Error dynamic elimination method and system for digital construction of special-shaped curtain wall

Abstract

This invention discloses a method and system for dynamic error elimination in the digital construction of irregular-shaped curtain walls. In the field of digital curtain wall construction technology, this invention first sets a unique digital identifier for each curtain wall panel within the BIM model and binds it to theoretical geometric data. Processing and installation batches are then divided based on installation dependencies and error propagation relationships. The first batch of panels is directly CNC machined based on theoretical data, while the remaining panels utilize compensation amounts to adjust tool paths for differentiated cutting at interfaces. After a single batch is installed on-site, point clouds are collected through 3D scanning and compared with the BIM model to extract measured deviations. A time-series prediction method is used to extrapolate the predicted deviations for subsequent panels, and an optimal compensation data package is generated and stored based on the adhesive joint adaptation conditions. This invention constructs a closed-loop control chain for modeling, processing, and inspection, proactively offsetting accumulated construction errors and solving the problems of error propagation and compensation lag in traditional processes, thereby improving the overall installation accuracy and sealing performance of irregular-shaped curtain walls.

Description

Technical Field

[0001] This invention relates to the field of digital construction technology for curtain walls, specifically to a method and system for dynamically eliminating errors in the digital construction of irregular-shaped curtain walls. Background Technology

[0002] With the widespread application of irregular curved curtain walls in various large public buildings, their construction precision directly determines the building's appearance and the safety performance of the building envelope. While digital processing and 3D inspection technologies have been gradually introduced into the construction of irregular curtain walls, the following shortcomings still exist in the dynamic control of errors throughout the entire process: Firstly, most existing technologies rely on theoretical models to uniformly complete the processing and manufacturing of all modules, without combining on-site installation sequence and error propagation laws for batch control. Installation deviations accumulate and are transmitted along the construction sequence, making it difficult to effectively control the alignment accuracy of downstream module splicing interfaces. Secondly, in the machining process, existing CNC machining mostly uses globally unified tool compensation parameters, which cannot make differentiated cutting corrections for the actual deviations of different splicing interfaces. This makes it difficult to adapt to the different characteristics of the interfaces of irregular plates, resulting in poor control of joint uniformity and sealing performance. Third, existing error correction methods are mostly based on post-installation adjustments using measured data from already installed areas, lacking a time-series prediction mechanism for deviations in subsequent installation batches. This makes it impossible to perform pre-compensation during the processing stage, resulting in a significant lag in error elimination. Therefore, there is an urgent need for a method and system for dynamic error elimination in the digital construction of irregular-shaped curtain walls. Summary of the Invention

[0003] To address the shortcomings of existing technologies, this invention provides a method and system for dynamic error elimination in the digital construction of irregular-shaped curtain walls, which solves the problems of error accumulation, compensation lag, and insufficient splicing accuracy in irregular-shaped curtain walls.

[0004] To achieve the above objectives, the present invention provides the following technical solution: a method for dynamic error elimination in the digital construction of irregularly shaped curtain walls, comprising: S1. In the BIM model, establish a unique digital identifier for each irregular curtain wall panel and bind its theoretical geometric data. At the same time, based on the specific location of the panel in the entire installation chain and the direction of error propagation, divide all panels into multiple installation batches. S2. For the first batch of plates, the CNC system directly calls the theoretical geometric data to perform machining. For non-first batch plates, the CNC system retrieves the optimal compensation amount corresponding to the digital identifier of the plate to be machined from the compensation data storage unit, and adjusts the tool path offset parameters of the corresponding plate according to the optimal compensation amount, so as to perform differentiated cutting on the splicing interface between plates. S3. After the current batch of panels is installed, use 3D scanning to obtain the point cloud of the installed curtain wall area, compare it with the theoretical BIM model, extract the measured deviation value of each panel, and use the deviation time series prediction method to calculate the estimated deviation of each panel in the next processing batch based on all the measured deviation data obtained. Then, combine the estimated deviation to determine the optimal compensation amount of each panel, generate a compensation data package indexed by the panel digital identifier, and store it in the compensation data storage unit.

[0005] As a further aspect of the present invention, the specific rules for establishing a unique digital identifier for each irregularly shaped curtain wall panel are as follows: In the BIM parametric modeling environment, three data items are recorded for each irregular curtain wall panel: the parent surface number to which the panel belongs, the row index and column index of the panel in the parent surface mesh division, and the design thickness value of the panel. The above three data items are concatenated into a string in a fixed order of "mother surface number - row index - column index - thickness", and this string is used as the unique numerical identifier for this section.

