Refrigeration plant pipeline construction deviation detection method and system based on point cloud mapping

Abstract

The application discloses a refrigeration plant pipeline construction deviation detection method and system based on point cloud mapping, belongs to the field of machine vision and three-dimensional space calculation technology, and comprises the following steps: extracting a theoretical pipeline center axis; performing a bounding box clipping on measured point clouds to output target point clouds; constructing a space transformation matrix based on a geometric centroid and a principal component direction vector to generate registered point clouds; fitting an actual pipeline center axis by using scattered point clusters; calculating the perpendicular distance and intersection angle of the actual and theoretical axis to generate a deviation data set; and comparing the deviation with a threshold value to output a warning signal. The application adopts the space bounding box clipping, transformation matrix registration and axis fitting technology, and can realize the automatic high-precision measurement and calculation and out-of-bound early warning of pipeline geometric deviation.

Description

Technical Field

[0001] This invention relates to the fields of machine vision and three-dimensional spatial computing technology, and in particular to a method and system for detecting construction deviations in chiller room pipelines based on point cloud mapping. Background Technology

[0002] During the construction of large-scale building electromechanical engineering projects, the interior of the chiller room contains a dense network of chilled water pipes and related components. The accuracy of the spatial routing of these pipes directly determines the hydraulic balance of the chiller room and the operating conditions of the chilled water pumps. With the comprehensive implementation of green and energy-saving building standards, there is an objective engineering and technical demand for non-contact and high-precision spatial location verification of chiller room pipeline construction deviations.

[0003] In related technologies, Chinese invention patent application CN121953873A discloses a method for measuring the center perpendicularity of a CNC machining center. The method includes: defining a virtual measurement reference plane in the workspace, and analyzing the spindle posture and planning the measurement path based on this plane. As the measuring device moves along the path, it simultaneously collects distance data between itself and the surface to be measured, and calculates the spatial angle deviation. Crucially, this method correlates the worktable flatness error data with this deviation to generate a comprehensive error correction vector, which is used to compensate for the theoretical spindle posture. The center perpendicularity measurement value is then obtained through comparison and calculation. This method eliminates the interference of errors inherent in the physical reference plane through a virtual reference and separates the worktable surface shape error through data fusion, thereby improving the accuracy and reliability of perpendicularity measurement.

[0004] Traditional pipeline construction deviation detection mainly relies on manual sampling and calculation using surveying tools such as total stations, or on obtaining global point clouds using conventional 3D laser scanning technology, followed by manual visual comparison and experience evaluation by on-site engineers in the post-processing platform by cropping the pipeline's outer contour.

[0005] The aforementioned traditional methods have technical drawbacks in both underlying logic and actual data flow execution. Manual sampling and calculation suffer from sparse sampling density and inaccurate spatial attitude capture, making it easy to miss local deviations. Conventional point cloud comparison methods lack automated spatial mapping benchmarks at the data flow level. It is impossible to establish an accurate spatial coordinate system relationship between the massive and disordered field point cloud and the ideal design model. As a result, the extraction of the pipeline centerline is limited by surface noise and occlusion interference, making it difficult to convert geometric deviations into deep resistance operation indicators such as pump energy efficiency loss. This leads to delayed deviation detection results and an inability to guide energy-saving control in the data center. Summary of the Invention

[0006] To address the aforementioned issues, this invention provides a method and system for detecting construction deviations in chiller room pipelines based on point cloud mapping. It employs core technologies such as spatial bounding box clipping combined with three-dimensional principal component direction vectors to construct a spatial transformation matrix for point cloud registration and fitting of spatially scattered point clusters to the pipeline's central axis. This enables automated calculation of pipeline spatial geometric deviations and objective results in determining energy consumption exceeding limits, triggering early warnings, and implementing closed-loop control.

[0007] The above objectives can be achieved through the following approach: A method for detecting construction deviations in chiller room pipelines based on point cloud mapping includes: receiving chiller room pipeline design documents; extracting the theoretical pipeline spatial feature endpoints contained in the chiller room pipeline design documents; connecting the theoretical pipeline spatial feature endpoints to generate a theoretical pipeline central axis; controlling a laser scanning device to acquire a set of measured surface point clouds at the chiller room pipeline construction site; performing a spatial bounding box clipping operation on the measured surface point cloud set to output a target pipeline point cloud set; extracting the three-dimensional geometric centroid and principal component direction vectors of the target pipeline point cloud set; constructing a spatial transformation matrix based on the three-dimensional geometric centroid and the principal component direction vectors and the theoretical pipeline central axis; multiplying the target pipeline point cloud set by the spatial transformation matrix to perform a spatial position transformation operation to generate registration points. The system generates a cloud of registration points; extracts spatially scattered point clusters corresponding to the outer surface of the pipeline from the cloud of registration points, and fits these spatially scattered point clusters to generate the actual pipeline centerline; calculates the spatial vertical distance and spatial intersection angle between the actual pipeline centerline and the theoretical pipeline centerline, and combines the spatial vertical distance and spatial intersection angle to generate a construction deviation dataset; receives a preset threshold configuration file, extracts the distance limit and angle limit contained in the preset threshold configuration file, subtracts the distance limit from the spatial vertical distance contained in the construction deviation dataset to generate a distance difference, and subtracts the angle limit from the spatial intersection angle to generate an angle difference; when the distance difference is greater than zero or the angle difference is greater than zero, a deviation exceeding the limit warning signal is generated and a signal output operation is performed.

[0008] Optionally, the step of performing a spatial bounding box clipping operation on the measured surface point cloud set to output the target pipeline point cloud set includes: extracting the theoretical pipeline center axis, extending a preset bounding box radius outward along the normal direction of the theoretical pipeline center axis to generate a cylindrical spatial bounding box that encloses the theoretical pipeline center axis; reading the three-dimensional spatial coordinates corresponding to the discrete data points contained in the measured surface point cloud set, and determining whether the three-dimensional spatial coordinates are located within the inner range of the cylindrical spatial bounding box; when the three-dimensional spatial coordinates are located within the inner range of the cylindrical spatial bounding box, extracting and aggregating the corresponding discrete data points to generate the target pipeline point cloud set.

[0009] Optionally, the method further includes: after generating the target pipeline point cloud set, reading a preset number of neighborhoods and a preset fluctuation coefficient, finding the number of reference discrete data points adjacent to each discrete data point in the target pipeline point cloud set, calculating the average Euclidean distance between each discrete data point and all the corresponding reference discrete data points; calculating the global average Euclidean distance and global standard deviation corresponding to all the discrete data points, multiplying the global standard deviation by the preset fluctuation coefficient to generate a floating penalty value, adding the floating penalty value to the global average Euclidean distance to generate a distance elimination threshold; performing a numerical comparison operation between the average Euclidean distance corresponding to each discrete data point and the distance elimination threshold, and deleting the corresponding discrete data point from the target pipeline point cloud set when the average Euclidean distance is greater than the distance elimination threshold.

[0010] Optionally, constructing a spatial transformation matrix based on the three-dimensional geometric centroid, the principal component direction vector, and the theoretical pipeline central axis includes: extracting the spatial direction vector and the coordinates of the spatial midpoint of the theoretical pipeline central axis; calculating the spatial rotation angle and the spatial rotation axis between the principal component direction vector and the spatial direction vector; generating a target rotation matrix based on the spatial rotation angle and the spatial rotation axis; subtracting the three-dimensional geometric centroid from the coordinates of the spatial midpoint to generate a target translation vector; and performing a concatenation operation on the target rotation matrix and the target translation vector to generate the spatial transformation matrix.

[0011] Optionally, the step of fitting the actual pipeline centerline based on the spatially scattered point clusters includes: extracting the spatial distribution covariance matrix of the spatially scattered point clusters; performing eigenvalue decomposition on the spatial distribution covariance matrix to generate an initial cylinder normal vector; dividing the spatially scattered point clusters into multiple local scattered point cluster segments along the direction of the initial cylinder normal vector; constructing a two-dimensional projection plane perpendicular to the initial cylinder normal vector; performing orthogonal projection on the multiple local scattered point cluster segments onto the two-dimensional projection plane to generate corresponding multiple sets of planar data points; performing algebraic circle equation fitting on the multiple sets of planar data points to calculate multiple cross-sectional center coordinates; performing reverse spatial projection on the multiple cross-sectional center coordinates along the direction of the initial cylinder normal vector to generate multiple three-dimensional spatial axis nodes; and sequentially connecting the multiple three-dimensional spatial axis nodes according to their spatial position to generate the actual pipeline centerline.

[0012] Optionally, calculating the spatial vertical distance and spatial intersection angle between the actual pipeline centerline and the theoretical pipeline centerline includes: reading a preset step size value; extracting multiple test sampling points on the actual pipeline centerline according to the preset step size value; constructing a spatial perpendicular line passing through each of the test sampling points and perpendicular to the theoretical pipeline centerline; extracting the intersection coordinates of the spatial perpendicular line and the theoretical pipeline centerline to generate a projection point; calculating the line segment length between the test sampling point and the projection point to generate the spatial vertical distance; extracting the spatial tangent vector of the actual pipeline centerline; extracting the local direction vector of the theoretical pipeline centerline at the corresponding projection point position; calculating the dot product value between the spatial tangent vector and the local direction vector; and performing an inverse cosine function operation based on the dot product value to generate the spatial intersection angle.

[0013] Optionally, the method further includes: after extracting the projection landing point, extracting a first set of spatial coordinates of the test sampling point, extracting a second set of spatial coordinates of the projection landing point, subtracting the second set of spatial coordinates from the first set of spatial coordinates to generate a deviation direction vector; reading a preset three-dimensional engineering reference coordinate system, extracting the component values ​​of the deviation direction vector on the three orthogonal coordinate axes of the three-dimensional engineering reference coordinate system; reading an orientation mapping table established based on sign rules, determining and generating a pipeline offset orientation based on the positive and negative signs of the component values ​​matching the orientation mapping table, and splicing the pipeline offset orientation to the construction deviation dataset.

[0014] Optionally, the method further includes: after splicing and generating the construction deviation dataset, reading the chiller room fluid parameter configuration file, extracting the refrigerant density, pipe inner diameter value, and rated flow rate of the chiller water pump contained in the chiller room fluid parameter configuration file; performing a cross-sectional conversion operation based on the pipe inner diameter value to generate the pipe cross-sectional area, dividing the rated flow rate of the chiller water pump by the pipe cross-sectional area to calculate and generate the refrigerant fluid reference flow velocity; extracting the spatial intersection angle contained in the construction deviation dataset, reading a pre-configured local resistance mapping table, inputting the spatial intersection angle into the local resistance mapping table to perform a matching query operation, and outputting the corresponding local resistance coefficient; and converting the refrigerant flow... The system performs a square operation on the reference flow velocity to generate a velocity square value. It then multiplies the local resistance coefficient, the refrigerant density, and the velocity square value, and divides the result by two to calculate an additional pressure loss value. This additional pressure loss value is multiplied by the rated flow rate of the refrigerant pump to generate an additional energy consumption assessment value. A preset energy-saving red line configuration file is read, and the preset energy-saving allowable limit contained in the configuration file is extracted. The additional energy consumption assessment value is subtracted from the preset energy-saving allowable limit to generate an energy consumption assessment difference. When the energy consumption assessment difference is greater than zero, a pipeline correction and rework instruction is generated and output to a preset monitoring terminal.

