Construction safety early warning method and system based on mixed reality

Abstract

The invention belongs to the technical field of building construction safety management, and provides a construction safety early warning method and system based on mixed reality, and the method comprises the steps: converting the longitude and latitude of geographic coordinate system data into plane coordinates, and converting the plane coordinates into a local coordinate system of a spatial mapping grid formed by simultaneous positioning and a map construction technology; through coordinate transformation, geographic coordinate system data can be accurately mapped into a model space coordinate system of the mixed reality equipment, the limitation that external high-precision positioning data are difficult to integrate in a traditional mixed reality technology is broken through, and high-precision equipment positioning is realized; meanwhile, an initial weight value is distributed to each grid unit in the space mapping grid, the weight is dynamically adjusted to change the size of the grid, dynamic management of space mapping grid data is achieved, detailed real space mapping grid data can be reserved after initial alignment is carried out through the space mapping capacity of mixed reality equipment, and the space mapping capacity of the mixed reality equipment is improved. And a data basis is provided for subsequent dynamic collision detection.

Description

Technical Field

[0001] The present invention belongs to the technical field of building construction safety management, and in particular relates to a construction safety early warning method and system based on mixed reality. Background Art

[0002] As construction projects continue to expand and become more complex, safety management at construction sites faces increasing challenges. Traditional safety management methods often rely on manual inspections and empirical judgments, making it difficult to detect and prevent potential safety hazards in a timely manner. In recent years, the rapid development of mixed reality technology and the Internet of Things technology has provided new possibilities for solving this problem.

[0003] The existing construction safety warning methods based on mixed reality technology mainly realize the alignment of virtual models with the actual environment in order to view the relative position of the model in the real space. However, in order to avoid interfering with the user's actual field of view, the real space mapping grid is usually hidden after the spatial alignment is completed. Therefore, this part of the data mapping generally no longer plays a role, affecting the subsequent dynamic collision detection and unable to continuously perform high-precision spatial analysis and risk assessment; in addition, the existing mixed reality equipment uses SLAM (simultaneous localization and mapping) technology for deep space recognition, which makes it difficult to effectively interact with other high-precision positioning equipment such as GPS-RTK systems, which limits its application in complex and large-scale construction environments and affects the positioning accuracy of the equipment. Summary of the invention

[0004] In order to solve the above problems, the present invention proposes a construction safety early warning method and system based on mixed reality. When mapping geographic coordinate system data to the spatial coordinate system of the mixed reality model, the present invention converts the longitude and latitude of the geographic coordinate system data into plane coordinates, and converts the plane coordinates into the local coordinate system of the spatial mapping grid formed by the simultaneous positioning and map construction technology; the real-time position data of the construction equipment obtained by the high-precision positioning system can be accurately mapped to the model space coordinate system of the mixed reality device through coordinate conversion, breaking through the limitation that the traditional mixed reality technology is difficult to integrate external high-precision positioning data, and realizing high-precision equipment positioning; at the same time, an initial weight value is assigned to each grid cell in the spatial mapping grid, and the weight is dynamically adjusted to change the grid size, so as to realize dynamic management of the spatial mapping grid data, and can retain detailed real space mapping grid data after initial alignment using the spatial mapping capability of the mixed reality device, so as to provide a data basis for subsequent dynamic collision detection, and can continuously perform high-precision spatial analysis and risk assessment without affecting the actual field of view of the user.

[0005] In order to achieve the above object, the present invention is implemented through the following technical solutions:

[0006] In a first aspect, the present invention provides a construction safety early warning method based on mixed reality, comprising:

[0007] Obtain surveying and mapping data of the construction site, as well as geometric dimension data and geographic coordinate system data of relevant mechanical equipment;

[0008] Use the acquired surveying and mapping data and geometric dimension data to build a mixed reality model;

[0009] Mapping geographic coordinate system data to the spatial coordinate system of the mixed reality model; wherein, during the mapping process, the longitude and latitude of the geographic coordinate system data are converted into plane coordinates, and the plane coordinates are converted into a spatial mapping grid local coordinate system formed by simultaneous positioning and mapping technology;

[0010] Assign an initial weight value to each grid cell in the space mapping grid, dynamically adjust the weight to change the grid size, and dynamically manage the space mapping grid data; perform collision detection based on the dynamically managed space mapping grid data to obtain collision detection results;

[0011] Conduct risk assessment and safety warning based on collision detection results.

[0012] Furthermore, the geographic coordinate system data is subjected to noise filtering and outlier detection; the converted position data is smoothed using a Kalman filter algorithm, and the position is predicted in combination with the device's motion model.

[0013] Further, in the spatial mapping grid, a known reference point is selected; and a reference point is obtained in the model space of the device in the mixed reality model to be used as a reference for alignment and transformation;

[0014] Calculate the plane coordinate difference between the origin of the coordinate system and the reference point, and apply it to other points; get the angle difference based on the known azimuth of the reference point in the two coordinate systems;

[0015] According to the information of the reference points, the spatial model and the virtual model are automatically synchronized and aligned, and the structural proportion of the virtual model is maintained. In the case of coordinate drift, the reference points are automatically calibrated and drift compensation is performed to achieve consistency between the virtual model and the spatial model.

[0016] Furthermore, the spatial data mapping is gridded and dynamically weighted, the maximum grid size and the corresponding first weight are set according to the minimum distance from the predetermined trajectory, and the second weight is dynamically set according to the distance to the motion device model; the closer the distance, the smaller the grid size, and the larger the first weight and the second weight.

