BIM-based hydroelectric engineering surveying control network layout optimization method and system

Abstract

The application relates to the technical field of building construction, in particular to a water and electricity engineering measurement control network layout optimization method and system based on BIM. The method comprises the following steps: constructing an initial candidate control network coordinate set; based on the initial candidate control network coordinate set and a BIM model, constructing a simulated total station sight line set, and removing the sight lines blocked by the entity grid; for each reserved simulated total station sight line, extracting its spatial geometric interference feature, converting the spatial geometric interference feature into a sight line connectivity weighting value by using a preset experience penalty coefficient, constructing a control network topology connectivity weighting matrix, and extracting an extreme risk cost representing the control network fracture risk; taking the minimization of the extreme risk cost as a target, driving an optimization algorithm to iteratively optimize the candidate control network coordinate set, and forcibly adsorbing the optimized coordinates to the BIM entity surface to output an optimal field control point coordinate set. Therefore, the matching degree of the digital network layout scheme and the physical construction environment is improved.

Description

Technical Field

[0001] This application relates to the field of building construction technology, specifically to a method and system for optimizing the layout of measurement and control networks for hydropower projects based on BIM. Background Technology

[0002] In hydropower engineering construction (such as underground mega-cave complexes and dam canyons), the survey control network is responsible for accurately transmitting absolute coordinate benchmarks to each construction work face. The calculation of elevation and plane coordinates in the field survey is highly dependent on the intersection of the total station's line of sight between control points.

[0003] Currently, BIM (Building Information Modeling) technology is widely used to assist in the pre-layout of control networks for field surveying. Existing 3D meshing software mainly determines line-of-sight by calculating whether there is a spatial intersection between the simulated survey line and the BIM structure mesh surface. However, hydropower construction sites have extreme spatial limitations and complex environmental interferences, and simple straight-line penetration analysis deviates from the physical error mechanisms of field surveying. In narrow caverns, the total station aiming at control points with large elevation differences will produce a large vertical elevation angle of the telescope, which will significantly amplify the mechanical clearance error of the instrument's vertical axis; at the same time, when the infrared ranging beam is extremely close to the rough concrete structure surface, it is easily affected by thermal refraction interference caused by temperature gradients. Existing meshing methods cannot quantify the above-mentioned spatial entity approximation phenomena into error constraints.

[0004] Furthermore, existing technologies lack a macroscopic assessment of the overall connectivity strength of the measurement network topology. The calculation of the absolute coordinate system of the control network relies on the redundant intersection of multiple lines of sight. In actual construction, the main lines of sight connecting different station clusters are easily blocked by large equipment or scaffolding temporarily parked on site. Once the main lines of sight are cut off, and the subsequent algorithms fail to identify the irreplaceable nature of these lines of sight in the coordinate transfer link in advance, it will directly lead to the paralysis of adjustment calculation tasks in large areas. Summary of the Invention

[0005] To address the aforementioned technical problems, the purpose of this application is to provide a BIM-based method and system for optimizing the layout of surveying and control networks in hydropower engineering. The specific technical solution adopted is as follows: In a first aspect, embodiments of this application provide a BIM-based method for optimizing the layout of a hydropower engineering survey control network, the method comprising the following steps: Obtain the BIM model of the hydropower project and the coordinates of the known starting control points to construct an initial candidate control network coordinate set; Based on the initial candidate control network coordinate set and BIM model, a simulated total station line-of-sight set is constructed, and lines of sight blocked by the physical grid are removed. For each retained simulated total station line of sight, its spatial geometric interference features are extracted, and the spatial geometric interference features are converted into line-of-sight connectivity weights using a preset empirical penalty coefficient. A topological connectivity weighting matrix for the control network is constructed based on the weighting values ​​of all line-of-sight connectivity. The matrix is ​​then subjected to eigenvalue decomposition to extract the extreme risk cost that characterizes the risk of control network failure. With the goal of minimizing extreme value risk costs, the optimization algorithm iteratively optimizes the candidate control network coordinate set and forcibly attaches the optimized coordinates to the BIM entity surface until the termination condition is met, and outputs the optimal field control point coordinate set.

[0006] In one embodiment, constructing the simulated total station line-of-sight set includes: The control points in the initial candidate control network coordinate set are combined in pairs, and the metric distance between each pair is calculated; and the nominal range parameters of the total station are obtained. The lines connecting the combinations whose measured distance is less than or equal to the nominal range parameter are used as effective simulated lines of sight to construct the simulated total station line of sight set.

[0007] In one embodiment, the removal of the line of sight obstructed by the solid mesh includes: Calculate the shortest distance between the line of sight of the simulated total station and the surface of the BIM entity grid, and deduct the preset reserved thickness of the construction scaffold to obtain the effective net distance from the line of sight to the surface of the structure. If the effective net distance is less than or equal to zero, the line of sight is discarded.

