High-precision survey lofting intelligent positioning system and method based on multi-source data fusion

Abstract

The invention relates to the technical field of building reconnaissance, in particular to a high-precision reconnaissance lofting intelligent positioning system and method based on multi-source data fusion, and the system comprises a data collection module, a fusion processing module, an intelligent positioning calculation module, a lofting path generation module, a wireless communication module and a visual terminal module. The data acquisition module synchronously acquires multi-source data; the data fusion processing module carries out space-time alignment and feature level fusion and detects the state of the sensor; the intelligent positioning resolving module outputs three-dimensional coordinates by adopting various algorithms; the lofting path generation module generates an optimized path according to the BIM model; the wireless communication module realizes high-speed data transmission and equipment communication; the visual terminal module provides a visual interaction interface. Multi-source data fusion is utilized, the positioning precision is improved, the reliability of the system is guaranteed through sensor failure detection, the system adapts to the complex environment, the lofting path is optimized, the overall performance of the system is enhanced, and an efficient and accurate solution is provided for building investigation lofting.

Description

Technical Field

[0001] The present invention relates to the technical field of building survey, and in particular to a high-precision survey and setting out intelligent positioning system and method based on multi-source data fusion. Background Art

[0002] In the field of building survey technology, with the continuous expansion of the scale and increasing complexity of modern construction projects, higher requirements are placed on the accuracy, efficiency and reliability of survey and layout positioning. However, existing positioning technologies and systems have many limitations and are difficult to meet actual engineering needs, specifically:

[0003] The accuracy of traditional positioning technology is limited: Common traditional positioning technologies, such as single GNSS (Global Navigation Satellite System) positioning, are easily blocked, reflected and interfered with in complex environments (such as urban high-rise areas and mountainous areas), resulting in a significant decrease in positioning accuracy, and cannot meet the requirements of centimeter-level or even higher accuracy for building survey and layout. When surveying in densely built-up areas in cities, GNSS signals will be blocked by surrounding high-rise buildings, resulting in signal loss or multipath effects, making the positioning error up to several meters or even larger, seriously affecting the accuracy of survey and layout;

[0004] Insufficient data collection and processing: Traditional data collection methods often rely on a single type of sensor, which obtains limited information and cannot fully reflect the complex environment of the construction site. In addition, the data collected by different sensors lack effective synchronization and fusion in time and space, and the data processing efficiency is low, making it difficult to provide real-time and accurate positioning data support. For example, only using laser scanning to obtain terrain data cannot obtain the absolute position information of the equipment in real time. At the same time, due to the lack of a unified data processing mechanism, the data of different sensors cannot work together, resulting in inaccurate and untimely positioning results.

[0005] Lack of intelligence in layout path planning: Most existing layout path planning methods are based on simple rules or experience, and fail to fully consider complex factors at the construction site, such as obstacle distribution, construction safety requirements, and the motion characteristics of equipment. This makes the planned path possibly not the optimal path, which not only increases construction time and cost, but also may pose safety hazards. When planning the layout path, obstacles formed by temporarily stacked materials and equipment at the construction site are not taken into account, resulting in frequent detours for layout equipment, reducing work efficiency.

[0006] Poor system stability and reliability: In actual construction environments, sensors are easily affected by various factors, such as bad weather (heavy rain, sandstorms, etc.), electromagnetic interference (interference from electrical equipment at the construction site), etc., which can lead to data anomalies or sensor failure. Traditional systems lack effective fault-tolerant mechanisms and adaptive adjustment capabilities when faced with these problems. The system stability and reliability are poor and the continuity and accuracy of survey and layout work cannot be guaranteed. When encountering bad weather that causes abnormal data from a sensor, traditional systems may experience positioning errors or be unable to locate, seriously affecting construction progress.

[0007] Therefore, in order to solve the above problems, a high-precision survey and layout intelligent positioning system and method based on multi-source data fusion is proposed. Summary of the invention

[0008] The purpose of the present invention is to provide a high-precision survey and layout intelligent positioning system and method based on multi-source data fusion to solve the problems raised in the above background technology.

[0009] To achieve the above object, the present invention provides the following technical solutions:

[0010] High-precision survey and layout intelligent positioning system based on multi-source data fusion, including:

[0011] Data acquisition module: including laser scanning unit, GNSS positioning unit, inertial measurement unit and visual sensor unit, synchronously collects 3D laser point cloud data, satellite positioning data, inertial navigation data and visual image data of the construction site, and outputs raw data streams with time stamp alignment;

[0012] Data fusion processing module: connect to the data acquisition module and execute in sequence:

[0013] Multi-source data spatiotemporal alignment unit: uses Beidou timestamp and ORB-SLAM3 time axis dual reference alignment algorithm to establish a unified spatiotemporal coordinate system;

[0014] Feature-level fusion unit: executes the improved hybrid Kalman filter-graph attention network model to generate a fused point cloud map;

[0015] Sensor failure detection unit: monitors the health status of each sensor in real time and dynamically adjusts the fusion strategy;

[0016] Intelligent positioning solution module: receives the fused point cloud map and connects to the inertial measurement unit of the data acquisition module. The configuration is:

[0017] The absolute coordinates are calculated through the RTK / PPK dual-mode differential positioning algorithm;

[0018] Calculate relative positioning coordinates through the VIO-SLAM algorithm assisted by laser point cloud;

[0019] Apply the robust estimation model to eliminate the influence of abnormal observations and output three-dimensional space coordinates;

[0020] Lofting path generation module: communicates with the intelligent positioning solution module, including:

[0021] BIM model analysis unit: converts the engineering BIM model into three-dimensional path constraints in the construction coordinate system;

[0022] Path planning unit: Generates the layout path based on the multi-objective optimization algorithm;

[0023] Wireless communication module: connected with each module, configured as:

[0024] 5G communication unit: upload data to the cloud platform at a rate of ≥1Gbps;

[0025] LoRa self-organizing network unit: establish a Mesh network in the signal blocking area to achieve direct communication between devices;

[0026] Visual terminal module: receives the layout path and real-time positioning data, including:

[0027] AR display unit: superimposes and displays theoretical / actual position deviation vectors;

[0028] Dynamic deviation correction unit: performs three-level feedback control to achieve real-time path correction.

[0029] As a preferred solution, the feature-level fusion unit adopts an improved hybrid Kalman filter-graph attention network model, and its state update formula is: ,in, for Fusion state estimates at all times; , is the state transfer matrix of online calibration; A Kalman gain matrix that is dynamically adjusted based on sensor confidence; is the graph network coupling coefficient; is the cross-modal attention weight; is the trainable parameter matrix; for The state value at the moment; for Control input at all times; for The measured value at the moment; is the measurement matrix; is the set of neighbor nodes of sensor node i; Neighbor node The status value of .

[0030] As a preferred solution, the calculation of cross-modal attention weights adopts a multi-sensor feature cross-validation algorithm: ,in, , is the query / key vector projection matrix; is the Euclidean distance of sensor nodes; is the environmental interference attenuation factor; , Encode vector for sensor type; For sensor nodes The eigenvector of is the dimension of the feature vector, ; is the normalized exponential function.

[0031] As a preferred solution, the multi-objective optimization algorithm executed by the path planning unit includes: ;

[0032] Constraints: ;

[0033] in, , is the total length of the path; , is the maximum curvature of the path; , is the safety scoring function; , is the dynamic weight coefficient; To plan a route; For path Positioning accuracy; For path Maximum slope of is the minimum value of the safety score; For the path The curvature of the moment; For the path Safety-related distances of points; is the number of points on the path.

[0034] As a preferred solution, when the sensor failure detection unit executes the fault-tolerant control strategy:

[0035] When the first When a sensor fails, the fusion algorithm is automatically adjusted to: ,in, is the failure indicator factor, and , Indicates A sensor completely failed; It is the set of neighbor nodes that exclude failed sensors.

