An absolute laser tracking-based method for controlling electromechanical installation details of a hydroelectric project

Abstract

The application discloses a kind of water and electricity engineering electromechanical installation detail control methods based on absolute laser tracking, and the measuring control point is set in advance on the component to be installed using laser tracking from the component to be installed acquisition mark point, obtains the three-dimensional coordinate parameters of mark point and measuring control point, as original electromechanical installation description data;Remote read the original electromechanical installation description data, convert into point cloud data and upload cloud end;The point cloud data is segmented to obtain point cloud model;Based on the point cloud model, solve installation description parameter, complete fitting according to installation description parameter to obtain installation offset, to determine installation detail guidance scheme, realizes installation detail adjustment according to installation detail guidance scheme.The application combines absolute laser tracking technology and proposes a kind of point cloud data analysis method, provides new ideas and new direction for the efficient, high accuracy guidance scheme design of electromechanical.

Description

Technical Field

[0001] This invention belongs to the interdisciplinary field of absolute laser tracking technology and electromechanical installation detail control based on sampling point data, specifically involving a method for electromechanical installation detail control of hydropower projects based on absolute laser tracking technology. Background Technology

[0002] Currently, the installation of metal structures in hydropower projects involves setting up monitoring points and manually conducting continuous or periodic inspections using equipment such as total stations and dial gauges. The accuracy of the installation is then analyzed and feedback is used to control the coordinate deviations of these monitoring points. In this process, the accuracy of the sampling data and the timeliness of the feedback are crucial for controlling the metal structure installation process. However, traditional methods for controlling the installation of electromechanical metal structures can only perform random checks of key parameters during installation, making it difficult to achieve rapid, automated, and intelligent sensing and feedback control of the installation process data. Emerging absolute laser tracking technology can solve this problem. By employing absolute laser tracking, the position and angle of gates can be precisely adjusted, ensuring perfect matching with upstream and downstream water conservancy and dam structures, improving the precise coordination with the hoisting and control systems, and achieving precise water flow control. For pressure steel pipes, laser tracking technology can ensure the continuity and stability of their installation and reduce the risk of leakage at joints. Therefore, this invention combines absolute laser tracking technology to study detailed control methods for the installation of metal structures in hydropower projects.

[0003] Given the complex installation site conditions, the quality of sampling point data is affected by various factors such as personnel movement, obstruction, and instrument parameters. Traditional manual detection methods are difficult to achieve satisfactory results in terms of efficiency and automation. As an emerging real-scene modeling technology, absolute laser tracking has matured in the field of positioning and mapping. Many scholars have conducted relevant research on the application of absolute laser trackers in engineering fields and the automatic extraction and analysis of point cloud data. Reference [1] verified that laser tracking technology can be fully applied to the welding requirements of complex components through the ten-point calibration method and the adjustment of multivariable parameters. Reference [2] proposed a method for rapid mapping of building facades in narrow blocks using laser vision tracking scanning technology. Compared with the traditional total station method and the traditional ground-based three-dimensional laser scanning method, laser vision tracking scanning technology has the characteristics of high data portability, good integrity, and high automation. Reference [3] used a laser rangefinder to obtain the relevant position information of the robot and the position information of obstacles. The extended Kalman filter algorithm was used to track the robot's movement path, which can obtain high-precision robot movement path tracking results and avoid obstacles on the movement path in real time. Reference [4] uses three-dimensional laser scanning technology to monitor large buildings such as tunnels and bridges, and uses the random sampling consensus algorithm to fit straight lines, calculate the central axis based on the geometric position, and realize the extraction of cross-section of tunnel scanning point cloud. Reference [5] proposes a boundary extraction algorithm for rectangular windows based on ground three-dimensional laser scanning data, which can quickly obtain rich details of building facades and has great application advantages in the three-dimensional fine modeling of digital cities.

[0004] In summary, the use of an absolute laser tracker during the installation of hydropower metal structures enables high-speed detection, scanning, and real-time control, reliably completing tasks such as online guided assembly and processing, and point cloud data acquisition. Furthermore, by analyzing the point cloud data, a guidance plan for on-site installation can be quickly formulated.