[0008] In one embodiment, the step of converting spatial geometric interference features into line-of-sight connectivity weights includes: Extract the elevation difference and horizontal projection distance between the two ends of each simulated total station line of sight, and calculate the telescope's vertical pitch angle; Based on the vertical pitch angle and the preset angle penalty coefficient, the angle penalty term is calculated; Based on the effective net distance and the preset refraction penalty coefficient, the refraction penalty term is calculated; The line-of-sight connectivity weighting value is negatively correlated with both the angle penalty term and the refraction penalty term.

[0009] In one embodiment, constructing the control network topology connectivity weighting matrix includes: Construct an N×N control network topology connectivity weighting matrix, where N is the number of control points; the matrix is ​​a Laplace matrix, the diagonal elements of the matrix are the sum of the line-of-sight connectivity weights of all connecting edges connected to the corresponding control points, and the off-diagonal elements are the negative values ​​of the line-of-sight connectivity weights of the connecting edges between two corresponding control points.

[0010] In one embodiment, the extreme risk cost characterizing the risk of control network failure is extracted, including: After the matrix completes eigenvalue decomposition, the second smallest non-zero eigenvalue and its corresponding eigenvector are extracted; for any line of sight, the absolute value of the difference between the corresponding components of the two endpoints of the line of sight in the eigenvector is calculated, which is used as the breakage risk value of the line of sight. The extreme risk cost is determined based on the numerical distribution of the breakage risk value for each retained line of sight.

[0011] In one embodiment, the maximum value among the breakage risk values ​​of all retained single lines of sight is determined as the extreme risk cost; if the second smallest feature value is equal to zero, the control network is determined to be disconnected, and a preset maximum penalty value is assigned.

[0012] In one embodiment, the step of forcibly snapping the optimized coordinates to the surface of the BIM entity includes: Using the coordinates of the undetermined measurement control point after translation during the optimization iteration as the emission source, a simulated probe ray is emitted along the preset projection direction of the nearest solid mesh surface in the BIM model; the coordinates of the collision intersection point between the simulated probe ray and the solid mesh surface of the BIM model are used as the coordinates of the undetermined measurement control point after adsorption.

[0013] In one embodiment, the optimal set of field control point coordinates is sent to the field terminal, which controls the total station to calculate the horizontal azimuth and zenith distance based on the coordinates, drives the servo motor to rotate and emit a visible laser spot to indicate the position of the measuring point on the solid surface.

[0014] Secondly, embodiments of this application also provide a BIM-based hydropower engineering survey control network layout optimization system, including a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor executes the computer program to implement the steps of any of the methods described above.

[0015] This application has at least the following beneficial effects: This application first transforms spatial geometric interference features into line-of-sight connectivity weights using a preset empirical penalty coefficient. This converts high-elevation-angle mechanical errors and wall-attached infrared thermal refraction interference in surveying fieldwork into calculable quantifiable weights, guiding the subsequent optimization process to proactively avoid extreme observation lines of sight and reduce errors in field data acquisition. Second, by constructing a control network topology connectivity weight matrix for feature decomposition and extracting extreme value risk costs, the macroscopic intersection intensity assessment of the survey network is transformed into a vulnerability algebraic calculation of the map matrix. Utilizing extreme value risk costs to drive the optimization algorithm can eliminate over-reliance on a single main line of sight. The network scheme enhances the topological robustness of the macroscopic structure of the measurement network, eliminates the potential for the entire control network to be paralyzed due to obstructed views of the main trunk lines at the construction site, and finally, by forcibly attaching the optimized coordinates to the surface of the BIM entity, spatial constraints oriented towards the physical entity are established, avoiding the problem of coordinate points being suspended or embedded in walls due to pure digital spatial optimization. This ensures that each generated control point has a physical bearing surface for placing reflective instruments, forming a network layout scheme that meets the closed-loop requirements of field operations, improving the matching degree between the digital network layout scheme and the physical construction environment, and enhancing the on-site engineering feasibility of the control points. Attached Figure Description

[0016] Figure 1 A flowchart illustrating the steps of a BIM-based hydropower engineering survey control network layout optimization method provided in one embodiment of this application. Detailed Implementation

[0017] The following, in conjunction with the accompanying drawings, details the specific scheme of the BIM-based hydropower engineering survey control network layout optimization method and system provided in this application.

[0018] Please see Figure 1 The document illustrates a flowchart of a BIM-based hydropower engineering survey control network layout optimization method according to an embodiment of this application. The method includes the following steps: S1: Obtain the BIM model of the hydropower project and the coordinates of the known starting control points to construct the initial candidate control network coordinate set.

[0019] First, this embodiment receives 3D construction drawings and hardware parameter files for a hydropower project from an external source, performs a BIM model import operation, and loads the 3D Building Information Model (BIM) file of the hydropower project site. This 3D BIM file consists of a large number of spatial triangular mesh faces, used to represent the 3D boundaries of concrete structures such as the walls, arches, and surfaces of underground caverns and dams.