[0036] As a preferred solution, the dynamic correction unit performs three-level control, including:

[0037] Primary Control , The interface displays a yellow warning box;

[0038] Secondary Control , generate local path correction instructions and vibrate alarms;

[0039] Three-level control , start the system self-check and freeze the positioning output until manual confirmation;

[0040] in, is the deviation between the theoretical position and the measured position.

[0041] As a preferred solution, when multiple devices work together, a distributed consensus algorithm is used to synchronize positioning data: ,in, is the convergence rate parameter, For equipment exist Positioning data at the moment; For equipment exist Positioning data at the moment; For equipment Neighboring devices exist Positioning data at the moment; is the Kalman gain; For equipment The measured value of is the measurement matrix.

[0042] The high-precision survey and layout intelligent positioning method based on multi-source data fusion is carried out by using a high-precision survey and layout intelligent positioning system based on multi-source data fusion.

[0043] It can be seen from the technical solutions provided by the present invention that the high-precision survey and layout intelligent positioning system and method based on multi-source data fusion provided by the present invention have the following beneficial effects:

[0044] Achieve high-precision positioning: Multi-source data fusion processing, the system integrates multiple units such as laser scanning, GNSS positioning, inertial measurement and visual sensors, collects rich data and fuses them; uses Beidou timestamp and ORB-SLAM3 time axis dual reference alignment algorithm to establish a unified space-time coordinate system, laying the foundation for accurate positioning; the improved hybrid Kalman filter-graph attention network model deeply fuses data, mines data correlation and complementarity, generates accurate fused point cloud maps, and provides reliable environmental information for positioning; uses RTK / PPK dual-mode differential positioning algorithm and laser point cloud-assisted VIO-SLAM algorithm, combined with anti-error estimation model, effectively eliminates the influence of abnormal observation values, and can accurately solve three-dimensional space coordinates in complex environments, meeting the strict requirements of building survey and layout for high-precision positioning;

[0045] Enhance system reliability: The sensor failure detection unit monitors the health status of each sensor in real time and automatically adjusts the fusion strategy once a sensor failure is detected. When a sensor fails, adjust the fusion algorithm to reduce the impact of the failed sensor, and continue the fusion processing based on the data of other normal sensors to ensure the continuity and reliability of data fusion, maintain the stable operation of the system, avoid positioning errors or system failures caused by sensor failures, and ensure the smooth progress of survey and layout work;

[0046] Improve environmental adaptability: In complex construction environments, the system responds to problems such as signal blocking and electromagnetic interference through multi-source data fusion; when satellite signals are blocked, inertial measurement units and visual sensor data assist in positioning; when electromagnetic interference affects GNSS positioning, laser scanning data is not interfered with and can still provide reliable environmental information; the wireless communication module integrates 5G and LoRa self-organizing network technologies. 5G achieves high-speed data transmission in areas with good signals, and LoRa self-organizing networks establish Mesh networks in signal blocking areas to ensure communication between devices, so that the system can adapt to different construction environments and improve adaptability and robustness;

[0047] Optimize the layout path planning: The layout path generation module generates efficient and safe layout paths with the help of BIM model analysis and multi-objective optimization algorithms; the BIM model analysis unit converts the engineering BIM model into three-dimensional path constraints in the construction coordinate system, which contains rich design and construction information; based on these conditions, the path planning unit comprehensively considers factors such as path length, curvature and safety, and finds the optimal path through multi-objective optimization algorithms and dynamic adjustment of weight coefficients, thereby reducing the equipment moving distance and time, improving the layout efficiency and ensuring construction safety;

[0048] Provide intuitive interactive experience: The AR display and dynamic correction functions of the visualization terminal module provide users with an intuitive and convenient interactive experience; the AR display unit superimposes the display of theoretical / measured position deviation vectors, so that users can intuitively understand the position deviation without complex calculation and analysis; the dynamic correction unit performs three-level feedback control according to the deviation, promptly reminds users and provides correction instructions to ensure the accuracy of the layout; this intuitive interaction method lowers the operating threshold, reduces dependence on professional technicians, and improves work efficiency and quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 It is a schematic diagram of the overall structure of the high-precision survey and layout intelligent positioning system and method based on multi-source data fusion of the present invention. DETAILED DESCRIPTION

[0050] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0051] In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0052] like Figure 1 As shown, the embodiment of the present invention provides a high-precision survey and layout intelligent positioning system based on multi-source data fusion, including a data acquisition module, a data fusion processing module, an intelligent positioning solution module, a layout path generation module, a wireless communication module and a visualization terminal module (wherein the wireless communication module is a communication tool between the data acquisition module, the data fusion processing module, the intelligent positioning solution module, the layout path generation module and the visualization terminal module, so in Figure 1 not reflected in the report).

[0053] In this embodiment, the data acquisition module includes a laser scanning unit, a GNSS positioning unit, an inertial measurement unit and a visual sensor unit, which synchronously collects three-dimensional laser point cloud data, satellite positioning data, inertial navigation data and visual image data of the construction site, and outputs a raw data stream with time stamp alignment;

[0054] Furthermore, the data acquisition module occupies a fundamental and key position in the high-precision survey and layout intelligent positioning system based on multi-source data fusion. It is like the "sensing tentacles" of the system, responsible for obtaining various key data from the construction site, and providing accurate and rich information support for subsequent data processing, positioning solution and layout operations; by integrating a variety of advanced sensor units, the module realizes the synchronous acquisition of multi-source data and outputs the original data stream aligned with the timestamp, ensuring the timeliness and consistency of the data, and laying a solid foundation for the high-precision operation of the entire positioning system. Specifically:

[0055] 1. Sensor unit composition:

[0056] Laser scanning unit: It uses advanced laser scanning technology to quickly and accurately obtain 3D laser point cloud data of the construction site. Its working principle is to emit a laser beam and measure the time it takes for the laser to reflect back, calculate the distance between the scanning point and the device, and thus construct a 3D point cloud model of the construction site scene. For example, at a large-scale construction site, the laser scanning unit can perform detailed scans of the terrain, building outlines, and various obstacles at the construction site to obtain high-precision 3D spatial information. These point cloud data not only contain the location information of the object, but also reflect the shape and surface features of the object, providing rich environmental details for subsequent positioning and path planning. The unit has the characteristics of high resolution and large scanning range, which can meet the survey needs of construction sites of different sizes, and can accurately obtain the required 3D data regardless of open sites or complex indoor environments.

[0057] GNSS positioning unit: Integrates global navigation satellite system (GNSS) technology, such as GPS, Beidou, etc., to collect satellite positioning data; it calculates the precise location information of the equipment on the earth, including longitude, latitude and altitude, by receiving satellite signals; in the process of building survey and layout, the GNSS positioning unit provides the system with an absolute positioning reference to ensure the accuracy and reliability of positioning; for example, in the initial positioning stage of field sites or large-scale construction projects, the GNSS positioning unit can quickly determine the approximate location of the measuring equipment and provide the starting coordinates for subsequent precise measurement and layout operations; the unit has high-precision positioning capabilities, combined with RTK (real-time dynamic differential positioning) / PPK (post-processing dynamic differential positioning) technology, it can achieve centimeter-level or even millimeter-level positioning accuracy, meeting the requirements of high-precision survey and layout; at the same time, it also has good anti-interference ability, and can stably receive satellite signals in complex electromagnetic environments to ensure the continuity of positioning;

[0058] Inertial measurement unit: It is composed of sensors such as accelerometers and gyroscopes, and is used to measure the acceleration and angular velocity information of the device, and then obtain inertial navigation data. The inertial measurement unit plays an important auxiliary role in the positioning system, especially when the satellite signal is blocked or lost, it can infer the position and attitude changes of the device through inertial measurement data. For example, in areas with weak satellite signals such as inside buildings or canyons, the inertial measurement unit can record the movement state of the device in real time, and combine it with previous positioning information to achieve continuous positioning tracking. It can quickly respond to the movement changes of the device, provide high-frequency measurement data, and provide real-time motion information for positioning solution, which helps to improve the accuracy and real-time performance of positioning. In addition, the inertial measurement unit is also small in size, light in weight, and low in power consumption, which makes it easy to integrate into various measurement devices.