[0006] As a further aspect of the present invention, the theoretical geometric data includes the overall outline dimensions of the plate, surface control parameters, the design joint width between each edge of the plate and adjacent plates, and the position coordinates of the connection and fixing points on the back of the plate.

[0007] As a further aspect of the present invention, the specific operation of dividing all modules into multiple installation batches is as follows: In the BIM model, each panel is assigned an installation sequence number that reflects its placement order, according to the construction and installation process of the irregular curtain wall. Traverse all adjacent block pairs that have joint interfaces, determine the error propagation direction based on the relationship between their sequence numbers, and thus organize the error propagation relationship between all blocks into a directed relationship structure. In the constructed directed relational structure, retrieve all modules that do not have any upstream dependent modules, assign them to the first installation batch, and remove modules that have been assigned to the batch from the relational structure. In the remaining structure, re-search for modules without upstream dependencies and designate them as the next installation batch. Repeat the removal and retrieval operations until all modules are assigned to the corresponding batches. The resulting batch sequence is the execution order of subsequent CNC batch processing and on-site batch installation.

[0008] As a further aspect of the present invention, the specific rule for determining the direction of error propagation is as follows: the plate with the smaller serial number is taken as the source of error in the pair, and the plate with the larger serial number is taken as the receiver of error.

[0009] As a further aspect of the present invention, the specific operation of adjusting the toolpath offset parameters of the corresponding plate according to the optimal compensation amount is as follows: The CNC system extracts the optimal compensation amount corresponding to each splicing interface of the current processing block from the compensation data packet, and writes the optimal compensation amount corresponding to each interface into the tool offset variable pre-allocated for that interface in the CNC system; In the CNC machining program of the plate, the cutting path segment of each splicing interface is pre-segmented and marked. When the tool radius compensation command is called, the dedicated offset variable of the corresponding interface is used as the compensation value input parameter. During machining, the tool travels along the path to a certain joint interface segment, automatically reads the current value of the bias variable bound to that segment, and drives the tool to complete the cutting of that interface segment under the condition of radial bias of that value.

[0010] As a further aspect of the present invention, the specific steps for extracting the measured deviation values ​​of each segment are as follows: After the current batch of panels is installed and fixed, a 3D laser scanner is used to perform a full field-of-view scan of the installed curtain wall area to obtain the original point cloud containing the surface shape information of the installed curtain wall. The unique digital identifier of each panel in the installed batch and its theoretical spatial boundary, i.e. the connection sequence of each corner point of the panel outline in the global coordinate system, are called from the BIM model. Using the theoretical boundary as a clipping template, the points located within the projection range of the theoretical boundary of each plate are marked in the original point cloud. Points belonging to the same plate are stored separately as a subset of the point cloud, and the subset of the point cloud is associated with the digital identifier of the plate and stored in the point cloud cache. For each point cloud subset in the point cloud cache, retrieve the theoretical three-dimensional coordinates of the splicing interface endpoint on the block from the BIM model, as well as the geometric extension direction of the two adjacent theoretical edges with the endpoint as the starting and ending point. Using the theoretical endpoint as a reference point, two strip-shaped neighborhoods are defined in the point cloud subset, each closely adjacent to the two theoretical edges. All points within the two neighborhoods are extracted, and spatial straight line fitting is performed on the two neighborhood point sets respectively to obtain two fitted straight lines. The spatial intersection of the two fitted lines is taken as the actual three-dimensional coordinates of the interface endpoint. After obtaining the actual coordinates of all interface endpoints one by one, the difference vector between the actual coordinates and theoretical coordinates of each endpoint in the global coordinate system is calculated. The difference vectors of each endpoint are arranged in sequence as the measured deviation value of the segment, and the deviation value is bound to the segment's digital identifier and written into the measured deviation database.

