batch traceability record, and output the component identification data; Obtain component positioning data, hoisting posture data and installation image data corresponding to the same prefabricated wall component and perform time alignment. Convert the aligned component positioning data to the on-site coordinate reference and establish a viewpoint association between the installation image data and the designed installation position to generate a component space sample. Based on the component spatial samples, the component entity, its orientation, and spatial relationship with the designed installation position are generated, and the construction site digital twin model is constructed using the spatial relationship.

[0007] Furthermore, the process of converting the aligned component positioning data to the on-site coordinate reference is as follows: Obtain the on-site measured coordinates of the control points at the construction site and the design coordinates of the corresponding control points in the overall design plan. Determine the coordinate transformation parameters based on the correspondence between the on-site measured coordinates and the design coordinates of the same control point. The coordinate transformation parameters include plane translation, coordinate axis rotation angle, elevation correction, and scale correction. The coordinates of the component center point, the coordinates of the component end, and the elevation of the component bottom edge in the aligned component positioning data are converted to the unified coordinate system of the construction site according to the coordinate transformation parameters. The converted coordinates are then checked against the coordinates of the site control points. When the deviation of the closure check is less than the allowable deviation threshold of the coordinate transformation, the component positioning coordinates under the site coordinate reference are generated.

[0008] Furthermore, the step of determining the component state sequence includes: Obtain the component unique code and design component number from the component identification data, and read the component location mapping table used to record the correspondence between the design component number and the design installation location; using the design component number as an index, query the corresponding design installation location in the component location mapping table, and write the component unique code, the design component number and the design installation location into the location association record; According to the time sequence of production, transportation, arrival, stacking, hoisting, installation and acceptance, the node status records corresponding to the location association records are read, and the node type, collection time and node identifier of each node are merged into a node status. The components are then sorted according to the preset order of the node types to generate the component status sequence.

[0009] Furthermore, the process of annotating the component state sequence in the digital twin model of the construction site includes: Read the current node status to be written from the component status sequence. The current node status includes the current node type, the current node acquisition time, and the current node identifier. Read the labeled node status from the component entity status field of the construction site digital twin model. The labeled node status includes the labeled node type, the labeled node acquisition time, and the status field update time. According to the preset process sequence of production node, transportation node, arrival node, stacking node, hoisting node, installation node and acceptance node, determine the current process sequence value corresponding to the current node type and the marked process sequence value corresponding to the marked node type respectively; When the current process sequence value is greater than the marked process sequence value, and the current node acquisition time is newer than the status field update time, the current node status is written into the component entity status field, and the status update time is recorded. When the current process sequence value is less than the marked process sequence value, or the current node acquisition time is earlier than the marked node acquisition time, the component entity status field is retained, and a status rollback record is generated.

[0010] Furthermore, the step of determining the installation offset includes: Read the node type, node acquisition time and node identifier in the component status sequence, filter the prefabricated wall components whose node type is hoisting node or installation node and whose node identifier passes the verification, and generate a list of components to be compared. Obtain the component positioning data corresponding to the list of components to be compared, convert the component positioning data to the on-site coordinate reference, and read the design center point, design axis direction and design elevation from the design installation position to generate the design position reference; The actual center point, actual axis direction, and actual elevation under the site coordinate reference are compared with the design position reference to determine the horizontal offset, angular offset, and elevation offset. The horizontal offset, angular offset, and elevation offset are then statistically analyzed to output the installation offset.

[0011] Further steps to determine the level of construction anomaly include: The system acquires the attitude angle deviation, lifting point force change and lifting stabilization time from the lifting attitude data, acquires the joint width, edge occlusion status and surface damage status from the installation image data, and reads the horizontal offset, angle offset and elevation offset from the installation offset to generate anomaly judgment parameters. The anomaly determination parameters are compared with the pose threshold table, image threshold table, and offset threshold table respectively to determine the pose anomaly result, image anomaly result, and offset anomaly result; The relative levels of the posture anomaly results, the image anomaly results, and the offset anomaly results are determined to ascertain the construction anomaly level.

[0012] Furthermore, the process of determining the level of construction anomaly also includes: Obtain pre-configured anomaly level determination rules, which include the level value ranges corresponding to pose anomaly parameters, image anomaly parameters, and offset anomaly parameters, respectively; and convert the pose anomaly results, image anomaly results, and offset anomaly results into pose candidate levels, image candidate levels, and offset candidate levels, respectively, according to the anomaly level determination rules. When at least two of the candidate levels of the attitude candidate level, the image candidate level, and the offset candidate level are consistent, the consistent candidate level is determined as the construction anomaly level; When the attitude candidate level, the image candidate level, and the offset candidate level are all inconsistent, the candidate level with the highest level is determined as the construction anomaly level.

[0013] Furthermore, the process of generating component review instructions, hoisting adjustment instructions, or acceptance archiving instructions includes: Read the instruction type, execution object, and execution time limit corresponding to the construction anomaly level from the instruction mapping table, and obtain the unique component code, current location, and responsible work team corresponding to the prefabricated wall component; When the construction anomaly level matches the review instruction, a component review instruction is generated that records the unique code of the component and the current location; When the construction anomaly level matches the adjustment instruction, a hoisting adjustment instruction is generated that records the responsible work team and the attitude adjustment direction; When the construction anomaly level matches the archiving instruction, an acceptance archiving instruction is generated.