[0005] [1] Research on the application of laser tracking technology in welding of engineering machinery [J]. Laser & Infrared, 2023, 53(05):699-705. [2] A method for surveying the facades of buildings in a street based on laser visual tracking scanning technology [J]. Heilongjiang Engineering, 2023, 37(02):22-26+33. DOI:10.19352 / j.cnki.issn1671-4679.2023.02.004. [3] Robot movement path tracking and real-time obstacle avoidance based on laser ranging technology [J]. Automation and Instrumentation, 2023, 38(03):30-34. DOI:10.19557 / j.cnki.1001-9944.2023.03.008. [4] Cross-section extraction of tunnel point cloud based on RANSAC algorithm [J]. Bulletin of Surveying and Mapping, 2021(09):120-123. [5] A method for extracting window boundaries based on three-dimensional laser scanning data [J]. Exploration Science and Technology (2):5. Summary of the Invention

[0006] This invention aims to propose a method for controlling the electromechanical installation details of hydropower projects based on absolute laser tracking. It combines the point cloud data obtained by preliminary calibration and transformation using the reference points of absolute laser tracking, and uses the point cloud data to segment the installation detail control model. The model segmentation results are then used for preliminary calibration and fitting analysis, which efficiently determines a high-accuracy installation detail guidance scheme.

[0007] This invention is achieved using the following technical solution: This invention proposes a method for detailed control of electromechanical installation in hydropower engineering based on absolute laser tracking. The method includes the following steps: Using laser tracking, marker points are collected from the component to be installed. Measurement control points are pre-set on the component to be installed to obtain the three-dimensional coordinate parameters of the marker points and measurement control points, which serve as the original electromechanical installation description data. The original electromechanical installation description data is remotely read, converted into point cloud data, and uploaded to the cloud. The point cloud data is segmented to obtain a point cloud model; The installation description parameters are solved based on the point cloud model. The installation offset is obtained by fitting the installation description parameters to determine the installation detail guidance scheme. The installation details are then adjusted according to the installation detail guidance scheme.

[0008] Furthermore, the point cloud model further includes a point cloud planar model and a point cloud spherical model.

[0009] Furthermore, the process of solving the installation description parameters based on the point cloud model, fitting the installation offset based on the installation description parameters to determine the installation detail guidance scheme, and adjusting the installation details according to the installation detail guidance scheme further includes: solving the installation description parameters based on the point cloud planar model and the point cloud spherical model, simulating the shape and position of the electromechanical installation to obtain the installation offset, comparing the offset to determine the installation detail guidance scheme, and feeding it back to the mobile receiving platform of the on-site installation personnel.

[0010] Furthermore, the model segmentation process of the point cloud planar model further includes the following steps: The iteration begins by randomly selecting the 3D coordinates of three points in the point cloud to form a point cloud plane. The distance from all points to the point cloud plane is calculated. If the distance is less than a threshold, the points are considered to be in the same point cloud plane. When the number of points in the same point cloud plane exceeds a set value, the current point cloud plane is saved, and all points in the current point cloud plane are marked as matched. When the number of points found in the point cloud plane after the iteration is less than the set value, the iteration terminates, and the target point cloud plane model is obtained.

[0011] Furthermore, this includes saving the point cloud planar model with the most detected marker points separately and removing it from the point cloud data.

[0012] Furthermore, the model segmentation process of the point cloud spherical model further includes the following steps: The iteration begins by detecting the plane with the most points in the remaining point cloud data. Elements that make up the sphere model are selected from the remaining point cloud data. Elements exceeding a threshold are classified as outliers, and elements within the threshold are interior points. The iteration continues until the sphere model corresponding to the most interior points is obtained, thus obtaining the target point cloud spherical model.

