Method, medium and device for constructing mr model of linear engineering

Abstract

The present application relates to the technical field of engineering construction, in particular to a MR model construction method for linear engineering, medium and equipment, the method comprising the following steps: constructing a lightweight BIM model, establishing a seven-parameter Boolean sand model, and aligning key control points in the BIM model with corresponding key control points in the field; based on the obtained final key control point sequence, laying out relay control points, and measuring the three-dimensional coordinates of the relay control points; based on the constructed model, calculating the long-distance offset correction starting point of the point set selection, calculating the MR model coordinates of the control points from the starting point, and calculating the offset of the control points; for the MR model segment whose offset exceeds the threshold, the MR model coordinates of the control points are corrected according to the distance. The present application solves the problems that when the MR model construction method is applied to long and large linear engineering, it is difficult to accurately align the virtual and real coordinate systems in a long distance range, and the long distance model is easy to deviate and distort, and the traditional correction method is low in efficiency and poor in adaptability.

Description

Technical Field

[0001] This invention relates to the field of engineering construction technology, and in particular to a method, medium and equipment for constructing MR models of linear engineering projects. Background Technology

[0002] Long-distance linear engineering projects are generally characterized by long extension distances, complex terrain and landforms along the route, diverse structural types, and dynamic changes in the construction environment. Their construction management and quality monitoring require extremely high precision. In traditional engineering management, tools such as two-dimensional drawings and simple BIM models have problems such as information fragmentation, insufficient visualization, and inaccurate integration of virtual and real data, making it difficult to meet the high-precision requirements in complex scenarios.

[0003] MR (Mixed Reality) technology, as a core technology for virtual-real fusion, can overlay virtual engineering models with real construction scenes in real time, providing intuitive guidance for engineers. However, when existing MR model building methods are applied to long and linear projects, there are problems such as difficulty in accurately aligning the virtual and real coordinate systems over long distances, and the models are prone to distortion over long distances. Traditional correction methods are also inefficient and have poor adaptability. Summary of the Invention

[0004] The main objective of this invention is to provide a method, medium, and device for constructing MR models for linear engineering projects. This invention aims to solve the technical problems of existing MR model construction methods when applied to long linear engineering projects, such as difficulty in accurately aligning virtual and real coordinate systems over long distances, easy distortion of long-distance models, and low efficiency and poor adaptability of traditional correction methods.

[0005] To achieve the above objectives, this invention proposes a method for constructing an MR model for linear engineering, comprising the following steps: S1. Set initial key control points, add key control points and merge neighboring points based on the initial key control points to obtain the sequence of final key control points, and determine the three-dimensional coordinates of the final key control points. S2. Construct a lightweight BIM model. Based on the key control points selected in S1, select multiple key control points to add to the model solution point set and mark them as selected. Establish a seven-parameter Bursa model and align the key control points in the BIM model with the corresponding key control points on site. S3. Based on the final key control point sequence obtained in S1, set up relay control points and determine the three-dimensional coordinates of the relay control points; S4. Based on the model solution point set constructed in S2, select the long-distance offset correction starting point, calculate the MR model coordinates of the control points from the starting point, and calculate the offset of the control points. For MR model segments with offsets exceeding the threshold, correct the MR model coordinates of the control points according to the distance. S5. If there are no subsequent control points, the offset correction ends; otherwise, return to S4 to perform offset correction for subsequent segments.

[0006] A further improvement of the linear engineering MR model construction method of the present invention is that, when setting the initial key control points, an initial spacing value is set. Starting from the beginning of the project, along the linear engineering section every... Set an initial critical control point; the endpoint is insufficient. If the endpoint is used as the last initial key control point, an equidistant initial control network is formed.

[0007] The method for constructing the MR model of linear engineering in this invention is further improved in that, when merging neighboring points, all supplementary key control points are traversed, and if the distance between the supplementary key control points and any initial key control point before or after the supplementary key control points is less than or equal to the distance between the supplementary key control points and the initial key control points before or after the supplementary key control points, the method further improves the method by finding that the distance between the supplementary key control points and any initial key control points before or after the If the distance to the first and last two initial key control points is greater than 3, then delete the added point. When the value is 3, the supplementary point is retained, and the final key control point sequence is obtained after merging.