[0015] Optionally, the method further includes: when the energy consumption assessment difference is not greater than zero and the spatial vertical distance contained in the construction deviation dataset is greater than zero, extracting the current spatial coordinate information of the actual pipeline centerline; replacing the corresponding theoretical pipeline spatial feature endpoints in the chiller room pipeline design file with the current spatial coordinate information, and rendering to generate an as-built updated pipeline model; reading a preset friction coefficient along the pipeline, and measuring and extracting the actual pipeline physical total length from the as-built updated pipeline model; dividing the actual pipeline physical total length by the pipe inner diameter value to calculate the length-to-diameter ratio value, and then... The aspect ratio, friction coefficient, refrigerant density, and velocity square are multiplied together, and the result is divided by two to calculate the friction loss value. The friction loss value and the additional pressure loss value are added together to generate the chiller room hydraulic balance parameters. Based on the chiller room hydraulic balance parameters, a preset pump frequency conversion mapping library is queried to extract the corresponding target operating frequency value. Based on the target operating frequency value, a pump frequency conversion control reference dataset is generated and sent to the chiller room operation control terminal.

[0016] Based on the same inventive concept, this invention also provides a point cloud mapping-based system for detecting construction deviations in chiller room pipelines, the system comprising: The design analysis module is used to receive the chiller room pipeline design file, extract the theoretical pipeline spatial feature endpoints contained in the chiller room pipeline design file, and generate the theoretical pipeline center axis based on the connection of the theoretical pipeline spatial feature endpoints; The point cloud clipping module is used to control the laser scanning equipment to collect the measured surface point cloud set at the construction site of the chiller room pipeline, and to perform a spatial bounding box clipping operation on the measured surface point cloud set to output the target pipeline point cloud set. The spatial registration module is used to extract the three-dimensional geometric centroid and principal component direction vector of the target pipeline point cloud set, construct a spatial transformation matrix based on the three-dimensional geometric centroid and the principal component direction vector and the theoretical pipeline central axis, and perform a spatial position transformation operation by multiplying the target pipeline point cloud set by the spatial transformation matrix to generate a registered point cloud set. The axis fitting module is used to extract spatially scattered point clusters corresponding to the outer surface of the pipeline in the registration point cloud set, and to generate the actual pipeline center axis based on the spatially scattered point clusters. The deviation calculation module is used to calculate the spatial vertical distance and spatial intersection angle between the actual pipeline center axis and the theoretical pipeline center axis, and to combine the spatial vertical distance and the spatial intersection angle to generate a construction deviation dataset; The over-limit early warning module is used to receive a preset threshold configuration file, extract the distance limit and angle limit contained in the preset threshold configuration file, subtract the distance limit from the spatial vertical distance contained in the construction deviation dataset to generate a distance difference, subtract the angle limit from the spatial intersection angle to generate an angle difference, and generate a deviation over-limit early warning signal and perform signal output operation when the distance difference is greater than zero or the angle difference is greater than zero.

[0017] Compared with the prior art, the present invention has the following advantages: 1. By extracting the central axis of the theoretical pipeline and constructing a cylindrical spatial bounding box, the measured surface point cloud set is clipped. Then, a spatial transformation matrix is ​​constructed using the 3D geometric centroid of the target pipeline point cloud set and the principal component direction vectors to perform spatial position transformation. This underlying computational mechanism overcomes the drawback of traditional laser scanning methods where the on-site point cloud and design model cannot be accurately correlated. It eliminates background noise interference, achieves automatic registration of the measured pipeline and the theoretical pipeline under the same coordinate reference, and improves the computational efficiency and mapping accuracy of point cloud extraction.

[0018] 2. By extracting the spatial distribution covariance matrix of scattered point clusters and performing eigenvalue decomposition to generate the initial cylinder normal vector, the scattered point clusters are orthogonally projected onto a two-dimensional projection plane for algebraic circle equation fitting. Then, a reverse spatial projection is performed to generate three-dimensional spatial axis nodes, which are then connected to generate the actual pipeline centerline. This dimensionality reduction and cross-sectional centerline reconstruction process solves the problem of traditional manual selection of the outer contour being easily affected by local occlusion and deformation interference, improving the accuracy of capturing the actual spatial trajectory of the pipeline.

[0019] 3. By calculating the spatial intersection angle between the actual pipeline centerline and the theoretical pipeline centerline, and then multiplying this angle sequentially by the friction coefficient of the refrigeration pipeline and the density of the refrigeration fluid, the additional head loss is calculated. This, combined with the pump's rated flow rate, generates an additional energy consumption assessment value. This data transfer process breaks through the limitations of traditional detection methods that only involve geometric surface comparison, objectively transforming spatial morphological deviations into fluid dynamic resistance and pump energy consumption penalties, providing accurate energy-saving benchmarks for determining whether to issue corrective commands.

[0020] 4. When the energy consumption assessment difference does not exceed the standard, the current spatial coordinate information of the actual pipeline centerline is directly extracted to replace the corresponding endpoints in the design document to render the as-built updated pipeline model. Based on this, the hydraulic balance parameters of the chiller room are recalculated and sent to the pump frequency conversion control terminal. This mechanism eliminates unnecessary rework for objective deviations within the tolerance range. By using the reverse update of the digital twin model, it breaks down the data barrier between deviation verification and subsequent operation control, realizing the adaptive correction of the pump frequency conversion control strategy.

[0021] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures pointed out in the description, claims and drawings. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1 This is a flowchart illustrating the method for detecting construction deviations in chiller room pipelines based on point cloud mapping, according to an embodiment of the present invention.

[0024] Figure 2This is a schematic diagram of a cylindrical space bounding box trimming model of a method and system for detecting construction deviations in chiller room pipelines based on point cloud mapping, according to an embodiment of the present invention.

[0025] Figure 3 This is a schematic diagram illustrating the local Euclidean distance statistics and rejection threshold determination of the method and system for detecting construction deviations in chiller room pipelines based on point cloud mapping, according to an embodiment of the present invention.

[0026] Figure 4 This is a schematic diagram of the orthogonal projection of scattered point clusters and the fitting of the cross-sectional center of a method and system for detecting construction deviations in chiller room pipelines based on point cloud mapping, according to an embodiment of the present invention.

[0027] Figure 5 This is a schematic diagram of the spatial vertical distance and intersection angle offset model of the method and system for detecting construction deviations of chiller room pipelines based on point cloud mapping, according to an embodiment of the present invention.

[0028] Figure 6 This is a schematic diagram of the pipeline energy consumption assessment and judgment threshold comparison model of the method and system for detecting construction deviations in chiller room pipelines based on point cloud mapping, according to an embodiment of the present invention.

[0029] Figure 7 This is a schematic diagram of the structure of the chiller room pipeline construction deviation detection system based on point cloud mapping according to an embodiment of the present invention. Detailed Implementation

[0030] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0031] Reference Figure 1 One embodiment of the present invention proposes a method for detecting construction deviations in chiller room pipelines based on point cloud mapping. It employs spatial bounding box clipping combined with three-dimensional principal component direction vectors to construct a spatial transformation matrix for point cloud registration and spatial random point cluster fitting of the pipeline central axis as core technical means. This method can achieve the objective results of automatically calculating pipeline spatial geometric deviations and determining energy consumption exceeding limits for early warning and closed-loop control.

[0032] The method described in this embodiment specifically includes: The system receives a chiller room pipeline design document, extracts the theoretical pipeline spatial feature endpoints contained in the design document, and connects these endpoints to generate a theoretical pipeline central axis. It then controls a laser scanning device to acquire a set of measured surface point clouds from the chiller room pipeline construction site, performs a spatial bounding box clipping operation on the measured surface point cloud set, and outputs a target pipeline point cloud set. The system extracts the 3D geometric centroid and principal component direction vectors of the target pipeline point cloud set, constructs a spatial transformation matrix based on the centroid and principal component direction vectors and the theoretical pipeline central axis, multiplies the target pipeline point cloud set by the spatial transformation matrix to perform a spatial position transformation operation, and generates a registration point cloud set. Finally, it extracts the registration point cloud set. The system generates a spatially scattered cluster of points on the outer surface of the pipeline, and fits these clusters to generate the actual pipeline centerline. It then calculates the spatial vertical distance and spatial intersection angle between the actual pipeline centerline and the theoretical pipeline centerline, and combines these to generate a construction deviation dataset. Finally, it receives a preset threshold configuration file, extracts the distance and angle limits contained within it, subtracts the distance limits from the spatial vertical distance in the construction deviation dataset to generate a distance difference, and subtracts the angle limits from the spatial intersection angle to generate an angle difference. When the distance difference is greater than zero or the angle difference is greater than zero, a deviation exceeding the limit warning signal is generated and a signal output operation is performed.

[0033] Specifically, the theoretical pipeline spatial feature endpoints are extracted from the chiller room pipeline design documents and connected to generate the theoretical pipeline central axis as a reference. A laser scanning device is used to collect the measured surface point cloud set of the pipelines on site, and spatial bounding box clipping technology is used to filter the massive point cloud and output the target pipeline point cloud set. The three-dimensional geometric centroid and principal component direction vectors of the target pipeline point cloud set are extracted and used together with the theoretical pipeline central axis to construct a spatial transformation matrix. Matrix multiplication is used to complete the spatial position transformation operation to obtain a precisely aligned registration point cloud set. In the data flow to the fitting stage, spatially scattered point clusters representing the pipeline's outer surface are extracted from the registration point cloud set, and the actual pipeline central axis is reconstructed through mathematical fitting. The spatial vertical distance and spatial intersection angle between the actual pipeline central axis and the theoretical pipeline central axis are measured in three-dimensional space to form a construction deviation dataset. This dataset is then subtracted from the distance and angle limits set in the preset threshold configuration file. Once any difference exceeds zero, a deviation exceeding the limit warning signal is triggered, and the terminal output action is executed to complete the overall monitoring closed loop. The core effectiveness of the initial technical solution is built upon a collaborative mechanism of three-dimensional spatial matrix transformation and inverse axis fitting. By leveraging spatial bounding box clipping combined with a spatial transformation matrix constructed from the three-dimensional geometric centroid and principal component direction vectors, it directly overcomes the data matching defects of traditional point cloud comparison, such as the isolation and disconnect between the on-site scanning coordinate system and the design drawing coordinate system, and the susceptibility to interference from cluttered background point clouds. This ensures automatic and high-precision spatial mapping of measured data to theoretical benchmarks. Inverse fitting of the actual pipeline centerline is performed on the registered spatially scattered point clusters, and based on this, spatial vertical distances and spatial intersection angles are extracted. This quantifies the originally blurred pipeline surface contour deviation into an absolute axis offset index, overcoming the subjective blind spots and accuracy bottlenecks of visual assessment under manual sampling. By comparing the multi-dimensional deviation index with the limits in the preset threshold configuration file, an automated and data-driven construction quality early warning capability is achieved, replacing experience-based judgment. This enables objective, verifiable, and closed-loop monitoring of the deviation status of concealed pipeline works in the chiller room.