[0011] As a further aspect of the present invention, the specific operation of calculating the estimated deviation of each segment in the next processing batch using the deviation time-series prediction method is as follows: The deviation data of each scanned and installed plate is extracted from the measured deviation database. The time series of the interface endpoint deviation of each plate is established according to the scanning timestamp, forming an error propagation state chain from the upstream plate to the downstream plate. Each node in the chain stores the current deviation vector of each interface endpoint of the corresponding plate. For each piece to be processed in the next batch, locate the nodes corresponding to all its direct upstream pieces in the error propagation state chain, and retrieve the current deviation vector of each interface endpoint of these upstream nodes. The deviation vector of each upstream endpoint is compared component by component with the design glue joint width vector of the interface: when a component of the deviation vector exceeds the corresponding glue joint width component, the excess part is transmitted to the corresponding endpoint of the plate to be processed as a forced displacement component along the interface normal; when the component of the deviation vector does not exceed the glue joint width component, the forced displacement component is zero. All forced displacement components are collected and applied to the theoretical positions of the plate to be processed according to their spatial locations to obtain the estimated deviation values ​​of each endpoint of the plate.

[0012] As a further aspect of the present invention, the specific rules for determining the optimal compensation amount for each segment based on the estimated deviation are as follows: For each joint of each panel in the next processing batch, obtain the design joint width value and the maximum allowable compression rate of the specified sealant from the BIM model; Read the estimated deviation value corresponding to the interface: When the estimated deviation points to the outside of the plate, calculate the absorbable deviation of the joint = design joint width × (1 - maximum allowable compression ratio), and take the estimated deviation value minus the absorbable deviation value as the optimal compensation amount; when the estimated deviation points to the inside of the plate, directly take the estimated deviation value as the optimal compensation amount.

[0013] The error dynamic elimination system for digital construction of irregularly shaped curtain walls includes: The batch division module establishes a unique digital identifier for each irregular curtain wall panel in the BIM model and binds it to its theoretical geometric data. At the same time, based on the specific location of the panel in the entire installation chain and the direction of error propagation, all panels are divided into multiple installation batches. The compensation cutting module allows the CNC system to directly call theoretical geometric data to perform machining for the first batch of plates. For non-first batch plates, the CNC system retrieves the optimal compensation amount corresponding to the digital identifier of the plate to be machined from the compensation data storage unit, and adjusts the tool path offset parameters of the corresponding plate according to the optimal compensation amount, so as to perform differentiated cutting on the splicing interface between plates. The predictive compensation module, after the current batch of panels is installed, uses 3D scanning to acquire the point cloud of the installed curtain wall area, compares it with the theoretical BIM model, extracts the measured deviation value of each panel, and calculates the estimated deviation of each panel in the next processing batch based on all the acquired measured deviation data using the deviation time series prediction method. Then, it combines the estimated deviation to determine the optimal compensation amount for each panel, generates a compensation data package indexed by the panel's digital identifier, and stores it in the compensation data storage unit.

[0014] This invention provides a method and system for dynamic error elimination in the digital construction of irregular-shaped curtain walls, which has the following advantages compared with the prior art: (1) This invention relies on the unique digital identifier of the panel and the directed division mechanism of the installation batch to sort out the error transmission link of the curtain wall construction and process the panels in batches according to the installation dependency relationship. The first batch of panels is processed based on the theoretical model, and the subsequent panels rely on the pre-compensation of measured deviations to block the accumulation of errors from the source of construction, solve the defects of traditional process without sequential control and continuous transmission of deviations, and improve the overall consistency of the curtain wall joints. (2) This invention extracts the measured deviation on site to extrapolate the estimated deviation of the downstream plate, and calculates the optimal cutting compensation amount in combination with the adhesive joint capacity. The error correction is brought forward to the CNC machining stage, so as to realize the early offsetting of the deviation and eliminate the problem of compensation lag. Attached Figure Description

[0015] Figure 1 This is a flowchart of the steps of the present invention; Figure 2 This is the system principle block diagram of the present invention. Detailed Implementation

[0016] 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, and 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.

[0017] like Figure 1 This invention provides a method for dynamically eliminating errors in the digital construction of irregularly shaped curtain walls; As an embodiment of this application, it includes: S1. In the BIM model, establish a unique digital identifier for each irregular curtain wall panel and bind its theoretical geometric data. At the same time, based on the specific location of the panel in the entire installation chain and the direction of error propagation, divide all panels into multiple installation batches. S2. For the first batch of plates, the CNC system directly calls the theoretical geometric data to perform machining. For non-first batch plates, the CNC system retrieves the optimal compensation amount corresponding to the digital identifier of the plate to be machined from the compensation data storage unit, and adjusts the tool path offset parameters of the corresponding plate according to the optimal compensation amount, so as to perform differentiated cutting on the splicing interface between plates. S3. After the current batch of panels is installed, use 3D scanning to obtain the point cloud of the installed curtain wall area, compare it with the theoretical BIM model, extract the measured deviation value of each panel, and use the deviation time series prediction method to calculate the estimated deviation of each panel in the next processing batch based on all the measured deviation data obtained. Then, combine the estimated deviation to determine the optimal compensation amount of each panel, generate a compensation data package indexed by the panel digital identifier, and store it in the compensation data storage unit.