[0059] Visual sensor unit: equipped with a high-resolution camera to collect visual image data on the construction site; the visual sensor unit can obtain rich texture information and scene images, providing an intuitive visual basis for positioning and environmental perception; for example, at the construction site, the visual sensor can take photos or videos of the surrounding environment, and use image recognition technology to identify objects such as buildings, roads, and signs to assist in positioning and path planning; it can also be used to monitor safety hazards during the construction process, such as illegal operations by personnel and equipment abnormalities; the visual sensor unit has real-time imaging and image transmission functions, and can quickly transmit the collected image data to the data fusion processing module for analysis and processing; at the same time, combined with computer vision algorithms, it can realize feature extraction and recognition of objects in the image, providing more valuable information for the system;

[0060] 2. Data collection and synchronization mechanism:

[0061] Synchronous acquisition: The laser scanning unit, GNSS positioning unit, inertial measurement unit and visual sensor unit in the data acquisition module work together to achieve synchronous acquisition of multi-source data; through precise clock synchronization technology, each sensor unit starts data acquisition at the same time to ensure that the collected data has temporal consistency; for example, during a survey and layout operation, while the laser scanning unit is scanning the construction site scene, the GNSS positioning unit obtains the satellite positioning data of the current position, the inertial measurement unit records the movement status of the equipment, and the visual sensor unit takes on-site images. All of these data are collected at the same time point, providing an accurate time reference for subsequent data fusion and analysis;

[0062] Timestamp alignment: To further ensure the time consistency of data, the data acquisition module adds an accurate timestamp to each set of collected data; the timestamp uses a high-precision timing system, such as the Beidou timestamp, to ensure the accuracy and uniqueness of the time; at the same time, combined with the ORB-SLAM3 time axis dual reference alignment algorithm, the data collected by different sensors are accurately aligned on the time axis to establish a unified space-time coordinate system; in this way, in the subsequent data processing process, the data of different sensors can be accurately matched and fused according to the timestamp, avoiding data errors and inconsistencies caused by time differences, and improving the accuracy and reliability of data fusion;

[0063] 3. Advantages of data acquisition module:

[0064] Multi-source data complementation: By integrating multiple types of sensor units, the data acquisition module can obtain rich and diverse information and achieve multi-source data complementation; the laser scanning unit provides high-precision three-dimensional spatial information, the GNSS positioning unit determines the absolute position, the inertial measurement unit tracks the motion state, and the visual sensor unit obtains texture and scene images; these different types of data complement each other and comprehensively describe the environment and equipment status of the construction site, providing a richer and more accurate information basis for high-precision survey and intelligent positioning; for example, in a complex construction site environment, when the GNSS signal is blocked, the inertial measurement unit and the visual sensor unit can assist in positioning to ensure the continuity and accuracy of positioning; the three-dimensional point cloud data obtained by the laser scanning unit is combined with the visual image data to more clearly identify and analyze objects and obstacles on the construction site, providing a more accurate environmental model for path planning;

[0065] Real-time and accuracy: The data acquisition module has high-speed data acquisition capabilities and precise time synchronization mechanisms, and can obtain construction site data in real time and accurately. Each sensor unit collects data at a high frequency, and ensures the timeliness and consistency of the data through timestamp alignment and synchronous transmission. This enables the system to reflect changes on the construction site in a timely manner and provide accurate data support for real-time positioning and decision-making. For example, during the construction process, when the construction site environment changes (such as new obstacles, equipment movement, etc.), the data acquisition module can quickly collect these change information and transmit it to the subsequent modules for processing in a timely manner, ensuring that the positioning system can adjust the positioning results and layout paths in real time, improving the safety and efficiency of construction.

[0066] Adaptability to complex environments: Each sensor unit in the module has been carefully designed and has good environmental adaptability; the laser scanning unit can work under different lighting conditions, the GNSS positioning unit has a certain anti-interference ability, the inertial measurement unit is not affected by external light and electromagnetic interference, and the visual sensor unit can still collect images normally under certain dust and bad weather conditions; this adaptability to complex environments enables the data acquisition module to operate stably in various construction sites, whether indoors or outdoors, day or night, sunny or bad weather, it can reliably collect the required data, ensuring the versatility and reliability of the positioning system;

[0067] As the front-end data acquisition component of the high-precision survey and layout intelligent positioning system based on multi-source data fusion, the data acquisition module provides comprehensive, accurate and real-time data support for the entire system through its advanced sensor unit composition, efficient data acquisition and synchronization mechanism and significant advantages. It is a key link in achieving high-precision survey and layout and intelligent positioning, and has effectively promoted the intelligent development of construction survey technology.

[0068] In this embodiment, the data fusion processing module is connected to the data acquisition module and executes in sequence:

[0069] Multi-source data spatiotemporal alignment unit: uses Beidou timestamp and ORB-SLAM3 time axis dual reference alignment algorithm to establish a unified spatiotemporal coordinate system;

[0070] Feature-level fusion unit: executes the improved hybrid Kalman filter-graph attention network model to generate a fused point cloud map;

[0071] Sensor failure detection unit: monitors the health status of each sensor in real time and dynamically adjusts the fusion strategy;

[0072] Furthermore, the data fusion processing module is the core hub of the high-precision survey and layout intelligent positioning system based on multi-source data fusion. It relies on various raw data collected by the data acquisition module and uses a series of advanced technologies and algorithms to achieve deep fusion and intelligent processing of multi-source data, providing key support for subsequent positioning solution and layout operations, and plays a decisive role in improving the positioning accuracy and reliability of the entire system. Specifically:

[0073] 1. Hardware and data support:

[0074] Multi-sensor data access device: This module has powerful data access capabilities and can be stably connected with laser scanning units, GNSS positioning units, inertial measurement units and visual sensor units; these sensors transmit the collected 3D laser point cloud data, satellite positioning data, inertial navigation data and visual image data to the data fusion processing module in real time, providing rich original information for data fusion; at the same time, the data access device has a high-speed data transmission interface to ensure the stability and timeliness of data transmission, avoid data loss or delay, and ensure the timeliness of data;

[0075] Data preprocessing and storage capabilities: After receiving multi-source data, the data fusion processing module first performs data preprocessing; it performs operations such as denoising and filtering on the original data to remove noise and outliers caused by sensor errors, environmental interference and other factors, thereby improving data quality; in addition, the module also has a large-capacity data storage capacity and can temporarily store preprocessed data for subsequent fusion calculation and analysis; these stored data not only provide support for the current survey and setting out tasks, but also provide historical data basis for subsequent review and data analysis;

[0076] 2. Functional module composition:

[0077] Multi-source data spatiotemporal alignment unit:

[0078] Implementation of dual-reference alignment algorithm: The Beidou timestamp and ORB-SLAM3 time axis dual-reference alignment algorithm are used to establish a unified space-time coordinate system; Beidou timestamp has the characteristics of high precision and high reliability, and can provide accurate time reference for all sensor data; ORB-SLAM3 time axis performs well in the field of visual positioning and has unique advantages in time processing of visual data; this unit combines the strengths of both, first uses Beidou timestamp to mark the time of various sensor data, and then finely calibrates the visual image data according to the ORB-SLAM3 time axis; through this process, it ensures that the data collected by different sensors are accurately matched in time and space, laying a solid space-time foundation for subsequent data fusion;

[0079] Time and space calibration and synchronization mechanism: In order to ensure the accuracy and stability of time and space alignment, the module has established a strict time and space calibration and synchronization mechanism; the time and space parameters of each sensor are calibrated regularly to ensure the accuracy of the timestamp and the consistency of the spatial coordinate system; during the data collection process, the time and space deviations of the sensor data are monitored in real time. Once the deviation exceeds the threshold, it is adjusted and synchronized immediately to ensure the consistency and accuracy of the data in time and space, providing a reliable data basis for subsequent fusion processing;

[0080] Feature-level fusion unit:

[0081] Improved model fusion data: Execute the improved hybrid Kalman filter-graph attention network model to deeply fuse multi-source data and generate a fused point cloud map; this model combines the advantages of Kalman filter in dynamic system state estimation and the ability of graph attention network to mine multimodal data features; through the state update formula , comprehensively considers the state value of the previous moment, the current control input and the measurement value, dynamically adjusts the Kalman gain matrix based on the sensor confidence, explores the correlation and complementarity between different sensor data, and generates a fused point cloud map containing rich environmental information, where, for Fusion state estimates at all times; , is the state transfer matrix of online calibration; A Kalman gain matrix that is dynamically adjusted based on sensor confidence; is the graph network coupling coefficient; is the cross-modal attention weight; is the trainable parameter matrix; for The state value at the moment; for Control input at all times; for The measured value at the moment; is the measurement matrix; is the set of neighbor nodes of sensor node i; Neighbor node The status value of

[0082] Cross-modal attention weight calculation: The cross-modal attention weight is calculated using a multi-sensor feature cross-validation algorithm, and the formula is: ; The algorithm comprehensively considers factors such as the Euclidean distance between sensor nodes, feature vector dimension, environmental interference attenuation factor, and sensor type encoding vector, and accurately calculates the cross-modal attention weight, so that the model can adaptively focus on the key information in different sensor data and improve the fusion effect. , is the query / key vector projection matrix; is the Euclidean distance of sensor nodes; is the environmental interference attenuation factor; , Encode vector for sensor type; For sensor nodes The eigenvector of is the dimension of the feature vector, ; is the normalized exponential function;

[0083] Sensor failure detection unit:

[0084] Real-time health monitoring: Real-time monitoring of the health status of each sensor. By analyzing the characteristics of the sensor data, data fluctuations, and comparing with historical data, the sensor can be judged to be working properly. Once abnormal fluctuations in sensor data are found, beyond a reasonable range, or there is an obvious conflict with other sensor data, the fault diagnosis process is immediately initiated to determine whether the sensor has failed and the extent of the failure.

[0085] Fault-tolerant control strategy: When the When a sensor fails, the fusion algorithm is automatically adjusted to: ,in, is the failure indicator factor, and , Indicates A sensor completely failed; In order to exclude the neighbor node set of the failed sensor, this adjustment ensures that even if some sensors fail, the system can still rely on the data of other normal sensors to continue fusion processing and maintain the stable operation of the system;

[0086] 3. Application scenario advantages:

[0087] Improve positioning accuracy: Through the spatiotemporal alignment and feature-level fusion of multi-source data, the advantages of different sensors are fully integrated to generate a more accurate fused point cloud map; in the intelligent positioning solution module, based on the fused precise data, the RTK / PPK dual-mode differential positioning algorithm and the laser point cloud-assisted VIO-SLAM algorithm are used to more accurately calculate the absolute coordinates and relative positioning coordinates, effectively eliminate the influence of abnormal observation values, and output high-precision three-dimensional space coordinates to meet the strict requirements of building survey and layout for positioning accuracy;

[0088] Enhanced system reliability: The sensor failure detection unit monitors the sensor status in real time and adjusts the fusion strategy in time when a sensor fails to ensure the continuity and reliability of data fusion. This enables the system to operate stably in complex construction environments and face emergencies such as sensor failures, avoiding positioning errors or system failures caused by sensor problems and ensuring the smooth progress of survey and layout work.

[0089] Adapt to complex environmental changes: At construction sites, the environment is complex and changeable, such as signal blocking, electromagnetic interference and other problems; the data fusion processing module can integrate multi-source data and utilize the advantages of different sensors in different environments. For example, when satellite signals are blocked, positioning is performed based on the data from the inertial measurement unit and visual sensor; in an electromagnetic interference environment, laser scanning data is not affected and can still provide reliable information for positioning; this multi-source data fusion method enables the system to better adapt to complex environmental changes and improve the system's adaptability and robustness.

[0090] Intelligent positioning solution module: receives the fused point cloud map and connects to the inertial measurement unit of the data acquisition module. The configuration is:

[0091] The absolute coordinates are calculated through the RTK / PPK dual-mode differential positioning algorithm;

[0092] Calculate relative positioning coordinates through the VIO-SLAM algorithm assisted by laser point cloud;

[0093] Apply the robust estimation model to eliminate the influence of abnormal observations and output three-dimensional space coordinates;

[0094] Furthermore, the intelligent positioning solution module is one of the core components of the high-precision survey and layout intelligent positioning system based on multi-source data fusion. It undertakes the key task of converting the fused data into accurate positioning information in the entire system; this module relies on the fused point cloud map generated by the data fusion processing module, combined with the data of the inertial measurement unit in the data acquisition module, and uses a variety of advanced positioning algorithms and models to achieve high-precision positioning of the target, providing accurate position basis for subsequent layout operations. Specifically:

[0095] 1. Data input and processing basis:

[0096] Fusion data access: The intelligent positioning solution module receives the fusion point cloud map from the data fusion processing module. This map integrates the data collected by the laser scanning unit, GNSS positioning unit, inertial measurement unit and visual sensor unit, and contains rich environmental information and features. The fusion point cloud map provides a comprehensive and accurate environmental model for positioning solution, enabling the module to more accurately determine the relationship between the target position and the surrounding environment. At the same time, the module is directly connected to the inertial measurement unit of the data acquisition module to obtain inertial navigation data in real time. These data record the acceleration and angular velocity information of the device, which can reflect the changes in the motion state of the device and provide dynamic motion parameters for positioning solution.

[0097] Data preprocessing mechanism: Before positioning and solving, the intelligent positioning and solving module preprocesses the input data; for the fused point cloud map, feature extraction and screening will be performed to highlight the key features related to positioning, such as significant terrain contours, feature points, etc., to reduce data redundancy and improve computing efficiency; for the data of the inertial measurement unit, filtering and denoising will be performed to remove noise signals caused by sensor errors or environmental interference to ensure data accuracy and stability; through these preprocessing steps, a high-quality data foundation is provided for the subsequent positioning algorithm to ensure the accuracy and reliability of positioning and solving;

[0098] 2. Positioning algorithm and model:

[0099] RTK / PPK dual-mode differential positioning algorithm: RTK (real-time dynamic differential positioning) / PPK (post-processing dynamic differential positioning) dual-mode differential positioning algorithm is used to solve the absolute coordinates; RTK technology uses the carrier phase observation values ​​between the base station and the mobile station for real-time differential processing, which can obtain centimeter-level positioning accuracy in real time in open field environments; in building survey and layout, when the satellite signal is good and there are suitable base stations around, the RTK mode can quickly determine the absolute position of the measuring equipment, providing a real-time positioning benchmark for subsequent measurement and layout operations; and PPK technology is in the measurement After the measurement is completed, the positioning accuracy is improved by post-processing the observation data collected by the base station and the mobile station. In some areas where the signal is easily blocked or interfered, such as between high-rise buildings in cities or in mountainous areas, RTK may not be able to obtain high-precision positioning results in real time. At this time, PPK technology can play its advantages. Through the fine processing of the data afterwards, it can effectively eliminate errors and obtain high-precision absolute coordinates to make up for the shortcomings of RTK in complex environments. The intelligent positioning solution module flexibly switches between RTK and PPK modes according to the actual environment and positioning requirements to ensure that accurate absolute coordinates can be obtained in various scenarios.