[0014] A digital management system for the construction process of prefabricated walls includes: The component acquisition and modeling unit is used to acquire component identification data, component positioning data, hoisting posture data, and installation image data generated at the production, transportation, arrival, stacking, hoisting, installation, and acceptance stages of prefabricated wall components, and to construct a digital twin model of the construction site based on the component positioning data, the hoisting posture data, and the installation image data. The association annotation unit is connected to the component acquisition and modeling unit. It is used to associate the prefabricated wall component with the design and installation location using the component identification data. According to the chronological order of the occurrence of production node, transportation node, arrival node, stacking node, hoisting node, installation node and acceptance node, the component state sequence of the prefabricated wall component is determined, and the component state sequence is annotated in the construction site digital twin model. The offset determination unit is connected to the associated annotation unit. It is used to filter the node status of the node type as hoisting node or installation node from the component status sequence, and determine the prefabricated wall component corresponding to the node status. In the construction site digital twin model, the positioning data of the corresponding component is compared with the design installation position to determine the installation offset. The instruction unit, connected to the offset determination unit, is used to determine the construction anomaly level based on the hoisting posture data, the installation image data, and the installation offset, and to generate component review instructions, hoisting adjustment instructions, or acceptance and archiving instructions according to the construction anomaly level.

[0015] Compared with the prior art, the advantages and positive effects of the present invention are as follows: In this invention, by collecting component identification, positioning, hoisting posture, and installation images, the production, transportation, arrival, stacking, hoisting, installation, and acceptance nodes are organized into a component status sequence in chronological order. The component identity and the designed installation position form a continuous correspondence, and the spatial state and process state of the component are synchronously presented in the digital twin model of the construction site. The hoisting or installation node triggers the comparison of positioning data with the design position. The installation offset, posture data, and image data jointly participate in the anomaly level determination. The review, adjustment, and archiving instructions form a closed-loop response according to the level difference, thereby reducing the breakpoints of manual recording and the bias of experience judgment, improving the timeliness of offset risk detection, and strengthening the traceability of construction quality and the consistency of acceptance basis. Attached Figure Description

[0016] Figure 1 This is a flowchart of the digital management method for the prefabricated wall construction process of the present invention; Figure 2This is a schematic diagram illustrating the entity association of components in the digital twin model of the construction site according to the present invention; Figure 3 This is a schematic diagram illustrating the effect of component state sequence annotation in this invention; Figure 4 This is a schematic diagram illustrating the installation offset comparison effect of the present invention; Figure 5 This is a schematic diagram illustrating the construction anomaly level and instruction feedback effect of the present invention. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Example

[0018] Please see Figures 1 to 5 This embodiment provides a digital management method and system for the construction process of prefabricated walls. In practical applications, such as during the continuous operation of component delivery, vehicle transfer, on-site material receiving, yard layout, hoisting, foundation placement, installation verification, and acceptance handover in prefabricated wall projects, components continuously generate identification, position, posture, and image records as the process flows. The on-site management terminal assigns these records to the corresponding component names according to the component codes and synchronizes the current status of the components to the digital twin model of the construction site. The method includes the following steps: S1: Obtain component identification data, component positioning data, hoisting posture data, and installation image data generated at the production, transportation, arrival, stacking, hoisting, installation, and acceptance stages of prefabricated wall components; and construct a digital twin model of the construction site based on the component positioning data, hoisting posture data, and installation image data. Component identification data comes from electronic component identification carriers, production records, and design component lists, including the component's unique code, production batch code, design component number, node source, and identification verification status. Component positioning data comes from on-site measurement terminals, positioning terminals, or hoisting positioning interfaces, including the component's center point position, end position, bottom elevation, acquisition time, positioning source, and coordinate verification status. Hoisting posture data comes from lifting equipment, lifting tool acquisition interfaces, and on-site operation terminals, including posture angle deviation, lifting point force changes, hoisting stabilization time, hoisting node identification, and posture acquisition status. Installation image data comes from fixed camera equipment, mobile inspection terminals, or acceptance shooting terminals, including image source, shooting angle, component edge area, joint area, surface condition area, and image resolution status.

[0019] The digital twin model of the construction site uses component entity records to carry on-site data. Each component entity is indexed by a unique component code, and is linked to the design component number, design installation location, on-site positioning record, hoisting posture record, installation image view record, node status field, and instruction status field. The component entities in the model are not standalone graphical displays, but rather unified records carrying on-site collected data, design location data, and process status data. Subsequent installation offset comparisons, anomaly level determinations, and handling instruction generation all read fields from this record.

[0020] S101: Read the electronic identification carrier on the prefabricated wall component, extract the component unique code, production batch code and design component number, perform a unique match between the component unique code and the design component number for each component, write the production batch code into the batch traceability record of the corresponding component, and output the component identification data. Electronic identification carriers are fixed to the surface of prefabricated wall components or embedded in the identification area. During component delivery, transportation handover, on-site registration, yard placement, hoisting commencement, installation confirmation, and acceptance verification, the operating terminal reads the electronic identification carrier and generates an original reading record. After the reading record enters the identification parsing cache, it first checks whether the coding field is complete, and then matches it with the component registration information in the production delivery record and the design component number in the design component list. If the match is successful, the component's unique code serves as the main index for the entire process, the design component number serves as the index for querying the design installation location, and the production batch code is written to the batch traceability record.

[0021] When the electronic identification carrier fails to read, a field is missing, a duplicate read occurs, or the component's unique code cannot match the design component number, the identification parsing cache writes an identification anomaly status. The identification anomaly status is saved along with the node status record, and the component is placed in the pending review queue. The positioning, attitude, and image records in the pending review queue retain their original source but are not used as acceptance archiving criteria; they are prioritized for component review instructions when subsequent instructions are generated.

[0022] S102: Obtain component positioning data, hoisting posture data and installation image data corresponding to the same precast wall component and perform time alignment, convert the aligned component positioning data to the site coordinate reference, and establish a viewpoint association between the installation image data and the designed installation position to generate a component space sample; Time alignment uses the component's unique code, node identifier, and acquisition time to determine data ownership. The on-site management system first aggregates positioning records, attitude records, and image records according to the component's unique code, then distinguishes between production, transportation, arrival, stacking, hoisting, installation, and acceptance processes according to the node identifier. The acquisition terminal's clock record and the platform's received record are jointly verified. In cases of clock drift, interface retransmission, or acquisition sequence conflicts, the record ownership is determined by the node identifier, source interface, and the status of the previous valid node, and the corrected status is written to the alignment log.