[0018] As a second embodiment of this application, it is implemented based on the first embodiment, except that this embodiment includes: S1. Establish a unique digital identifier for each irregularly shaped curtain wall panel in the BIM model and bind it to its theoretical geometric data; Irregular curtain walls are often different from standard curtain walls. Their panels are mostly one-piece molds, meaning that the geometry, curvature, and boundary dimensions of each panel may vary depending on the needs of the building's curved surface. They are easily confused by the naked eye or by traditional numbering. Establishing a unique digital identifier for each module in the BIM model is equivalent to giving it a digital ID card that accompanies it throughout its entire life cycle. The specific rules for establishing a unique digital identifier for each irregularly shaped curtain wall panel are as follows: In the BIM parametric modeling environment, three data items are recorded for each irregular curtain wall panel: the parent surface number to which the panel belongs, the row index and column index of the panel in the parent surface mesh division, and the design thickness value of the panel. In the parametric modeling process of irregular curtain walls, the entire building exterior surface is divided into several grid units, each grid unit uniquely corresponding to a panel. The parent surface number determines the building surface area to which the panel belongs (such as the east facade, the curved roof, etc.), while the row index and column index lock the unique grid position of the panel inside the surface. The above three data items are concatenated into a string in a fixed order of "mother surface number - row index - column index - thickness", and this string is used as the unique numerical identifier of the section. Since the mesh divisions of the same surface do not overlap, different plates cannot occupy the same row and column positions. Therefore, the combination of "parent surface number + row index + column index" naturally guarantees uniqueness. The addition of the thickness value further distinguishes between thin and thick plate variants designed for the same mesh position due to structural needs. The theoretical geometric data includes the overall outline dimensions of the plate, surface control parameters, the design joint width between each edge of the plate and adjacent plates, and the position coordinates of the connection and fixing points on the back of the plate. Overall outline dimensions refer to the basic quantities that describe the geometric range occupied by the plate as an independent entity in space, including but not limited to: the length and width of the plate in its designed unfolded state, i.e., the maximum dimension of the plate's shape in the main extension direction; and the thickness of the plate, i.e., the normal distance from the front surface to the back surface in the cross-sectional direction of the plate. Surface control parameters refer to the mathematical descriptive quantities that define the continuous bending morphology of a plate surface using parametric surface representation methods (such as Bézier surfaces), including: Control point coordinates are the three-dimensional coordinates of each vertex of the spatial lattice that makes up the control polygon of the surface. The control points determine the approximate shape and local concavity and convexity of the surface. The order of a surface refers to the polynomial degree that controls the surface in each parameter direction. The higher the order, the richer the details that the surface can express. Usually, irregularly shaped curtain wall panels are expressed using second- or third-order surfaces. The node vector is a one-dimensional array that defines the segmentation method of the parameter interval, which determines the segmentation method of the surface in the parameter space and the degree of convergence of the surface near the control points. Simultaneously, based on the specific location and error propagation attributes of each module in the entire installation chain, all modules are divided into multiple installation batches, and the execution order of each batch is determined. The specific operation is as follows: In the BIM model, each panel is assigned an installation sequence number that reflects its placement order, according to the construction and installation process of the irregular curtain wall. Traverse all adjacent block pairs that have joint interfaces, determine the error propagation direction based on the relationship between their sequence numbers, and thus organize the error propagation relationship between all blocks into a directed relationship structure. The specific rule for determining the direction of error propagation is as follows: the segment with the smaller serial number is taken as the source of error in the pair, and the segment with the larger serial number is taken as the receiver of error. In the constructed directed relational structure, retrieve all modules that do not have any upstream dependent modules, assign them to the first installation batch, and remove modules that have been assigned to the batch from the relational structure. In the remaining structure, re-search for modules without upstream dependencies and designate them as the next installation batch. Repeat the removal and retrieval operations until all modules are assigned to the corresponding batches. The resulting batch sequence is the execution order of subsequent CNC batch processing and on-site batch installation.