[0008] The method for constructing MR models of linear engineering in this invention is further improved in that S2 includes the following steps: S201. Construct a lightweight BIM model; S202. Select four key control points and add them to the model solution point set, and mark them as selected. Establish a seven-parameter Bursa model, with the following expression: ; in: The actual coordinates of the key control points. Virtual coordinates of key control points With seven unknown parameters, the virtual and real coordinates of four key control points were substituted into the model to obtain 12 equations. The estimated values ​​of the seven parameters were then calculated using the least squares estimation method.

[0009] A further improvement of the linear engineering MR model construction method of the present invention is that S2 further includes the following step: S203. The accuracy of the seven-parameter Bursa model is verified, and the residual threshold is set as follows: The virtual coordinates of the four key control points are substituted into the following model: ; in: These are predicted values ​​based on actual coordinates. Using virtual coordinates, the predicted values ​​of the actual coordinates are calculated, and the predicted residuals of the four key control points in the model's solution point set are calculated. : ; Wherein: if the maximum value of the prediction residuals of the four key control points is less than the residual threshold If the model generalization ability is positive, proceed to the model generalization ability test; otherwise, it indicates that the model accuracy is insufficient and return to S202. S204. Model Generalization Ability Test: Select any point from the unselected key control points as a redundant check point and mark it as selected; substitute its virtual coordinates into the model, predict its actual coordinates, and calculate the prediction residuals. If predicting residuals Less than the residual threshold If the prediction residual is correct, the virtual-to-real conversion model is completed; otherwise, it indicates that the model's generalization ability is insufficient. The point with the largest prediction residual is removed from the model solution point set, and redundant verification points are added to the model solution point set. Then, return to S202 to recalculate the model parameters.

[0010] The method for constructing MR models for linear engineering in this invention is further improved in that S4 specifically includes the following steps: S401. Based on the model solution point set constructed by S2, select one of the four key control points as the starting point for long-distance offset correction, and perform long-distance offset correction in both the front and rear directions respectively. S402. Starting from the origin, select five consecutive control points, including the origin. If there are fewer than five, select all five. Align the control points with the reflective target using the MR equipment. Align key control points with the high-precision target and relay control points with the medium-precision target. The MR equipment automatically calculates the MR model coordinates of the control points using its built-in algorithm and calculates the alignment error of the five control points. Alignment error The expression is as follows: ; in: These are the MR model coordinates of the control points. These are the actual coordinates of the control points.

[0011] The method for constructing the MR model of linear engineering in this invention is further improved in that S4 includes S403, calculating the offset of the segment containing the five key control points by distance. ,like If the offset is less than the offset threshold, the segment passes the offset detection and proceeds to S5. If greater than or equal to, proceed to S404; offset. The calculation formula is as follows: ; in: It is the first Alignment error of one point, It is the first The distance of each point from the starting point, It is the first Error weights for each point.

[0012] The method for constructing MR models for linear engineering in this invention is further improved in that S4 includes S404: for MR model segments whose offset exceeds a threshold, the MR model coordinates of the five key control points on that segment are corrected according to the distance; the offset correction formula is as follows: ; in: It is the first The MR model coordinates after point correction It is the first For each point, the MR model coordinates before correction are used. If the MR model coordinates of the control points before correction are smaller than the actual coordinates, a plus sign is used in the formula; otherwise, a minus sign is used. It is the first Distance gain coefficient at each point, It is the offset of that model segment. It is a rate control parameter. It is the distance from the control point to the starting point. It is the length of the MR model segment. It is a natural constant.

[0013] The present invention also provides a readable storage medium storing a computer program adapted to be loaded by a processor and executed as described above for the MR model construction method of linear engineering.

[0014] The present invention also provides a computer device, the computer device including a memory and a processor, the memory storing a computer program, which, when executed by the processor, performs the MR model construction method for linear engineering as described above.