[0013] Furthermore, the step of solving for the installation description parameters based on the point cloud model and completing the fitting based on the installation description parameters further includes: Point cloud planar model extraction is performed by randomly selecting the minimum number of elements that make up the point cloud planar model from the point cloud data. The installation description parameters are solved using the point cloud planar model equation (1). The installation description parameters are checked using the remaining point cloud data after removing the minimum number of elements that make up the point cloud planar model. The installation error of the model fitting is compared with the coordinates of the pre-set measurement control points. If it is within the range, it is classified as an interior point and the number of interior points is recorded. The above two steps of solving and checking are repeated. When the number of interior points in the model is greater than the number of interior points already saved, the installation description parameters of the model are updated to ensure that the model contains the maximum number of interior points. The process is iterated until the threshold is reached to obtain the final fitted installation offset.

[0014] Furthermore, the step of solving the installation description parameters based on the point cloud model and completing the fitting based on the installation description parameters further includes: extracting the point cloud spherical model, continuing to use the random sampling consensus algorithm to randomly select the minimum number of elements constituting the point cloud spherical model from the point cloud data, and solving the installation description parameters using the point cloud spherical model equation, with the key element being the sphere center coordinates (…). x,y,z ) and radius dCheck whether the remaining point cloud data after removing the minimum number of points that make up the point plane model are inside the model: compare the verification error with the set measurement control point coordinates. If they are within the range, they are classified as interior points and the number of interior points is recorded. Repeat this process multiple times and select the model with the most interior points as the best fitting result to obtain the final fitted installation offset.

[0015] Compared with existing technologies, the present invention can achieve the following beneficial technical effects: 1) A point cloud data model of the electromechanical installation site is established by combining absolute laser tracking technology. The point cloud model is segmented by random sampling consensus algorithm. The turbine unit and generator components are simulated according to the center coordinates of the target sphere. The offset is obtained by comparing with the specifications and design drawings. In this way, the guidance scheme for installation details can be quickly and accurately determined in the actual engineering site installation, and the automation and intelligence level of the installation process can be improved. 2) The proposed point cloud data analysis provides new ideas and directions for the design of electromechanical installation guidance schemes. Attached Figure Description

[0016] Figure 1 This is a detailed flowchart of a method for controlling the electromechanical installation details of hydropower projects based on absolute laser tracking, according to the present invention.

[0017] Figure 2 This is a flowchart of the installation detail control model segmentation process based on the random sampling consensus algorithm in this invention. Detailed Implementation

[0018] The specific embodiments of the present invention will now be described in further detail with reference to the accompanying drawings.

[0019] Numerous sampling points are set up around the electromechanical installation of water conservancy projects, and three-dimensional point cloud data is recorded using absolute laser tracking technology. Furthermore, a point cloud segmentation and automatic extraction model is established, and then the corresponding adjustment scheme is determined based on the offset of coordinate displacement relative to the design drawings and construction specifications. The results are fed back to the mobile terminal at the electromechanical installation site, thereby improving the installation quality of electromechanical details in hydropower projects.