[0019] S2. The installation accuracy control of irregular curtain walls is not something that can be achieved overnight. Instead, it should be broken down into two stages: “benchmark establishment” and “batch correction”. The first batch of panels serves as the starting point for the installation of the entire curtain wall area. Their position and orientation constitute the spatial basis for the entire subsequent structure, which is equivalent to the primary control network in measurement. At this stage, the core objective of the operation is to ensure that the first batch of panels themselves conform to the design as much as possible, rather than to adapt to an existing deviation that does not yet exist. Therefore, for the first batch of modules, the CNC system directly calls the theoretical geometric data to perform machining. The specific operation is as follows: Using the overall contour dimensions and surface control parameters of the plate, a three-dimensional surface model of the plate is reconstructed through the NURBS surface interpolation algorithm; The design seam width value of each edge is converted into the cutting offset distance of the corresponding edge, and the tool radius compensation register is driven to automatically expand or shrink the offset distance on the edge path to generate an edge cutting toolpath with seam compensation. Based on the three-dimensional coordinates of the fixed connection points, the cutting point sequence for drilling or milling is automatically generated on the reconstructed curved surface model. All toolpaths and point sequences are merged and compiled into machine tool motion control instructions, which then start the spindle and servo axes to perform machining. For non-first batch panels, since the system has entered the dynamic correction stage at this time, the offset and trend of the installed area relative to the theory can be known through the scanning feedback of the previous batch. At this time, processing according to the original theoretical value will produce gaps because the actual field conditions have changed. Therefore, for non-first batch of plates, the CNC system retrieves the optimal compensation amount corresponding to the digital identifier of the plate to be processed from the compensation data storage unit; Based on this optimal compensation amount, the toolpath offset parameters of the corresponding plates are adjusted to perform differentiated cutting at the splicing interface between plates. The specific operation is as follows: The CNC system extracts the optimal compensation amount corresponding to each splicing interface of the current processing block from the compensation data packet, and writes the optimal compensation amount corresponding to each interface into the tool offset variable pre-allocated for that interface in the CNC system; In the CNC machining program of the plate, the cutting path segment of each splicing interface is pre-segmented and marked. When the tool radius compensation command is called, the dedicated offset variable of the corresponding interface is used as the compensation value input parameter. During machining, the tool travels along the path to a certain joint interface segment, automatically reads the current value of the bias variable bound to that segment, and drives the tool to complete the cutting of that interface segment under the condition of radial bias of that value. Different interface segments read their own bias variable values, and the segments do not interfere with each other.

[0020] S3. After the current batch of panels is installed, use 3D scanning to obtain the point cloud of the installed curtain wall area, compare it with the theoretical BIM model, and extract the measured deviation value of each panel accordingly. Based on all the measured deviation data obtained, the estimated deviation of each segment in the next processing batch is calculated using the deviation time series prediction method. The specific operation is as follows: The deviation data of each scanned and installed module is extracted from the measured deviation database. The time series of the interface endpoint deviation of each module is established according to the scanning timestamp, forming an error propagation state chain from the upstream module to the downstream module. Each node in the chain stores the current deviation vector of each interface endpoint of the corresponding module. For the deviation sequence obtained after multiple batches of scanning of the same module, only the deviation value corresponding to the latest timestamp is retained as the current deviation state of the module. For each piece to be processed in the next batch, locate the nodes corresponding to all its direct upstream pieces in the error propagation state chain, and retrieve the current deviation vector of each interface endpoint of these upstream nodes. The deviation vector of each upstream endpoint is compared component by component with the design glue joint width vector of the interface: when a component of the deviation vector exceeds the corresponding glue joint width component, the excess part is transmitted to the corresponding endpoint of the plate to be processed as a forced displacement component along the interface normal; when the component of the deviation vector does not exceed the glue joint width component, the forced displacement component is zero. The design joint width vector refers to the width value of the designed reserved joint at the interface of two adjacent panels, expressed in vector form. The specific construction method is as follows: In the BIM model, the normal direction of the theoretical boundary of the two adjacent panels at the interface is taken at the midpoint of the joint. The design joint width value is used as the magnitude of the vector, and the normal direction is used as its direction, thereby generating a three-dimensional vector with magnitude and direction. The vector is decomposed into three scalar components in the three spatial coordinate axes. For example, in the global coordinate system, it can be represented as (△X, △Y, △Z), which correspond to the allowable spatial components of the joint width in the X-axis, Y-axis, and Z-axis directions, respectively. All forced displacement components are collected and applied to the theoretical positions of the plate to be processed according to their spatial locations to obtain the estimated deviation values ​​of each endpoint of the plate. Then, by combining the predicted deviation, the optimal compensation amount for each segment is determined, and a compensation data package indexed by the segment's numerical identifier is generated and stored in the compensation data storage unit. The specific rules for determining the optimal compensation amount for each segment based on the estimated deviation are as follows: For each joint of each panel in the next processing batch, obtain the design joint width value and the maximum allowable compression rate of the specified sealant from the BIM model; Read the estimated deviation value corresponding to the interface: When the estimated deviation points to the outside of the plate, calculate the absorbable deviation of the joint = design joint width × (1 - maximum allowable compression ratio), and take the estimated deviation value minus the absorbable deviation value as the optimal compensation amount; when the estimated deviation points to the inside of the plate, directly take the estimated deviation value as the optimal compensation amount.