[0020] Next, known control points are introduced as the starting reference for the network calculation. To ensure that the optimized measurement control network is consistent with the national large-scale absolute coordinate system and to avoid generating a floating, unrelated network with no practical engineering significance, the coordinates of at least two known starting control points, pre-calibrated by conventional geodetic surveying, need to be loaded. The positions of the known starting control points remain fixed in all subsequent algorithm iterations, serving only as anchor points for network connections.

[0021] Subsequently, environmental empirical parameters affecting line-of-sight and spatial interference are read and preset. The specific steps are as follows: Read the reserved thickness of the construction scaffolding as specified in the on-site survey and construction procedures. Since BIM models typically only contain permanent concrete structures, while in actual construction, working platforms or formwork systems are often erected on the side walls of the tunnel. To allow for the spatial extension of these temporary barriers in the purely digital space, this thickness parameter needs to be set. In this embodiment, the reserved thickness of the construction scaffolding is... The value is 0.8 meters.

[0022] Extract the nominal effective measurement range as specified in the total station hardware manual. This numerical value characterizes the farthest linear distance in space from which the instrument can effectively receive the returned signal after emitting infrared laser light. In this embodiment, this nominal effective range... The value is 1000 meters.

[0023] In order to transform the physical interference phenomenon into weights calculated using pure mathematical graph theory in subsequent steps, this embodiment will supplement the loading of two preset constants, namely the refraction penalty coefficient. With dimensionless angle penalty coefficient Dimensionless angular penalty coefficient The value range is 5~15, used to quantify the magnification factor of the total station's vertical axis mechanical error as a function of elevation angle. In this embodiment... The higher the accuracy required for the surveying and mapping, the larger the value should be. Refraction penalty coefficient. The unit is meters, and the value ranges from 0.1 to 0.5. It is used to characterize the sensitivity to thermal refraction interference near the wall. The larger the environmental temperature gradient, the larger the value. In this embodiment... rice.

[0024] After statically presetting the above parameters, this embodiment initiates the particle swarm optimization algorithm. Upon initiation, the algorithm randomly generates an initial set of 3D point clouds within the idle construction area of ​​the 3D building information model (i.e., the space not occupied by solid meshes). This initially generated 3D coordinate combination scheme is defined as the first... Coordinate sets of candidate control networks This coordinate set It contains a quantity of Undetermined measurement control points , where index number The range is to This coordinate set It will be passed as an input object to the next processing step.

[0025] S2, construct the total station simulated line of sight between the undetermined measuring point and the known measuring point.

[0026] Specifically, the current number Coordinate sets of candidate control networks All included One undetermined measurement control point This is combined with at least two pre-loaded known starting control points to form a complete set of coordinate points.

[0027] Next, this embodiment performs basic line construction based on the maximum distance limit. Using a double-loop algorithm, any two control points from the complete set of coordinate points are extracted sequentially, including the points between undetermined points and between an undetermined point and a known point. After extracting the three-dimensional coordinates of the two control points, the Pythagorean theorem is used to calculate the Euclidean distance between these two coordinate points in three-dimensional space. After calculating the spatial Euclidean distance, it is compared with a preset nominal effective range. Compare the numerical values.

[0028] If the Euclidean distance between lines in space is greater than or equal to This indicates that the distance between the two control points exceeds the hardware transmission and reception limits of the total station, making it impossible to generate valid measurement readings in the field. In this case, connecting the two control points is rejected.

[0029] If the Euclidean distance between straight lines in space is less than The two control points are determined to be within the hardware capabilities of the distance measuring instrument. A directed line segment connecting the two control points is generated in three-dimensional digital space, representing the effective simulated line of sight. This line segment represents the infrared distance measuring optical path emitted by the total station. The generated single line segment is defined as the first... Simulated total station line of sight ,in, Add a sequence number to the generated lines, and combine all the simulated total station lines of sight into a simulated total station line of sight set.

[0030] Subsequently, a physical obstacle occlusion penetration test was performed. In a 3D environment, simply having a distance less than the range does not guarantee actual line-of-sight, as the line of sight may penetrate opaque concrete walls or rock formations. Therefore, for each simulated total station line of sight generated above... The algorithm calls a 3D ray collision detection algorithm to calculate the shortest vertical spatial distance between the line segment and the surface of the 3D building information model's solid mesh.

[0031] After calculating the shortest vertical spatial distance, subtract the pre-set thickness of the construction scaffolding from this distance value. An initial clearance difference value is obtained. It is then determined whether this initial clearance difference value is less than or equal to zero. If the initial clearance difference value is less than or equal to zero, it indicates that the simulated light path is extremely close to the structure surface, its distance insufficient to avoid the 0.8-meter temporary scaffolding obstruction area, or that the line of sight itself has penetrated the BIM entity mesh. In this case, the line of sight is determined to be physically blocked in the actual site and is absolutely impassable. Therefore, a connection culling operation will be directly triggered, removing the first... Simulated total station line of sight Completely remove it from the current set of lines of sight in the mesh.