[0020] To achieve speed and accuracy in determining the installation guidance scheme and further improve installation quality, the technical solution adopted in this invention is a detailed control method for electromechanical installation in hydropower projects based on absolute laser tracking. Taking the installation of rotor and stator as an example, the method specifically includes the following steps: Step 1, Data Acquisition: Using a total station and the provided drawings, mark reference points and measurement control points based on the local plane coordinate system of the engineering site. Set up laser trackers sequentially around the cylindrical target, using an absolute laser tracker as the data acquisition platform to observe and acquire all the reference points and measurement control points at the center of the cylindrical target. Unify all measurement control points into the local plane coordinate system of the engineering site, i.e., the spatial rectangular coordinate system XOZ. Use the center reference points at the cylindrical target to achieve center calibration of the component to be installed, obtaining the initial calibration data for the center of the component to be installed. The present invention further includes preprocessing: pre-determining measurement control points on the electromechanical components of the hydropower project to be installed, and setting calibration targets at the upper and lower ends of the rotor and stator center bodies. For example, in this embodiment, cylindrical targets are used, and the cylindrical targets are equipped with central reference markers. The position and number of the cylindrical targets should ensure that the cylinder is in an upright state, and that the measuring stations set at appropriate distances (determined according to the on-site working space) in each direction can simultaneously observe at least 3 central reference markers at each of the upper and lower ends of the cylindrical center body. Step 2, Data Processing: Use a computer to remotely read the initial calibration data from the absolute laser tracker, upload it to the cloud and database, and use development tools in the background to interpret the raw data into a point cloud general format to obtain the initial calibration point cloud data of the electromechanical components of the hydropower project to be installed; specifically, the true distance between the target object and the instrument is directly measured through laser interferometric ranging technology, without relying on other reference points or accumulating errors from multiple measurements; Step 3: Model Segmentation: Point cloud segmentation based on a random sampling consensus algorithm is used to segment the initial calibration point cloud data of the hydropower engineering electromechanical components to be installed, obtained in Step 2. To address the segmentation problem of fragmented point clouds, a pre-segmentation method based on supervoxel clustering is used. For the under-segmentation problem caused by contact between different objects, a re-segmentation method based on point cloud slicing is proposed, improving the segmentation accuracy. The point cloud dataset of the hydropower engineering electromechanical components is obtained through model segmentation. By expanding and connecting the scattered point clouds using triangular meshes, a three-dimensional mesh model reflecting the true state of the objects can be obtained, fitting the three-dimensional models of the generator rotor and stator. Step 4: Based on the 3D model obtained by model segmentation and fitting, the 3D geometry and installation position of the electromechanical components can be obtained. Then, by comparing with the graphic and using the coordinate deviation value, the final installation detail guidance plan is determined and fed back to the mobile terminal receiving platform of the on-site installation personnel, so that the installation adjustment can be realized according to the comparison results.

[0021] The above process is described in detail below with reference to specific implementation methods: Using an absolute laser tracker as the data acquisition platform, it's necessary to consider the potential obstruction caused by moving construction machinery and personnel during the absolute laser tracking process. Therefore, a pre-planned field operation strategy is required to complete the point cloud data acquisition of the target area. Specifically: survey the site environment, design measurement stations and target points for the electromechanical components requiring data collection, ensuring the absolute laser tracker completes point cloud data acquisition at the designated locations; and rationally set the laser tracker parameters, including laser emission power, scanning angle, data sampling rate, filtering, and noise reduction settings, to ensure both high-quality point cloud acquisition and efficient operation, achieving rapid and high-precision point cloud data acquisition.

[0022] The raw point cloud data is uploaded to the cloud and stored in the database as a file stream. Based on the scan sequence number, the SDK tool reads the three-dimensional spatial information of the collected point cloud data and outputs it as a common point cloud format such as .txt or .xyz file, which facilitates subsequent point cloud analysis.

[0023] like Figure 2 The diagram shown is a detailed flowchart of the installation detail control model segmentation based on the random sampling consensus algorithm in this invention.

[0024] Point cloud segmentation technology based on random sampling consensus algorithm is used to divide the initial calibration point cloud data of the electromechanical components of the hydropower project to be installed into planar and spherical regions, resulting in a point cloud planar model and a point cloud spherical model, respectively. The point cloud is the point cloud data corresponding to these points; the measurement control points are the reference points used to define and constrain the entire measurement coordinate system. Specifically, the process is as follows: At the start of the iteration, the point cloud planar model segmentation process involves randomly selecting three points (any three points in the model) to form a plane. The distance from all points to this plane is calculated. If the distance is less than a threshold, the points are considered to be in the same plane. When more than n points are in the same plane, this plane is saved, and all points on this plane are marked as matched. The iteration terminates when fewer than n points are found on the plane after iteration, or when no more than 3 unmarked points can be found (n>3). The plane with the most detected points is saved separately and removed from the point cloud data. The purpose of this process is to find the largest possible point cloud plane composed of marker points. The process continues to detect the plane with the most unmarked points in the remaining point cloud data. The point cloud spherical model segmentation process involves selecting elements from the remaining point cloud data that can form a spherical model. All point cloud data are checked to ensure they conform to the above model. Elements exceeding a threshold are classified as outliers, and those within the threshold are inliers. This process is repeated until the spherical model with the most inliers is identified as the target model. The center coordinates and parameters of the spherical model are output as a .txt file. The goal of this process is to find the largest possible point cloud sphere composed of marker points; the remaining point cloud data refers to the point cloud after removing the minimum number of points that make up the point cloud planar model.