[0021] As a third embodiment of this application, this embodiment further discloses a method for extracting the measured deviation values ​​of each segment based on embodiments one and two, the specific content of which includes: After the current batch of panels is installed and fixed, a 3D laser scanner is used to perform a full field-of-view scan of the installed curtain wall area to obtain the original point cloud containing the surface shape information of the installed curtain wall. The unique digital identifier of each panel in the installed batch and its theoretical spatial boundary, i.e. the connection sequence of each corner point of the panel outline in the global coordinate system, are called from the BIM model. Using the theoretical boundary as a clipping template, the points located within the projection range of the theoretical boundary of each plate are marked in the original point cloud. Points belonging to the same plate are stored separately as a subset of the point cloud, and the subset of the point cloud is associated with the digital identifier of the plate and stored in the point cloud cache. For each point cloud subset in the point cloud cache, retrieve the theoretical three-dimensional coordinates of the splicing interface endpoint on the block from the BIM model, as well as the geometric extension direction of the two adjacent theoretical edges with the endpoint as the starting and ending point. Using the theoretical endpoint as a reference point, two strip-shaped neighborhoods are defined in the point cloud subset, each closely adjacent to the two theoretical edges. All points within the two neighborhoods are extracted, and spatial straight line fitting is performed on the two neighborhood point sets respectively to obtain two fitted straight lines. The spatial intersection of the two fitted lines is taken as the actual three-dimensional coordinates of the interface endpoint. After obtaining the actual coordinates of all interface endpoints one by one, the difference vector between the actual coordinates and theoretical coordinates of each endpoint in the global coordinate system is calculated. The difference vectors of each endpoint are arranged in sequence as the measured deviation value of the segment, and the deviation value is bound to the segment's digital identifier and written into the measured deviation database.

[0022] As an embodiment of this application, such as Figure 2 As shown, the error dynamic elimination system for digital construction of irregular-shaped curtain walls includes: The batch division module establishes a unique digital identifier for each irregular curtain wall panel in the BIM model and binds it to its theoretical geometric data. At the same time, based on the specific location of the panel in the entire installation chain and the direction of error propagation, all panels are divided into multiple installation batches. The compensation cutting module allows the CNC system to directly call theoretical geometric data to perform machining for the first batch of plates. For non-first batch plates, the CNC system retrieves the optimal compensation amount corresponding to the digital identifier of the plate to be machined from the compensation data storage unit, and adjusts the tool path offset parameters of the corresponding plate according to the optimal compensation amount, so as to perform differentiated cutting on the splicing interface between plates. The predictive compensation module, after the current batch of panels is installed, uses 3D scanning to acquire the point cloud of the installed curtain wall area, compares it with the theoretical BIM model, extracts the measured deviation value of each panel, and calculates the estimated deviation of each panel in the next processing batch based on all the acquired measured deviation data using the deviation time series prediction method. Then, it combines the estimated deviation to determine the optimal compensation amount for each panel, generates a compensation data package indexed by the panel's digital identifier, and stores it in the compensation data storage unit.

[0023] Some of the data in the above formulas are numerical calculations with dimensions removed, and the contents not described in detail in this specification are all prior art known to those skilled in the art.

[0024] The above embodiments are only used to illustrate the technical methods of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical methods of the present invention without departing from the spirit and scope of the technical methods of the present invention.