[0100] Laser point cloud-assisted VIO-SLAM algorithm: The relative positioning coordinates are calculated by the laser point cloud-assisted VIO-SLAM (Visual Inertial Simultaneous Localization and Mapping) algorithm. The VIO-SLAM algorithm combines the data of the visual sensor and the inertial measurement unit, uses visual information for environmental perception and feature matching, and predicts the motion state of the device with the help of the measurement data of the inertial measurement unit, thereby achieving simultaneous positioning and mapping. In the construction survey scene, when the satellite signal is blocked or cannot be obtained, the VIO-SLAM algorithm can play an important role. The assistance of the laser point cloud further enhances the positioning ability of the algorithm. The three-dimensional laser point cloud data obtained by the laser scanning unit provides more accurate environmental geometry information, which complements the visual image data. When performing feature matching and positioning calculations, the high-precision spatial information of the laser point cloud can help the algorithm more accurately determine the position and posture changes of the device, and improve the accuracy and stability of relative positioning. For example, in areas where satellite signals cannot be covered, such as inside buildings or underground spaces, the laser point cloud-assisted VIO-SLAM algorithm can calculate the position and posture of the measuring device relative to known environmental features in real time, and achieve accurate relative positioning.

[0101] Robust estimation model: The robust estimation model is used to eliminate the influence of abnormal observations. In the actual positioning measurement process, abnormal observations may occur due to environmental interference, sensor failure, etc. If these abnormal values ​​are not handled, they will seriously affect the positioning accuracy. The robust estimation model identifies abnormal observations by statistically analyzing the observation data and assigns them a smaller weight, thereby reducing the influence of abnormal values ​​on the positioning results. The model can adaptively adjust the weights and determine the contribution of each observation in the positioning solution based on the reliability of the data. For example, when a sensor data fluctuates abnormally, the robust estimation model will automatically reduce the weight of the data in the calculation to ensure that the final positioning result is not affected by it and output stable and reliable three-dimensional space coordinates.

[0102] 3. Application scenario advantages:

[0103] High-precision positioning guarantee: The intelligent positioning solution module uses a variety of advanced positioning algorithms and models, combined with the advantages of multi-source data, to achieve high-precision positioning; whether in open field sites, complex urban building environments or indoor spaces, it can provide accurate location information for survey and layout operations; in the survey and layout of large bridge construction, this module can accurately determine the position of bridge piers, with the error controlled at the centimeter level or even smaller, meeting the strict requirements of engineering construction for high-precision positioning and ensuring the accuracy and quality of engineering construction;

[0104] Strong adaptability to complex environments: In construction environments, there are often complex situations such as signal blocking and electromagnetic interference, and traditional single positioning methods are often difficult to meet the needs; the intelligent positioning solution module integrates multiple positioning technologies, such as using RTK / PPK to deal with open environments, and using laser point cloud-assisted VIO-SLAM to process signal blocking areas, to achieve reliable positioning in various complex environments; when conducting construction surveys in urban areas with dense high-rise buildings, even if the satellite signal is severely blocked, the position of the measurement point can still be accurately obtained through the fusion positioning of laser point cloud and visual and inertial data, ensuring the smooth progress of the survey work;

[0105] Balancing real-time performance and stability: When performing positioning solutions, the module can process data in real time and quickly output positioning results; whether it is real-time RTK positioning or relative positioning based on VIO-SLAM assisted by laser point cloud, it can meet the real-time requirements of on-site operations; at the same time, the application of the robust estimation model ensures the stability of the positioning results, and can continuously output reliable three-dimensional spatial coordinates even in the presence of abnormal observations, providing stable positioning support for survey and layout operations and improving work efficiency and reliability.

[0106] In this embodiment, the layout path generation module communicates with the intelligent positioning solution module, including:

[0107] BIM model analysis unit: converts the engineering BIM model into three-dimensional path constraints in the construction coordinate system;

[0108] Path planning unit: Generates the layout path based on the multi-objective optimization algorithm;

[0109] Furthermore, the layout path generation module is a key component of the high-precision survey and layout intelligent positioning system based on multi-source data fusion. It is responsible for planning a reasonable layout path according to engineering requirements and positioning information. It plays an important role in the entire survey and layout process and directly affects the efficiency and accuracy of the layout work. Specifically:

[0110] 1. Module composition and data interaction:

[0111] BIM model parsing unit: This unit is a bridge between engineering design and actual layout operations. It takes the engineering BIM (Building Information Model) model as the core input and transforms the rich design information contained in the BIM model into three-dimensional path constraints in the construction coordinate system through professional parsing algorithms. The BIM model contains detailed information such as the precise geometric shape, spatial position, and relationship between components of the building. The parsing unit extracts the coordinates of the layout points, the layout sequence, the connection relationship between the points, and various constraints in the construction process, such as the location of obstacles and the boundaries of the construction area. These constraints provide a solid foundation for subsequent path planning and ensure that the generated layout path meets the engineering design requirements. For example, in large-scale construction projects, the BIM model parsing unit can accurately obtain the design coordinates of the building foundation piles and the surrounding underground pipelines and other obstacle information, thereby providing accurate constraint basis for layout path planning.

[0112] Path planning unit: The path planning unit uses the three-dimensional path constraints output by the BIM model analysis unit as the basis, combined with the actual situation of the construction site, and uses a multi-objective optimization algorithm to generate the optimal layout path; this unit works closely with the intelligent positioning solution module to obtain positioning data in real time so that the current equipment position and movement direction can be considered when planning the path; during the planning process, it comprehensively considers multiple factors, such as the total length of the path, the maximum curvature of the path, the safety score, etc., to achieve multi-objective optimization; its objective function is ,in, , is the total length of the path, a shorter path can reduce the layout time and cost; , is the maximum curvature of the path. A smaller curvature can ensure smoother movement of the equipment and reduce equipment loss; , is the safety scoring function, which ensures that the path avoids dangerous areas and guarantees construction safety; , is the dynamic weight coefficient; For path planning; at the same time, strict constraints are set, such as Ensure positioning accuracy, Limit the path slope to avoid the equipment operating on a path with a too steep slope. Ensure path security, where For path Positioning accuracy; For path Maximum slope of is the minimum value of the safety score; For the path The curvature of the moment; For the path Safety-related distances of points; is the number of points on the path; by continuously optimizing these parameters, the optimal layout path that meets engineering requirements is generated;

[0113] 2. Algorithms and optimization strategies:

[0114] The core of the multi-objective optimization algorithm: The multi-objective optimization algorithm is the core of the path planning unit. In practical applications, there are often conflicts between objectives such as path length, curvature, and safety. For example, the shortest path may pass through a dangerous area, while a safe path may be longer. The algorithm uses dynamic weight coefficients to To balance the importance of different goals; dynamically adjust the weight coefficient according to the actual situation of the project and the different needs of the construction stage; in the early stage of construction, the safety requirements are higher, and the weight of the safety score function can be appropriately increased; in the later stage of construction, if more attention is paid to efficiency, the path length weight can be increased; in this way, the optimal balance between multiple goals is found to generate a layout path that meets actual needs;

[0115] Iterative optimization process: When generating the layout path, the path planning unit adopts the iterative optimization method. First, an initial path is generated according to the initial constraints and weight coefficients. Then, the path is continuously adjusted and optimized through the algorithm to gradually reduce the value of the objective function. In each iteration, the influence of the position change of each point on the path on the objective function is considered, such as adjusting the order or position of a certain layout point, and observing the changes in the path length, curvature and safety. After multiple iterations, the path gradually approaches the optimal solution. When encountering complex construction environments or multiple conflicting constraints, through repeated iterative optimization, a reasonable layout path that takes multiple factors into consideration can be found.

[0116] 3. Application scenario advantages:

[0117] Improve the efficiency of layout: By accurately analyzing the BIM model and optimizing the path planning, the layout path generation module can plan the shortest and most convenient layout path; reduce the moving distance and time of the equipment at the construction site, avoid unnecessary detours and repeated operations, and greatly improve the efficiency of the layout work; in the survey and layout of large sites, compared with the traditional manually planned path, the path generated by this module can save a lot of time and speed up the construction progress;

[0118] Ensure construction safety: The layout path generated by the module fully considers safety factors. Through the safety scoring function and related constraints, it ensures that the path avoids dangerous areas of the construction site, such as deep foundation pits, high-voltage equipment, and high-altitude working areas. This effectively reduces the safety risks during the construction process and ensures the safety of construction personnel and equipment. In some complex construction sites, the module can provide safe and reliable path planning for layout operations and reduce the occurrence of safety accidents.