[0034] Optionally, the step of performing a spatial bounding box clipping operation on the measured surface point cloud set to output the target pipeline point cloud set includes: extracting the theoretical pipeline center axis, extending a preset bounding box radius outward along the normal direction of the theoretical pipeline center axis to generate a cylindrical spatial bounding box that encloses the theoretical pipeline center axis; reading the three-dimensional spatial coordinates corresponding to the discrete data points contained in the measured surface point cloud set, and determining whether the three-dimensional spatial coordinates are located within the inner range of the cylindrical spatial bounding box; when the three-dimensional spatial coordinates are located within the inner range of the cylindrical spatial bounding box, extracting and aggregating the corresponding discrete data points to generate the target pipeline point cloud set.

[0035] Specifically, the theoretical pipeline's central axis is extracted, and a pre-defined bounding box radius is extended outward along the normal direction of this axis. This pre-defined bounding box radius is generated from the configuration file. The outward extension operation generates a cylindrical spatial bounding box that completely encloses the theoretical pipeline's central axis, serving as the clipping comparison boundary. The data reading thread reads the 3D spatial coordinates of each discrete data point in the measured surface point cloud set. To accurately determine whether the 3D spatial coordinates are within the cylindrical spatial bounding box, a point-to-space perpendicular distance determination mechanism is designed based on the principle of spatial vector projection. Its calculation design is based on the spatial projection theorem from a spatial coordinate point to a direction vector, and the corresponding calculation formula is as follows: ; ; In the formula, It represents the length of the projection line segment of the discrete data point above the central axis of the theoretical pipeline, and its dimension is millimeters. as well as as well as These represent the absolute coordinate differences between the three-dimensional spatial coordinates of the target discrete data point and the coordinates of the reference point of the theoretical pipeline center axis in the three orthogonal spatial coordinate axes, with the dimension being millimeters. as well as as well as These represent the dimensionless components of the unit direction vector of the theoretical pipeline's central axis on the three corresponding orthogonal spatial coordinate axes. D represents the absolute spatial perpendicular distance from the calculated target discrete data point to the theoretical pipeline's central axis, with dimensions in millimeters. The calculated absolute spatial vertical distance D is given by its square root. This distance is then transmitted to a logic comparator and compared with the preset bounding box radius. When the absolute spatial vertical distance D is not greater than the preset bounding box radius and the 3D spatial coordinates are determined to be within the inner range of the cylindrical bounding box, the logic comparator outputs a high-level trigger instruction to control the memory extractor to extract the corresponding discrete data points and write them into the target cache pool. All discrete data points in the cache pool are then aggregated to generate the target pipeline point cloud set and output to the next-level registration module.

[0036] like Figure 2 The figure visually illustrates the spatial filtering boundary formed by the radius of the pre-defined bounding box extending outward from the central axis of the theoretical pipeline in a two-dimensional cross-section. The figure clearly compares the geometric distribution of the target pipeline point cloud extracted within the inner range with the background free point cloud removed outside, thus confirming the physical boundary determination logic of the point cloud clipping algorithm.

[0037] The theoretical pipeline centerline is a spatial geometric reference line generated by connecting the coordinates of feature endpoints extracted from the chiller room design documents by the preceding feature extraction module. The preset bounding box radius is a pre-defined threshold boundary used to delineate the effective point cloud extraction range, obtained by the control module from the configuration file. The cylindrical spatial bounding box is a three-dimensional spatial filtering boundary radiating outward from the theoretical pipeline centerline to remove irrelevant environmental background point clouds. The measured surface point cloud set is a cluster of data points capturing the actual physical morphology of the chiller room pipeline's exterior from the laser scanning hardware. Discrete data points are basic digital elements with independent three-dimensional spatial coordinate information in the measured surface point cloud set. The target pipeline point cloud set is a high-value spatial point set output object that has undergone boundary filtering to remove background noise and is focused on the surface of the pipeline under test.

[0038] For example, the standard pipe diameter is retrieved and a 50 mm margin is added to generate a preset bounding box radius, which is then forcibly assigned a default size of 150 mm. A set of parameters for the theoretical pipeline centerline is read, and its reference point is set to be exactly at the origin of the spatial coordinate system, with its direction completely parallel to the Z-axis. The corresponding unit direction vector components are... Equal to 0, Equal to 0, The coordinates of a discrete data point received from a laser scanning device are equal to 1. The X-coordinate is 60 mm, the Y-coordinate is 80 mm, and the Z-coordinate is 200 mm. Based on algebraic logic, the three-axis differences between the coordinates of this discrete data point and the reference origin are calculated. Equals 60, Equals 80, The value equals 200. The central processing unit directly substitutes the above constant value into the projection length calculation formula to perform numerical derivation and obtain the result. The value of 200 mm is then loaded into the system register. The microprocessor inputs the coordinate difference and the projected line segment length into the spatial distance calculation formula to perform the underlying solution derivation and generate... The square root of the distance is calculated, and the absolute spatial vertical distance D is calculated to be 100 mm, with dimensions consistent with physical reality. The calculated 100 mm absolute spatial vertical distance is identified as less than the preset bounding box radius threshold of 150 mm and within the axial range. A register enable signal is then sent to the storage controller, pushing the index of the memory address of the discrete data point onto the reserved stack. After extracting and iteratively processing a single discrete data point, the remaining matching point arrays are assembled and spliced ​​to generate the target pipeline point cloud set, which is then sent to the 3D registration module for execution.

[0039] Optionally, the method further includes: after generating the target pipeline point cloud set, reading a preset number of neighborhoods and a preset fluctuation coefficient, finding the number of reference discrete data points adjacent to each discrete data point in the target pipeline point cloud set, calculating the average Euclidean distance between each discrete data point and all the corresponding reference discrete data points; calculating the global average Euclidean distance and global standard deviation corresponding to all the discrete data points, multiplying the global standard deviation by the preset fluctuation coefficient to generate a floating penalty value, adding the floating penalty value to the global average Euclidean distance to generate a distance elimination threshold; performing a numerical comparison operation between the average Euclidean distance corresponding to each discrete data point and the distance elimination threshold, and deleting the corresponding discrete data point from the target pipeline point cloud set when the average Euclidean distance is greater than the distance elimination threshold.

[0040] Specifically, after generating the target pipeline point cloud set, the preset neighborhood number and preset fluctuation coefficient are read and loaded into the cache. For each discrete data point residing in the target pipeline point cloud set, a three-dimensional tree index structure is used to diverge outward in three-dimensional space until a specified number of reference discrete data points with the closest spatial physical distance to the current target point are identified. Based on the principle of spatial geometric Euclidean distance, the average Euclidean distance between each discrete data point and all corresponding reference discrete data points is calculated. The calculation design is based on the arithmetic mean mechanism of the sum of squares of the differences in multidimensional spatial coordinates. The calculation formulas are as follows: ; ; ; ; In the formula, as well as as well as These represent the absolute coordinate differences between the currently processed discrete data point and a locked reference discrete data point on three orthogonal spatial coordinate axes, with the dimension in millimeters. This represents the sum of the squares of the differences between the three coordinate axes. The absolute Euclidean distance between two spatial data points calculated in a single operation is measured in millimeters. This represents the sum of the spatial Euclidean distances between the current discrete data point and all found reference discrete data points. This represents the preset number of dimensionless integer neighbors read into memory. This represents the average Euclidean distance corresponding to a specific discrete data point in the calculated output, with dimensions in millimeters. This average Euclidean distance is stored in a local feature database for subsequent global statistical operations.

[0041] Traverse the local feature database corresponding to all discrete data points, and calculate the global mean Euclidean distance and global standard deviation corresponding to all discrete data points based on the statistical dispersion measurement principle. The calculation formulas are as follows: ; ; ; ; ; ; In the formula, the letters This represents the total number of dimensionless discrete data points contained in the target pipeline point cloud set. (Letter combination) This represents the average Euclidean distance across all discrete data points. The sum of all terms. (Letters) This represents the calculated global average Euclidean distance, measured in millimeters. (Letter) It represents the absolute difference between the average Euclidean distance of a single data point and the global average. Represents the difference The square value. Letter combination. This represents the sum of the squared differences between all discrete data points. (Letter) Represents the population variance. (Letter) The global standard deviation generated by the operation is represented by millimeters.

[0042] After the global baseline data is generated, the global standard deviation is multiplied by a preset volatility coefficient to generate a floating penalty value. Then, the global average Euclidean distance is added to the floating penalty value to generate a distance removal threshold. The logic comparator performs a numerical comparison operation between the average Euclidean distance corresponding to each discrete data point and the calculated distance removal threshold. When the comparison instruction determines that the average Euclidean distance is greater than the distance removal threshold, the storage controller triggers a memory overwrite instruction to delete the corresponding abnormal discrete data point from the storage mapping table of the target pipeline point cloud set.

[0043] like Figure 3 The data histogram shows the statistical distribution characteristics of the average Euclidean distance of each discrete data point, and the vertical dashed line clearly marks the distance removal threshold generated by superimposing the global average Euclidean distance and the floating penalty value, which intuitively reflects the algorithm's mechanism for truncating and filtering long-tailed abnormal noise.

[0044] The preset neighborhood number is a dimensionless spatial topological feature parameter obtained by the system from the engineering acceptance configuration file, used to define the local search and comparison range. The preset fluctuation coefficient is a dispersion tolerance weight parameter pre-entered into the system memory by on-site technicians based on the signal-to-noise ratio of the scanning equipment. The reference discrete data point is the surrounding basic feature coordinate point that is closest in spatial distance to the currently processed target point in the target pipeline point cloud set, locked by a tree-indexing algorithm. The average Euclidean distance is an absolute physical length index that quantifies the density of the spatial distribution of a single target point and its surrounding local point group. The global average Euclidean distance is the statistical mean that reflects the average density and sparseness baseline state of the data point distribution on the surface of the entire pipeline point cloud. The global standard deviation is a dynamic statistical feature value that characterizes the degree of spatial dispersion of the local density of all data points on the pipeline surface relative to the overall average level. The floating penalty value is the threshold increment that measures the degree to which local environmental noise deviates from the mainstream data group of the pipeline body and converts it across dimensions into physical length dimensions. The distance removal threshold is a physical safety upper limit boundary dynamically and adaptively calculated by the system based on the overall macroscopic features of the pipeline and the local microscopic fluctuations, used to determine and permanently filter out outlier noise points.