[0017] Further, the trajectory distance weight of the grid unit is calculated using a preset trajectory weight function, and the grid size is adjusted; if the minimum distance from the unit to the trajectory is less than a preset threshold, the grid size is reduced, and the first weight is increased; the device model distance weight is calculated using a preset device distance weight function; if the minimum distance from the unit to the device model is less than the threshold of the device model, the grid size is reduced again, and the second weight is increased; the final weight value of the grid unit is calculated to be the product of the first weight and the second weight;

[0018] Sort the grid cells according to the final weight. The grid cells with higher weight have higher priority. The grid cells with final weight higher than the preset weight are placed at the front of the priority queue for priority loading. The weight and priority of the grid cells are updated in real time every time the device moves. For the grid cells in the direction of the device's advance, a rolling window is set, and the trajectory distance and device distance are dynamically recalculated within the rolling window, and their weights are adjusted.

[0019] Use grid cells whose final weights are higher than the preset weights for collision detection. When the device is moving, first load the data of grid cells whose final weights are higher than the preset weights, and ignore or delay grid cells with other weights. Each time it is called, only retrieve the data of grid cells along the trajectory whose final weights are higher than the preset weights. Set a cache mechanism within a certain range to cache the data of grid cells near the trajectory. When the device moves to a new position, recalculate the weights of each grid cell in the current scrolling window, and adjust the priority queue according to the updated weights.

[0020] Furthermore, the separating axis theorem is used for collision detection, including preset trajectory mode and real-time position mode;

[0021] In the preset trajectory mode, based on the construction plan, a three-dimensional trajectory of the equipment movement is generated in the modeling stage; a virtual space scan is performed on-site along the preset trajectory to obtain a space mapping grid; collision detection is run on the equipment based on the scanning results; in the real-time position mode, the geographic coordinate system data is used to continuously update the position of the moving equipment in the mixed reality space.

[0022] In a second aspect, the present invention further provides a construction safety early warning system based on mixed reality, comprising:

[0023] The data acquisition module is configured to: obtain surveying and mapping data of the construction site, as well as geometric dimension data and geographic coordinate system data of relevant mechanical equipment;

[0024] The mixed reality model building module is configured to: build a mixed reality model using the acquired surveying and mapping data and geometric dimension data;

[0025] A coordinate system data mapping module is configured to: map geographic coordinate system data to a spatial coordinate system of a mixed reality model; wherein, during the mapping process, the longitude and latitude of the geographic coordinate system data are converted into plane coordinates, and the plane coordinates are converted into a spatial mapping grid local coordinate system formed by simultaneous positioning and mapping technology;

[0026] The collision detection module is configured to: assign an initial weight value to each grid cell in the space mapping grid, dynamically adjust the weight to change the grid size, and dynamically manage the space mapping grid data; perform collision detection based on the dynamically managed space mapping grid data to obtain a collision detection result;

[0027] The safety warning module is configured to: perform risk assessment and safety warning based on the collision detection results.

[0028] In a third aspect, the present invention further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the construction safety warning method based on mixed reality described in the first aspect.

[0029] In a fourth aspect, the present invention further provides an electronic device comprising a memory, a processor, and a computer program stored in the memory and capable of running on the processor, wherein when the processor executes the program, the steps of the construction safety warning method based on mixed reality described in the first aspect are implemented.

[0030] In a fifth aspect, the present invention further provides a computer program product, which includes a computer program. When the computer program is executed by a processor, the steps of the construction safety warning method based on mixed reality described in the first aspect are implemented.

[0031] Compared with the prior art, the present invention has the following beneficial effects:

[0032] When the geographic coordinate system data is mapped to the spatial coordinate system of the mixed reality model in the present invention, the longitude and latitude of the geographic coordinate system data are converted into plane coordinates, and the plane coordinates are converted into the local coordinate system of the spatial mapping grid formed by the simultaneous positioning and map construction technology; the real-time position data of the construction equipment obtained by the high-precision positioning system can be accurately mapped to the model space coordinate system of the mixed reality device through coordinate conversion, breaking through the limitation that the traditional mixed reality technology is difficult to integrate external high-precision positioning data, and realizing high-precision equipment positioning; at the same time, an initial weight value is assigned to each grid cell in the spatial mapping grid, and the weight is dynamically adjusted to change the grid size, so as to realize dynamic management of the spatial mapping grid data, and after the initial alignment is performed using the spatial mapping capability of the mixed reality device, detailed real space mapping grid data can be retained to provide a data basis for subsequent dynamic collision detection, and high-precision spatial analysis and risk assessment can be continuously performed without affecting the user's actual field of view.

[0033] The present invention achieves millisecond-level real-time data processing and centimeter-level spatial positioning accuracy through the combination of mixed reality technology and high-precision GPS-RTK positioning, greatly improving the timeliness and accuracy of safety warnings. Using mixed reality technology, virtual safety information is seamlessly integrated into the actual construction environment, providing an intuitive and immersive safety management experience, greatly improving the efficiency of identifying safety hazards. It can predict potential safety risks and achieve a transition from passive response to active prevention. It is suitable for diverse engineering environments such as high-rise buildings, bridges, and tunnels, and improves the overall safety of different types of construction projects.

[0034] The remote collaboration function of the present invention breaks geographical restrictions, makes full use of expert resources, and significantly improves the efficiency of problem solving. It ensures the integrity and non-tamperability of safety records, and provides reliable data support for safety management. By identifying and preventing safety hazards in advance, potential accident losses and downtime costs are greatly reduced. Through innovative GPS-RTK data processing and coordinate conversion technology, real-time positioning of GPS-RTK-based devices in spatial mapping of MR mixed reality devices is achieved, which greatly improves the spatial perception ability in the mixed reality environment, which enables the system to accurately predict and evaluate potential risks in complex construction environments.

[0035] GPS-RTK data and mixed reality data have always been two independent coordinate systems, which has largely limited the application scenarios of mixed reality devices. The present invention can integrate the two, thereby greatly improving the application scenarios of mixed reality devices, allowing them to move out of indoors and be used on a larger scale.