[0119] Ensure the accuracy of layout: Based on the precise analysis of the BIM model and strict positioning accuracy constraints, the generated layout path can ensure that the positioning accuracy of the layout points is within the specified range; this makes the actual layout results highly consistent with the design requirements, avoids construction deviations caused by layout errors, and improves the quality of the project; in construction projects with extremely high precision requirements, such as bridges, high-rise buildings and other projects, this module can ensure the accurate position of each layout point, providing a reliable foundation for subsequent construction.

[0120] In this embodiment, the wireless communication module is connected to each module and configured as follows:

[0121] 5G communication unit: upload data to the cloud platform at a rate of ≥1Gbps;

[0122] LoRa self-organizing network unit: establish a Mesh network in the signal blocking area to achieve direct communication between devices;

[0123] Furthermore, the wireless communication module is an important component of the high-precision survey and layout intelligent positioning system based on multi-source data fusion. It is like the "nerve network" of the system, responsible for transmitting data between modules and between the system and external devices, ensuring the smooth flow of information, and plays a vital role in the efficient operation of the system. By integrating multiple communication technologies, the module can adapt to different construction environments and data transmission requirements, and provide stable and efficient data communication guarantees for survey and layout work. Specifically:

[0124] 1. Communication technology integration:

[0125] 5G communication unit: The 5G communication unit is an important component of the wireless communication module. It has the advantages of high speed, low latency and large capacity communication. In this system, the 5G communication unit uploads data to the cloud platform at a rate of ≥1Gbps. This high-speed data transmission capability enables the system to quickly upload a large amount of collected data (such as 3D laser point cloud data, visual image data, etc.) to the cloud for storage, analysis and processing. During the construction survey, a large amount of real-time data collected on site can be quickly transmitted to the cloud server through the 5G network. Professionals can analyze these data in real time on the cloud to find problems and make decisions in a timely manner. At the same time, the low latency feature ensures the timeliness of data transmission. For operations that require real-time response (such as remote control equipment adjustment and positioning, etc.), 5G communication can ensure that instructions are transmitted quickly and accurately, improving the real-time and response speed of the system. In addition, 5G's large-capacity communication capability can meet the needs of multiple devices communicating online at the same time and adapt to complex construction site environments.

[0126] LoRa self-organizing unit: Signal blocking is common at construction sites. For example, traditional communication methods may be affected in areas such as inside buildings, underground spaces, or areas blocked by large construction equipment. To address this problem, LoRa self-organizing units establish Mesh networks in signal blocking areas to achieve direct communication between devices. LoRa technology has the characteristics of low power consumption and long-distance transmission, and can achieve stable communication in complex environments. Through the self-organizing characteristics of the Mesh network, devices can automatically establish connections and transmit data. Even if some nodes fail or signals are blocked, data can be forwarded through other nodes to ensure communication reliability. For example, when surveying the interior of a large building, multiple positioning devices can establish a Mesh network through LoRa self-organizing units to achieve real-time data sharing and interaction, ensuring that each device can work together to improve the efficiency and accuracy of survey and layout.

[0127] 2. Data transmission and network management:

[0128] Data transmission optimization: The wireless communication module classifies and transmits data according to the type and urgency of the data; for data with high real-time requirements (such as real-time positioning data, equipment control instructions, etc.), it is preferentially transmitted through the 5G communication unit to ensure rapid data transmission and real-time response of the system; for some non-real-time data (such as historical data backup, large-scale point cloud data storage, etc.), it can be selected to be transmitted through 5G or LoRa self-organizing network units when the network load is low, so as to reasonably use network resources and avoid network congestion; at the same time, the module adopts data compression and encryption technology to reduce the amount of data transmission, improve transmission efficiency, and ensure data security while ensuring data integrity, so as to prevent data from being stolen or tampered with during transmission;

[0129] Network switching and management: During the actual construction process, the equipment may move in different areas, causing changes in the communication environment; the wireless communication module has an intelligent network switching function, which can automatically switch between 5G and LoRa self-organizing networks according to factors such as signal strength and network quality; when the equipment is in an area with good signals, it will give priority to using the 5G network for high-speed data transmission; when the equipment enters a signal-blocking area, it will automatically switch to the LoRa self-organizing network to ensure the continuity of communication; in addition, the module also manages the network in real time, monitors network connection status, signal strength, data transmission rate and other parameters, promptly detects and resolves network failures, and ensures the stability of system communication;

[0130] 3. Application scenario advantages:

[0131] Adapt to complex construction environments: The construction site environment is complex, and signal interference and obstruction are prominent. The wireless communication module can achieve reliable communication in different environments by integrating 5G and LoRa self-organizing network technology. Whether it is using the high-speed communication advantage of 5G in an open field, or using the flexible networking capability of LoRa self-organizing network in a signal-blocked area, it can ensure that the data transmission between the modules of the system is not affected, ensuring the smooth progress of the survey and layout work. In urban construction projects, there are many high-rise buildings around, and the signal is easily blocked. At this time, the LoRa self-organizing network unit can play an important role to ensure smooth communication between equipment. In open fields with good signals, the 5G communication unit can quickly transmit large amounts of data and improve work efficiency.

[0132] Support remote monitoring and collaboration: With the high speed and low latency characteristics of 5G communication, the wireless communication module supports remote monitoring and collaboration functions. Managers and technical experts can remotely view the survey and layout of the construction site in real time through the cloud platform, obtain real-time positioning data, equipment status information, etc., and remotely collaborate with on-site staff. When encountering complex problems, experts can remotely guide on-site operations, solve problems in a timely manner, and improve work efficiency and quality. This remote monitoring and collaboration method breaks geographical restrictions, makes full use of the professional resources of all parties, and improves the overall efficiency of the system.

[0133] Improve system integration and flexibility: The integrated design of the wireless communication module integrates multiple communication technologies, reduces the external connection devices of the system, and improves the system integration; at the same time, its flexible network configuration and adaptive capabilities enable the system to be quickly adjusted and deployed according to different construction needs and environments; the wireless communication module can be easily applied in construction projects of different sizes and types, providing convenient conditions for the wide application of the system.

[0134] In this embodiment, the visualization terminal module receives the layout path and real-time positioning data, including:

[0135] AR display unit: superimposes and displays theoretical / actual position deviation vectors;

[0136] Dynamic deviation correction unit: performs three-level feedback control to achieve real-time path correction;

[0137] Furthermore, the visualization terminal module is a key part directly facing users in the high-precision survey and layout intelligent positioning system based on multi-source data fusion. As a window for interaction between the system and users, it presents data and provides operation functions in an intuitive and convenient visualization way, which significantly improves user experience and work efficiency, making survey and layout work more intelligent and efficient. Specifically:

[0138] 1. Hardware and data interaction basis:

[0139] Data receiving interface: Visual terminal module, which has stable data interaction with intelligent positioning solution module and layout path generation module; this interface can efficiently receive real-time positioning data and layout path data, ensuring that the information obtained by the terminal is always synchronized with the actual situation on site; whether it is the real-time position change of the positioning equipment on the construction site, or the layout path dynamically adjusted according to the project requirements, it can be transmitted to the visual terminal module in a timely and accurate manner, providing users with the latest work information;

[0140] Display and interactive hardware: The module is equipped with a high-performance display device with high resolution and good color reproduction, which can clearly and delicately display various data information; at the same time, combined with interactive hardware such as touch screen, users can interact with the terminal naturally through simple touch operations, realize data viewing, analysis and call of system functions, which greatly improves the convenience and efficiency of operation;

[0141] 2. Analysis of core functions:

[0142] AR display unit: The AR display unit is one of the core functions of the visualization terminal module. It uses augmented reality technology to integrate virtual information with real scenes for display. In survey and layout work, the unit can overlay and display the theoretical / measured position deviation vector. Users can intuitively see the deviation between the actual measured position and the theoretical layout position through AR devices (such as AR glasses or tablets that support AR functions). This intuitive visualization presentation method allows users to quickly determine whether the current position is accurate, as well as the direction and distance that need to be adjusted without complex calculations and analysis. When laying out the foundation pile position of a building, construction workers wearing AR devices can directly see the deviation vector between the current pile position measurement point and the designed pile position, clearly understand the deviation size and direction, and quickly make adjustments to improve layout accuracy and work efficiency.