[0023] The component spatial sample is an intermediate record before being written into the digital twin model of the construction site. It includes aligned positioning records, attitude and direction records, image viewpoint records, and design location indexes. When establishing a viewpoint association between the installation image and the designed installation location, the location of the shooting equipment, the shooting direction, and the component edge area in the image are first read. Then, the area covered by the image is correlated with the design centerline, end boundary, and joint area in the designed installation location. If the image is blurry, the edges are obscured, or the shooting viewpoint does not cover the key installation area, an image pending review status is written into the spatial sample. This status continues to be passed to the anomaly level determination stage.

[0024] S103: Generate component entities, orientation directions, and spatial relationships with the designed installation positions based on component spatial samples, and construct a digital twin model of the construction site using the spatial relationships; The component entity consists of a unique component code, a design component number, on-site positioning coordinates, attitude direction records, and an image view index. The attitude direction is derived from the hoisting attitude acquisition records and the component end coordinate identification results, and is used to describe the orientation changes of the component during hoisting and positioning. Spatial relationships are used to connect the component entity, the designed installation location, the actual positioning record, and the installation image view; fields include the design location index, the on-site location index, the image view index, the node status index, and the command status index.

[0025] When building the model, the design installation location layer is loaded first, then the component entities are written into the actual component layer, and the design locations are connected to the actual components through spatial relationships. When an existing component entity is written again, the component's unique code and node identifier are compared first; if it is a node update, the corresponding node field is written; if it is a duplicate report, the source record is retained and a duplicate collection identifier is generated. Original collection records, time alignment records, coordinate transformation records, and entity update records are all saved to the traceability log, which can be retrieved by the component's unique code during subsequent review.

[0026] S104: Obtain the on-site measured coordinates of the control points at the construction site and the design coordinates of the corresponding control points in the overall design plan. Determine the coordinate transformation parameters based on the correspondence between the on-site measured coordinates and the design coordinates of the same control points. The coordinate transformation parameters include plane translation, coordinate axis rotation angle, elevation correction, and scale correction. On-site control points are formed through surveying operations, while design control points originate from the overall design plan. Planar translation is used to record the positional difference between the on-site coordinate origin and the design coordinate origin; coordinate axis rotation angle is used to record the directional difference between the on-site coordinate direction and the design coordinate direction; elevation correction is used to record the datum difference between the on-site elevation datum and the design elevation datum; and scale correction is used to record the proportional correction relationship between the surveyed coordinates and the coordinates on the design drawings. These fields are written into the coordinate transformation parameter record, which includes the source of the control points, the time of parameter generation, the verification status, and the applicable construction area.

[0027] If a control point is missing, a control point with the same name cannot be matched, the control point status is invalid, or the coordinate source fails the permission verification, the new parameters will not be written to the effective record. At this time, the newly collected component positioning data will enter the conversion cache, and the cache record will retain the component's unique code, the original positioning source, and the reason for confirmation. After the control point correspondence is verified, the component positioning data in the conversion cache will then enter the on-site coordinate reference conversion process.

[0028] S105: Convert the coordinates of the component center point, the coordinates of the component end, and the elevation of the component bottom edge in the aligned component positioning data to the unified coordinate system of the construction site according to the coordinate transformation parameters, and perform a closure check between the converted coordinates and the coordinates of the site control points. When the closure check deviation is within the allowable state corresponding to the allowable deviation threshold of the coordinate transformation, generate the component positioning coordinates under the site coordinate reference.

[0029] The coordinates of the component's center point are used to determine the main position of the component entity in the model, the coordinates of the component's ends are used to determine the component's axial direction, and the elevation of the component's bottom edge is used to determine the component's installation height. The conversion process reads the time-aligned positioning record and the effective coordinate conversion parameter record, converts the component's center point, end position, and bottom edge elevation to the unified coordinate system of the construction site, and writes the parameter version and conversion status into the conversion result.

[0030] The allowable deviation threshold for coordinate transformation is stored in the coordinate rule base, derived from project measurement control rules, component installation quality requirements, and measurement equipment calibration records. After the transformed coordinates are closed-loop verified against the coordinates of the on-site control points, the result in the allowable state is written to the actual positioning field of the component entity. When the verification state is under review, the transformation result is not included in the acceptance archiving path; the original positioning record, control point record, parameter version, and review reason are all written to the coordinate review record for use by component review instructions.

[0031] S2: Using component identification data, prefabricated wall components are associated with the design and installation locations. The component status sequence of the prefabricated wall components is determined according to the chronological order of the occurrence of production nodes, transportation nodes, arrival nodes, stacking nodes, hoisting nodes, installation nodes, and acceptance nodes. The component status sequence is then marked in the digital twin model of the construction site. The component status sequence is used to record the process flow of a single component from factory to acceptance. The sequence uses the component's unique code as an index and records the node type, node acquisition time, node identifier, node verification status, location association status, and current spatial status. After this sequence is written into the model, on-site personnel can view the current process of the component through the component entity. In the offset determination process, the components corresponding to hoisting nodes and installation nodes can also be directly filtered.

[0032] S201: Obtain the unique component code and design component number from the component identification data, and read the component location mapping table used to record the correspondence between the design component number and the design installation location; using the design component number as an index, query the corresponding design installation location in the component location mapping table, and write the unique component code, design component number and design installation location into the location association record; The component location mapping table is generated from the analysis of design data, recording the design component number, design installation location, design center point, design axis direction, design elevation, relationship with adjacent components, and design version status. The location association process first checks the design version status, then uses the design component number to query the corresponding design installation location. Once the query is successful, the component's unique code, design component number, and design installation location are written to the location association record.

[0033] If the corresponding design component number is not found in the component location mapping table, if the same design component number corresponds to multiple unconfirmed installation locations, or if the design version is inconsistent with the on-site construction version, the location association record is written to the "Pending Verification" status. The "Pending Verification" status is passed along with the component status sequence. Even if positioning data, attitude data, and installation images have been collected subsequently, the component will first enter the verification path and will not directly enter the acceptance archiving path.