[0036] The present invention realizes dynamic management based on the dynamic division and weight assignment of the space mapping grid, can adapt to the dynamic changes of the equipment on the construction site in real time, significantly improves the real-time and accuracy of collision detection, and enables the system to cope with the rapidly changing environment of the construction site. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] The drawings in the specification that constitute a part of this embodiment are used to provide a further understanding of this embodiment. The schematic embodiments of this embodiment and their descriptions are used to explain this embodiment and do not constitute improper limitations on this embodiment.

[0038] Figure 1 This is an overall architecture diagram of the system of Embodiment 1 of the present invention;

[0039] Figure 2 The flowchart of GPS-RTK data processing and coordinate conversion of Embodiment 1 of the present invention is as follows;

[0040] Figure 3 This is a flow chart of dynamic management optimization of MR device space mapping grid according to Embodiment 1 of the present invention;

[0041] Figure 4 This is a schematic diagram of space partition optimization during dynamic collision detection in Example 1 of the present invention. DETAILED DESCRIPTION

[0042] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0043] It should be noted that the following detailed descriptions are exemplary and are intended to provide further explanation of the present application. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present application belongs.

[0044] Embodiment 1:

[0045] At present, some technologies have tried to apply emerging technologies to the field of construction safety management. For example, using BIM technology and Revit secondary development technology to manage high-rise prefabricated building construction sites. It can predict possible dangers before construction and formulate corresponding countermeasures; however, this method is mainly limited to static model analysis and lacks real-time dynamic monitoring capabilities. By constructing a three-dimensional scene that integrates the underground engineering site and the geotechnical engineering information model, the risk analysis capability is improved; however, it is mainly aimed at underground engineering and does not make full use of high-precision positioning technology, and its application in complex ground construction environments is limited. Use drones to obtain remote sensing images, obtain equipment positioning coordinates through target recognition networks, and generate inspection paths in virtual roaming three-dimensional space; although this method improves inspection efficiency, it is mainly aimed at the inspection of fixed equipment and has limited real-time safety warning capabilities in dynamic construction environments. However, these existing technologies still have shortcomings in real-time, accuracy and interactivity; for example, although traditional video surveillance systems can observe construction sites in real time, it is difficult to conduct in-depth and accurate spatial analysis and risk prediction; and although simple IoT sensor networks can collect a large amount of data, they lack intuitive analysis results and visualization and human-computer interaction capabilities compared with scenes.

[0046] As described in the background technology, existing mixed reality technologies are mainly used to align virtual models with the actual environment in order to view the relative position of the model in the real space. However, in order to avoid interfering with the user's actual field of view, these systems usually hide the real space mapping grid after completing the spatial alignment, so this part of the data mapping generally no longer plays a role; in addition, because existing mixed reality devices use SLAM technology for deep space recognition, it is difficult to effectively interact with other high-precision positioning devices, which limits their application in complex, large-scale construction environments.

[0047] In order to solve at least one of the above problems, the present embodiment provides a construction safety warning method based on mixed reality, which is a construction site safety warning system that combines mixed reality technology, high-precision positioning technology and real-time spatial interaction to achieve more accurate, real-time and reliable safety management; by innovatively combining mixed reality technology, high-precision positioning technology and real-time spatial interaction, accurate monitoring, dynamic risk assessment and real-time warning of the construction site are achieved, aiming to overcome the limitations of existing technologies and provide a more intelligent and efficient construction safety management solution.

[0048] Mainly, after the initial alignment using the spatial mapping capability of the mixed reality device, the detailed real-space mapping grid data is innovatively retained; after optimization, these data become the basis for subsequent dynamic collision detection, allowing the system to continue to perform high-precision spatial analysis and risk assessment without affecting the user's actual field of view; and the real-time position data of construction equipment obtained by the high-precision GPS-RTK (Real-Time Kinematic) positioning system is innovatively accurately mapped to the model space coordinate system of the mixed reality device through a specially designed coordinate conversion algorithm, breaking through the limitation of traditional mixed reality technology that is difficult to integrate external high-precision positioning data, and achieving high-precision equipment positioning. Through the above settings, this embodiment can realize real-time three-dimensional interaction between real physical space, design model space and dynamic equipment position, greatly improving the accuracy, real-time and reliability of construction safety management. Figure 1 As shown, the system corresponding to the method in this embodiment shows the relationship between each functional module and the data flow. The figure clearly shows how this system integrates mixed reality technology, GPS-RTK high-precision positioning, Internet of Things devices and various intelligent algorithms to form a complete construction site safety warning system; the main modules include design model creation module, mechanical equipment construction module, high-precision positioning data processing module, mixed reality space processing module, dynamic collision detection module, intelligent warning and visualization module, remote collaboration module and automatic recording module. The figure also shows the direction of data flow, explains the interaction between the modules, and highlights the real-time and integrated nature of the system. The main steps of the method in this embodiment are:

[0049] S1. Design model:

[0050] Optionally, integrate high-precision surveying and mapping data and building information modeling (BIM) design solutions to create accurate 3D construction models. Support multi-source data fusion, including laser scanning point clouds, drone aerial images, etc.; use parametric modeling technology to ensure high accuracy and editability of the model; have real-time update capabilities, and can dynamically adjust the model according to the construction progress.

[0051] Corresponding to step S1, a design model creation module is set to implement the content in step S1. Integrate high-precision surveying and mapping data and BIM design solutions to create accurate three-dimensional construction models; optionally, use three-dimensional modeling software and point cloud data obtained by laser scanning technology to perform three-dimensional modeling; support real-time updates, can reflect construction progress and environmental changes, and provide basic data for mixed reality application modules and collision detection.

[0052] S2. Mechanical equipment modeling:

[0053] Optionally, create a 3D device model with precise geometry and dynamic motion characteristics. Use high-precision CAD data to build the device's geometric model; integrate the device's kinematic and dynamic characteristics to achieve complex motion simulation; support multi-level LOD (Level of Detail) models to meet different computing needs.