[0025] Establish a set of analysis rules that conform to the site conditions, and simulate the on-site electromechanical system to obtain the offset from the design drawings and specifications. Specifically, a detailed model of the target sphere is created and combined with the electromechanical 3D model at the sampling point position. The center coordinates of the sphere are fitted based on the extracted target sphere point cloud data to control the installation details. The deviation is obtained by comparing with the design drawings and specifications. The deviation results are summarized into an adjustment plan for the existing installation details, and the adjustment plan is sent to the mobile client of the on-site installation personnel.

[0026] To address the lack of automation and efficiency in current research on the detailed control of electromechanical installation in hydropower projects, a method for detailed control of electromechanical installation based on absolute laser tracking is proposed. The specific implementation of the technical solution is as follows: [See details]. Figure 1 This is a flowchart illustrating the specific process of the present invention.

[0027] First, absolute laser tracking is used to collect details of the electromechanical installation to ensure that the collected point cloud data contains sufficiently rich information for further analysis. The main scanning area is first determined, then the scanning station location is roughly determined within the scanning area, the absolute laser tracker is placed, the instrument parameters are set, and data acquisition begins.

[0028] Second, the absolute laser tracker data is remotely read using a computer and encapsulated layer by layer in the form of data packets on the source computer. These packets are then forwarded through switches and routers via IP and MAC addresses, and finally sent to the target computer for point cloud analysis and cloud backup. The raw electromechanical point cloud data is then converted into a common point cloud format file using software development tools.

[0029] Third, the point cloud data is analyzed and processed, and the parameters of the mathematical model are calculated from a series of data containing outliers using the random sample consensus algorithm. The essence of the random sample consensus algorithm in fitting point cloud data is to match the point cloud with different original geometric shapes, and it is usually applied to shape detection or extraction. However, since the object processed in this invention is a parametric geometric shape or model, such as a plane or sphere, it can also be called a model segmentation method. Compared with the least squares fitting algorithm, the random sample consensus algorithm is less affected by noise and has higher fitting accuracy. The collected spatial point cloud information is rich, and model segmentation is performed using the PCL point cloud library.

[0030] i. Fitting based on point cloud planar model, the process includes: extracting point cloud planar model, that is, randomly selecting the minimum number of elements to form point cloud planar model from point cloud data using random sampling consensus algorithm (planar model fitting requires at least three points), solving for installation description parameters using point cloud planar model equation (1), the key element being three-dimensional coordinates ( x,y,z The process involves checking whether the remaining point cloud data after removing the minimum number of points constituting the plane model is inside the model: comparing the verification error with a set threshold (i.e., the coordinates of the pre-set measurement control points). If it is within the threshold range, it is classified as an interior point, and the number of interior points is recorded. The above two steps of solving and checking are repeated iteratively until the threshold is reached. When the number of interior points in the model is greater than the number of saved interior points, the installation description parameters of the model are updated. The model with the most interior points corresponding to the installation description parameters is selected as the best fitting result, and the final fitted installation offset of the model is obtained. The fitted plane interior points are removed from the point cloud data, and the next verification is performed. The purpose of this process is to use the point cloud plane model to perform installation detail control fitting and obtain a precise installation detail guidance scheme.

[0031] ii. Fitting based on the point cloud spherical model. This process includes: after the planar model is verified, the point cloud spherical model is extracted, that is, the random sampling consensus algorithm is used to randomly select the minimum number of elements that make up the point cloud spherical model from the point cloud data, and the installation description parameters are solved using the point cloud spherical model equation (2). The key element is the sphere center coordinates ( x,y, z ) and radius d The process involves checking whether the remaining point cloud data, after removing the minimum number of points constituting the planar model, is within the model: comparing the verification error with a set threshold (i.e., the coordinates of pre-defined measurement control points). If within the threshold range, the points are classified as interior points, and the number of interior points is recorded. This process is repeated multiple times, selecting the model with the most interior points corresponding to the installation description parameters as the best fit result, thus obtaining the final fitted installation parameters. Finally, the model with the most interior points is selected as the best fit result, yielding the final fitted installation offset. The purpose of this process is to use a point cloud spherical model for installation detail control fitting to obtain a correct installation detail guidance scheme.