[0034] S202: According to the time sequence of production, transportation, arrival, stacking, hoisting, installation and acceptance, read the node status records corresponding to the location association records, merge the node type, collection time and node identifier of each node into a node status, and sort them according to the preset order of node types to generate a component status sequence. Node status records originate from the production management interface, transportation handover interface, arrival registration interface, stacking management interface, hoisting operation interface, installation confirmation interface, and acceptance record interface. The node type indicates the process step, the acquisition time indicates the order in which the record was created or received by the platform, and the node identifier is used to trace the source of the node record. When a node status is generated, the node type, acquisition time, node identifier, component unique code, and location association status are merged into a status record.

[0035] During sorting, nodes are first organized according to the process sequence, and then valid update records for the same type of node are determined according to the acquisition time. When a production node is missing, the batch traceability record is read as the source to fill the gap; when a transportation node or arrival node is missing, the sequence is written to the node gap status and acceptance archiving is blocked; when there are duplicate records for hoisting or installation nodes, the records that have passed verification and are valid updates are retained, and the replaced records are written to the node change log. The sorted component status sequence is written to the component entity status field and output to the offset determination stage.

[0036] S203: Read the current node status to be written from the component status sequence. The current node status includes the current node type, the current node acquisition time, and the current node identifier. Read the labeled node status from the component entity status field of the construction site digital twin model. The labeled node status includes the labeled node type, the labeled node acquisition time, and the status field update time. The current node status comes from the component status sequence generated this time, while the labeled node status comes from the existing component entity status fields in the model. Before updating the status, the current node type, current node acquisition time, and current node identifier are read first, and then the labeled node type, labeled node acquisition time, and status field update time are read. Both sets of fields are used together for process advancement verification.

[0037] When the component entity status field is empty, the current node status is written as the initial status. If historical node records are not yet complete after system startup, the earliest valid node that has been collected and verified is used as the initial status. For subsequent nodes belonging to processes before the marked status, the node traceability record is written but the current status is not overwritten; for nodes belonging to processes after the marked status, the process sequence comparison is performed.

[0038] S204: According to the preset process sequence of production node, transportation node, arrival node, stacking node, hoisting node, installation node and acceptance node, determine the current process sequence value corresponding to the current node type and the marked process sequence value corresponding to the marked node type respectively; The preset process sequence is stored in the process rule base. The rule base returns the process sequence relationship, node admission conditions, and node rollback handling rules according to node type. The current node type and the labeled node types are queried in the rule base to obtain the process sequence identifier used for status update verification. This identifier is only used to determine whether the node status has progressed along the construction process and is not written into the acceptance record.

[0039] When a node type cannot be found in the process rule base, the current node status is written as an abnormal node type status, and an abnormal status log is generated. When the process rule base version is updated, the status of marked nodes retains the original rule version, and the current node status is compared with the currently effective rule version. The component entity status field synchronously records the rule version, facilitating the review of status changes.

[0040] S205: When the current process sequence value is after the marked process sequence value, and the current node acquisition time is newer than the status field update time, write the current node status into the component entity status field and record the status update time. When the current node meets the process advancement requirements and the data acquisition time meets the update requirements, the component entity status field is written with the current node type, current node identifier, status update time, and node source interface. When the current node is a hoisting node or an installation node, the component entity is synchronously written with a comparable status, and the offset determination process adds the component to the candidate comparison range accordingly.

[0041] A status update log is generated during status writing, recording the component's unique code, status before writing, status after writing, node source, and verification result. If the status update log writing fails, the component entity's status field is set to pending confirmation, and the current node status enters the status rewrite queue. Upon successful rewrite, the status is restored to valid; if the rewrite is not confirmed, the component is not included in the acceptance archiving path.

[0042] S206: When the current process sequence value is before the marked process sequence value, or the current node acquisition time is earlier than the marked node acquisition time, retain the component entity status field and generate a status rollback record.

[0043] Status rollback records are used to save information such as old nodes overwriting new nodes, procedures before a node enters its current state being added, and duplicate reporting by an interface. Record fields include the component's unique code, the status of the labeled node, the current node status, the rollback reason, the source interface, and the processing status. After generating a status rollback record, the component entity status fields in the model are not overwritten.

[0044] When the rollback is due to historical data addition, the node status is written to the node traceability record; when it is due to duplicate reporting, it is written to the data collection duplicate log; when it is due to process abnormality rollback, the rollback record is entered into the component review instruction generation path. This process ensures that the component status always progresses along the construction process sequence, avoiding the reading of invalid node statuses during subsequent offset comparisons.

[0045] S3: Filter the node states of the node type as hoisting node or installation node from the component state sequence, and determine the prefabricated wall components corresponding to the node states. In the digital twin model of the construction site, compare the positioning data of the corresponding components with the design installation position to determine the installation offset. Installation offset is used to record the difference between the actual position of a component and its designed installation position. Fields include horizontal offset, angular offset, elevation offset, offset source, offset verification status, and offset record version. Installation offsets under hoisting nodes are used to determine the hoisting adjustment direction, and installation offsets under installation nodes are used to determine the installation review and acceptance archiving status. Only components with valid node identifiers, valid position association records, and valid coordinate transformation status are included in the valid offset comparison path.

[0046] S301: Read the node type, node acquisition time and node identifier in the component status sequence, filter the prefabricated wall components whose node type is hoisting node or installation node and whose node identifier has passed the verification, and generate a list of components to be compared. The screening process first reads the node types from the component status sequence, retaining hoisting nodes and installation nodes, and then verifies whether the node identifiers can form a consistent traceability among the node status records, component entity status fields, and source interface records. Components that pass the verification are written into the list of components to be compared. The list fields include the component's unique code, design component number, node type, node identifier, node acquisition time, and location association status.