[0015] The technical solution of the present invention has the following beneficial effects: 1. Scientific and efficient selection of key control points lays the foundation for high-precision alignment: This invention adopts a three-level selection strategy of "initial control network + feature supplementation + neighboring point merging". This strategy ensures uniform coverage across the entire line through an equidistant initial control network, accurately captures key sections reflecting core engineering features such as abrupt changes in axis curvature and structural connections, and eliminates redundant data by merging neighboring points, avoiding computational redundancy and error interference caused by excessive control point density. Combined with high-precision reflective targets and total station measurements, the obtained three-dimensional coordinates of key control points are highly accurate and reliable, providing a precise and efficient real-world benchmark for subsequent alignment of the virtual and real coordinate systems. This solves the pain point of traditional control point selection, which involves choosing between incomplete coverage and excessive redundancy.

[0016] 2. Controllable accuracy of virtual-real coordinate system transformation, adaptable to stringent engineering requirements: This invention relies on the high-precision transformation characteristics of the seven-parameter Bursa model and ensures transformation accuracy through a "four-point solution + dual verification" mechanism. The solution of four key control points can construct a complete coordinate transformation equation, and the least squares estimation method can effectively reduce the impact of measurement errors and accurately solve for the seven unknown parameters. The dual verification of model accuracy and generalization ability not only ensures the transformation accuracy within the solution point set, but also verifies the model's adaptability to key control points along the entire line through redundant verification points, avoiding misjudgment of accuracy caused by a single verification, and strictly controlling the virtual-real coordinate system alignment residuals within the allowable range of engineering.

[0017] 3. Long-distance offset correction with dynamic intelligence ensures overall model stability: (1) The present invention proposes to deploy relay control points at a density of 1 / 10 of the spacing between key control points, and to build a high-density, full-coverage error monitoring network between key control points. Compared with the traditional sparse monitoring point layout, it can capture the small offset changes of MR models over long distances more accurately and in real time.

[0018] (2) The present invention uses a distance-weighted formula to calculate the offset, which fully considers the weight difference of the distance between different control points and the correction starting point on the offset, and avoids the offset judgment deviation caused by traditional equalization calculation.

[0019] (3) The present invention adopts a correction strategy of “segmented processing + dynamic iteration”, taking five consecutive control points as a correction unit, avoiding the overall distortion of the model caused by one-time correction of the entire line; by adjusting the correction rate through the distance gain coefficient, and dynamically adjusting the correction amplitude by combining the distance of the control point from the starting point and the length of the model segment, a flexible correction of “fine adjustment near the starting point and accurate correction far from the starting point” is achieved, which not only ensures the effectiveness of the offset correction, but also maintains the smooth connection between the segments of the model.

[0020] (4) This invention uses a closed-loop process of “starting point setting - error calculation - offset judgment - correction execution - subsequent segment iteration” to gradually advance from the correction starting point in both directions. After each correction is completed, the farthest control point of that segment is used as the new starting point to continue the correction, ensuring that all segments of the entire line have been accurately verified and corrected. Ultimately, the MR model and the real scene are highly matched over a long distance. The offset error is controlled within the allowable threshold of the project throughout the process, ensuring the reliability of the model in scenarios such as construction layout and quality inspection. Detailed Implementation

[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0022] In this invention, the use of terms such as "first," "second," etc., is for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this invention, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0023] In this invention, unless otherwise explicitly specified and limited, the terms "connection," "fixed," etc., should be interpreted broadly. For example, "fixed" can mean a fixed connection, a detachable connection, or an integral part; it can mean a mechanical connection or an electrical connection; it can mean a direct connection or an indirect connection through an intermediate medium; it can mean the internal communication of two components or the interaction between two components, unless otherwise explicitly limited. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0024] Furthermore, the technical solutions of the various embodiments of the present invention can be combined with each other, but only if they are feasible for those skilled in the art. If the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such combination of technical solutions does not exist and is not within the scope of protection claimed by the present invention.