[0054] Corresponding to step S2, a design mechanical equipment building module is set to implement the content in step S2. Optionally, a parametric modeling technology is used, including the kinematic and dynamic characteristics of the equipment, which can simulate complex equipment motion trajectories, including multi-degree-of-freedom motion, and provide a dynamic equipment model for collision detection.

[0055] S3, high-precision positioning data processing:

[0056] Optionally, receive and process GPS-RTK data from construction equipment. Specifically:

[0057] S3.1, data reception preprocessing:

[0058] Optionally, receive GPS-RTK data in real time via a low-latency communication protocol such as MQTT and perform preliminary noise filtering and outlier detection.

[0059] S3.2, Data smoothing and prediction:

[0060] Optionally, a Kalman filter algorithm is used to smooth the converted position data and combine it with the device's motion model to perform short-term position prediction to compensate for possible communication delays.

[0061] Corresponding to step S3, a high-precision positioning data processing module is set and designed to implement the content in step S3. Optionally, in order to achieve real-time and stable data transmission between GPS-RTK data and the mixed reality system; adopt a multi-mode communication strategy, including Wi-Fi, Bluetooth 5.0 and LoRa technology, etc., use extended Kalman filtering for data smoothing and prediction, maintain stable communication in complex construction environments, and provide stable and high-precision device location data for mixed reality applications.

[0062] S4. Conversion between GPS-RTK data and spatial mapping grid coordinate system of mixed reality device:

[0063] Optionally, the data in the GPS geographic coordinate system (WGS84) is converted to the spatial mapping grid local coordinate system of the mixed reality device. This coordinate conversion process involves multiple steps, including projection conversion and coordinate system conversion. The specific steps are as follows:

[0064] S4.1, Conversion from WGS84 longitude and latitude to projected coordinate system:

[0065] Optionally, GPS-RTK generates WGS84 longitude and latitude coordinates. Convert to a plane coordinate (x) using a projection method suitable for the local area (such as UTM or Gauss-Kruger projection) gps ,y gps , z gps ). The projection method is a mature ellipsoid-plane coordinate system projection conversion method in the GIS field, which will not be described here.

[0066] S4.2. Conversion of the projected coordinate system to the local coordinate system of the spatial mapping grid formed by the SLAM technology of the mixed reality device:

[0067] S4.2.1. Define the reference points in the SLAM coordinate system:

[0068] Select reference points: Select known reference points in the SLAM scan space mapping grid of the MR device.

[0069] Positioning and registration: Get the reference point (x ref ,y ref , z gps ) will be used as a basis for alignment and transformation.

[0070] Set reference position: Set the reference point as the reference in the SLAM coordinate system.

[0071] S4.2.2. Calculate the translation, rotation and scaling factors of the spatial model:

[0072] Translation: Calculate the plane coordinate difference between the origin of the SLAM coordinate system and the GPS reference point, and apply it to other points (x transformed ,y transformed , z transformed ).

[0073] The translation formula is as follows:

[0074] x transformed =x gps -x ref

[0075] y transformed =y gps -y ref

[0076] Z transformed =z gps -z ref

[0077] Rotation: Based on the known azimuths of the reference points in the two coordinate systems, calculate their angular difference:

[0078] x′=x transformed cos(θ)-y transformed sin(θ)

[0079] y′=x transformed sin(θ)+y transformed cos(θ)

[0080] Scaling factor: Considering that there is no systematic scale difference in the principle of SLAM technology, no scaling transformation is performed.

[0081] S4.2.3. Synchronize virtual model and spatial model:

[0082] Based on the information of the reference points, the spatial model and the virtual model are automatically synchronized and aligned, and the structural proportion of the virtual model is maintained.

[0083] Real-time adjustment of relative position: In the case of SLAM coordinate drift, automatic calibration of reference points is used to compensate for drift to achieve consistency between the virtual model and the spatial model.

[0084] Through this conversion and processing step, GPS-RTK data can be accurately mapped to the spatial mapping grid coordinate system of the MR device, ensuring accurate association between the spatial model, virtual model and geographic location. This method can not only solve the accuracy problem of short-distance applications, but also achieve high-precision coordinate alignment over a larger range. In practical applications, the appropriate projection method and conversion parameters can be selected according to the specific location and requirements of the construction site to ensure the highest positioning accuracy.

[0085] Corresponding to step S4, a spatial mapping grid coordinate system conversion module between GPS-RTK data and mixed reality equipment is set to implement the content in step S4. Optionally, in order to realize the coordinate conversion between GPS-RTK data and mixed reality system, a projection method suitable for the construction site is used to perform projection transformation, and then an innovative coordinate conversion method is used to convert the projection coordinates into mixed reality space coordinates, which can convert GPS-RTK data into MR virtual model space and provide device position data of the real coordinates of the motion device for mixed reality applications.

[0086] Figure 2 As shown in the figure, a complete process of GPS-RTK data from reception to final conversion into a local engineering coordinate system is demonstrated, including GPS satellites, RTK base stations, GPS-RTK receivers on construction equipment, data processing servers, and coordinate conversion processes, which clearly demonstrates how the system can achieve centimeter-level positioning accuracy. Figure 3 As shown in FIG. 1 , the dynamic management optimization process of the spatial mapping grid of the MR device is shown, which shows how to optimize the data by dynamically adjusting the weight and grid size in the overall optimization scheduling of massive spatial mapping data.

[0087] S5. Mixed reality space processing:

[0088] This step is the core step, responsible for processing and integrating various spatial data. Its innovative functions are as follows:

[0089] S5.1. Dynamic spatial data management:

[0090] Optionally, a new spatial data management method has been developed for collision detection needs: gridding and dynamic weighting of spatial data mapping, setting the maximum grid size A1 and the corresponding weight a1 according to the minimum distance from the predetermined trajectory, and dynamically setting the weight b1 according to the distance to the motion device model. The closer the distance, the smaller the grid size A1, and the larger the weights a1 and b1. During collision detection, fast query and update of verification data in the massive data of spatial mapping are achieved, greatly reducing the amount of query data.