[0047] If a node identifier fails verification, the node acquisition time conflicts with the status field update time, or the location association status is pending verification, the component will not be added to the list of valid components to be compared, but rather to the list of components to be verified. The list of components to be verified retains the original location data, node status, and reasons for the anomaly, and will be invoked by the component verification command later.

[0048] S302: Obtain the component positioning data corresponding to the component list to be compared, convert the component positioning data to the on-site coordinate reference, and read the design center point, design axis direction and design elevation from the design installation position to generate the design position reference; Component positioning data is read from the actual positioning field of the component entity. If the positioning record has not yet been converted to the site coordinate datum, the conversion is completed by calling the effective coordinate transformation parameter record and closure verification result; if the positioning record has been converted and the parameter version is still valid, the component positioning coordinates under the site coordinate datum are read directly. The design center point, design axis direction, and design elevation are read from the component position mapping table and written to the design position datum along with the design version status.

[0049] The design location benchmark is used for subsequent offset comparison. Fields include the design center point, design axis direction, design elevation, design version status, and association with adjacent components. If the design version status is inconsistent, the design location is missing, or the component positioning data is still in the coordinate verification stage, the design location benchmark is written to a pending confirmation status. The offset comparison will not output a valid installation offset, and an offset verification record will be generated.

[0050] S303: Compare the actual center point, actual axis direction, and actual elevation under the site coordinate reference with the design position reference to determine the horizontal offset, angular offset, and elevation offset. Then, calculate the horizontal offset, angular offset, and elevation offset and output the installation offset.

[0051] The actual center point comes from the coordinate transformation result of the component's center point; the actual axis direction comes from the direction record formed by the coordinates of the component's end; and the actual elevation comes from the elevation transformation result of the component's bottom edge. The horizontal offset records the planar deviation of the actual center point relative to the design center point; the angular offset records the directional deviation of the actual axis direction relative to the design axis direction; and the elevation offset records the height deviation of the actual elevation relative to the design elevation. Each offset field retains the positioning source, design source, and coordinate transformation version.

[0052] When multiple valid offset records are generated for the same component at hoisting and installation nodes, the current installation offset is determined based on the current valid node status in the component's status sequence, and the remaining records are written to the offset history record. The offset history record is used to review spatial changes before and after hoisting adjustments and does not overwrite the current installation offset. The current installation offset is written to the component entity offset field and passed to the construction anomaly level determination process.

[0053] S4: Determine the construction anomaly level based on hoisting posture data, installation image data, and installation offset, and generate component review instructions, hoisting adjustment instructions, or acceptance and archiving instructions according to the construction anomaly level.

[0054] The construction anomaly level is determined jointly by posture anomaly results, image anomaly results, and offset anomaly results. Lifting posture data reflects the posture of the component during lifting and the stress state of the lifting equipment; installation image data reflects the joint, obstruction, and surface condition of the component after it is in place; and installation offset reflects the difference between the actual position and the designed position of the component. After each of these three types of data undergoes rule comparison, they are merged into a single construction anomaly level. This level directly determines the subsequent generation of review, adjustment, or archiving instructions.

[0055] S401: Obtain the attitude angle deviation, lifting point force change and lifting stabilization time from the lifting attitude data, obtain the joint width, edge occlusion status and surface damage status from the installation image data, and read the horizontal offset, angle offset and elevation offset from the installation offset to generate anomaly judgment parameters. The anomaly detection parameters consist of attitude fields, image fields, offset fields, data source status, and field validity status. Attitude angle deviation records the deviation of the component's hoisting direction from the target direction; lifting point force change records the force change of the lifting equipment; and hoisting stabilization time records the stability of the component's attitude. Joint width records the gap between adjacent components or the connection between a component and the foundation; edge occlusion status records whether the component's edge is identifiable in the image; and surface damage status records the identification result of surface damage. Horizontal offset, angular offset, and elevation offset are read from the installation offset record.

[0056] Before being included in the anomaly detection parameters, fields undergo validity checks. When the attitude field is missing, a "Position Data Missing" status is entered; when the image is blurry, occluded, or cannot be parsed, an "Image Pending Review" status is entered; when the offset field cannot be generated due to unconfirmed coordinate transformation or design location, an "Offset Pending Confirmation" status is entered. Anomaly detection parameters with missing or pending confirmation statuses do not enter the acceptance archiving instruction path but instead enter the review-guided detection process.

[0057] S402: Compare the anomaly determination parameters with the pose threshold table, image threshold table and offset threshold table respectively to determine the pose anomaly result, image anomaly result and offset anomaly result; The attitude threshold table stores the judgment boundaries corresponding to hoisting operation control rules, hoisting equipment calibration records, and component hoisting process requirements. Fields cover attitude angle deviation, hoisting point stress changes, and hoisting stabilization time. The image threshold table stores the judgment boundaries corresponding to installation quality inspection rules, joint inspection requirements, and surface defect identification rules. Fields cover joint width, edge occlusion status, and surface damage status. The offset threshold table stores the judgment boundaries corresponding to design installation allowable deviation rules, measurement control rules, and acceptance inspection rules. Fields cover horizontal offset, angular offset, and elevation offset.

[0058] During comparison, attitude-related fields are only entered into the attitude threshold table, image-related fields are only entered into the image threshold table, and offset-related fields are only entered into the offset threshold table. A normal result is generated when the field status falls into the normal category, a review result is generated when it falls into the review category, an adjustment result is generated when it falls into the adjustment category, and a review-pending result is generated when the field is missing or pending confirmation. Each abnormal result is written to an anomaly determination record, which also saves the rule source, rule version, and field validity status.

[0059] S403: Determine the hierarchical relationship between the attitude anomaly results, image anomaly results, and offset anomaly results to determine the construction anomaly level; The hierarchy of anomaly levels is determined according to the anomaly level judgment rules. These rules are stored in the construction anomaly rule library and include the anomaly result category, candidate level, handling instruction type, and status flow. When attitude anomalies, image anomalies, and offset anomalies enter the level judgment process, the rule version is verified first, and then each anomaly result is converted into a candidate level. If any result has a pending review status, the corresponding candidate level is written into the review guidance status.