Claims

1. A method for dynamic error elimination in the digital construction of irregularly shaped curtain walls, characterized in that, include: S1. In the BIM model, establish a unique digital identifier for each irregular curtain wall panel and bind its theoretical geometric data. At the same time, based on the specific location of the panel in the entire installation chain and the direction of error propagation, divide all panels into multiple installation batches. S2. For the first batch of plates, the CNC system directly calls the theoretical geometric data to perform machining. For non-first batch plates, the CNC system retrieves the optimal compensation amount corresponding to the digital identifier of the plate to be machined from the compensation data storage unit, and adjusts the tool path offset parameters of the corresponding plate according to the optimal compensation amount, so as to perform differentiated cutting on the splicing interface between plates. S3. After the current batch of panels is installed, use 3D scanning to obtain the point cloud of the installed curtain wall area, compare it with the theoretical BIM model, extract the measured deviation value of each panel, and use the deviation time series prediction method to calculate the estimated deviation of each panel in the next processing batch based on all the measured deviation data obtained. Then, combine the estimated deviation to determine the optimal compensation amount of each panel, generate a compensation data package indexed by the panel digital identifier, and store it in the compensation data storage unit.

2. The method for dynamic error elimination in the digital construction of irregularly shaped curtain walls according to claim 1, characterized in that, The specific rules for establishing a unique digital identifier for each irregularly shaped curtain wall panel are as follows: In the BIM parametric modeling environment, three data items are recorded for each irregular curtain wall panel: the parent surface number to which the panel belongs, the row index and column index of the panel in the parent surface mesh division, and the design thickness value of the panel. The above three data items are concatenated into a string in a fixed order of "mother surface number - row index - column index - thickness", and this string is used as the unique numerical identifier of the section.

3. The method for dynamic error elimination in the digital construction of irregularly shaped curtain walls according to claim 1, characterized in that, The theoretical geometric data includes the overall outline dimensions of the plate, surface control parameters, the design joint width between each edge of the plate and adjacent plates, and the position coordinates of the connection and fixing points on the back of the plate.

4. The method for dynamic error elimination in the digital construction of irregularly shaped curtain walls according to claim 1, characterized in that, The specific steps for dividing all modules into multiple installation batches are as follows: In the BIM model, each panel is assigned an installation sequence number that reflects its placement order, according to the construction and installation process of the irregular curtain wall. Traverse all adjacent block pairs that have joint interfaces, determine the error propagation direction based on the relationship between their sequence numbers, and thus organize the error propagation relationship between all blocks into a directed relationship structure. In the constructed directed relational structure, retrieve all modules that do not have any upstream dependent modules, assign them to the first installation batch, and remove modules that have been assigned to the batch from the relational structure. In the remaining structure, re-search for modules without upstream dependencies and designate them as the next installation batch. Repeat the removal and retrieval operations until all modules are assigned to the corresponding batches. The resulting batch sequence is the execution order of subsequent CNC batch processing and on-site batch installation.

5. The method for dynamic error elimination in the digital construction of irregularly shaped curtain walls according to claim 4, characterized in that, The specific rule for determining the direction of error propagation is as follows: the plate with the smaller serial number is taken as the source of error in the pair, and the plate with the larger serial number is taken as the receiver of error.

6. The method for dynamic error elimination in the digital construction of irregularly shaped curtain walls according to claim 1, characterized in that, The specific steps for adjusting the toolpath offset parameters of the corresponding plate based on the optimal compensation amount are as follows: The CNC system extracts the optimal compensation amount corresponding to each splicing interface of the current processing block from the compensation data packet, and writes the optimal compensation amount corresponding to each interface into the tool offset variable pre-allocated for that interface in the CNC system; In the CNC machining program of the plate, the cutting path segment of each splicing interface is pre-segmented and marked. When the tool radius compensation command is called, the dedicated offset variable of the corresponding interface is used as the compensation value input parameter. During machining, the tool travels along the path to a certain joint interface segment, automatically reads the current value of the bias variable bound to that segment, and drives the tool to complete the cutting of that interface segment under the condition of radial bias of that value.