[0025] This invention proposes a method for constructing MR models for linear engineering, comprising the following steps: I. Alignment of Real and Virtual Coordinate Systems: 1. Selection of key control points: S1. Set initial key control points, add key control points and merge neighboring points based on the initial key control points to obtain the sequence of final key control points, and determine the three-dimensional coordinates of the final key control points. S101, Initial Control Network: Set initial spacing values Starting from the beginning of the project, along the linear engineering section every... Set an initial critical control point; the endpoint is insufficient. If the endpoint is used as the last initial key control point, an equidistant initial control network is formed.

[0026] S102. Supplementing Key Control Points: Between every two adjacent initial key control points in the initial control network, if there is an important location reflecting changes in alignment and structural characteristics, a key control point shall be added at that location. If multiple locations exist, the most important one shall be selected based on engineering experience. If no key control point exists, no key control point shall be added. Important locations reflecting changes in alignment and structural characteristics refer to cross-sections where the alignment axis or structure type undergoes significant changes, such as points of abrupt changes in horizontal and vertical curvature of the axis, bridge-tunnel-roadbed, dam-culvert-channel, pipe-gallery-well, and other connecting cross-sections.

[0027] S103. Merge neighboring points: Traverse all supplementary key control points, and if the distance between a point and an initial key control point before or after it is less than or equal to... If the distance to the first and last two initial key control points is greater than 3, then delete the added point. When the value is 3, the supplementary point is retained, and the final key control point sequence is obtained after merging.

[0028] S104. Determine the three-dimensional coordinates of key control points: Set up high-precision reflective targets at each key control point, and use a total station to measure the three-dimensional coordinates to obtain the actual three-dimensional coordinates, which will serve as the real reference for subsequent alignment of the virtual and real coordinate systems.

[0029] 2. Construct a virtual-real coordinate system transformation model: S2. Construct a lightweight BIM model. Based on the key control points selected in S1, select multiple key control points to add to the model solution point set and mark them as selected. Establish a seven-parameter Bursa model and align the key control points in the BIM model with the corresponding key control points on site. S201. Construct a lightweight BIM model: Construct a BIM model for long linear engineering projects, simplify the BIM model, retain only the core information (including the main mechanical component models of the engineering structure and the necessary skeleton of the BIM model), and delete redundant components (referring to non-main mechanical components of the engineering structure and decorative components of the BIM model). S202. Calculate the parameters of the virtual-to-real conversion model: Based on the selected key control points, randomly select four key control points to add to the model solution point set and mark them as selected. Establish a seven-parameter Bursa model, aligning the key control points in the BIM model with their corresponding key control points on site. The seven-parameter Bursa model is a method for achieving high-precision conversion between different coordinate systems; the model contains seven unknown parameters. The expression for the seven-parameter Bursa model is as follows: ; in: The actual coordinates of the key control points. Virtual coordinates of key control points With seven unknown parameters, the virtual and real coordinates of four key control points were substituted into the model to obtain 12 equations. The estimated values ​​of the seven parameters were then calculated using the least squares estimation method.

[0030] S203. Verify the accuracy of the seven-parameter Bursa model: Set the residual threshold to... The virtual coordinates of the four key control points are substituted into the following model: ; in: These are predicted values ​​based on actual coordinates. Using virtual coordinates, the predicted values ​​of the actual coordinates are calculated, and the predicted residuals of the four key control points in the model's solution point set are calculated. : ; Wherein: if the maximum value of the prediction residuals of the four key control points is less than the residual threshold If the model generalization ability is positive, proceed to the model generalization ability test; otherwise, it indicates that the model accuracy is insufficient and return to S202. S204. Model Generalization Ability Test: Select any point from the unselected key control points as a redundant check point and mark it as selected; substitute its virtual coordinates into the model, predict its actual coordinates, and calculate the prediction residuals. If predicting residuals Less than the residual threshold If the prediction residual is correct, the virtual-to-real conversion model is completed; otherwise, it indicates that the model's generalization ability is insufficient. The point with the largest prediction residual is removed from the model solution point set, and redundant verification points are added to the model solution point set. Then, return to S202 to recalculate the model parameters.