[0060] The construction anomaly level is written to the component entity's anomaly status field, and an anomaly judgment log is generated. The log content includes the component's unique code, node type, source of anomaly judgment parameters, anomaly result category, rule version, and output level. If log writing fails, the construction anomaly level is set to a pending confirmation status, the instruction generation process does not output acceptance archiving instructions, and the record is placed in the log rewrite queue. After log rewrite confirmation, the instruction generation process continues.

[0061] S404: Obtain the pre-configured anomaly level determination rules, which include the level value ranges corresponding to the attitude anomaly parameters, image anomaly parameters, and offset anomaly parameters, respectively; according to the anomaly level determination rules, convert the attitude anomaly results, image anomaly results, and offset anomaly results into attitude candidate levels, image candidate levels, and offset candidate levels, respectively. The anomaly level determination rules consist of rule version, applicable nodes, parameter categories, level value range, handling instruction type, and effective status. Attitude anomaly results, image anomaly results, and offset anomaly results are queried in the rule base according to parameter categories and converted into attitude candidate levels, image candidate levels, and offset candidate levels, respectively. Candidate levels serve as intermediate states before the generation of construction anomaly levels, representing the handling intensity corresponding to each data source.

[0062] When the rule base does not return a valid level, the rule version is invalid, or the parameter category does not match the node type, the corresponding candidate level is written to the rule pending confirmation status, and the component entity is simultaneously set to the review-oriented status. The review-oriented status enters the component review instruction generation process, and the original posture record, image parsing results, and offset record are retained as the review basis.

[0063] S405: When at least two of the candidate levels of attitude, image and offset are consistent, the consistent candidate level is determined as the construction anomaly level. When candidate levels are consistent, the level merging process first confirms that the rule sources corresponding to the consistent levels are all valid, and then writes the consistent level into the construction anomaly level field. The data sources involved in the consistency judgment are written into the level source record. The level source record is used for subsequent instruction assembly, and the instruction generation stage uses it to reference component positions, attitude adjustment directions, image verification points, and responsible work teams.

[0064] When the consistency level is in the review-oriented state, the construction anomaly level will be handled in the review category; when the consistency level is in the adjustment-oriented state, the construction anomaly level will be handled in the hoisting adjustment category; when the consistency level is in the archiving-oriented state and there are no fields to be confirmed, the construction anomaly level will be handled in the acceptance archiving category. After the handling direction is determined, the instruction unit generates the corresponding instruction according to the instruction mapping table.

[0065] S406: When the attitude candidate level, image candidate level, and offset candidate level are all inconsistent, the candidate level with the highest level shall be determined as the construction anomaly level; When candidate levels are inconsistent, the candidate level with the most stringent handling requirements is selected as the construction anomaly level according to the order of handling intensity in the anomaly level determination rules. The selected candidate level is written into the construction anomaly level field, and the remaining candidate levels are written into the level conflict record. The level conflict record saves the differences between different data sources for review or adjustment.

[0066] When the selected candidate level corresponds to a review-type handling, a component review instruction is output; when it corresponds to an adjustment-type handling, a hoisting adjustment instruction is output; when it corresponds to an archiving-type handling and none of the candidate levels have fields to be confirmed, an acceptance archiving instruction is output. This rule ensures that a unique handling path can still be formed even when the three types of data are inconsistent.

[0067] S407: Read the instruction type, execution object, and execution time limit corresponding to the construction anomaly level in the instruction mapping table, and obtain the unique component code, current location, and responsible work team corresponding to the prefabricated wall component; The instruction mapping table is used to convert construction anomaly levels into on-site executable instructions. Fields include the construction anomaly level, instruction type, execution target, execution time limit, feedback requirements, and archiving status. The unique component code is read from the component identification data, the current location is read from the current valid component positioning coordinates in the model, and the responsible work team is read from the hoisting operation record, installation operation record, or construction task assignment record. These fields are combined to form the instruction generation input record.

[0068] When the execution object cannot be matched, the responsible work team is missing, the current location is in coordinate verification status, or the instruction mapping table does not return a valid instruction type, the instruction generation input record is written to the instruction pending confirmation status. The instruction pending confirmation status does not generate an acceptance archiving instruction; instead, it enters the management terminal's pending processing queue and saves the component's unique code, anomaly judgment log, and missing field type.

[0069] S408: When the construction anomaly level matches the review instruction, generate a component review instruction that records the component's unique code and current location; The component verification instruction is used for on-site verification of component identity, design location, and actual installation status. Fields include the component's unique code, current location, design component number, design installation location, verification reason, source of verification fields, and receipt status. Verification reasons can originate from anomaly status, location pending verification, image pending verification, offset pending confirmation, status rollback records, or level conflict records. After generation, the component verification instruction is written to the instruction queue, and its status is synchronized to the component entity.

[0070] After the review receipt is confirmed, the review results are written back according to the field source. After component identity is confirmed, the component identification data and location association record are updated; after design location is confirmed, the component location mapping table version status is updated; after image status is confirmed, the installation image parsing status is updated; after offset status is confirmed, the installation offset record is updated. The written-back fields then re-enter the anomaly level determination process.

[0071] S409: When the construction anomaly level matches the adjustment instruction, generate a hoisting adjustment instruction that records the responsible work team and the attitude adjustment direction; The hoisting adjustment command is used to guide adjustments during the hoisting or positioning phase. Fields include the responsible work team, the component's unique code, current position, attitude adjustment direction, offset source, adjustment reason, and confirmation status. The attitude adjustment direction is determined based on the hoisting attitude record and installation offset record, indicating the required on-site orientation and attitude state of the component. The hoisting adjustment command is written to the hoisting operation queue and associated with the component entity in the model.