7. The method for dynamic error elimination in the digital construction of irregular-shaped curtain walls according to claim 1, characterized in that, The specific steps for extracting the measured deviation values ​​of each segment are as follows: After the current batch of panels is installed and fixed, a 3D laser scanner is used to perform a full field-of-view scan of the installed curtain wall area to obtain the original point cloud containing the surface shape information of the installed curtain wall. The unique digital identifier of each panel in the installed batch and its theoretical spatial boundary, i.e. the connection sequence of each corner point of the panel outline in the global coordinate system, are called from the BIM model. Using the theoretical boundary as a clipping template, the points located within the projection range of the theoretical boundary of each plate are marked in the original point cloud. Points belonging to the same plate are stored separately as a subset of the point cloud, and the subset of the point cloud is associated with the digital identifier of the plate and stored in the point cloud cache. For each point cloud subset in the point cloud cache, retrieve the theoretical three-dimensional coordinates of the splicing interface endpoint on the block from the BIM model, as well as the geometric extension direction of the two adjacent theoretical edges with the endpoint as the starting and ending point. Using the theoretical endpoint as a reference point, two strip-shaped neighborhoods are defined in the point cloud subset, each closely adjacent to the two theoretical edges. All points within the two neighborhoods are extracted, and spatial straight line fitting is performed on the two neighborhood point sets respectively to obtain two fitted straight lines. The spatial intersection of the two fitted lines is taken as the actual three-dimensional coordinates of the interface endpoint. After obtaining the actual coordinates of all interface endpoints one by one, the difference vector between the actual coordinates and theoretical coordinates of each endpoint in the global coordinate system is calculated. The difference vectors of each endpoint are arranged in sequence as the measured deviation value of the segment, and the deviation value is bound to the segment's digital identifier and written into the measured deviation database.

8. The method for dynamic error elimination in the digital construction of irregularly shaped curtain walls according to claim 1, characterized in that, The specific steps for calculating the estimated deviation of each segment in the next processing batch using the deviation time series prediction method are as follows: The deviation data of each scanned and installed plate is extracted from the measured deviation database. The time series of the interface endpoint deviation of each plate is established according to the scanning timestamp, forming an error propagation state chain from the upstream plate to the downstream plate. Each node in the chain stores the current deviation vector of each interface endpoint of the corresponding plate. For each piece to be processed in the next batch, locate the nodes corresponding to all its direct upstream pieces in the error propagation state chain, and retrieve the current deviation vector of each interface endpoint of these upstream nodes. The deviation vector of each upstream endpoint is compared component by component with the design glue joint width vector of the interface: when a component of the deviation vector exceeds the corresponding glue joint width component, the excess part is transmitted to the corresponding endpoint of the plate to be processed as a forced displacement component along the interface normal. When the component of the deviation vector does not exceed the component of the joint width, the forced displacement component is zero; All forced displacement components are collected and applied to the theoretical positions of the plate to be processed according to their spatial locations to obtain the estimated deviation values ​​of each endpoint of the plate.

9. The method for dynamic error elimination in the digital construction of irregularly shaped curtain walls according to claim 1, characterized in that, The specific rules for determining the optimal compensation amount for each segment based on the predicted deviation are as follows: For each joint of each panel in the next processing batch, obtain the design joint width value and the maximum allowable compression rate of the specified sealant from the BIM model; Read the estimated deviation value corresponding to the interface: When the estimated deviation points to the outside of the plate, calculate the absorbable deviation of the joint = design joint width × (1 - maximum allowable compression ratio), and take the estimated deviation value minus the absorbable deviation value as the optimal compensation amount; when the estimated deviation points to the inside of the plate, directly take the estimated deviation value as the optimal compensation amount.

10. A dynamic error elimination system for digital construction of irregularly shaped curtain walls, used to execute the dynamic error elimination method for digital construction of irregularly shaped curtain walls as described in any one of claims 1-9, characterized in that, include: The batch division module establishes a unique digital identifier for each irregular curtain wall panel in the BIM model and binds it to its theoretical geometric data. At the same time, based on the specific location of the panel in the entire installation chain and the direction of error propagation, all panels are divided into multiple installation batches. The compensation cutting module allows the CNC system to directly call theoretical geometric data to perform machining for the first batch of plates. For non-first batch plates, the CNC system retrieves the optimal compensation amount corresponding to the digital identifier of the plate to be machined from the compensation data storage unit, and adjusts the tool path offset parameters of the corresponding plate according to the optimal compensation amount, so as to perform differentiated cutting on the splicing interface between plates. The predictive compensation module, after the current batch of panels is installed, uses 3D scanning to acquire the point cloud of the installed curtain wall area, compares it with the theoretical BIM model, extracts the measured deviation value of each panel, and calculates the estimated deviation of each panel in the next processing batch based on all the acquired measured deviation data using the deviation time series prediction method. Then, it combines the estimated deviation to determine the optimal compensation amount for each panel, generates a compensation data package indexed by the panel's digital identifier, and stores it in the compensation data storage unit.

Images