[0091] S5.2. Multi-source data fusion:

[0092] Optionally, compared with conventional mixed reality applications, in addition to the spatial mapping data and virtual design models from the mixed reality device, the real-time position data of the device after GPS-RTK conversion of the motion device is also integrated.

[0093] Compared with the existing technology, the innovation of this embodiment is to achieve dynamic management optimization of spatial mapping data and real-time fusion of multi-source data. In particular, the real space mapping is retained and continuously updated, which not only ensures that it does not interfere with the user's field of view, but also provides a basis for subsequent dynamic collision detection, significantly improving the system's adaptability in complex and changing construction environments.

[0094] Corresponding to step S5, a mixed reality space processing module is set to implement the content in step S5. Optionally, the dynamic weight method and grid size adjustment are used to dynamically manage the SLAM space mapping grid; the position of the motion device after the coordinate conversion is merged with the space mapping grid; it has the ability to dynamically optimize and manage high-precision space perception data and multivariate data fusion, dynamically manage space mapping grid data, provide basic data for collision verification, and is also the core interactive interface of the system.

[0095] S6. Dynamic collision detection:

[0096] Figure 4The figure shows the spatial partition optimization strategy during dynamic collision detection, including the grid division diagram of different risk level areas during dynamic collision detection. Optionally, real-time collision risk assessment is performed based on the fused spatial data, model data and real-time data of sports equipment. The Separating Axis Theorem (SAT) is used for collision detection. For complex shapes, the convex hull decomposition technique is used to decompose them into multiple convex polyhedrons, and then SAT is applied. Specifically, there are two innovative working modes:

[0097] S6.1, Preset track mode:

[0098] This mode requires the use of MR equipment to scan the scene first, generate a spatial mapping grid, and then run collision detection inside the MR device.

[0099] Trajectory generation: Based on the construction plan, accurate 3D trajectories of equipment movement are generated during the modeling phase.

[0100] Space scanning: Scan the virtual space along the preset trajectory to obtain the space mapping grid.

[0101] Virtual collision check: Based on the scanning results, collision detection is run on the MR device, which is an offline detection.

[0102] S6.2, Real-time location mode:

[0103] Real-time monitoring: GPS-RTK data is used to continuously update the position of the motion device in the mixed reality space. At this time, the MR device does not need to be on site, which improves the level of safe operation.

[0104] Collision detection: Combines the device's current position, speed, acceleration, and motion trajectory to provide a collision risk warning.

[0105] Unlike the prior art, this embodiment is able to handle highly dynamic construction environments. By combining preset trajectories and real-time location data, it can not only predict potential collision risks, but also make real-time adjustments based on the actual movement of the equipment, providing more accurate and timely safety warnings.

[0106] Corresponding to step S6, a real-time location mode module is set to implement the content in step S6. Optionally, the spatial mapping grid and preset trajectory scanned by the hybrid device are used for offline collision risk inspection; the mature separation axis method is used to perform collision assessment in combination with GPS-RTK data and dynamically optimized spatial grid data; the collision risk of static obstacles and dynamic equipment can be processed simultaneously, spatial data and equipment models are received, and risk assessment results are output.

[0107] S7, Intelligent early warning and visualization:

[0108] Optionally, based on the collision detection results, multi-level visual warnings and operation suggestions are generated and intuitively presented to the operator through a mixed reality device.

[0109] Corresponding to step S7, an intelligent warning and visualization module is set to implement the content in step S7. Optionally, in order to convert the collision detection results into intuitive visual warnings and operation suggestions, color coding and dynamic icons are used to generate virtual obstacle avoidance paths, dynamically adjust information display according to the user's perspective, provide non-intrusive warning information, minimize interference with actual work, receive collision detection results, and output visual warnings.

[0110] S8. Remote collaboration:

[0111] Optionally, it supports real-time audio and video communication and virtual annotation sharing between on-site operators and remote experts.

[0112] Corresponding to step S8, a remote collaboration module is set to implement the content in step S8. Optionally, support multi-person real-time mixed reality collaboration, use WebRTC technology to achieve low-latency audio and video transmission, adopt a distributed architecture, support multiple users to operate a shared virtual environment at the same time, allow remote experts to participate in on-site decision-making in real time, provide instant guidance, and be closely integrated with the mixed reality application module to share on-site views and virtual models.

[0113] S9, Automatic recording:

[0114] Optionally, record key data and events during system operation and generate structured security reports.

[0115] Corresponding to step S9, a remote collaboration module is set to implement the content in step S9. Optionally, natural language processing technology is used to automatically generate report text, organize and associate safety data, store key safety records, provide comprehensive and traceable safety management records, support subsequent analysis and auditing, and integrate data from all other modules to generate a comprehensive report.

[0116] This embodiment dynamically adjusts the grid size according to the minimum distance from the predetermined track and the distance from the device model, so that the areas close to the track and the device have a higher resolution, ensuring that these areas are processed more finely and updated in real time; while for areas far away from the track or the device, it is sufficient to maintain a larger grid unit, thereby reducing unnecessary detail processing; in this way, the device resources and computing power can be concentrated on the most important parts. Through gridding and weight assignment, high-weight grid units can be retrieved first; since these high-weight units contain data close to the track or device, low-weight grid units can be directly skipped during collision detection, greatly reducing the scope and total amount of data query. The dynamic management method allows the grid size and weight to be adjusted in real time during the movement of the device; as the device moves along the track, the system automatically increases the grid density of the high-weight area, thereby always ensuring that the spatial data around the device and the track has a high degree of accuracy, which is suitable for large-scale real-time collision detection, because the change of dynamic weight can be updated in real time with the device position. By combining grid size and weight, a more efficient spatial index structure can be established for data retrieval; after the grid cells are sorted by weight, the high-weight grid data can be placed in a priority position, and with index structures such as KD tree or R tree, the data query speed can be further improved; at the same time, the data of low-weight grids can be temporarily placed in secondary storage to save memory. Dynamic grid adjustment under weight control can greatly reduce the processing volume of high-resolution data and retain dense grids only in key areas; this can not only reduce memory usage, but also reduce the computational overhead of CPU and GPU; during device processing, only the part of the grid data closest to the device and trajectory needs to be processed, which improves the efficiency of collision detection.