[0045] For example, a point cloud filtering configuration file is read from the electromechanical scanning specification database to forcibly assign a preset neighborhood number constant value of 4 and a preset fluctuation coefficient weight value of 1.5. A total of N discrete data points (5 in total) are pushed into the target pipeline point cloud set cache pool. The physical distances of the four reference discrete data points surrounding the first discrete data point are found to be equal to 10 mm, and the sum S equals 40, resulting in an average Euclidean distance A of 40 / 4 = 10 mm. Similarly, the average Euclidean distances of the second, third, and fourth discrete data points are calculated to be 12 mm, 8 mm, and 10 mm respectively. Due to interference from pipeline reflections, the spatial distances of the four neighbors of the fifth discrete data point are all equal to 50 mm, resulting in an average Euclidean distance of 200 / 4 = 50 mm. All the above average Euclidean distances are then aggregated and globally summed. The result is 10 + 12 + 8 + 10 + 50 = 90. Dividing this by the total number of points (5) generates the global average Euclidean distance M, which is 90 / 5 = 18 millimeters. The processor sequentially performs the difference calculation, and the difference D between each point and the average distance is 10 - 18 = -8, 12 - 18 = -6, 8 - 18 = -10, 10 - 18 = -8, and 50 - 18 = 32. The processor then performs the squaring operation C, obtaining the corresponding values ​​of 64, 36, 100, 64, and 1024. The squared differences at each point are then summed. Using 1288 / 5, the variance V is calculated to be 257.6. Performing a 0.5 power root operation on this variance generates the global standard deviation F. The floating penalty value is calculated to be 1.5 * 16.05 = 24.075 mm, and the distance removal threshold is 18 + 24.075 = 42.075 mm. The data comparator pushes the average Euclidean distance of the fifth discrete data point, 50 mm, into the numerical comparison instruction. It determines that 50 is greater than 42.075, and then controls the memory pointer to permanently delete the fifth reflective noise point from the target pipeline point cloud set, outputting a clean point cloud sequence.

[0046] Optionally, constructing a spatial transformation matrix based on the three-dimensional geometric centroid, the principal component direction vector, and the theoretical pipeline central axis includes: extracting the spatial direction vector and the coordinates of the spatial midpoint of the theoretical pipeline central axis; calculating the spatial rotation angle and the spatial rotation axis between the principal component direction vector and the spatial direction vector; generating a target rotation matrix based on the spatial rotation angle and the spatial rotation axis; subtracting the three-dimensional geometric centroid from the coordinates of the spatial midpoint to generate a target translation vector; and performing a concatenation operation on the target rotation matrix and the target translation vector to generate the spatial transformation matrix.

[0047] Specifically, the theoretical pipeline centerline generated by the computation is extracted, and the coordinates of its two end feature points are read to perform a difference operation to extract the spatial direction vector and the coordinates of the spatial midpoint of the theoretical pipeline centerline. To determine the degree of spatial attitude deviation of the actual point cloud relative to the design baseline, attitude alignment calculation is performed based on the spatial geometric algebra principle of three-dimensional vector cross product and inner product. The corresponding underlying solution formulas are as follows: ; ; ; ; In the formula, as well as as well as These represent the dimensionless component values ​​of the principal component direction vectors extracted and output by the preceding principal component analysis module along the three axes of the three-dimensional orthogonal coordinate system. as well as as well as These represent the dimensionless component values ​​of the extracted spatial direction vector along the corresponding three axes. as well as as well as These represent the dimensionless component values ​​of the spatial rotation axis center in the corresponding three axes, respectively. This set of values ​​is fed to the attitude solver to determine the rotation reference axis. W represents the dimensionless cosine value generated by the sum of the inner products of the two direction vectors. The inverse trigonometric function library is called to perform an inverse cosine mapping operation on the cosine value W to generate a dimensionless spatial rotation angle. The spatial rotation angle and spatial rotation axis center are extracted as input parameters and imported into the underlying logic algorithm of the Rodrigues rotation transformation to generate the target rotation matrix to indicate the point cloud attitude deflection. Based on attitude alignment, to achieve absolute coincidence of physical spatial positions, distance compensation calculation is performed based on the principle of spatial coordinate translation, and the corresponding formula is as follows: ; ; ; In the formula, as well as as well as These represent three orthogonal absolute physical location coordinates of the midpoint in space, all in millimeters. as well as as well as These represent the three absolute physical coordinates of the three-dimensional geometric centroid, which are statistically output by the preceding module, and all have the dimension of millimeters. as well as as well as These represent the three axial translation components of the calculated target translation vector, each with dimensions in millimeters. The generated target rotation matrix representing attitude deflection and the target translation vector representing position compensation are pushed into the matrix concatenation register to perform homogeneous concatenation, outputting a four-row, four-column spatial transformation matrix, which is then sent out.

[0048] The theoretical pipeline centerline is a reference geometric straight line generated by connecting the feature endpoints extracted from the computer room design documents. The spatial direction vector is a dimensionless feature array indicating the absolute design orientation of the theoretical pipeline in the three-dimensional architectural coordinate system. The spatial midpoint coordinates are the three-dimensional absolute positioning point representing the geometric center of the physical length of the theoretical pipeline. The principal component direction vector is a dimensionless feature array extracted from the measured point cloud to represent the principal axis orientation of the overall data distribution of the pipeline on site, generated by a preceding principal component analysis algorithm. The three-dimensional geometric centroid is the coordinate reference point representing the absolute physical center of gravity of the measured point cloud in three-dimensional space. The spatial rotation angle is the radian or angular parameter quantifying the torsional deviation of the measured pipeline point cloud relative to the design reference in three-dimensional space. The spatial rotation axis is the reference spatial reference vector controlling the spatial rotation correction action of the three-dimensional point cloud. The target rotation matrix is ​​a three-row, three-column mathematical control operator used to perform three-dimensional attitude deflection correction of the point cloud. The target translation vector is a linear distance control command driving the centroid of the measured point cloud to move towards the midpoint of the design reference in three dimensions. The spatial transformation matrix is ​​a four-row, four-column homogeneous transformation mathematical matrix that integrates rotation and translation operations. It is used to drive the target pipeline point cloud set to complete a one-time spatial position and attitude transformation.

[0049] For example, the standard Rodriguez rotation algorithm component is actively retrieved from an open-source 3D point cloud processing function library as a preset pose solver. The storage controller pushes the principal component direction vector components to the processor in the following order: Equals 1, Equal to 0, Equal to 0, and simultaneously, the components of the spatial direction vector of the central axis of the theoretical pipeline are, in order: Equal to 0, Equals 1, The result is equal to 0. The processor hardware multiplier performs a cross product formula to numerically calculate the spatial rotation axis component. , , Simultaneously, the inner product is calculated to obtain the cosine parameters. Calculating the arccosine of 0 generates a dimensionless spatial rotation angle of 1.5708 radians, equivalent to 90 degrees. The preset attitude solver generates a 3x3 target rotation matrix based on 1.5708 radians and the rotation axis vector along the Z-axis of the coordinate system, with its main diagonal elements being 0, 0, and 1 respectively. The coordinates of the midpoint in space are then read. Equal to 100 millimeters, Equal to 200 millimeters, Equal to 300 millimeters, read the three-dimensional geometric centroid coordinates Equal to 50 millimeters, Equal to 100 millimeters, Equal to 300 millimeters. The components of the target translation vector are obtained by applying the translation algebra formula. millimeters millimeters The translation dimensions are all aligned to millimeters. The target translation vector is appended as the fourth column to the target rotation matrix, and a fourth row consisting of 0s and 0s and 1s is added to the bottom to complete the assembly loop of the fourth-order homogeneous space transformation matrix and push it directly into the hardware buffer.

[0050] Optionally, the step of fitting the actual pipeline centerline based on the spatially scattered point clusters includes: extracting the spatial distribution covariance matrix of the spatially scattered point clusters; performing eigenvalue decomposition on the spatial distribution covariance matrix to generate an initial cylinder normal vector; dividing the spatially scattered point clusters into multiple local scattered point cluster segments along the direction of the initial cylinder normal vector; constructing a two-dimensional projection plane perpendicular to the initial cylinder normal vector; performing orthogonal projection on the multiple local scattered point cluster segments onto the two-dimensional projection plane to generate corresponding multiple sets of planar data points; performing algebraic circle equation fitting on the multiple sets of planar data points to calculate multiple cross-sectional center coordinates; performing reverse spatial projection on the multiple cross-sectional center coordinates along the direction of the initial cylinder normal vector to generate multiple three-dimensional spatial axis nodes; and sequentially connecting the multiple three-dimensional spatial axis nodes according to their spatial position to generate the actual pipeline centerline.

[0051] Specifically, spatially scattered point clusters are extracted, and the spatial distribution covariance matrix is ​​extracted and eigenvalue decomposition is performed on it to generate an initial cylinder normal vector. Along the direction of the initial cylinder normal vector, the spatially scattered point clusters are segmented into multiple local scattered point cluster segments. To achieve dimensionality reduction mapping, a two-dimensional projection plane perpendicular to the initial cylinder normal vector is constructed. Orthogonal projection operations are then performed on the multiple local scattered point cluster segments onto the two-dimensional projection plane to generate corresponding sets of planar data points. The calculation design of the orthogonal projection is based on the perpendicular mapping mechanism from spatial points to a plane, and the corresponding calculation formula is as follows: ; Greek letters in the formula Represents the 3D coordinate vector of a point on a reduced plane generated by orthogonal projection, with dimensions in millimeters, and the Greek letter. Represents the three-dimensional coordinate vector of the original spatial points within a local scattered point cluster segment, with dimensions in millimeters, and Greek letters. The scalar value representing the dot product of the original spatial point's coordinate vector and normal vector, with dimensions in millimeters, represented by the Greek letter. The initial cylinder normal vector is represented by a dimensionless unit space vector. After generating a set of planar data points using the projection operation described above, it is added to the fitting queue. Algebraic circle equation fitting operations are performed on multiple sets of planar data points to calculate the coordinates of the center of multiple cross-sections. The coordinates of the center of each cross-section are then subjected to a reverse spatial projection operation along the direction of the initial cylinder normal vector to generate multiple three-dimensional spatial axis nodes. These three-dimensional spatial axis nodes are then connected sequentially according to their spatial positions to generate the actual pipeline center axis, which is output to the deviation comparison module.

[0052] like Figure 4 The diagram shows the annular discrete distribution of scattered point clusters in three-dimensional space after orthogonal projection along the normal vector of the initial cylinder onto the two-dimensional projection plane, as well as the geometric circle boundary of the pipeline outer contour generated by algebraic equation fitting and the coordinates of the center of the centered section circle.