[0031] II. Long-distance offset correction: 3. Relay control point deployment: S3. Based on the final key control point sequence obtained in S1, set up relay control points and determine the three-dimensional coordinates of the relay control points; S301. Deploy relay control points: Based on the final critical control point sequence obtained in S1, evenly deploy relay control points between every two critical control points, and set the relay control point spacing value. ; S302. Determine the three-dimensional coordinates of relay control points: Set up medium-precision reflective targets at each relay control point, and use a total station to measure the three-dimensional coordinates to obtain the actual three-dimensional coordinates, which serve as the actual benchmark for long-distance offset correction.

[0032] 4. Dynamic long-distance offset correction: S4. Based on the model solution point set constructed in S2, select the long-distance offset correction starting point, calculate the MR model coordinates of the control points from the starting point, and calculate the offset of the control points. For MR model segments with offsets exceeding the threshold, correct the MR model coordinates of the control points according to the distance. S401. Select the starting point for long-distance offset correction: Based on the model solution point set constructed in S2, select any one of the four key control points as the starting point for long-distance offset correction, and perform long-distance offset correction in both the front and rear directions respectively. S402. Alignment Error Calculation: Starting from the starting point, take five consecutive control points, including the starting point; if there are fewer than five, take all five. Align the control points with the reflective target using the MR equipment. Critical control points are aligned with high-precision targets, and relay control points are aligned with medium-precision targets. The MR equipment automatically calculates the MR model coordinates of the control points using its built-in algorithm, and calculates the alignment error of the five control points. Alignment error The expression is as follows: ; in: These are the MR model coordinates of the control points. These are the actual coordinates of the control points.

[0033] S403, Offset Calculation: Calculate the offset of the segment containing the five key control points based on distance. ,like If the offset is less than the offset threshold, the segment passes the offset detection and proceeds to S5. If greater than or equal to, proceed to S404; offset. The calculation formula is as follows: ; in: It is the first Alignment error of one point, It is the first The distance of each point from the starting point, It is the first Error weights for each point.

[0034] S404, Offset Correction: For MR model segments with offsets exceeding the threshold, the MR model coordinates of the five key control points on that segment are corrected based on the distance (which is the mileage value of the linear engineering project); the offset correction formula is as follows: ; in: It is the first The MR model coordinates after point correction It is the first For each point, the MR model coordinates before correction are used. If the MR model coordinates of the control points before correction are smaller than the actual coordinates, a plus sign is used in the formula; otherwise, a minus sign is used. It is the first Distance gain coefficient at each point, It is the offset of that model segment. It is a rate control parameter. It is the distance from the control point to the starting point. It is the length of the MR model segment. It is a natural constant.

[0035] S5. Subsequent Segment Offset Detection and Correction: If there are no subsequent control points, the offset correction ends; otherwise, it will return to S4 to perform offset correction for subsequent segments.

[0036] The present invention also provides a readable storage medium storing a computer program adapted to be loaded by a processor and executed as described above for the MR model construction method of linear engineering.

[0037] The present invention also provides a computer device, the computer device including a memory and a processor, the memory storing a computer program, which, when executed by the processor, performs the MR model construction method for linear engineering as described above.

[0038] The following describes in detail the specific implementation of the present invention in conjunction with the first phase of the reinforcement project of a key embankment in a certain lake.

[0039] The first phase of the dike reinforcement project for a key embankment of a certain lake is a typical long-distance linear water conservancy project. The total length of the dike exceeds 100km, encompassing various complex terrains including plains, depressions, and river-lake confluence zones. It includes over 30 culverts and 8 pumping stations, and features numerous abrupt changes in horizontal and vertical curvature along the axis, as well as distinctive cross-sections such as dike-culvert and dike-pump connections. This places extremely high demands on the high-precision fusion and long-distance stability of the MR model. This embodiment employs the method of this invention to construct an MR model for scenarios such as engineering construction layout, structure installation accuracy detection, and dike construction quality monitoring.