[0032] By performing the deletion operation, the illegal connection will no longer participate in the construction of the subsequent graph Laplacian matrix, objectively ensuring the authenticity of the matrix topology. If the initial clearance difference is greater than zero, the line of sight is determined to be unobstructed. This value is retained, and the final value after retention is defined as the first value. Simulated total station line of sight Effective net distance from line of sight to the surface of the structure The unit is meters.

[0033] After completing pairwise comparisons and eliminations of all coordinate points in the complete set, for the current point... Coordinate sets of candidate control networks The final output is a set of lines of sight that are confirmed to be unobstructed and have the potential for effective distance measurement.

[0034] S3, extract the spatial geometric interference features of the effective line of sight.

[0035] In on-site surveying of hydropower projects, besides being directly blocked by scaffolding, total stations also face two spatial physical interference phenomena that easily introduce errors. Firstly, when the instrument is aimed at a control point with a large elevation difference, the telescope needs to be rotated upwards or downwards at a large angle, which magnifies the instrument's built-in vertical axis mechanical clearance error. Secondly, the previously calculated effective net distance from the line of sight to the surface of the structure... If this distance is too small, it means that the infrared beam is extremely close to the rough concrete sidewall, and under the influence of the uneven temperature gradient inside the cavity, severe thermal refraction and bending will occur. Therefore, this embodiment needs to extract the vertical pitch angle to quantify the aforementioned mechanical vertical axis error amplification effect.

[0036] Specifically: Regarding the reserved first Simulated total station line of sight Extract the 3D coordinates of the control points at the starting and ending points of the line segment. After extracting the coordinate data from both ends, calculate the coordinate difference between the starting and ending points along the Z-axis (i.e., the vertical direction), and take the absolute value of this difference to obtain the absolute value of the elevation coordinate difference. Simultaneously, the coordinate difference between the starting point and the ending point on the XY plane (i.e., the absolute horizontal plane) is calculated, and the projected distance on the horizontal plane is calculated using the Pythagorean theorem. .

[0037] Using the two-parameter arctangent function atan2, the absolute value of the elevation coordinate difference is used. The first parameter is the horizontal projection spacing. The second parameter is used to directly calculate the telescope's vertical elevation angle. .

[0038] Thus, for each valid connection in the current candidate control network scheme, two key geometric interference quantities are bound: the effective net distance from the line of sight to the surface of the structure. Vertical elevation angle of the telescope .

[0039] S4 is transformed into line-of-sight connectivity weights based on empirical penalty coefficients.

[0040] Because traditional surveying error theory has extremely strict alignment requirements for unit dimensions when calculating coordinate variance, directly adding interferometric values ​​with different dimensions would cause the algorithm to completely collapse in terms of physical model interpretation. Therefore, this embodiment introduces a dimensionless penalty coefficient mechanism to construct an empirical objective function specifically for evaluating the topological robustness of graph-theoretic networks. This dimensionality reduction transforms incalculable multi-source physical interference into a unified-scale digital weight feedback.

[0041] Specifically, regarding the first Simulated total station line of sight The aforementioned extracted telescope vertical elevation angle Compared with the preset angle penalty coefficient Multiply the products, then square the result to obtain the first penalty term, denoted as the angle penalty term; simultaneously, apply the preset refraction penalty coefficient. Divide by the aforementioned effective net distance from the line of sight to the surface of the structure. Then, the result of the division operation is squared to obtain the second penalty term, which is denoted as the refraction penalty term.

[0042] Then, the first penalty term and the second penalty term are added together. To prevent computational overflow due to a zero denominator during the reciprocal step when both the first and second penalty terms happen to approach zero simultaneously, a preset zero-prevention parameter tuning factor is added again to the sum. Finally, the reciprocal of the summed denominator is taken.

[0043] The final calculation result is defined as the line of sight specific to this simulated total station. Line-of-sight connectivity weighting The specific expression is:

[0044] in, This is a preset dimensionless angle penalty coefficient; This represents the telescope's vertical elevation angle caused by the height difference between the two endpoints of the line of sight. In the formula... The first penalty is applied when the elevation difference between two control points increases sharply, specifically the pitch angle. As the value increases, the penalty term will increase exponentially. This mathematical change directly indicates that when the instrument is used for observation, the telescope needs to be raised extremely high, and the tiny wobble error of the mechanical vertical axis will be magnified many times over. This observation posture is not recommended in engineering.

[0045] This is the preset refraction penalty coefficient; This represents the actual remaining air gap between the line of sight and the nearest concrete surface, after deducting the thickness of the scaffolding; that is, the effective clear distance. In the formula... This is the second penalty item. When the line of sight is extremely close to the side wall, the effective clearance... As the distance decreases, the net distance in the denominator causes the penalty term to increase dramatically. This mathematical change directly reflects the severe impact of thermal refraction caused by the temperature gradient of the concrete surface on the beam attached to the wall, resulting in distortion of the measured distance data.