[0053] The spatial distribution covariance matrix is ​​a feature matrix describing the discrete state of spatially scattered point clusters in three orthogonal dimensions. The initial cylinder normal vector is a reference unit vector representing the pipeline orientation, generated through eigenvalue decomposition. Local scattered point cluster segments are independent 3D data blocks formed by truncating and extracting the point cloud along the normal. The 2D projection plane is a mathematical surface perpendicular to the normal vector, used to support the dimensionality-reduced coordinates. The planar data point set is a dimensionality-reduced point set that presents a ring distribution on the 2D plane after eliminating the normal height difference. The cross-section center coordinates are the origin of the location, representing the geometric center of a specific slice level, calculated through algebraic fitting. The 3D spatial axis nodes are the absolute location coordinates of the cross-section center coordinates in real 3D space after height inverse compensation. The actual pipeline center axis is a 3D polyline reflecting the real spatial trajectory of the pipeline on site, formed by sequentially connecting the various 3D spatial axis nodes.

[0054] For example, a preset truncation length of 100 mm is read from a pre-defined slicing parameter configuration file and forcibly assigned. The processor retrieves spatially scattered point clusters and extracts the spatial distribution covariance matrix, then extracts the upward-oriented initial cylinder normal vector through eigenvalue decomposition. Its three orthogonal axis components are 0, 0, and 1, respectively. The point cloud is truncated along this normal to generate a local scattered point cluster segment, and the three-dimensional coordinates of a specific original spatial point within this cluster are read. The three orthogonal axes have values ​​of 30, 40, and 150, respectively. Calculate the vector at this point and the normal vector of the initial cylinder. The scalar value of the dot product Millimeters. Substituting the above variables directly into the projection mapping formula to expand the underlying solution, the planar dimension-reduced coordinate components along the three axes are derived as follows: , , All units are millimeters. Continuously aggregating and reducing the dimensionality of the point set generates a planar data point set for this level. Algebraic circle fitting is then performed on this set, resulting in all axial components of the cross-sectional center coordinates of the slice being 0 millimeters. The cross-sectional center coordinates are then aligned along the normal vector... The original height is compensated by 150 mm to generate the first three-dimensional spatial axis node with three axial components of 0, 0, and 150 mm respectively. The process is iteratively repeated to calculate the second adjacent three-dimensional spatial axis node with a higher height, whose three axial components are 0, 0, and 250 mm respectively. Based on the spatial order of increasing orthogonal height, the two nodes are connected by line segments to complete the fitting and generate the actual pipeline centerline that runs through the pipeline entity.

[0055] Optionally, calculating the spatial vertical distance and spatial intersection angle between the actual pipeline centerline and the theoretical pipeline centerline includes: reading a preset step size value; extracting multiple test sampling points on the actual pipeline centerline according to the preset step size value; constructing a spatial perpendicular line passing through each of the test sampling points and perpendicular to the theoretical pipeline centerline; extracting the intersection coordinates of the spatial perpendicular line and the theoretical pipeline centerline to generate a projection point; calculating the line segment length between the test sampling point and the projection point to generate the spatial vertical distance; extracting the spatial tangent vector of the actual pipeline centerline; extracting the local direction vector of the theoretical pipeline centerline at the corresponding projection point position; calculating the dot product value between the spatial tangent vector and the local direction vector; and performing an inverse cosine function operation based on the dot product value to generate the spatial intersection angle.

[0056] Specifically, the preset step size value is read and distributed. An equidistant cutting operation is performed on the actual pipeline's central axis according to the preset step size value to extract multiple test sampling points, which reside in the system memory array. For each test sampling point, a spatial perpendicular line is constructed passing through that point and perpendicular to the theoretical pipeline's central axis. The intersection coordinates of this spatial perpendicular line and the theoretical pipeline's central axis are extracted to generate a projection point, which is stored in the local feature database. The length of the line segment between the test sampling point and the projection point is calculated to generate the spatial perpendicular distance. This distance calculation is based on the Euclidean distance metric between two points in three-dimensional space, and the specific calculation formula is as follows: ; ; ; ; ; In the formula as well as as well as These represent the three orthogonal physical coordinate components of the test sampling point in the three-dimensional spatial coordinate system, and all have the dimension of millimeters. as well as as well as These represent the three orthogonal physical coordinate components of the projection point in the same coordinate system, and all have the dimension of millimeters. as well as as well as These represent the absolute coordinate differences between two points on their respective coordinate axes. Q represents the sum of the squares of the coordinate differences across the three axes. The distance, measured in millimeters, represents the vertical spatial distance generated by the calculation and is output to the deviation comparison register.

[0057] The spatial tangent vector at the current sampling position is extracted along the extended trajectory of the actual pipeline's central axis, and the local direction vector at the corresponding projection point of the theoretical pipeline's central axis is extracted simultaneously. The attitude calculation module calculates the dot product between the spatial tangent vector and the local direction vector, and performs an inverse cosine function operation based on the dot product to generate the spatial intersection angle. The angle calculation is based on the equivalent mapping principle of the spatial vector dot product and the cosine of the angle, and the specific calculation formula is as follows: ; In the formula as well as as well as These represent the dimensionless component values ​​of the spatial tangent vector in the three orthogonal dimensions. as well as as well as These represent the dimensionless components of the local direction vector in the corresponding three dimensions. C represents the calculated dot product value, which is a dimensionless algebraic scalar. The hardware coprocessor uses the dot product value C as an input variable to generate the spatial intersection angle. The data encapsulation module packages and combines the spatial perpendicular distance and the spatial intersection angle to generate a construction deviation dataset and distributes it.

[0058] like Figure 5 The diagram depicts the spatial pose of the actual pipeline's central axis deviating from the theoretical pipeline's central axis using a digital geometric model. It illustrates the process of generating projection points by sending spatial vertical lines from the test sampling points to the theoretical reference plane, and demonstrates a spatial intersection angle mapping model that characterizes the attitude deflection.

[0059] The preset step size is a physical length reference value obtained by the system from the computer room acceptance standard configuration file, used to control the sampling density. The test sampling point is a discrete position reference origin extracted from the actual pipeline centerline at a constant interval, used to perform point-by-point deviation calculation. The spatial perpendicular line is a three-dimensional mathematical auxiliary line that passes through a specific test sampling point and forms an orthogonal geometric relationship with the theoretical reference axis. The projection point is the three-dimensional spatial absolute positioning coordinate reference point generated by the intersection of the spatial perpendicular line and the theoretical pipeline centerline. The spatial vertical distance is a quantitative characterization of the absolute physical translation error of the actual pipeline axis deviating from the theoretical design centerline. The spatial tangent vector is a dimensionless directional feature array describing the instantaneous three-dimensional trajectory of the actual pipeline centerline at a specific sampling point. The local direction vector is a dimensionless standard orientation reference feature array of the theoretical pipeline centerline in the corresponding mapping area. The dot product value is a dimensionless algebraic product reflecting the degree of directional coincidence between two three-dimensional vectors in physical space. The spatial intersection angle is an absolute physical angle index generated by inverse trigonometric function mapping, used to measure the severity of the actual pipeline attitude deflection. The construction deviation dataset is a comprehensive data array that aggregates pipeline translation error and deflection error, which is directly used for subsequent energy consumption assessment and evaluation under operating conditions.

[0060] For example, the default truncation constant is read from the built-in electromechanical construction quality assessment configuration table, thus actively assigning a preset step size of 500 mm. The first test sampling point is extracted by moving 500 mm along the starting end on the center axis of the fitted actual pipeline, and its three-dimensional spatial coordinate components are... Equal to 60 millimeters, Equal to 80 millimeters, Equal to 100 millimeters. Construct orthogonal spatial perpendiculars and extract the corresponding projection points on the theoretical pipeline's central axis, and their three-dimensional spatial coordinate components. Equal to 0 millimeters, Equal to 0 millimeters, Equal to 100 millimeters. Receive the above data, perform distance calculations, and derive the coordinate difference. millimeters millimeters Millimeters. Performing a sum of squares yields Q equal to 10000. The coprocessor performs a power-0.5 calculation on 10000, yielding a spatial perpendicular distance D equal to 100 millimeters. The actual pipeline spatial tangent vector components at the test sampling point are as follows: Equal to 0, It equals 0.1. Equals 0.995. The theoretical pipeline local direction vector components extracted simultaneously at the projection landing point are as follows: Equal to 0, Equal to 0, Equal to 1. Performing algebraic multiplication and summation, the dot product value C is derived to be 0.995, a dimensionless cosine value. An angle mapping derivation is performed on 0.995 to derive a spatial intersection angle of 0.1 radians. The data bus encapsulates the vertical distance of 100 mm and the intersection angle of 0.1 radians into the construction deviation dataset, entering the subsequent early warning and evaluation loop.

[0061] Optionally, the method further includes: after extracting the projection landing point, extracting a first set of spatial coordinates of the test sampling point, extracting a second set of spatial coordinates of the projection landing point, subtracting the second set of spatial coordinates from the first set of spatial coordinates to generate a deviation direction vector; reading a preset three-dimensional engineering reference coordinate system, extracting the component values ​​of the deviation direction vector on the three orthogonal coordinate axes of the three-dimensional engineering reference coordinate system; reading an orientation mapping table established based on sign rules, determining and generating a pipeline offset orientation based on the positive and negative signs of the component values ​​matching the orientation mapping table, and splicing the pipeline offset orientation to the construction deviation dataset.

[0062] Specifically, after extracting the projection landing point, the first set of spatial coordinates of the test sampling points is extracted from the cached feature library, and the second set of spatial coordinates of the projection landing point is also extracted. The coordinate values ​​of each axis of the first spatial coordinate set are subtracted from the coordinate values ​​of the second spatial coordinate set accordingly to generate a deviation direction vector representing the relative spatial displacement. This step is based on the spatial vector subtraction operation mechanism between two points in a three-dimensional Cartesian coordinate system, and the specific calculation formula is as follows: ; ; ; In the formula as well as as well as These represent the three orthogonal coordinate components of the first spatial coordinate set in the three-dimensional coordinate system, and all have the dimension of millimeters. as well as as well as These represent the three orthogonal coordinate components of the second spatial coordinate set in the same coordinate system, and all have the dimension of millimeters. as well as as well as These values ​​represent the calculated deviation direction vector components on three orthogonal coordinate axes, all in millimeters. This set of values ​​is used to determine the specific offset orientation. A preset 3D engineering reference coordinate system is read, and the component values ​​of the deviation direction vector on the three orthogonal coordinate axes of this system are extracted. An azimuth mapping table based on sign rules is read, and the positive and negative signs of the extracted component values ​​are matched against the azimuth mapping table. The pipeline offset azimuth is determined through low-level table lookup and comparison. The data bus then stitches the pipeline offset azimuth to the previously generated construction deviation dataset to supplement the spatial orientation early warning features.