[0143] Dynamic deviation correction unit: Based on the deviation information provided by the AR display unit, the dynamic deviation correction unit performs three-level feedback control to achieve real-time path correction; when the deviation value Δ between the theoretical position and the measured position is in different ranges, the system will adopt different control strategies;

[0144] First level control: When hour, The interface displays a yellow warning box. At this time, the deviation is in a relatively small range, but it still needs attention. The appearance of the yellow warning box reminds the user that there is a certain deviation in the current position, and it needs attention and timely adjustment to avoid further expansion of the deviation.

[0145] Secondary control: When When the system is in operation, the system not only displays the deviation information on the AR interface, but also generates local path correction instructions and notifies the user through vibration alarms. The vibration alarm can attract the user's high attention, while the local path correction instructions provide the user with specific operation guidance, helping the user to quickly adjust the device position and return the measurement point to the correct layout path.

[0146] Level 3 control: When the deviation is large, it may have a great impact on the quality of the project. At this time, the system starts the system self-check and freezes the positioning output until manual confirmation. The system self-check can timely discover possible problems, such as sensor failure, positioning algorithm abnormality, etc., and freezing the positioning output avoids the further transmission and use of erroneous data. The manual confirmation link ensures that the positioning and layout work will continue only after the problem is properly handled, ensuring the accuracy and safety of the project.

[0147] in, is the deviation between the theoretical position and the measured position;

[0148] 3. Application scenario advantages:

[0149] Improve construction accuracy and efficiency: The AR display and dynamic deviation correction function of the visualization terminal module enable construction personnel to grasp the deviation between the measured position and the theoretical position in real time and intuitively, and make timely adjustments; this greatly reduces the rework and correction work caused by position deviation, and improves construction accuracy and efficiency; in large-scale construction projects, accurate layout work can ensure that the position and size of each part of the building meet the design requirements, avoid structural problems caused by layout errors, shorten the construction period, and reduce construction costs;

[0150] Lowering the threshold for operation and training costs: With an intuitive visual interface and simple operation methods, even relatively inexperienced construction workers can quickly get started; they do not need to deeply understand complex positioning data and algorithms, but only need to operate according to the deviation information and system prompts displayed by AR; this lowers the threshold for operation and reduces dependence on professional technicians, while also reducing training costs, allowing new employees to adapt to their jobs more quickly;

[0151] Enhanced construction safety and quality control: The three-level control strategy of the dynamic deviation correction unit, especially the system self-check and frozen positioning output functions that are activated when the deviation is large, effectively avoids construction accidents caused by incorrect operation or equipment failure, and ensures construction safety; at the same time, real-time deviation monitoring and correction ensure that the construction process always meets the design requirements, improves the level of project quality control, and provides a strong guarantee for the quality and safety of construction projects.

[0152] In this embodiment, when multiple devices work together, a distributed consensus algorithm is used to synchronize positioning data: ,in, is the convergence rate parameter, For equipment exist Positioning data at the moment; For equipment exist Positioning data at the moment; For equipment Neighboring devices exist Positioning data at the moment; is the Kalman gain; For equipment The measured value of is the measurement matrix;

[0153] Furthermore, in the high-precision survey and layout intelligent positioning system based on multi-source data fusion of the present invention, the collaborative work of multiple devices is the key link to improve the overall performance of the system and achieve efficient and accurate survey and layout; the collaborative work of multiple devices covers data synchronization, task allocation and collaboration, and fault tolerance processing. Through these mechanisms, each device cooperates closely to ensure the stable operation of the positioning system. Specifically:

[0154] Data Synchronization:

[0155] Principle of distributed consensus algorithm: When multiple devices work together, a distributed consensus algorithm is used to synchronize positioning data. The formula is: , where the convergence rate parameter Determines the speed of data synchronization, which controls the frequency of information exchange between devices; when When it is larger, the information between devices is updated faster, but it may increase the network burden; conversely, the information update speed is slow, but the network pressure is less; in actual application, it needs to be reasonably adjusted according to the network conditions and positioning accuracy requirements;

[0156] Data synchronization process: In actual scenarios, each device continuously collects data and exchanges positioning information with adjacent devices through the algorithm;

[0157] Task allocation and collaboration:

[0158] Task allocation strategy: The system allocates tasks based on the functional characteristics and actual working status of each device. For example, laser scanning equipment is good at obtaining high-precision three-dimensional point cloud data. When conducting complex terrain or building structure surveys, it is mainly responsible for collecting detailed spatial data. GNSS positioning equipment has high positioning accuracy and can obtain absolute coordinates. It plays a key role in open areas or determining the overall position benchmark. Inertial measurement units and visual sensor units assist in positioning and supplement environmental information when satellite signals are interfered with or in indoor environments.

[0159] Collaborative operation process: Taking a large bridge construction project as an example, in the early stage of site survey, the GNSS positioning equipment first determines the approximate location and provides reference coordinates for other equipment; then, the laser scanning equipment scans the bridge site area to obtain three-dimensional point cloud data of the terrain and existing buildings; during the construction process, if the satellite signal is blocked, the inertial measurement unit and the visual sensor unit begin to play a role to ensure the continuity and accuracy of positioning; the data collected by each device is transmitted to the data fusion processing module in real time, and after fusion processing, it provides more accurate data for the intelligent positioning solution module, thereby realizing high-precision survey and layout operations;

[0160] Fault tolerance processing:

[0161] Fault detection and notification: In collaborative work, each device monitors the status of itself and adjacent devices in real time. Once a device fails (such as sensor failure, communication interruption, etc.), the faulty device will immediately send a fault notification to other devices. At the same time, the system's sensor failure detection unit will also assist in determining device failures through abnormal changes in monitoring data, and take corresponding measures in a timely manner.

[0162] Implementation of fault-tolerant strategies: When a device fails, the system adjusts its working strategy according to the type and severity of the failure. If only a sensor fails, the system will adjust the data fusion algorithm based on the strategy of the sensor failure detection unit to reduce reliance on the data from the faulty sensor and rely on other normal sensors to maintain positioning and data collection functions. If a device fails completely, the system will reallocate tasks and reasonably allocate the tasks of the faulty device to other normal devices to ensure that the entire system can continue to operate and complete the survey and layout tasks.

[0163] The high-precision survey and layout intelligent positioning method based on multi-source data fusion relies on a highly integrated and powerful system. Through the collaborative collection, deep fusion, intelligent solution, path planning and real-time feedback correction of multi-source data, high-precision positioning in the process of building survey and layout is achieved, which significantly improves the efficiency and quality of engineering construction. The specific steps include:

[0164] Step S1, multi-source data acquisition and preprocessing: using the laser scanning unit, GNSS positioning unit, inertial measurement unit and visual sensor unit in the data acquisition module, synchronously collect the three-dimensional laser point cloud data, satellite positioning data, inertial navigation data and visual image data of the construction site; these sensors are like the "perception tentacles" of the system, acquiring information on the construction site from different dimensions; the collected data will be timestamped to ensure the consistency of each data source in the time dimension, providing an accurate time reference for subsequent fusion processing; at the same time, the raw data will be preprocessed by denoising, filtering and other operations to remove noise and outliers caused by sensor errors, environmental interference, etc., to improve data quality;

[0165] Step S2, data fusion and spatiotemporal alignment: The data fusion processing module uses the Beidou timestamp and ORB-SLAM3 time axis dual reference alignment algorithm to establish a unified spatiotemporal coordinate system and accurately match multi-source data in time and space; uses the improved hybrid Kalman filter-graph attention network model to perform feature-level fusion of multi-source data; through this model, the characteristics and associations of each sensor data are comprehensively considered, the key information in the data is mined, and a fused point cloud map is generated; in the fusion process, the calculation of cross-modal attention weights uses a multi-sensor feature cross-validation algorithm to ensure that the model can adaptively focus on the key information in different sensor data and improve the fusion effect;