[0117] In this embodiment, the setting of dynamic grids and weights can ensure that the device only calls high-priority grid data when needed, and a large amount of grid data far away from the device and track does not need to be loaded and processed; as the device moves, the range of loaded data is dynamically adjusted to achieve efficient data retrieval. Grid data sorted by weight can be stored in a high-priority cache for easy and fast access; each time a query is made, only high-weight grid data in the current moving direction needs to be retrieved, and low-weight grid cells will not be frequently accessed. The grid size and weight change dynamically to ensure that the closer the area to the device and track, the faster the retrieval speed. It can not only effectively reduce the massive amount of data during the query, but also make the data retrieval meet the requirements of the device's movement direction. Therefore, the setting of dynamic weights can automatically increase when accuracy is required, and maintain a coarse grid when accuracy is not required, thereby improving the response speed of the overall system.

[0118] In summary, in this embodiment, the SLAM spatial data management method based on dynamic gridding and weight assignment can indeed significantly improve the retrieval efficiency of spatial data during collision detection; the advantage is that it effectively utilizes equipment resources, flexibly responds to dynamically changing detection needs, and has an efficient spatial indexing and retrieval mechanism; by reducing the amount of data queries and reducing computing overhead, this method is very suitable for application scenarios with high real-time requirements. Combined with high-precision positioning and advanced collision detection algorithms, real-time collision risk assessment can be performed. This predictive safety management method can significantly reduce the incidence of safety accidents and improve construction efficiency. Through real-time fusion and intelligent visualization of multi-source data, it provides construction management personnel with unprecedented spatial perception capabilities and decision-making support, which not only improves the safety of daily construction, but also provides powerful tool support for complex operations and emergency handling.

[0119] Embodiment 2:

[0120] This embodiment provides a construction safety early warning method based on mixed reality. On the basis of Embodiment 1, it illustrates how the system uses the spatial mapping grid data of the mixed reality device to dynamically manage and optimize data scheduling to achieve dynamic collision detection:

[0121] S1. Initialize spatial data gridding:

[0122] Divide the spatial mapping data generated by SLAM into initial grid cells. The initial grid size can be preset based on the spatial scene size and device performance. For example, set a global default grid size. Assign an initial weight value to each grid cell. The initial weight value can be set to zero or a small value so that it can be adjusted according to the specific situation in subsequent steps.

[0123] S2. Define key variables and thresholds:

[0124] Optionally, define two key variables: trajectory minimum distance weight (a1) and device model minimum distance weight (b1). The trajectory minimum distance weight (a1) is used to measure the closest distance from the grid cell to the predetermined trajectory, and the device model minimum distance weight (b1) is used to measure the closest distance from the grid cell to the device model. Set the maximum and minimum size thresholds for the grid cells. The closer the distance, the smaller the grid size; the farther the distance, the larger the grid size. According to these thresholds, adjust the grid size to A1. Define two weight functions: trajectory weight function (f(a1)) and device distance weight function (g(b1)), which are used to update the grid weight in real time according to the trajectory distance and device distance. The weight function can use an inverse proportional function, such as $f(a1)=$, where $k$ is the weight adjustment factor and $∈$ is a small value to prevent division by zero.

[0125] S3, dynamic grid size and weight adjustment:

[0126] For each grid cell, calculate the minimum distance d from the cell to the trajectory t and the minimum distance d to the device model m . Use the trajectory weight function $f(d_t)$ to calculate the trajectory distance weight $a1$ of the grid cell and adjust the grid size A1. t If the distance is less than the preset threshold, the grid size A1 is reduced and the weight a1 is increased. Using the device distance weight function g(d m )) Calculate the device model distance weight b1. Similarly, if d m If the grid size A1 is smaller than the threshold of the device model, the grid size A1 is reduced again and the weight b1 is increased. The final weight value W = a1*b1 of the grid cell is calculated to ensure that the grid cells close to the track and the device are given higher weights.

[0127] S4. Prioritization based on weight:

[0128] The grid cells are sorted according to the final weight W. The grid cells with higher weights have higher priorities. The high-priority grid cells are placed at the front of the priority queue so that they can be loaded first during retrieval. The weights and priorities of the grid cells are updated in real time every time the device moves. For the grid cells in the direction of the device's advance, a rolling window can be set to dynamically recalculate the trajectory distance and device distance within the window and adjust their weights.

[0129] S5. Data retrieval and call optimization:

[0130] Use high-weight grid cells in the priority queue for collision detection. When the device is moving, the data of high-priority grid cells is loaded first, and the loading of low-weight grid cells is ignored or delayed. Each time it is called, only the high-weight grid data around the trajectory is retrieved, thereby reducing the amount of data retrieval. A cache mechanism can be set within a certain range to cache the grid cell data near the trajectory to ensure that frequently retrieved data is loaded quickly. When the device moves to a new position, the weights of each grid cell in the current scrolling window are recalculated, and the priority queue is adjusted according to the updated weights. This ensures that the data closest to the device and trajectory can be called at any time.

[0131] S6. Data update and rolling window management:

[0132] The grid data of the area in front of the device's direction of movement is dynamically updated through the rolling window mechanism. Each time the device moves along the track, the grid cells that have left are removed from the priority queue, and the grid cells that have newly entered the window range in front are added to the priority queue. When the weight of a grid cell changes significantly, an event can be triggered to recalculate the resolution and weight of the grid cell to ensure that the data around the device and track has higher precision. If the device speed increases, the size of the rolling window is appropriately expanded to ensure that more grid data is loaded in advance; when the device slows down, the window size is reduced to reduce the frequency of data updates, thereby further optimizing the calculation efficiency while ensuring accuracy.