[0046] Based on the above formula, when a certain simulated total station line of sight... In a BIM environment, when facing unfavorable spatial interference conditions such as high elevation differences, large upward angles, or close proximity to side walls, increasing either the first or second penalty term will cause the total value in the denominator of the formula to expand rapidly. This expansion of the denominator directly affects the line-of-sight connectivity weighting value obtained after calculating the reciprocal. It exhibits a precipitous decline and decay.

[0047] Line-of-sight connectivity weighting The smaller the value, the lower the confidence level of the spatial connection. Although it is not theoretically completely obscured (visible), the confidence level of its actual carrying of reliable observation data is extremely low, and its graph theory weight as a network connection edge is also reduced to the lowest level.

[0048] S5, construct the topological connectivity weighting matrix and extract the network fracture risk value.

[0049] At surveying sites, the intersections of underground caverns often become congested with numerous muck trucks and temporary structures. If the sole main line of sight (the "throat") connecting the two major survey areas in the control network is blocked at this point, it will not only cause the failure of a local survey point but also completely sever the coordinate reference transmission link for the entire rear survey area, leading to a global adjustment solution collapse. Traditional geometric network layout algorithms only know which lines of sight are transparent, but cannot determine which line of sight plays an irreplaceable pivotal role in the overall structure.

[0050] This embodiment introduces the Laplace matrix analysis method from graph theory, stripping away specific spatial coordinates and directional partial derivatives, and independently calculating the network's connectivity robustness solely based on the topological edges and connection weights between measurement points. This dimensionality reduction operation from physical geometry to algebraic graph makes the evaluation of the resilience of complex control networks computationally feasible through pure digital means.

[0051] First, the weighting matrix is ​​constructed. This is performed for the current evaluation cycle. Coordinate sets of candidate control networks Count the total number of control points contained in this coordinate set, assuming the total number is . Initialize and construct a dimension with size in memory. (Right now OK A blank real symmetric algebraic square matrix (columns). This square matrix is ​​defined as the control network topology connectivity weighting matrix. The elements in this matrix are indexed by row. With column index Perform positioning marking, using Represents a matrix element at a specific position, where, and The range of values ​​is all within to Between, respectively with this The sequence numbering of each control point must be strictly corresponding.

[0052] Subsequently, weighting values ​​were assigned based on line-of-sight connectivity. Fill the blank matrix with the values. The corresponding position: for the main diagonal position of the matrix, that is, the position where the row and column indices are the same, The execution weights are accumulated, and all endpoints in the control network that are connected to the first endpoint are retrieved. Simulated total station lines of sight for each control point. Next, line-of-sight connectivity weights are assigned to these retrieved lines of sight. Perform a full summation. Use the summed values ​​as matrix elements. The values ​​are filled in. The larger the value of the diagonal element of the matrix, the more reliable lines of sight are connected to the control point, that is, the stronger the intersection support of the control point in the network.

[0053] For the non-main diagonal positions of this matrix, i.e., positions with different row and column indices. Perform negative padding on the connected edge weights. Determine the first... The control point and the first If there is a direct, valid line of sight between the control points, the system determines whether such a line of sight exists between the two points. If the determination is true, a direct line of sight exists between the two points, and the line-of-sight connectivity weight value belonging to this line of sight is extracted. Take the negative of the value and use this negative value as a matrix element. Fill in, i.e., command If the determination result is false, meaning there is no direct connection between the two points, whether it exceeds the distance measurement limit or was eliminated in a previous step due to obstruction by a structure, the matrix element is directly removed. The numerical value assigned is .

[0054] After completing the numerical filling of all diagonal and off-diagonal elements, a complete control network topology connectivity weighting matrix was successfully constructed. This matrix conforms to the definition of a graph-theoretic Laplace matrix in terms of mathematical structure.

[0055] Furthermore, the eigenvalue decomposition module of the real symmetric matrix from a mature numerical computation function library is invoked; in this embodiment, it is an iterative decomposition function based on the QR algorithm. The topological connectivity weighting matrix of the control network is then applied. After performing the decomposition operation, a set of quantities is extracted. A set of eigenvalues ​​is generated, and these eigenvalues ​​are sorted in ascending order of numerical value.

[0056] In graph theory, there is a strict mathematical correspondence between the eigenvalues ​​of the Laplacian matrix and the connectivity of its corresponding graph. The smallest eigenvalue is always equal to... This corresponds to a constant eigenvector where all components are equal. The second smallest eigenvalue in the sequence is the most crucial indicator representing the connectivity and resilience of the entire network graph.

[0057] Therefore, from the ascending set of eigenvalues, the second smallest eigenvalue is accurately extracted, and this second smallest eigenvalue is defined as the control network topology connectivity strength eigenvalue. .

[0058] After extracting this value, a crucial pre-legality check is performed: that is, to determine the validity of the value. Is the value equal to .