[0063] The first spatial coordinate set is a three-axis coordinate array containing the absolute physical location of actual measured points in three-dimensional space, generated by the preceding sampling and extraction steps. The second spatial coordinate set is a three-axis coordinate array containing the absolute physical location of theoretical reference points in three-dimensional space, output by the preceding spatial perpendicular line mapping operation. The deviation direction vector is a mathematical entity of a relatively directed line segment that accurately describes the spatial offset of the actual pipeline from the theoretical design center. The preset three-dimensional engineering reference coordinate system is an orthogonal reference frame established during the initialization phase of the building information model for unifying the global spatial positioning of the chiller room. The component value is a one-dimensional absolute length scalar with positive and negative direction attributes extracted by projecting the three-dimensional spatial vector onto the three orthogonal reference axes. The orientation mapping table is a two-dimensional matching logic matrix pre-configured in the system configuration database that translates the algebraic positive and negative sign states of each reference axis into human-readable spatial orientation labels. The pipeline offset orientation is a textual indication result that intuitively indicates the pipeline's deviation from the design direction, output by multi-dimensional symbol comparison logic. The construction deviation dataset is a comprehensive data collection carrier that continuously accumulates various geometric errors and directional characteristics of pipelines in the system's main memory, thereby providing full digital support for the evaluation of working conditions.

[0064] For example, the digital ledger of architectural drawings and specifications injected during the project construction preparation phase is read, and the preset three-dimensional engineering reference coordinate system and orientation mapping table are directly extracted. This ledger specification forcibly assigns a basic symbol rule where the X-axis positive sign maps to the east, the Y-axis negative sign maps to the south, and the Z-axis negative sign maps to the downward direction. The first spatial coordinate set of a test sampling point in a test state is extracted, and its coordinate components with all three directions having millimeter dimensions are recorded. as well as as well as Simultaneously extract the second spatial coordinate set of the corresponding projection point, record its design endpoint coordinate components, and assign theoretical values ​​to them sequentially. as well as as well as The algorithm receives all variables and substitutes them into the spatial vector subtraction formula to perform low-level derivation calculations, directly calculating the component values ​​of the deviation direction vector on each reference axis. millimeters millimeters Millimeters. The signs of these three components are extracted and identified: the X-axis has a negative sign, the Y-axis is zero, and the Z-axis has a negative sign. The extracted combination of negative and zero values ​​is input into a location mapping table in memory for a mapping query, matching and determining the pipeline offset direction as west-down. Upon receiving a successful match flag, the storage controller responds with a write-back command, directly appending the west-down direction text description feature to the tail of the construction deviation dataset.

[0065] Optionally, the method further includes: after splicing and generating the construction deviation dataset, reading the chiller room fluid parameter configuration file, extracting the refrigerant density, pipe inner diameter value, and rated flow rate of the chiller water pump contained in the chiller room fluid parameter configuration file; performing a cross-sectional conversion operation based on the pipe inner diameter value to generate the pipe cross-sectional area, dividing the rated flow rate of the chiller water pump by the pipe cross-sectional area to calculate and generate the refrigerant fluid reference flow velocity; extracting the spatial intersection angle contained in the construction deviation dataset, reading a pre-configured local resistance mapping table, inputting the spatial intersection angle into the local resistance mapping table to perform a matching query operation, and outputting the corresponding local resistance coefficient; and converting the refrigerant flow... The system performs a square operation on the reference flow velocity to generate a velocity square value. It then multiplies the local resistance coefficient, the refrigerant density, and the velocity square value, and divides the result by two to calculate an additional pressure loss value. This additional pressure loss value is multiplied by the rated flow rate of the refrigerant pump to generate an additional energy consumption assessment value. A preset energy-saving red line configuration file is read, and the preset energy-saving allowable limit contained in the configuration file is extracted. The additional energy consumption assessment value is subtracted from the preset energy-saving allowable limit to generate an energy consumption assessment difference. When the energy consumption assessment difference is greater than zero, a pipeline correction and rework instruction is generated and output to a preset monitoring terminal.

[0066] Specifically, the system reads the refrigeration room fluid parameter configuration file and extracts the refrigerant density, pipe inner diameter, and chilled water pump rated flow rate contained within it. Based on the pipe inner diameter, a cross-sectional conversion operation is performed to generate the pipe cross-sectional area. The calculation design is based on the standard circular cross-sectional area geometric area derivation mechanism, and the corresponding calculation formula is as follows: ; Greek letters in the formula Represents the extracted pipe inner diameter value, with the physical dimension in meters, and the Greek letter. The calculated pipe cross-sectional area is expressed in square meters. The reference flow velocity of the refrigerant fluid is calculated by dividing the rated flow rate of the chilled water pump by the pipe cross-sectional area. The design calculation is based on the proportional mapping relationship between the fluid volumetric flow rate and the flow cross-sectional area, as shown in the specific formula below: ; Greek letters in the formula Represents the rated flow rate of a chilled water pump, with the physical dimension being cubic meters per second, represented by the Greek letter. The calculated reference flow velocity of the refrigerant fluid is expressed in meters per second. The data processing engine simultaneously extracts the spatial intersection angles from the construction deviation dataset, reads a pre-configured local resistance mapping table, inputs the spatial intersection angles into the local resistance mapping table to perform a matching query, and outputs the corresponding local resistance coefficients associated with the mapping. The reference flow velocity of the refrigerant fluid is multiplied by itself to generate a squared velocity value. Then, the local resistance coefficient, the refrigerant fluid density, and the squared velocity value are multiplied together, and the result of this multiplication is directly divided by two to calculate the additional pressure loss value. This step is based on the classic physical mechanism in fluid mechanics that local resistance loss is proportional to dynamic head. The specific calculation formula is as follows: ; Greek letters in the formula Represents the dimensionless local drag coefficient of the query output, represented by the Greek letter. Represents the density of the refrigerant, with the physical dimension of kilograms per cubic meter, represented by the Greek letter. The value representing the additional pressure loss generated by the solution is expressed in Pascals. This additional pressure loss value is then multiplied by the rated flow rate of the chilled water pump to calculate the additional energy consumption assessment value. This calculation is based on the physical power equation, which converts the work done by the fluid overcoming resistance into mechanical energy consumption. The corresponding formula is as follows: ; Greek letters in the formula This represents the calculated additional energy consumption assessment value, with the absolute physical dimension in watts. A preset energy-saving red line configuration file is read, and the preset energy-saving allowable limit contained in the configuration file is extracted. A subtractor is used to subtract the preset energy-saving allowable limit from the additional energy consumption assessment value to generate an energy consumption assessment difference. The control chip continuously monitors this difference. When the energy consumption assessment difference is greater than zero, an interrupt request is triggered to generate a pipeline correction and rework command, which is then output to a preset monitoring terminal via the communication bus.

[0067] like Figure 6 The bar chart shown quantifies and compares the ideal baseline energy consumption state, the additional energy consumption assessment value generated by the superposition of additional head loss caused by pipeline space deviation, and the preset energy-saving allowable limit red line defined by engineering specifications. It intuitively presents the absolute value discrimination boundary of the pipeline correction and rework command triggered by the underlying hardware comparison circuit.

[0068] The refrigeration room fluid parameter configuration file is a low-level digital structure ledger containing pipeline media attributes and equipment design parameters. The refrigerant density is an absolute physical constant characterizing the mass per unit volume of the water flow medium within the pipeline network. The pipe inner diameter is a characteristic dimensional parameter guiding the physical width of the water flow passage within the pipeline. The rated flow rate of the chilled water pump is the absolute total volume of fluid medium delivered by the pump drive motor per unit time at its rated operating point. The pipe cross-sectional area is the absolute physical area of ​​the internal annular cross-section perpendicular to the fluid flow direction. The refrigerant reference velocity is the average one-dimensional linear velocity of the fluid medium running in an ideal, straight pipe section without deviation. The local resistance mapping table is a two-dimensional relational database stored in the system memory that maps the spatial angle of pipeline deflection to the characteristic values ​​of the fluid resistance coefficient. The local resistance coefficient is a dimensionless characteristic multiplier that quantitatively measures the degree of local disturbance and stripping caused by pipeline geometric deflection to the smooth flow of the internal fluid. The square value of the velocity is the quadratic algebraic mapping result of the reference velocity, used to characterize the dynamic energy potential of fluid particles. The additional pressure loss value is the absolute hydrostatic pressure difference caused by vortex dissipation when fluid passes through pipe network nodes with construction angle deviations. The additional energy consumption assessment value is the absolute mechanical power cost that the chilled water pump must forcibly output to overcome the additional pressure loss caused by the previous construction deviations. The energy-saving red line configuration file is a digital carrier of engineering rules entered into the system platform to constrain the overall energy efficiency of the chiller room from the tolerance limit. The preset energy-saving allowable limit value is the limit safety red line index pre-assigned by the design institute, representing the allowable overflow power of local deviations in the chiller room. The energy consumption assessment difference is the excess margin separated from the measured energy consumption cost by arithmetic difference between the measured energy consumption cost and the system tolerance limit boundary. The pipeline correction and rework instruction is the low-level control signal automatically coded and generated after the system hard judgment identifies the energy efficiency window to drive the physical pipeline cutting and re-welding action on site. The preset monitoring terminal is a 3D visualization electronic Kanban hardware that receives the control signal and presents the coordinates of the rework point with sound and light.

[0069] For example, parameters are retrieved from the initial HVAC data dictionary of the project, and a default constant value of 1000 kg / m³ is actively assigned to the refrigerant fluid density. Simultaneously, a pipe inner diameter of 0.2 m and a rated flow rate of 0.031416 m / s for the chilled water pump are assigned. The processor substitutes the pipe inner diameter of 0.2 m into the cross-sectional area conversion formula to calculate the pipe cross-sectional area. Square meters. The reference flow rate of the refrigerant was then derived by dividing the rated flow rate by the cross-sectional area. Meters per second. The actual spatial intersection angle of the deflection, read from the pre-existing construction deviation dataset, is 1.57 radians. This radian is input into the local resistance mapping table for key-value comparison, and the corresponding local resistance coefficient, a dimensionless algebraic value of 0.8, is extracted. The hardware multiplier squares the baseline flow velocity of 1.0, still yielding a velocity square value of 1.0. Multiple multiplications followed by division by two are then performed to calculate the additional pressure loss value. Pascal. The additional pressure loss value is directly multiplied by the pump's rated flow rate to perform an energy efficiency calculation, generating an additional energy consumption assessment value. Watts. The energy-saving red line configuration file imported by the project supervisor is retrieved. This file forcibly sets a preset energy-saving allowable limit of 10.0 watts. The logic subtractor performs arithmetic subtraction on the two values, deriving an energy consumption assessment difference of 12.5664 - 10.0 = 2.5664 watts. The comparator hardware circuit keenly detects that the value 2.5664 is mathematically greater than the zero threshold, broadcasting a pipeline correction and rework instruction to the bus, triggering an alarm on the preset monitoring terminal.