[0133] S7. Collision detection execution:

[0134] During the detection process, high-weight grid cells in the window are prioritized for collision detection. The detailed data in the grid cells is used to determine the collision between the model and the spatial data. If a potential collision occurs in a high-weight area, the grid can be refined and the data encrypted. The accuracy of the collision location is ensured by encrypting the scan, while other unimportant areas are not refined.

[0135] This example demonstrates how to combine high-precision GPS-RTK data with mixed reality technology to achieve accurate spatial perception, dynamic collision detection, and intelligent early warning, which can provide unprecedented safety management capabilities for complex construction environments.

[0136] Embodiment 3:

[0137] This embodiment provides a construction safety early warning method based on mixed reality. On the basis of Embodiment 1 and Embodiment 2, the industrial applicability is introduced. The construction safety early warning system based on mixed reality in this embodiment has wide industrial applicability and can significantly improve the safety management level of various large-scale construction projects and infrastructure construction projects; the application can not only effectively reduce the occurrence rate of construction safety accidents, but also improve construction efficiency and management level; specifically, the industrial applicability is reflected in the following aspects:

[0138] Safety management of large-scale construction projects: Application scenarios: complex construction projects such as high-rise buildings, large commercial complexes, and sports stadiums. Specific applications: Safety monitoring of aerial work: real-time monitoring of the location of aerial workers and equipment, and early warning of potential falling risks. Optimization of hoisting operations: Through virtual simulation and real-time monitoring, the hoisting path of large equipment is optimized to reduce the risk of collision. Benefit analysis: It can reduce the accident rate of aerial work by more than 30% and improve the installation efficiency of large equipment by about 20%.

[0139] Infrastructure construction safety control: Application scenarios: linear engineering projects such as bridges, tunnels, highways, and railways. Specific applications: Tunnel construction safety: monitor the stability of the excavation surface and warn of possible landslide risks. Precise positioning of bridge construction: ensure the accurate installation of large components and reduce construction errors. Benefit analysis: It can increase the safety factor of tunnel construction by about 25%, and the installation accuracy of bridge components is improved to the millimeter level.

[0140] Underground engineering risk management: Application scenarios: subway, underground pipe gallery, deep foundation pit engineering, etc. Specific applications: Underground space visualization: Through mixed reality technology, the underground pipelines and geological structures are intuitively displayed. Earthwork excavation monitoring: Real-time monitoring of excavation depth and settlement of surrounding buildings. Benefit analysis: Reduce underground pipeline damage accidents by about 40% and improve underground engineering construction efficiency by about 15%.

[0141] Industrial plant and equipment installation: Application scenarios: chemical plants, power stations, large manufacturing workshops, etc. Specific applications: Equipment installation guidance: Assist the precise installation of complex equipment through mixed reality technology. Pipeline wiring optimization: Real-time inspection of pipeline layout to avoid conflicts and interference. Benefit analysis: Reduce equipment installation time by about 25% and reduce pipeline system design changes by about 35%.

[0142] Construction site personnel safety management: Application scenario: Daily safety management of various construction sites. Specific application: Personal protective equipment (PPE) inspection: Automatically identify whether workers are wearing helmets, safety belts, etc. correctly. Dangerous area warning: When workers approach dangerous areas, warnings are issued through mixed reality devices. Benefit analysis: Reduce personnel safety accidents by about 50% and improve safety management efficiency by about 30%.

[0143] Emergency response and rescue: Application scenarios: construction accident sites or construction sites affected by natural disasters. Specific applications: Virtual site survey: Remote experts can conduct real-time site assessments through mixed reality technology. Escape route planning: In an emergency, provide on-site personnel with the best evacuation route guidance. Benefit analysis: Increase emergency response speed by about 40% and improve decision-making quality in emergencies.

[0144] Construction progress and quality management: Application scenarios: Full life cycle management of various construction projects. Specific applications: Comparison of actual progress and plan: Visually display the difference in construction progress through mixed reality technology. Quality inspection assistance: Combined with BIM model, real-time quality verification and deviation analysis. Benefit analysis: Improve the completion rate of construction period by about 20% and reduce quality defects by about 30%.

[0145] Training and skills improvement: Application scenarios: new employee training, skills upgrading, safety education, etc. Specific applications: Virtual construction simulation: high-risk operation training in a safe virtual environment. Equipment operation training: immersive equipment operation guidance is provided through mixed reality technology. Benefit analysis: shorten the training cycle by about 35% and improve the training effect and skill mastery by about 40%.

[0146] Economic and social benefits: Direct economic benefits: reduce the rate of safety accidents, reduce related compensation and downtime losses; improve construction efficiency, shorten construction period, and reduce labor costs; optimize resource allocation, reduce material waste and rework. Indirect economic benefits: enhance corporate reputation and enhance market competitiveness; reduce insurance costs and obtain more favorable insurance terms. Attract high-quality talents and improve employee satisfaction and loyalty. Social benefits: improve the safety of construction projects and protect the safety of workers and the public; promote technological progress and digital transformation in the construction industry; provide technical support and practical experience for smart city construction.

[0147] To sum up, this embodiment has broad application prospects and significant economic and social benefits in various types of construction projects and infrastructure construction. It can not only effectively improve construction safety, but also play an important role in improving management efficiency, optimizing construction processes, improving collaboration effects, etc., and provide strong support for the sustainable development of the construction industry.