[0040] I. Alignment of Real and Virtual Coordinate Systems: 1. Selection of key control points: S101. Initial Control Network Layout: Based on the project length, terrain complexity, and accuracy requirements, determine the initial spacing between key control points. =500m. Starting from the starting point of the project, initial key control points are set up every 500m along the dike axis. When the distance between the end point and the previous initial key control point is less than 500m, the end point is set as the last initial key control point, forming an equidistant initial control network covering the entire line; S102. Supplementing key control points: Check the sections between adjacent initial key control points one by one, accurately identify key sections that reflect the core characteristics of the project, such as abrupt changes in the curvature of the axis horizontal and vertical planes, the connection between the dike and the culvert / pump station, and changes in the type of protective structure. Supplement key control points at these locations. When there are multiple similar locations in the same section, select the most representative core section as the supplementary point based on engineering experience. S103. Merge neighboring points: Traverse and check all supplementary points. If the distance between a supplementary point and any initial key control point before or after it is less than or equal to... If the distance to the first critical control point is 167m, then the redundant supplementary point is deleted; if the distance to the first and last two initial critical control points is greater than 167m, then it is retained, thus forming a concise critical control point sequence with complete key information. S104. Determine the three-dimensional coordinates: Set up high-precision reflective targets at each final key control point, and use a total station to measure the three-dimensional coordinates. After data adjustment, obtain the accurate real three-dimensional coordinates of each point, which serve as the core benchmark for aligning the virtual and real coordinate systems.

[0041] 2. Construct a virtual-real coordinate system transformation model: S201. Construct a lightweight BIM model: Based on detailed engineering design drawings, construct a model using professional BIM software, covering core structures such as embankments, culverts, and pumping stations. Optimize the model for lightweighting, removing redundant information such as temporary facilities and decorative components, retaining only key dimensions, axis positions, and structural outlines to ensure smooth model loading and operation. S202. Calculate the transformation model parameters: Select four points from the final key control point sequence as the model solution point set and mark them. Establish a seven-parameter Bursa model. Substitute the BIM virtual coordinates and the measured real three-dimensional coordinates of these four points into the model. Solve for the seven unknown parameters (3 translation parameters, 3 rotation parameters, and 1 scale parameter) using the least squares estimation method to achieve the initial association between the virtual and real coordinate systems. S203. Model Accuracy Verification: Based on the engineering construction accuracy requirements, a residual threshold of 3mm is set. Substitute the virtual coordinates of the four solution points into the model after preliminary solution, and calculate the residual between the predicted and measured values ​​of the actual coordinates of each point. If the maximum residual of all points is less than 3mm, the model accuracy meets the standard; if it does not meet the standard, return to select a new set of solution points and calculate the model parameters. S204. Model generalization ability test: Select one key control point from the unincluded solution point set as a redundant check point and mark it. Substitute its virtual coordinates into the model to calculate the predicted real coordinate value. If the prediction residual is less than 3mm, it indicates that the model has good generalization ability and the virtual-to-real conversion model construction is completed. If the residual does not meet the standard, remove the point with the largest residual in the original solution point set, include the redundant check point in the solution point set, and return to recalculate the model parameters until the requirements are met.

[0042] II. Long-distance offset correction: 3. Relay control point deployment: S301. Deploy relay control points: Based on the final critical control point sequence, evenly deploy relay control points between every two adjacent critical control points, and set the spacing between relay control points. =50m, forming a high-density auxiliary monitoring network; S302. Measurement of three-dimensional coordinates: Medium-precision reflective targets are set up at each relay control point, and a total station is used to measure the three-dimensional coordinates to obtain the actual three-dimensional coordinates of each point, which serve as an auxiliary reference for long-distance offset correction.

[0043] 4. Dynamic long-distance offset correction: S401. Select the correction starting point: Select one of the four solution points from the virtual-to-real conversion model as the starting point for long-distance offset correction. Based on this, carry out correction work step by step in both directions before and after the dike.