[0070] Optionally, the method further includes: when the energy consumption assessment difference is not greater than zero and the spatial vertical distance contained in the construction deviation dataset is greater than zero, extracting the current spatial coordinate information of the actual pipeline centerline; replacing the corresponding theoretical pipeline spatial feature endpoints in the chiller room pipeline design file with the current spatial coordinate information, and rendering to generate an as-built updated pipeline model; reading a preset friction coefficient along the pipeline, and measuring and extracting the actual pipeline physical total length from the as-built updated pipeline model; dividing the actual pipeline physical total length by the pipe inner diameter value to calculate the length-to-diameter ratio value, and then... The aspect ratio, friction coefficient, refrigerant density, and velocity square are multiplied together, and the result is divided by two to calculate the friction loss value. The friction loss value and the additional pressure loss value are added together to generate the chiller room hydraulic balance parameters. Based on the chiller room hydraulic balance parameters, a preset pump frequency conversion mapping library is queried to extract the corresponding target operating frequency value. Based on the target operating frequency value, a pump frequency conversion control reference dataset is generated and sent to the chiller room operation control terminal.

[0071] Specifically, the data flow link is continuously monitored. When the received energy consumption assessment difference is not greater than zero and the spatial vertical distance contained in the construction deviation dataset is detected to be greater than zero, it is determined that the geometric shape of the chiller room pipeline has undergone a physical offset that does not exceed the energy consumption red line. This triggers a data interception command to extract the current spatial coordinate information of the actual pipeline center axis. The 3D rendering engine uses the current spatial coordinate information to directly replace the corresponding theoretical pipeline spatial feature endpoints in the chiller room pipeline design file. Based on the replaced new nodes, the topology is connected and rendered to generate the as-built updated pipeline model. The preset friction coefficient is read, and the actual pipeline physical total length is measured and extracted from the as-built updated pipeline model using a 3D measurement operator. The actual pipeline physical total length is divided by the retained pipe inner diameter value to calculate the length-to-diameter ratio. This calculation is based on the dimensionless geometric mapping mechanism for measuring the relative length of pipe segments in fluid mechanics. The specific calculation formula is as follows: ; Greek letters in the formula Represents the actual total physical length of the pipeline measured and extracted, with absolute physical dimensions in meters, using the Greek letter [unclear]. Represents the inner diameter of the pipe residing in memory, with the absolute physical dimension in meters, represented by the Greek letter. The calculated aspect ratio is a dimensionless, purely algebraic parameter. The aspect ratio, friction coefficient, refrigerant density, and velocity square are multiplied together, and the result is divided by two to calculate the friction loss. This calculation is based on the classical Darcy's law, which describes the dissipation of mechanical energy in a straight pipe due to viscous friction. The corresponding formula is as follows:

[0072] Greek letters in the formula Represents the dimensionless friction coefficient along the line of measurement, represented by the Greek letter. Represents the density of the refrigerant, with the physical dimension of kilograms per cubic meter, represented by the Greek letter. The Greek letter represents the reference velocity of the refrigeration fluid derived from the preceding module, and its square is used to characterize the kinetic head. The value represents the friction loss generated by the solution, with the physical dimension being Pascal. The adder then takes over the data bus and adds the friction loss value to the additional pressure loss value generated in the previous process to generate the hydraulic balance parameters for the chiller room. This step is based on the fluid energy conservation principle, where the total pressure drop of the entire pipe network equals the arithmetic superposition of friction loss and local friction loss, as shown in the following formula: ; Greek letters in the formula Represents the extracted additional pressure loss value, with dimensions in Pascals and Greek letters. The parameters represent the hydraulic balance parameters of the chiller room obtained by superposition and summation, with absolute physical dimensions in Pascals. Based on the chiller room hydraulic balance parameters, a preset pump frequency conversion mapping library is queried from the hard drive to extract the corresponding target operating frequency values. A pump frequency conversion control reference dataset is generated based on the target operating frequency values ​​and then sent to the chiller room operation control terminal via the industrial bus to perform adaptive control of physical fluid conditions.

[0073] The current spatial coordinate information is a 3D positioning data matrix representing the spatial position of the pipeline center trajectory in a 3D absolute coordinate system after construction is completed. The as-built updated pipeline model is a digital twin geometry that incorporates real-world geometric offset data to accurately reflect the physical state of the pipeline network after completion. The preset friction coefficient is a dimensionless empirical constant representing the viscous resistance effect of the pipe wall, pre-determined based on the pipe material and the Reynolds number of the fluid flow regime. The actual total physical length of the pipeline is the absolute physical spatial distance that the fluid medium must traverse in a single pipeline loop. The length-to-diameter ratio is a dimensionless characteristic parameter measuring the relative proportion of the longitudinal extension length of the pipeline to the transverse flow diameter. The friction loss is an index of the irreversible kinetic energy to heat dissipation generated by the fluid overcoming the internal friction between fluid layers in a uniform, straight pipe section. The hydraulic balance parameters of the chiller room comprehensively reflect the total mechanical resistance required to overcome the fluid circulation of the entire system due to the actual geometric distortion of the pipeline network and friction along the pipeline. The pre-set pump frequency conversion mapping library is a two-dimensional digital database that maps specific system pressure drop resistance values ​​to the optimal stator speed of the pump motor. The target operating frequency value is an absolute physical quantity for electrical control that directly determines the alternating cycle of the rotating magnetic field of the pump motor stator. The pump frequency conversion control reference dataset is a structured industrial communication digital packet that encapsulates the target frequency control command and equipment start / stop check code. The chiller room operation control terminal is a programmable automated drive cabinet that directly parses the command packet and uses the frequency converter hardware to adjust the mechanical power output of the pump.

[0074] For example, if the detected energy consumption assessment difference is -5.0 watts and the spatial vertical distance fed back by the construction deviation dataset is 100.0 mm, it is determined that the energy consumption difference does not exceed the zero warning line while the physical displacement is greater than the zero limit. The processing enable pin is then raised to extract the current spatial coordinate information of the actual pipeline's center axis, replacing the design drawing's reference coordinates, and thus reconstructing the as-built updated pipeline model. Based on the national building water supply and drainage design standards, the preset friction coefficient along the pipeline is automatically assigned a default constant of 0.02. The measurement operator scans the as-built updated pipeline model to measure the actual pipeline's total physical length. Equals 50.0 meters. Dividing the actual total physical length of the pipeline by the preceding pipe inner diameter of 0.2 meters, the length-to-diameter ratio is derived from the bottom layer. The computational circuit then performs friction loss calculation, multiplying the aspect ratio of 250.0 by the coefficient of friction of 0.02, the density of 1000.0 kg / m³ by the square of the flow velocity of 1.0, and then dividing by two to generate the friction loss value. Pascal. Combine this friction loss value with the additional pressure loss value of 400.0 Pascals issued by the front-end module to generate the hydraulic balance parameters for the chiller room. Pascal. The control core directly addresses the preset pump frequency conversion mapping library using the comprehensive pressure drop value of 2900.0, finds and extracts the target operating frequency value of 45.0 Hz required to overcome this level of resistance. The 45.0 Hz control signal is integrated to generate a pump frequency conversion control reference dataset and sent in full-duplex along the Ethernet communication link to the chiller room operation control terminal. The underlying inverter hardware implements adaptive compensation for pump power under scenarios of increased fluid impedance.

[0075] Reference Figure 7 Based on the same inventive concept, this invention also provides a point cloud mapping-based system for detecting construction deviations in chiller room pipelines, the system comprising: The design analysis module is used to receive the chiller room pipeline design file, extract the theoretical pipeline spatial feature endpoints contained in the chiller room pipeline design file, and generate the theoretical pipeline center axis based on the connection of the theoretical pipeline spatial feature endpoints; The point cloud clipping module is used to control the laser scanning equipment to collect the measured surface point cloud set at the construction site of the chiller room pipeline, and to perform a spatial bounding box clipping operation on the measured surface point cloud set to output the target pipeline point cloud set. The spatial registration module is used to extract the three-dimensional geometric centroid and principal component direction vector of the target pipeline point cloud set, construct a spatial transformation matrix based on the three-dimensional geometric centroid and the principal component direction vector and the theoretical pipeline central axis, and perform a spatial position transformation operation by multiplying the target pipeline point cloud set by the spatial transformation matrix to generate a registered point cloud set. The axis fitting module is used to extract spatially scattered point clusters corresponding to the outer surface of the pipeline in the registration point cloud set, and to generate the actual pipeline center axis based on the spatially scattered point clusters. The deviation calculation module is used to calculate the spatial vertical distance and spatial intersection angle between the actual pipeline center axis and the theoretical pipeline center axis, and to combine the spatial vertical distance and the spatial intersection angle to generate a construction deviation dataset; The over-limit early warning module is used to receive a preset threshold configuration file, extract the distance limit and angle limit contained in the preset threshold configuration file, subtract the distance limit from the spatial vertical distance contained in the construction deviation dataset to generate a distance difference, subtract the angle limit from the spatial intersection angle to generate an angle difference, and generate a deviation over-limit early warning signal and perform signal output operation when the distance difference is greater than zero or the angle difference is greater than zero.

[0076] It should be noted that the electrical connections between the various units described above do not necessarily represent direct or indirect connections. Any indirect connection method can be applied to the embodiments of the present invention as long as it achieves the purpose of the present invention. The above descriptions are merely exemplary embodiments of the present invention and should not be construed as limiting the scope of the present invention.

[0077] All equivalent changes and modifications made in accordance with the teachings of this invention are still within the scope of this invention. Those skilled in the art will readily conceive of other embodiments of this invention upon considering the specification and the disclosure of practical truth. This application is intended to cover any variations, uses, or adaptations of this invention that follow the general principles of this invention and include common knowledge or conventional techniques in the art not described herein.