[0166] Step S3, intelligent positioning solution: The intelligent positioning solution module receives the fused point cloud map, combines the data of the inertial measurement unit, and uses a variety of algorithms to achieve high-precision positioning; uses the RTK / PPK dual-mode differential positioning algorithm to solve the absolute coordinates, and flexibly switches the RTK (real-time dynamic differential positioning) and PPK (post-processing dynamic differential positioning) modes according to the satellite signal conditions at the construction site to ensure that accurate absolute coordinates can be obtained in various scenarios; uses the laser point cloud-assisted VIO-SLAM algorithm to calculate relative positioning coordinates. When the satellite signal is blocked or lost, such as inside a building or in a canyon, the algorithm achieves continuous positioning and tracking by fusing visual and inertial data; applies an anti-error estimation model to eliminate the influence of abnormal observations, ensure the stability and reliability of the positioning results, and finally output accurate three-dimensional spatial coordinates;

[0167] Step S4, staking path generation and optimization: The BIM model parsing unit in the staking path generation module converts the engineering BIM model into three-dimensional path constraints in the construction coordinate system, extracts the coordinates of the staking points, the staking order, the connection relationship between the points, and the constraints in the construction process; the path planning unit generates the staking path based on these constraints using a multi-objective optimization algorithm; the algorithm comprehensively considers factors such as the total length of the path, the maximum curvature of the path, and the safety score, and dynamically adjusts the weight coefficient to find the optimal balance between multiple objectives to generate the optimal staking path that meets the engineering requirements;

[0168] Step S5, real-time feedback and dynamic deviation correction: the visualization terminal module receives the layout path and real-time positioning data; the AR display unit superimposes and displays the theoretical / measured position deviation vector, allowing the user to intuitively understand the deviation between the current position and the theoretical layout position; the dynamic deviation correction unit performs three-level feedback control according to the deviation value; when the deviation value is small, the AR interface displays a yellow warning box; when the deviation value is slightly larger, a local path correction instruction is generated and a vibration alarm is issued; when the deviation value is too large, the system self-check is started and the positioning output is frozen until manual confirmation; through this real-time feedback and dynamic deviation correction mechanism, it is ensured that the layout process always maintains high precision;

[0169] Step S6, multi-device collaboration and data synchronization: When multiple devices work together, a distributed consensus algorithm is used to synchronize positioning data to ensure the real-time and consistency of the positioning data of each device; tasks are assigned according to the functional characteristics and actual working status of each device, such as laser scanning equipment is responsible for collecting spatial data, and GNSS positioning equipment determines absolute coordinates, etc.; devices collaborate with each other, share data in real time, and jointly complete survey and layout tasks; when a device fails, the system can detect and adjust the working strategy in time, and ensure that the entire system can still operate normally by adjusting the data fusion algorithm or reallocating tasks.

[0170] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. High-precision survey and layout intelligent positioning system based on multi-source data fusion, characterized by: include: Data acquisition module: including laser scanning unit, GNSS positioning unit, inertial measurement unit and visual sensor unit, synchronously collects 3D laser point cloud data, satellite positioning data, inertial navigation data and visual image data of the construction site, and outputs raw data streams with time stamp alignment; Data fusion processing module: connected to the data acquisition module, and executes in sequence: Multi-source data spatiotemporal alignment unit: uses Beidou timestamp and ORB-SLAM3 time axis dual reference alignment algorithm to establish a unified spatiotemporal coordinate system; Feature-level fusion unit: executes the improved hybrid Kalman filter-graph attention network model to generate a fused point cloud map; Sensor failure detection unit: monitors the health status of each sensor in real time and dynamically adjusts the fusion strategy; Intelligent positioning solution module: receives the fused point cloud map and connects to the inertial measurement unit of the data acquisition module, and is configured as follows: The absolute coordinates are calculated through the RTK / PPK dual-mode differential positioning algorithm; Calculate relative positioning coordinates through the VIO-SLAM algorithm assisted by laser point cloud; Apply the robust estimation model to eliminate the influence of abnormal observations and output three-dimensional space coordinates; Lofting path generation module: communicates with the intelligent positioning solution module, including: BIM model analysis unit: converts the engineering BIM model into three-dimensional path constraints in the construction coordinate system; Path planning unit: Generates the layout path based on the multi-objective optimization algorithm; Wireless communication module: connected with each module, configured as: 5G communication unit: upload data to the cloud platform at a rate of ≥1Gbps; LoRa self-organizing network unit: establish a Mesh network in the signal blocking area to achieve direct communication between devices; Visual terminal module: receives the layout path and real-time positioning data, including: AR display unit: superimposes and displays theoretical / actual position deviation vectors; Dynamic deviation correction unit: performs three-level feedback control to achieve real-time path correction.

2. The high-precision survey and setting-out intelligent positioning system based on multi-source data fusion according to claim 1 is characterized by: The feature-level fusion unit adopts an improved hybrid Kalman filter-graph attention network model, and its state update formula is: ,in, for Fusion state estimates at all times; , is the state transfer matrix of online calibration; A Kalman gain matrix that is dynamically adjusted based on sensor confidence; is the graph network coupling coefficient; is the cross-modal attention weight; is the trainable parameter matrix; for The state value at the moment; for Control input at all times; for The measured value at the moment; is the measurement matrix; is the set of neighbor nodes of sensor node i; Neighbor node The status value of .

3. The high-precision survey and setting-out intelligent positioning system based on multi-source data fusion according to claim 2 is characterized by: The cross-modal attention weight is calculated using a multi-sensor feature cross-validation algorithm: ,in, , is the query / key vector projection matrix; is the Euclidean distance of sensor nodes; is the environmental interference attenuation factor; , Encode vector for sensor type; For sensor nodes The eigenvector of is the dimension of the feature vector, ; is the normalized exponential function.

4. The high-precision survey and setting-out intelligent positioning system based on multi-source data fusion according to claim 1 is characterized by: The multi-objective optimization algorithm executed by the path planning unit includes: ； Constraints: ; in, , is the total length of the path; , is the maximum curvature of the path; , is the safety scoring function; , is the dynamic weight coefficient; To plan a route; For path Positioning accuracy; For path Maximum slope of is the minimum value of the safety score; For the path The curvature of the moment; For the path Safety related distance of each point; is the number of points on the path.

5. The high-precision survey and setting-out intelligent positioning system based on multi-source data fusion according to claim 2 is characterized by: When the sensor failure detection unit executes the fault-tolerant control strategy: When the first When a sensor fails, the fusion algorithm is automatically adjusted to: ,in, is the failure indicator factor, and , Indicates A sensor completely failed; It is the set of neighbor nodes that exclude failed sensors.

6. The high-precision survey and setting-out intelligent positioning system based on multi-source data fusion according to claim 1 is characterized by: The dynamic correction unit performs three-level control, including: Primary Control , The interface displays a yellow warning box; Secondary Control , generate local path correction instructions and vibrate alarms; Three-level control , start the system self-check and freeze the positioning output until manual confirmation; in, is the deviation between the theoretical position and the measured position.

7. The high-precision survey and setting-out intelligent positioning system based on multi-source data fusion according to claim 1 is characterized by: When multiple devices work together, a distributed consensus algorithm is used to synchronize positioning data: ,in, is the convergence rate parameter, For equipment exist Positioning data at the moment; For equipment exist Positioning data at the moment; For equipment Neighboring devices exist Positioning data at the moment; is the Kalman gain; For equipment The measured value of is the measurement matrix.

8. A high-precision survey and layout intelligent positioning method based on multi-source data fusion, characterized by: The high-precision survey and layout intelligent positioning method is based on the high-precision survey and layout intelligent positioning system based on multi-source data fusion described in any one of claims 1-7.

Images