[0044] S402. Alignment Error Calculation: Starting from the starting point, select 5 consecutive control points (including key control points and relay control points) including the starting point. Use an MR device (in this embodiment, a Philips MR device (model: Ingenia Eliton X)) to align the reflective targets of each control point (key control points correspond to high-precision targets, and relay control points correspond to medium-precision targets). The device's built-in algorithm automatically calculates the MR model coordinates of each control point. Combined with the actual 3D coordinates, calculate the alignment error of each control point.

[0045] S403, Offset Calculation: Set the offset threshold to 8mm, and calculate the overall offset of the segment using a distance-weighted formula (fully considering the alignment error of each control point and the distance weight from the starting point). If the offset is less than 8mm, the segment passes the detection and proceeds directly to subsequent segment correction; if the offset is ≥8mm, targeted correction is required.

[0046] S404, Offset Correction: For model segments with excessive offset, adjust the MR model coordinates of the five control points of that segment according to the distance-related correction formula. The formula sign is determined based on the direction of deviation between the model coordinates before correction and the actual coordinates. The correction rate is adjusted using a distance gain coefficient (based on engineering experience). The correction magnitude for each control point is dynamically adjusted based on the distance from the starting point and the total length of the segment, ensuring that the corrected model accurately matches the actual scene and avoiding over-correction that could lead to model distortion.

[0047] S5. Correction of Subsequent Segments: If the current segment is the end of the project and there are no subsequent control points, the correction work is completed; if there are still subsequent segments, the control point farthest from the starting point in the current segment is taken as the correction starting point for the next segment, and the alignment error calculation step is returned to proceed in sequence until the offset correction of all segments of the entire line is completed.

[0048] The MR model constructed using the method described in this embodiment achieves high-precision, long-distance, and stable fusion of the virtual model with the real-world scene. The offset error is controlled within 8mm over long distances, meeting the high-precision requirements of engineering construction layout and quality inspection scenarios, effectively improving engineering construction efficiency and management level. It is applicable to long-distance linear projects such as dikes, railways, highways, and pipelines, especially for water conservancy dike projects with complex terrain, diverse structures, and high spatial positioning accuracy requirements. It can achieve high-precision MR model construction, providing technical support for engineering construction, quality monitoring, and progress management.

[0049] The above description is only a preferred embodiment of the present invention and does not limit the scope of the present invention. All equivalent structural transformations made using the present invention under the inventive concept of the present invention, or direct / indirect applications in other related technical fields, are included within the protection scope of the present invention.