[0059] If the determination result is equal to In mathematical graph theory, this strictly represents the current coordinate system layout as not forming a closed connection. That is, within this coordinate system, there exist isolated breakpoints (islands) with no visual connection to other survey areas, effectively rendering the coordinate system fragmented. Faced with this ineffective network layout, a preset maximum constant value is directly assigned as a maximum penalty value; in this embodiment, it is set to [value missing]. As an extreme form of elimination penalty, the application to this coordinate set will be terminated immediately. The subsequent evaluation cycle.

[0060] If the judgment result is greater than This indicates that the network as a whole is connected. Continuing with the output of the decomposition module, we extract the topological connectivity strength feature value of this non-zero control network. A set of strictly bound one-dimensional eigenvectors is denoted as eigenvector. .

[0061] This feature vector Includes Each independent component value corresponds sequentially to one of the following in the network diagram: Each measurement point is numbered. In spectral theory, these component values ​​perform a relaxed binary classification of the graph. If a connecting edge lies at the cut edge that divides the network graph into two independent parts, then the eigenvector components corresponding to the two endpoints of this edge will exhibit the largest numerical difference. Utilizing this hidden data pattern, all valid simulated total station lines of sight specific to this network are re-traversed. Regarding the first Find the control point corresponding to the starting point of the line segment by tracing the line line. and the corresponding number of the endpoint control point. .

[0062] Subsequently, in one-dimensional feature vectors In the middle, accurately extract the index position corresponding to The first component value, and the corresponding index position. The second component value. After obtaining these two component values, subtract the difference between them, and take the absolute value of the result. The absolute value of the difference is defined as the single line-of-sight breakage risk value. .

[0063] Risk value of single line of sight breakage The numerical value objectively and quantitatively characterizes the risk level of the connecting line: when the risk value of this single line of sight breaking is... The larger the value, the greater the disparity in the connection between the points at both ends of the line of sight in the graph topology, indicating that the line of sight is more likely to act as a narrow bridge or vital artery connecting two different clusters of measuring points. Once the line of sight is blocked by temporary facilities such as large trucks during on-site construction, the impact on the network connectivity and coordinate calculation of the entire area will be devastating.

[0064] Finally, since the survival rate of the field control network is determined by the weakest and most vulnerable control line of sight (i.e., the shortest plank in the barrel effect), the current line of sight must be considered as the basis for determining the survival rate. Coordinate sets of candidate control networks The calculated risk values ​​for the breakage of all individual lines of sight are as follows. Perform a descending sort operation on this set of numerical sequences. After the comparison, extract the first (i.e., the largest) extreme indicator in the sequence. Define this largest numerical indicator as the extreme value risk cost. .

[0065] S6 is based on risk feedback-driven measurement point translation and projection adsorption.

[0066] This step receives the extreme risk cost, representing the highest potential fracture risk in a single cycle, output from step S5. The extreme value is used as the objective function output of the particle swarm optimization algorithm, i.e., the fitness. Since the cost of this algebraic extreme value comes from the complex eigenvectors of the graph matrix, it exhibits nonlinearity and lacks partial derivative gradients in spatial directions, making conventional gradient descent methods based on derivatives unsuitable. Therefore, this embodiment utilizes the gradient-free global heuristic search mechanism of the particle swarm optimization algorithm, using a cost-based survival-of-the-fittest principle to drive the measurement network to gradually move towards a more stable three-dimensional entity space.

[0067] Specifically, extract the corresponding current number Coordinate sets of candidate control networks Extreme value risk cost This value is set as the penalty fitness value for the current algorithm-optimized particle. The larger this value, the more easily the mesh layout is paralyzed and collapses when it encounters sudden obstacles, and the algorithm will tend to eliminate it or force it to significantly modify its coordinate position during subsequent evolution.

[0068] Then, a termination check is performed to determine the current iteration round. Has the preset maximum number of iterations been reached? In this embodiment, the maximum number of iterations will be... Set to 500 times.

[0069] If the iteration rounds are determined Not yet reached According to the evolutionary rules of the particle swarm optimization algorithm, the above-mentioned penalty fitness value is substituted into the gradient-free particle velocity update formula. This formula takes into account the particle's own historical best position and the best position currently explored by the population, and calculates the translational velocity and direction vector of each control point in three-dimensional space (X / Y / Z axis).

[0070] Based on the calculated three-dimensional translational velocity vector, drive the current... Coordinate sets of candidate control networks All undetermined measurement control points Perform three-dimensional coordinate translation and displacement operations within the spatial boundaries represented by the three-dimensional building information model.