Claims

1. A method for detecting construction deviations in chiller room pipelines based on point cloud mapping, characterized in that, include: Receive the chiller room pipeline design file, extract the theoretical pipeline spatial characteristic endpoints contained in the chiller room pipeline design file, and generate the theoretical pipeline center axis based on the connection of the theoretical pipeline spatial characteristic endpoints; The laser scanning equipment is controlled to collect the measured surface point cloud set of the refrigeration room pipeline construction site, and the spatial bounding box clipping operation is performed on the measured surface point cloud set to output the target pipeline point cloud set; Extract the three-dimensional geometric centroid and principal component direction vector of the target pipeline point cloud set. Construct a spatial transformation matrix based on the three-dimensional geometric centroid and the principal component direction vector and the theoretical pipeline central axis. Multiply the target pipeline point cloud set by the spatial transformation matrix to perform a spatial position transformation operation and generate a registration point cloud set. Extract the spatially scattered point clusters corresponding to the outer surface of the pipeline from the registration point cloud set, and generate the actual pipeline center axis based on the spatially scattered point clusters; Calculate the spatial vertical distance and spatial intersection angle between the actual pipeline centerline and the theoretical pipeline centerline, and combine the spatial vertical distance and spatial intersection angle to generate a construction deviation dataset; Receive a preset threshold configuration file, extract the distance limit and angle limit contained in the preset threshold configuration file, subtract the distance limit from the spatial vertical distance contained in the construction deviation dataset to generate a distance difference, subtract the angle limit from the spatial intersection angle to generate an angle difference, and generate a deviation over-limit warning signal and execute signal output operation when the distance difference is greater than zero or the angle difference is greater than zero.

2. The method for detecting construction deviations in chiller room pipelines based on point cloud mapping according to claim 1, characterized in that, The step of performing a spatial bounding box clipping operation on the measured surface point cloud set to output the target pipeline point cloud set includes: Extract the central axis of the theoretical pipeline, and extend it outward along the normal direction of the central axis of the theoretical pipeline by a predetermined bounding box radius to generate a cylindrical space bounding box that encloses the central axis of the theoretical pipeline. Read the three-dimensional spatial coordinates corresponding to the discrete data points contained in the measured surface point cloud set, and determine whether the three-dimensional spatial coordinates are located within the inner range of the cylindrical space bounding box; When the three-dimensional spatial coordinates are located within the inner range of the cylindrical spatial bounding box, the corresponding discrete data points are extracted and aggregated to generate the target pipeline point cloud set.

3. The method for detecting construction deviations in chiller room pipelines based on point cloud mapping according to claim 2, characterized in that, The method further includes: After generating the target pipeline point cloud set, the preset number of neighborhoods and preset fluctuation coefficient are read. The target pipeline point cloud set is used to find the number of reference discrete data points adjacent to each discrete data point in the neighborhood, and the average Euclidean distance between each discrete data point and all the corresponding reference discrete data points is calculated. Calculate the global average Euclidean distance and global standard deviation corresponding to all the discrete data points, multiply the global standard deviation by the preset fluctuation coefficient to generate a floating penalty value, and add the global average Euclidean distance to the floating penalty value to generate a distance elimination threshold. A numerical comparison operation is performed between the average Euclidean distance corresponding to each discrete data point and the distance removal threshold. When the average Euclidean distance is greater than the distance removal threshold, the corresponding discrete data point is deleted from the target pipeline point cloud set.

4. The method for detecting construction deviations in chiller room pipelines based on point cloud mapping according to claim 1, characterized in that, The construction of the spatial transformation matrix based on the three-dimensional geometric centroid, the principal component direction vector, and the theoretical pipeline central axis includes: Extract the spatial direction vector and spatial midpoint coordinates of the theoretical pipeline's central axis; Calculate the spatial rotation angle and spatial rotation axis between the principal component direction vector and the spatial direction vector, and generate the target rotation matrix based on the spatial rotation angle and spatial rotation axis; Subtracting the three-dimensional geometric centroid from the coordinates of the midpoint in space generates the target translation vector; The spatial transformation matrix is ​​generated by concatenating the target rotation matrix and the target translation vector.

5. The method for detecting construction deviations in chiller room pipelines based on point cloud mapping according to claim 4, characterized in that, The process of generating the actual pipeline centerline based on the spatially scattered point cluster includes: Extract the spatial distribution covariance matrix of the spatially scattered point clusters, and perform eigenvalue decomposition on the spatial distribution covariance matrix to generate the initial cylinder normal vector; Along the direction of the initial cylinder normal vector, the spatial scattered point cluster is divided into multiple local scattered point cluster segments; Construct a two-dimensional projection plane perpendicular to the normal vector of the initial cylinder, and perform orthogonal projection operations on the multiple local scattered point clusters onto the two-dimensional projection plane to generate corresponding multiple sets of planar data points; Perform algebraic circle equation fitting operations on the multiple sets of planar data points respectively to calculate and generate the coordinates of the center of multiple cross-sections; The coordinates of the center of each of the multiple cross sections are respectively subjected to reverse spatial projection along the direction of the normal vector of the initial cylinder to generate multiple three-dimensional spatial axis nodes. The multiple three-dimensional spatial axis nodes are then connected in sequence according to their spatial positions to generate the actual pipeline center axis.

6. The method for detecting construction deviations of chiller room pipelines based on point cloud mapping according to claim 1, characterized in that, The calculation of the spatial perpendicular distance and spatial intersection angle between the actual pipeline centerline and the theoretical pipeline centerline includes: Read the preset step size value, and extract multiple test sampling points on the center axis of the actual pipeline according to the preset step size value; Construct a spatial perpendicular line passing through each of the test sampling points and perpendicular to the central axis of the theoretical pipeline, and extract the coordinate points of the intersection of the spatial perpendicular line and the central axis of the theoretical pipeline to generate the projection point; The spatial vertical distance is generated by calculating the length of the line segment between the test sampling point and the projection landing point. Extract the spatial tangent vector of the actual pipeline centerline, extract the local direction vector of the theoretical pipeline centerline at the corresponding projection point, calculate the dot product between the spatial tangent vector and the local direction vector, and generate the spatial intersection angle based on the dot product value by performing an inverse cosine function operation.

7. The method for detecting construction deviations in chiller room pipelines based on point cloud mapping according to claim 6, characterized in that, The method further includes: After extracting the projection landing point, extract the first set of spatial coordinates of the test sampling point, extract the second set of spatial coordinates of the projection landing point, and subtract the second set of spatial coordinates from the first set of spatial coordinates to generate the deviation direction vector. Read the preset three-dimensional engineering reference coordinate system and extract the component values ​​of the deviation direction vector on the three orthogonal coordinate axes of the three-dimensional engineering reference coordinate system; The azimuth mapping table established based on symbol rules is read, and the pipeline offset azimuth is determined by matching the positive and negative signs of the component values ​​with the azimuth mapping table. The pipeline offset azimuth is then spliced ​​into the construction deviation dataset.

8. The method for detecting construction deviations of chiller room pipelines based on point cloud mapping according to claim 7, characterized in that, The method further includes: After splicing and generating the construction deviation dataset, the refrigeration room fluid parameter configuration file is read, and the refrigeration fluid density, pipe inner diameter value and refrigeration water pump rated flow rate contained in the refrigeration room fluid parameter configuration file are extracted. Based on the inner diameter of the pipe, a cross-sectional conversion operation is performed to generate the cross-sectional area of ​​the pipe. The rated flow rate of the chilled water pump is divided by the cross-sectional area of ​​the pipe to calculate and generate the reference flow velocity of the chilled fluid. Extract the spatial intersection angles contained in the construction deviation dataset, read the pre-configured local resistance mapping table, input the spatial intersection angles into the local resistance mapping table to perform a matching query operation, and output the corresponding local resistance coefficients; The reference flow velocity of the refrigerant fluid is squared to generate a flow velocity square value. The local resistance coefficient, the density of the refrigerant fluid, and the flow velocity square value are multiplied together, and the result of the multiplication is divided by two to calculate the additional pressure loss value. The additional pressure loss value is multiplied by the rated flow rate of the cooling water pump to generate the additional energy consumption assessment value. Read the preset energy-saving red line configuration file, extract the preset energy-saving allowable limit contained in the energy-saving red line configuration file, subtract the preset energy-saving allowable limit from the additional energy consumption assessment value to generate an energy consumption assessment difference, and when the energy consumption assessment difference is greater than zero, generate a pipeline correction and rework instruction and output the pipeline correction and rework instruction to the preset monitoring terminal.

9. The method for detecting construction deviations in chiller room pipelines based on point cloud mapping according to claim 8, characterized in that, The method further includes: When the energy consumption assessment difference is not greater than zero and the spatial vertical distance contained in the construction deviation dataset is greater than zero, the current spatial coordinate information of the actual pipeline center axis is extracted. The current spatial coordinate information is used to replace the corresponding theoretical pipeline spatial feature endpoints in the chiller room pipeline design file, and an as-built updated pipeline model is rendered. Read the preset friction coefficient along the pipeline and measure and extract the actual total physical length of the pipeline from the as-built updated pipeline model; The length-to-diameter ratio is calculated by dividing the actual total physical length of the pipeline by the inner diameter of the pipeline. The length-to-diameter ratio, the friction coefficient, the density of the refrigerant fluid, and the square of the flow velocity are multiplied together, and the result of the multiplication is divided by two to calculate the pressure loss along the pipeline. The friction loss value and the additional pressure loss value are added together to generate the hydraulic balance parameters for the chiller room. Based on the hydraulic balance parameters of the chiller room, a preset pump frequency conversion mapping library is queried to extract the corresponding target operating frequency value. Based on the target operating frequency value, a pump frequency conversion control reference dataset is generated and sent to the chiller room operation control terminal.

10. A point cloud mapping-based system for detecting construction deviations in chiller room pipelines, applied to the point cloud mapping-based method for detecting construction deviations in chiller room pipelines as described in any one of claims 1-9, characterized in that, The system includes: The design analysis module is used to receive the chiller room pipeline design file, extract the theoretical pipeline spatial feature endpoints contained in the chiller room pipeline design file, and generate the theoretical pipeline center axis based on the connection of the theoretical pipeline spatial feature endpoints; The point cloud clipping module is used to control the laser scanning equipment to collect the measured surface point cloud set at the construction site of the chiller room pipeline, and to perform a spatial bounding box clipping operation on the measured surface point cloud set to output the target pipeline point cloud set. The spatial registration module is used to extract the three-dimensional geometric centroid and principal component direction vector of the target pipeline point cloud set, construct a spatial transformation matrix based on the three-dimensional geometric centroid and the principal component direction vector and the theoretical pipeline central axis, and perform a spatial position transformation operation by multiplying the target pipeline point cloud set by the spatial transformation matrix to generate a registered point cloud set. The axis fitting module is used to extract spatially scattered point clusters corresponding to the outer surface of the pipeline in the registration point cloud set, and to generate the actual pipeline center axis based on the spatially scattered point clusters. The deviation calculation module is used to calculate the spatial vertical distance and spatial intersection angle between the actual pipeline center axis and the theoretical pipeline center axis, and to combine the spatial vertical distance and the spatial intersection angle to generate a construction deviation dataset; The over-limit early warning module is used to receive a preset threshold configuration file, extract the distance limit and angle limit contained in the preset threshold configuration file, subtract the distance limit from the spatial vertical distance contained in the construction deviation dataset to generate a distance difference, subtract the angle limit from the spatial intersection angle to generate an angle difference, and generate a deviation over-limit early warning signal and perform signal output operation when the distance difference is greater than zero or the angle difference is greater than zero.

Images