Claims

1. A method for constructing an MR model for linear engineering, characterized in that, Includes the following steps: S1. Set initial key control points, add key control points and merge neighboring points based on the initial key control points to obtain the sequence of final key control points, and determine the three-dimensional coordinates of the final key control points. S101, Initial Control Network: Set initial spacing values Starting from the beginning of the project, along the linear engineering section every... Set an initial critical control point; the endpoint is insufficient. When the endpoint is reached, the endpoint is taken as the last initial key control point, forming an equidistant initial control network; S102. Supplementing key control points: If there is a position between every two adjacent initial key control points in the initial control network that reflects changes in the alignment and structural characteristics, then a key control point shall be added at that position. If there are multiple key control points, then one shall be selected. If there are no key control points, then no key control point shall be added. S103. Merge neighboring points: Traverse all supplementary key control points, and if the distance between a point and an initial key control point before or after it is less than or equal to... If the distance to the first and last two initial key control points is greater than 3, then delete the added point. When the value is 3, the supplementary point is retained, and the merging is completed to obtain the final key control point sequence. S2. Construct a lightweight BIM model. Based on the key control points selected in S1, randomly select four key control points to add to the model solution point set and mark them as selected. Establish a seven-parameter Bursa model and align the key control points in the BIM model with the corresponding key control points on site. S3. Based on the final key control point sequence obtained in S1, set up relay control points and determine the three-dimensional coordinates of the relay control points; S4. Based on the model solution point set constructed in S2, select the long-distance offset correction starting point, calculate the MR model coordinates of the control points from the starting point, and calculate the offset of the control points. For MR model segments with offsets exceeding the threshold, correct the MR model coordinates of the control points according to the distance; specifically including the following steps: S401. Based on the model solution point set constructed by S2, select one of the four key control points as the starting point for long-distance offset correction, and perform long-distance offset correction in both the front and rear directions respectively. S402. Starting from the origin, select five consecutive control points, including the origin. If there are fewer than five, select all five. Align the control points with the reflective target using the MR equipment. Align key control points with the high-precision target and relay control points with the medium-precision target. The MR equipment automatically calculates the MR model coordinates of the control points using its built-in algorithm and calculates the alignment error of the five control points. Alignment error The expression is as follows: ； in: These are the MR model coordinates of the control points. The actual coordinates of the control points; S403. Calculate the offset of the segment containing the five key control points by distance. ,like If the offset is less than the offset threshold, the segment passes the offset detection and proceeds to S5. If greater than or equal to, proceed to S404; offset. The calculation formula is as follows: ； in: It is the first Alignment error of one point, It is the first The distance of each point from the starting point, It is the first Error weights for each point; S404. For MR model segments whose offset exceeds the threshold, correct the MR model coordinates of the five key control points on that segment based on the distance; the offset correction formula is as follows: ； in: It is the first The MR model coordinates after point correction It is the first For each point, the MR model coordinates before correction are used. If the MR model coordinates of the control points before correction are smaller than the actual coordinates, a plus sign is used in the formula; otherwise, a minus sign is used. It is the first Distance gain coefficient at each point, It is the offset of that model segment. It is a rate control parameter. It is the distance from the control point to the starting point. It is the length of the MR model segment. It is a natural constant; S5. If there are no subsequent control points, the offset correction ends; otherwise, return to S4 to perform offset correction for subsequent segments.

2. The method for constructing an MR model for linear engineering according to claim 1, characterized in that, S2 includes the following steps: S201. Construct a lightweight BIM model; S202. Select four key control points and add them to the model solution point set, and mark them as selected. Establish a seven-parameter Bursa model, with the following expression: ； in: The actual coordinates of the key control points. Virtual coordinates of key control points With seven unknown parameters, the virtual and real coordinates of four key control points were substituted into the model to obtain 12 equations. The estimated values ​​of the seven parameters were then calculated using the least squares estimation method.

3. The method for constructing an MR model of a linear engineering project according to claim 2, characterized in that, S2 also includes the following steps: S203. The accuracy of the seven-parameter Bursa model is verified, and the residual threshold is set as follows: The virtual coordinates of the four key control points are substituted into the following model: ； in: These are predicted values ​​based on actual coordinates. Using virtual coordinates, the predicted values ​​of the actual coordinates are calculated, and the predicted residuals of the four key control points in the model's solution point set are calculated. : ； Wherein: if the maximum value of the prediction residuals of the four key control points is less than the residual threshold If the model generalization ability is positive, proceed to the model generalization ability test; otherwise, it indicates that the model accuracy is insufficient and return to S202. S204. Model Generalization Ability Test: Select any point from the unselected key control points as a redundant check point and mark it as selected; substitute its virtual coordinates into the model, predict its actual coordinates, and calculate the prediction residuals. If predicting residuals Less than the residual threshold If the prediction residual is correct, the virtual-to-real conversion model is completed; otherwise, it indicates that the model's generalization ability is insufficient. The point with the largest prediction residual is removed from the model solution point set, and redundant verification points are added to the model solution point set. Then, return to S202 to recalculate the model parameters.

4. A readable storage medium, characterized in that, The readable storage medium stores a computer program that is adapted to be loaded by a processor and executed as the MR model construction method for linear engineering according to any one of claims 1-3.

5. A computer device, characterized in that, The computer device includes a memory and a processor. The memory stores a computer program, which, when executed by the processor, performs the MR model construction method for linear engineering as described in any one of claims 1-3.

Images