[0071] Because simply translating in three-dimensional space according to the velocity vector can easily cause newly generated control points to detach from the structure surface and float in mid-air, or be incorrectly embedded inside solid walls. Such points cannot find a solid physical bearing surface to mount the total station and reflecting prism during actual construction, leading to layout failure. Therefore, to ensure that the coordinates generated after each translation are legal points with solid installation conditions, a layer of enhanced physical projection snap-fit ​​constraint needs to be invoked after the translation operation. The specific operation is as follows: The 3D spatial ray casting algorithm is invoked, using the translated, suspended control point as the emission source, to emit a simulated probe ray along the vertical direction of the 3D coordinate system (usually downward along the Z-axis or close to the side wall normal). The collision points between this ray and the valid floor or corridor grid surfaces with entity attributes in the imported 3D Building Information Model (BIM) are calculated. After obtaining the collision points, the coordinates of the intersection points on the entity surfaces are used as update replacement values, forcibly snapping to and overwriting the original suspended 3D coordinates of the measurement point.

[0072] After completing the coordinate update and correction based on surface projection adsorption, the particle position adjustment for this round ends, and the process jumps back to step S1. The updated 3D coordinates are then used to reconnect the lines, and the next round (the first round) begins. Interference assessment and matrix feature extraction of (wheels).

[0073] If determining the iteration rounds achieve Once 500 optimization iterations are completed, the algorithm immediately terminates and exits. At this point, the entire population record of the completed iterations is traversed, and the extreme value risk cost corresponding to all retained candidate control networks is retrieved and compared. The algorithm extracts the historical minimum value record. It then extracts the final 3D coordinate point cloud set that perfectly corresponds to the risk cost of this historical minimum. This outstanding point cloud coordinate set passes the constraint based on physical entity adsorption and exhibits the lowest graph theory breakage risk in feature evaluation. This final coordinate set is defined and renamed the Optimal Field Control Point Coordinate Set. .

[0074] S7, coordinate transformation and spot layout indication in the field.

[0075] This step receives the optimal set of field control point coordinates generated above. It is responsible for converting the three-dimensional rectangular coordinates into digital sequence commands of time and angle required by the field instruments to drive the underlying servo motors. Ultimately, it guides the field personnel to complete the setting up of measurement points by directing the total station to execute physical mechanical movements and laser indications.

[0076] First, because the built-in motor servo system of the total station cannot directly recognize and respond to discrete absolute three-dimensional coordinates (X, Y, Z), the instrument's underlying physical drive only accepts control of relative horizontal rotation and vertical pitch angles. Therefore, a coordinate axis transformation based on analytical geometric perspective must be performed by the system terminal. The specific operation is as follows: According to the undetermined measurement control points The optimal set of field control point coordinates is extracted sequentially based on the initial numerical sequence of the network entry. The extracted ordered 3D coordinate data is then combined and packaged into a lightweight standard measurement task instruction data packet. This data packet is then directly transmitted via a general wireless mobile network to the field control tablet or smart terminal device carried by the field surveying personnel.

[0077] In the physical construction area of ​​a hydropower project, field workers first set up the total station at any pre-fixed, known starting control point and complete the routine instrument leveling and station setup operations. Subsequently, the field intelligent terminal reads the three-dimensional coordinates of the current field station center point of the total station via Bluetooth or serial communication cable.

[0078] The field intelligent terminal calls the internally pre-built three-dimensional spatial analysis function library, and uses the three-dimensional coordinates of the current total station center and the three-dimensional coordinates of a remote target control point passed in the command package to perform coordinate axis difference calculation: that is, to calculate the coordinate difference between the target point and the station center in the X-axis direction and the coordinate difference in the Y-axis direction respectively. The field intelligent terminal performs quadrant determination and conversion based on the coordinate difference between the target point and the station center in the X and Y axes, and calculates the relative horizontal azimuth angle from 0 to 360°.

[0079] Simultaneously, based on the three-dimensional coordinates of the total station center and the three-dimensional coordinates of a remote target control point transmitted in the command packet, the field intelligent terminal calculates the coordinate height difference between the two in the Z-axis direction, divides the height difference by the spatial slant distance, and finally obtains the inverse cosine function to calculate the absolute zenith distance required by the total station telescope.

[0080] After completing the above geometric transformation, the field intelligent terminal will convert the calculated relative horizontal azimuth and zenith distance into digital pulse sequences, i.e. motor drive level pulse commands, through the underlying communication protocol encoding conversion, and directly send them to the hardware control motherboard inside the total station.

[0081] After receiving the level pulse command, the horizontal and vertical axis stepper motors inside the total station drive the hardware mechanical mechanism to rotate precisely on both axes according to the number of pulses, so that the center of the crosshairs of the total station's telescope is accurately aligned with the target angle position planned by the algorithm.

[0082] Once the mechanical rotation stops and locks, the total station automatically activates the visible infrared laser pointing function (laser point-finding function) integrated into the telescope. A visible red laser beam is emitted from the telescope, reaching and forming a clear and conspicuous red spot on the surface of the distant concrete structure. The location of this red spot is the optimal measuring point, selected through complex evaluation to effectively resist line-of-sight interruptions and fully meet the line-of-sight requirements. At this point, on-site personnel go directly to the specific location illuminated by the red spot to manually mark the control point's center cross or install the reflective prism base using expansion bolts.