[0032] 1. The equation of the plane is as follows: (1) A , B , C The normal vector of a plane has three components. Its direction determines the orientation of the plane in space (the normal vector is perpendicular to the plane).

[0033] x, y, z: The three-dimensional coordinates (x-axis, y-axis, z-axis) of a point (or fitted point) in the point cloud.

[0034] D: The intercepts of the plane on the x, y, and z axes, which are the values ​​of the points where the plane intersects the coordinate axes when the point is located (usually normalized to 0 or 1).

[0035] Overall meaning: The plane passes through the normal vector. A , B , C and intercept D To determine a fixed position in space.

[0036] 2. Equation of a Sphere: (2) a, b, c: The coordinates of the sphere's center on the x, y, and z axes. The center of the sphere is the "center point" of the sphere in space.

[0037] d: The square of the sphere's radius (many fitting algorithms directly use the square to avoid taking the square root). The actual radius is √d.

[0038] Overall meaning: A sphere is defined by its center (a, b, c) and radius √d.

[0039] Fourth, a correlation method for determining the installation detail control scheme based on absolute laser tracking and point cloud analysis sampling point results. First, according to specifications, design drawings, and target sphere dimensions, a detailed model of the electromechanical components and marker points (target sphere) is created. A combined model based on the actual location of the sampling points is then used as a reference model for comparison. Next, after segmenting the measured point cloud data acquired through absolute laser tracking, the target sphere's center coordinates are used to replace the measured 3D spatial sampling point positioning information. This is then compared with the precise model's 3D spatial information of the target sphere to determine the positional deviation between the two. Then, based on the magnitude of the offset, a guidance scheme for on-site installation details is determined, and this scheme is abstracted into an adjustment strategy relative to the world coordinate system. Finally, output feedback is provided to the on-site installation work, displaying and reminding users of adjustment distances and relative orientations.

[0040] During the electromechanical installation of hydropower plants, real-time feedback allows for the direct monitoring of control parameters such as rotor and stator roundness and levelness, pipe seat verticality, inner and outer shell flange end face distance, inner and outer guide ring coaxiality of the flow guiding mechanism, upstream inner and outer flange distance, elevation difference at the four corners of the flow channel cover's foundation frame, and vertical shaft flange levelness. Installation workers can precisely control the installation position and angle of these components, ensuring maximum compatibility and stability with the overall unit structure. This not only improves installation efficiency but, more importantly, guarantees the quality of turbine unit installation details, laying a solid foundation for stable operation and performance improvement.

[0041] In summary, the use of an absolute laser tracker during the installation of hydropower metal structures enables high-speed detection, scanning, and real-time control, reliably completing tasks such as online guided assembly and processing, and point cloud data acquisition. Furthermore, by analyzing the point cloud data, a guidance plan for on-site installation can be quickly formulated.

[0042] It should be understood that the embodiments and examples discussed herein are merely illustrative and may be modified or altered by those skilled in the art, and all such modifications and alterations should fall within the scope of protection of the appended claims.

[0043] It should be noted that although the present invention has been shown and described with reference to specific exemplary embodiments thereof, those skilled in the art should understand that the present invention is not limited to the above embodiments, and all modifications thereof fall within the scope of protection of the present invention.

Claims

1. A method for detailed control of electromechanical installation in hydropower engineering based on absolute laser tracking, characterized in that, The method includes the following steps: Using laser tracking, marker points are collected from the component to be installed. Measurement control points are pre-set on the component to be installed to obtain the three-dimensional coordinate parameters of the marker points and measurement control points, which serve as the original electromechanical installation description data. The original electromechanical installation description data is remotely read, converted into point cloud data, and uploaded to the cloud. The point cloud data is segmented to obtain a point cloud model; The installation description parameters are solved based on the point cloud model. The installation offset is obtained by fitting the installation description parameters to determine the installation detail guidance scheme. The installation details are then adjusted according to the installation detail guidance scheme.