[0072] After the hoisting adjustment command is issued, hoisting attitude data and component positioning data are re-collected on-site. The re-collected data is then written back to the component entity according to the component's unique code, the installation offset is updated, and the construction anomaly level is re-evaluated. If the adjustment command is not confirmed, the receipt is not closed, or the re-collected data is still in the adjustment guidance state, the command status remains unclosed, and the management terminal continues to display the adjustment task for that component.

[0073] S410: When the construction anomaly level matches the archiving instruction, generate an acceptance archiving instruction.

[0074] The acceptance archiving instruction is used to write the component installation status into the acceptance record. Fields include the component's unique code, design component number, design installation location, component status sequence, installation offset, construction anomaly level, image record index, acceptance node identifier, and archiving status. Before generating the acceptance archiving instruction, it is confirmed that the component identification data, location association records, component status sequence, installation offset, and anomaly judgment log are all valid, and that there are no unclosed component review instructions or hoisting adjustment instructions.

[0075] After the acceptance archiving instruction is generated, it is written to the acceptance archiving record, and the archiving status is simultaneously fed back to the digital twin model of the construction site. The corresponding component entity is updated to the acceptance node status. Batch traceability records, node status records, coordinate transformation records, offset records, anomaly judgment logs, and instruction receipt records together form a full-process traceability chain for the component. If the archiving write fails, the component entity retains the installation node status, the acceptance archiving instruction enters the archiving rewrite queue, and the component entity status field is updated only after the archiving rewrite is confirmed.

[0076] The digital management system for prefabricated wall construction, used in conjunction with the aforementioned method, includes a component acquisition and modeling unit, an association and annotation unit, an offset determination unit, and an instruction unit. The component acquisition and modeling unit is responsible for accessing component identification data, component positioning data, hoisting posture data, and installation image data. It performs time alignment, coordinate transformation, image perspective association, and writing to the digital twin model of the construction site, outputting component entities, spatial relationships, and positioning records under the site coordinate reference. The association and annotation unit reads component identification data, component position mapping tables, and node status records, generates position association records and component status sequences, and writes valid node statuses into the component entity status field. The offset determination unit reads the component entities corresponding to hoisting and installation nodes, calls the positioning coordinates under the site coordinate reference and the design position reference, forms the installation offset, and writes it into the offset record. The instruction unit reads hoisting posture data, installation image data, and installation offsets, calls the posture threshold table, image threshold table, offset threshold table, anomaly level judgment rules, and instruction mapping table, generates component review instructions, hoisting adjustment instructions, or acceptance archiving instructions, and feeds back the instruction status and receipt status to the construction site digital twin model.

[0077] During system operation, each unit is connected through a unique component code, design component number, node identifier, status field, and instruction status field. The component acquisition and modeling unit only outputs acquisition records and model fields, without generating disposal instructions; the association and annotation unit only maintains the component process status and does not determine installation offset; the offset determination unit only outputs the installation offset and does not determine the final disposal; the instruction unit only generates instructions and record receipts based on confirmed fields and does not rewrite the original acquisition records. When interface transmission fails, field parsing fails, permission verification fails, or log writing fails, the relevant record enters a pending confirmation state, the acceptance archiving path is paused, and the pending confirmation state is restored through review instructions, rewrite queues, or rule verification before re-entering the corresponding processing stage.

[0078] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A digital management method for the construction process of prefabricated walls, characterized in that, include: Acquire component identification data, component positioning data, hoisting posture data, and installation image data generated at the production, transportation, arrival, stacking, hoisting, installation, and acceptance stages of prefabricated wall components; and construct a digital twin model of the construction site based on the component positioning data, the hoisting posture data, and the installation image data. The prefabricated wall components are associated with the designed installation location using the component identification data. The component status sequence of the prefabricated wall components is determined according to the chronological order of the occurrence of production nodes, transportation nodes, arrival nodes, stacking nodes, hoisting nodes, installation nodes, and acceptance nodes. The component status sequence is then marked in the construction site digital twin model. Node states of type hoisting node or installation node are selected from the component state sequence, and the prefabricated wall components corresponding to the node states are determined. The positioning data of the corresponding components are compared with the designed installation position in the construction site digital twin model to determine the installation offset. Based on the hoisting posture data, the installation image data, and the installation offset, the construction anomaly level is determined, and a component review instruction, hoisting adjustment instruction, or acceptance archiving instruction is generated according to the construction anomaly level.

2. The digital management method for the construction process of prefabricated walls according to claim 1, characterized in that, The steps for constructing the digital twin model of the construction site include: Read the electronic identification carrier on the prefabricated wall component, extract the component's unique code, production batch code, and design component number, match the component's unique code with the design component number one-to-one, write the production batch code into the corresponding component's batch traceability record, and output the component identification data; Obtain component positioning data, hoisting posture data and installation image data corresponding to the same prefabricated wall component and perform time alignment. Convert the aligned component positioning data to the on-site coordinate reference and establish a viewpoint association between the installation image data and the designed installation position to generate a component space sample. Based on the component spatial samples, the component entity, its orientation, and spatial relationship with the designed installation position are generated, and the construction site digital twin model is constructed using the spatial relationship.

3. The digital management method for the construction process of prefabricated walls according to claim 2, characterized in that, The process of converting the aligned component positioning data to the on-site coordinate reference is as follows: Obtain the on-site measured coordinates of the control points at the construction site and the design coordinates of the corresponding control points in the overall design plan. Determine the coordinate transformation parameters based on the correspondence between the on-site measured coordinates and the design coordinates of the same control point. The coordinate transformation parameters include plane translation, coordinate axis rotation angle, elevation correction, and scale correction. The coordinates of the component center point, the coordinates of the component end, and the elevation of the component bottom edge in the aligned component positioning data are converted to the unified coordinate system of the construction site according to the coordinate transformation parameters. The converted coordinates are then checked against the coordinates of the site control points. When the deviation of the closure check is less than the allowable deviation threshold of the coordinate transformation, the component positioning coordinates under the site coordinate reference are generated.