[0083] By controlling the light spot with the above-mentioned instructions and assisting in the physical action combination of manually placing the prism, this embodiment completely breaks down the data barrier from the upstream virtual three-dimensional environment constraints to the downstream physical on-site operation execution, and completes the final technical closed loop from algorithm-generated planning to engineering entity layout construction of the optimal measurement control network.

[0084] Based on the same inventive concept as the above methods, this application also provides a BIM-based hydropower engineering survey control network layout optimization system, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of any one of the above-described BIM-based hydropower engineering survey control network layout optimization methods.

Claims

1. A BIM-based optimization method for the layout of surveying and control networks in hydropower engineering, characterized in that, The method includes the following steps: Obtain the BIM model of the hydropower project and the coordinates of the known starting control points to construct an initial candidate control network coordinate set; Based on the initial candidate control network coordinate set and BIM model, a simulated total station line-of-sight set is constructed, and lines of sight blocked by the physical grid are removed. For each retained simulated total station line of sight, its spatial geometric interference features are extracted. These features are then converted into line-of-sight connectivity weighting values ​​using a preset empirical penalty coefficient. This process includes: extracting the elevation difference and horizontal projection distance between the two endpoints of each retained simulated total station line of sight; calculating the telescope's vertical pitch angle; calculating an angle penalty term based on the vertical pitch angle and a preset angle penalty coefficient; and calculating a refraction penalty term based on the effective net distance and a preset refraction penalty coefficient. The line-of-sight connectivity weighting values ​​are negatively correlated with both the angle penalty term and the refraction penalty term. Constructing a control network topology connectivity weighting matrix based on all line-of-sight connectivity weights includes: constructing a control network topology connectivity weighting matrix of dimension N×N, where N is the number of control points; the matrix is ​​a Laplace matrix, the diagonal elements of the matrix are the sum of the line-of-sight connectivity weights of all connecting edges connected to the corresponding control points, and the off-diagonal elements are the negative values ​​of the line-of-sight connectivity weights of the connecting edges between two corresponding control points; The matrix is ​​subjected to eigenvalue decomposition to extract the extreme risk cost characterizing the control network failure risk. This includes: after the matrix is ​​decomposed, the non-zero second smallest eigenvalue and its corresponding eigenvector are extracted; for any line of sight, the absolute value of the difference between the corresponding components of the two endpoints of the line of sight in the eigenvector is calculated as the failure risk value of the line of sight; based on the numerical distribution of the failure risk values ​​of each retained line of sight, the extreme risk cost is determined; wherein, the maximum value among all the failure risk values ​​of the retained single lines of sight is determined as the extreme risk cost; if the second smallest eigenvalue is equal to zero, the control network is determined to be disconnected, and a preset maximum penalty value is assigned. With the goal of minimizing extreme value risk costs, the optimization algorithm iteratively optimizes the candidate control network coordinate set and forcibly attaches the optimized coordinates to the BIM entity surface until the termination condition is met, and outputs the optimal field control point coordinate set.

2. The method for optimizing the layout of the measurement and control network for hydropower engineering based on BIM as described in claim 1, characterized in that, The construction of the simulated total station line-of-sight set includes: The control points in the initial candidate control network coordinate set are combined in pairs, and the metric distance between each pair is calculated; and the nominal range parameters of the total station are obtained. The lines connecting the combinations whose measured distance is less than or equal to the nominal range parameter are used as effective simulated lines of sight to construct the simulated total station line of sight set.

3. The method for optimizing the layout of the measurement and control network for hydropower engineering based on BIM as described in claim 1, characterized in that, The process of removing the line of sight obstructed by the solid mesh includes: Calculate the shortest distance between the line of sight of the simulated total station and the surface of the BIM entity grid, and deduct the preset reserved thickness of the construction scaffold to obtain the effective net distance from the line of sight to the surface of the structure. If the effective net distance is less than or equal to zero, the line of sight is discarded.

4. The method for optimizing the layout of the measurement and control network for hydropower projects based on BIM as described in claim 1, characterized in that, The step of forcibly attaching the optimized coordinates to the surface of the BIM entity includes: Using the coordinates of the undetermined measurement control point after translation during the optimization iteration as the emission source, a simulated probe ray is emitted along the preset projection direction of the nearest solid mesh surface in the BIM model; the coordinates of the collision intersection point between the simulated probe ray and the solid mesh surface of the BIM model are used as the coordinates of the undetermined measurement control point after adsorption.

5. The method for optimizing the layout of the measurement and control network for hydropower projects based on BIM as described in claim 1, characterized in that, The optimal set of field control point coordinates is sent to the field terminal. The field terminal controls the total station to calculate the horizontal azimuth and zenith distance based on the coordinates, drives the servo motor to rotate and emits a visible laser spot to indicate the position of the measuring point on the solid surface.

6. A BIM-based hydropower engineering survey control network layout optimization system, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method as described in any one of claims 1-5.

Images