2. The method for detailed control of electromechanical installation in hydropower engineering based on absolute laser tracking according to claim 1, characterized in that, The point cloud model further includes a point cloud planar model and a point cloud spherical model.

3. The method for detailed control of electromechanical installation in hydropower engineering based on absolute laser tracking according to claim 2, characterized in that, The process of solving for installation description parameters based on the point cloud model, fitting the installation offset based on the installation description parameters to determine the installation detail guidance scheme, and adjusting the installation details based on the installation detail guidance scheme further includes: solving for installation description parameters based on the point cloud planar model and the point cloud spherical model, simulating the shape and position of the electromechanical installation to obtain the installation offset, comparing the offset to determine the installation detail guidance scheme, and feeding it back to the mobile receiving platform of the on-site installation personnel.

4. The method for detailed control of electromechanical installation in hydropower engineering based on absolute laser tracking according to claim 1, characterized in that, The model segmentation process of the point cloud planar model further includes the following steps: The iteration begins by randomly selecting the 3D coordinates of three points in the point cloud to form a point cloud plane. The distance from all points to the point cloud plane is calculated. If the distance is less than a threshold, the points are considered to be in the same point cloud plane. When the number of points in the same point cloud plane exceeds a set value, the current point cloud plane is saved, and all points in the current point cloud plane are marked as matched. When the number of points found in the point cloud plane after the iteration is less than the set value, the iteration terminates, and the target point cloud plane model is obtained.

5. The method for detailed control of electromechanical installation in hydropower engineering based on absolute laser tracking according to claim 1, characterized in that, Further steps include saving the point cloud plane model with the most detected marker points separately and removing it from the point cloud data.

6. The method for detailed control of electromechanical installation in hydropower engineering based on absolute laser tracking according to claim 1, characterized in that, The model segmentation process of the point cloud spherical model further includes the following steps: The iteration begins by detecting the plane with the most points in the remaining point cloud data. Elements that make up the sphere model are selected from the remaining point cloud data. Elements exceeding a threshold are classified as outliers, and elements within the threshold are interior points. The iteration continues until the sphere model corresponding to the most interior points is obtained, thus obtaining the target point cloud spherical model.

7. The method for detailed control of electromechanical installation in hydropower engineering based on absolute laser tracking according to claim 1, characterized in that, The step of solving the installation description parameters based on the point cloud model and completing the fitting based on the installation description parameters further includes: Point cloud planar model extraction is performed by randomly selecting the minimum number of elements that make up the point cloud planar model from the point cloud data. The installation description parameters are solved using the point cloud planar model equation (1). The installation description parameters are checked using the remaining point cloud data after removing the minimum number of elements that make up the point cloud planar model. The installation error of the model fitting is compared with the coordinates of the pre-set measurement control points. If it is within the range, it is classified as an interior point and the number of interior points is recorded. The above two steps of solving and checking are repeated. When the number of interior points in the model is greater than the number of interior points already saved, the installation description parameters of the model are updated to ensure that the model contains the maximum number of interior points. The process is iterated until the threshold is reached to obtain the final fitted installation offset.

8. The method for detailed control of electromechanical installation in hydropower engineering based on absolute laser tracking according to claim 1, characterized in that, The step of solving the installation description parameters based on the point cloud model and completing the fitting based on the installation description parameters further includes: extracting the point cloud spherical model, continuing to use the random sampling consensus algorithm to randomly select the minimum number of elements constituting the point cloud spherical model from the point cloud data, and solving the installation description parameters using the point cloud spherical model equation. The key element is the coordinates of the sphere center (…). x,y,z ) and radius d Check whether the remaining point cloud data after removing the minimum number of points that make up the point plane model are inside the model: compare the verification error with the set measurement control point coordinates. If they are within the range, they are classified as interior points and the number of interior points is recorded. Repeat this process multiple times and select the model with the most interior points as the best fitting result to obtain the final fitted installation offset.

Images