4. The digital management method for the construction process of prefabricated walls according to claim 1, characterized in that, The steps for determining the component state sequence include: Obtain the component unique code and design component number from the component identification data, and read the component location mapping table used to record the correspondence between the design component number and the design installation location; using the design component number as an index, query the corresponding design installation location in the component location mapping table, and write the component unique code, the design component number and the design installation location into the location association record; According to the time sequence of production, transportation, arrival, stacking, hoisting, installation and acceptance, the node status records corresponding to the location association records are read, and the node type, collection time and node identifier of each node are merged into a node status. The components are then sorted according to the preset order of the node types to generate the component status sequence.

5. The digital management method for the construction process of prefabricated walls according to claim 1, characterized in that, The process of annotating the component state sequence in the digital twin model of the construction site includes: Read the current node status to be written from the component status sequence. The current node status includes the current node type, the current node acquisition time, and the current node identifier. Read the labeled node status from the component entity status field of the construction site digital twin model. The labeled node status includes the labeled node type, the labeled node acquisition time, and the status field update time. According to the preset process sequence of production node, transportation node, arrival node, stacking node, hoisting node, installation node and acceptance node, determine the current process sequence value corresponding to the current node type and the marked process sequence value corresponding to the marked node type respectively; When the current process sequence value is greater than the marked process sequence value, and the current node acquisition time is newer than the status field update time, the current node status is written into the component entity status field, and the status update time is recorded. When the current process sequence value is less than the marked process sequence value, or the current node acquisition time is earlier than the marked node acquisition time, the component entity status field is retained, and a status rollback record is generated.

6. The digital management method for the construction process of prefabricated walls according to claim 1, characterized in that, The steps for determining the installation offset include: Read the node type, node acquisition time and node identifier in the component status sequence, filter the prefabricated wall components whose node type is hoisting node or installation node and whose node identifier passes the verification, and generate a list of components to be compared. Obtain the component positioning data corresponding to the list of components to be compared, convert the component positioning data to the on-site coordinate reference, and read the design center point, design axis direction and design elevation from the design installation position to generate the design position reference; The actual center point, actual axis direction, and actual elevation under the site coordinate reference are compared with the design position reference to determine the horizontal offset, angular offset, and elevation offset. The horizontal offset, angular offset, and elevation offset are then statistically analyzed to output the installation offset.

7. The digital management method for the construction process of prefabricated walls according to claim 1, characterized in that, The steps for determining the level of construction anomaly include: The system acquires the attitude angle deviation, lifting point force change and lifting stabilization time from the lifting attitude data, acquires the joint width, edge occlusion status and surface damage status from the installation image data, and reads the horizontal offset, angle offset and elevation offset from the installation offset to generate anomaly judgment parameters. The anomaly determination parameters are compared with the pose threshold table, image threshold table, and offset threshold table respectively to determine the pose anomaly result, image anomaly result, and offset anomaly result; The relative levels of the posture anomaly results, the image anomaly results, and the offset anomaly results are determined to ascertain the construction anomaly level.

8. The digital management method for the construction process of prefabricated walls according to claim 7, characterized in that, The process of determining the construction anomaly level also includes: Obtain pre-configured anomaly level determination rules, which include the level value ranges corresponding to pose anomaly parameters, image anomaly parameters, and offset anomaly parameters, respectively; and convert the pose anomaly results, image anomaly results, and offset anomaly results into pose candidate levels, image candidate levels, and offset candidate levels, respectively, according to the anomaly level determination rules. When at least two of the candidate levels of the attitude candidate level, the image candidate level, and the offset candidate level are consistent, the consistent candidate level is determined as the construction anomaly level; When the attitude candidate level, the image candidate level, and the offset candidate level are all inconsistent, the candidate level with the highest level is determined as the construction anomaly level.

9. The digital management method for the construction process of prefabricated walls according to claim 1, characterized in that, The process of generating component review instructions, hoisting adjustment instructions, or acceptance archiving instructions includes: Read the instruction type, execution object, and execution time limit corresponding to the construction anomaly level from the instruction mapping table, and obtain the unique component code, current location, and responsible work team corresponding to the prefabricated wall component; When the construction anomaly level matches the review instruction, a component review instruction is generated that records the unique code of the component and the current location; When the construction anomaly level matches the adjustment instruction, a hoisting adjustment instruction is generated that records the responsible work team and the attitude adjustment direction; When the construction anomaly level matches the archiving instruction, an acceptance archiving instruction is generated.

10. A digital management system for the construction process of prefabricated walls, characterized in that, The system is used to implement the digital management method for the construction process of prefabricated walls as described in any one of claims 1-9, including: The component acquisition and modeling unit is used to acquire component identification data, component positioning data, hoisting posture data, and installation image data generated at the production, transportation, arrival, stacking, hoisting, installation, and acceptance stages of prefabricated wall components, and to construct a digital twin model of the construction site based on the component positioning data, the hoisting posture data, and the installation image data. The association annotation unit is connected to the component acquisition and modeling unit. It is used to associate the prefabricated wall component with the design and installation location using the component identification data. According to the chronological order of the occurrence of production node, transportation node, arrival node, stacking node, hoisting node, installation node and acceptance node, the component state sequence of the prefabricated wall component is determined, and the component state sequence is annotated in the construction site digital twin model. The offset determination unit is connected to the associated annotation unit. It is used to filter the node status of the node type as hoisting node or installation node from the component status sequence, and determine the prefabricated wall component corresponding to the node status. In the construction site digital twin model, the positioning data of the corresponding component is compared with the design installation position to determine the installation offset. The instruction unit, connected to the offset determination unit, is used to determine the construction anomaly level based on the hoisting posture data, the installation image data, and the installation offset, and to generate component review instructions, hoisting adjustment instructions, or acceptance and archiving instructions according to the construction anomaly level.

Images