[0148] Embodiment 4:

[0149] This embodiment provides a construction safety early warning system based on mixed reality, including:

[0150] The data acquisition module is configured to: obtain surveying and mapping data of the construction site, as well as geometric dimension data and geographic coordinate system data of relevant mechanical equipment;

[0151] The mixed reality model building module is configured to: build a mixed reality model using the acquired surveying and mapping data and geometric dimension data;

[0152] A coordinate system data mapping module is configured to: map geographic coordinate system data to a spatial coordinate system of a mixed reality model; wherein, during the mapping process, the longitude and latitude of the geographic coordinate system data are converted into plane coordinates, and the plane coordinates are converted into a spatial mapping grid local coordinate system formed by simultaneous positioning and mapping technology;

[0153] The collision detection module is configured to: assign an initial weight value to each grid cell in the space mapping grid, dynamically adjust the weight to change the grid size, and dynamically manage the space mapping grid data; perform collision detection based on the dynamically managed space mapping grid data to obtain a collision detection result;

[0154] The safety warning module is configured to: perform risk assessment and safety warning based on the collision detection results.

[0155] The working method of the system is the same as the construction safety early warning method based on mixed reality in Example 1, and will not be repeated here.

[0156] Embodiment 5:

[0157] This embodiment provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the steps of the construction safety early warning method based on mixed reality described in Example 1 are implemented.

[0158] Embodiment 6:

[0159] This embodiment provides an electronic device, including a memory, a processor, and a computer program stored in the memory and capable of running on the processor. When the processor executes the program, the steps of the construction safety warning method based on mixed reality described in Example 1 are implemented.

[0160] Embodiment 7:

[0161] This embodiment provides a computer program product, which includes a computer program. When the computer program is executed by a processor, the steps of the construction safety early warning method based on mixed reality described in Example 1 are implemented.

[0162] The above description is only a preferred embodiment of the present embodiment and is not intended to limit the present embodiment. For those skilled in the art, the present embodiment may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present embodiment shall be included in the protection scope of the present embodiment.

Claims

1. A construction safety early warning method based on mixed reality, characterized in that: include: Obtain surveying and mapping data of the construction site, as well as geometric dimension data and geographic coordinate system data of relevant mechanical equipment; Use the acquired surveying and mapping data and geometric dimension data to build a mixed reality model; Mapping geographic coordinate system data to the spatial coordinate system of the mixed reality model; wherein, during the mapping process, the longitude and latitude of the geographic coordinate system data are converted into plane coordinates, and the plane coordinates are converted into a spatial mapping grid local coordinate system formed by simultaneous positioning and mapping technology; Assign an initial weight value to each grid cell in the space mapping grid, dynamically adjust the weight to change the grid size, and dynamically manage the space mapping grid data; perform collision detection based on the dynamically managed space mapping grid data to obtain collision detection results; Conduct risk assessment and safety warning based on collision detection results.

2. A construction safety early warning method based on mixed reality as claimed in claim 1, characterized in that: Noise filtering and outlier detection are performed on geographic coordinate system data. The converted location data is smoothed using the Kalman filter algorithm, and the location is predicted in combination with the device's motion model.

3. A construction safety early warning method based on mixed reality as claimed in claim 1, characterized in that: Selecting known reference points in the spatial mapping grid; and obtaining reference points in the device's model space in the mixed reality model to use as a basis for alignment and transformation; Calculate the plane coordinate difference between the origin of the coordinate system and the reference point, and apply it to other points; get the angle difference based on the known azimuth of the reference point in the two coordinate systems; According to the information of the reference points, the spatial model and the virtual model are automatically synchronized and aligned, and the structural proportion of the virtual model is maintained. In the case of coordinate drift, the reference points are automatically calibrated and drift compensation is performed to achieve consistency between the virtual model and the spatial model.

4. The construction safety early warning method based on mixed reality as claimed in claim 1, characterized in that: Gridding and dynamically weighting the spatial data mapping, setting a maximum grid size and a corresponding first weight according to a minimum distance from a predetermined trajectory, and dynamically setting a second weight according to a distance to a motion device model; The closer the distance, the smaller the grid size, and the larger the first weight and the second weight.

5. A construction safety early warning method based on mixed reality as claimed in claim 4, characterized in that: The trajectory distance weight of the grid cell is calculated using the preset trajectory weight function, and the grid size is adjusted; if the minimum distance from the cell to the trajectory is less than the preset threshold, the grid size is reduced and the weight first is increased; the device model distance weight is calculated using the preset device distance weight function; If the minimum distance from the cell to the device model is less than the threshold of the device model, the grid size is reduced again and the second weight is increased; Calculate the final weight value of the grid unit as the product of the first weight and the second weight; Sort the grid cells according to the final weight, and the grid cells with higher weights have higher priorities; The grid cells with final weights higher than the preset weights are placed at the front of the priority queue for priority loading; the weights and priorities of the grid cells are updated in real time every time the device moves; for the grid cells in the direction of the device's advance, a rolling window is set, and the trajectory distance and device distance are dynamically recalculated within the rolling window, and their weights are adjusted; Use grid cells whose final weights are higher than the preset weights for collision detection. When the device is moving, first load the data of grid cells whose final weights are higher than the preset weights, and ignore or delay grid cells with other weights. Each time it is called, only retrieve the data of grid cells along the trajectory whose final weights are higher than the preset weights. Set a cache mechanism within a certain range to cache the data of grid cells near the trajectory. When the device moves to a new position, recalculate the weights of each grid cell in the current scrolling window, and adjust the priority queue according to the updated weights.

6. The construction safety early warning method based on mixed reality as claimed in claim 1, characterized in that: Collision detection using the separating axis theorem, including preset trajectory mode and real-time position mode; In the preset trajectory mode, a three-dimensional trajectory of equipment movement is generated during the modeling phase based on the construction plan; Scan the virtual space along the preset trajectory to obtain the space mapping grid; run collision detection on the device based on the scanning results; in the real-time position mode, use the geographic coordinate system data to continuously update the position of the sports device in the mixed reality space.

7. A construction safety early warning system based on mixed reality, characterized in that: include: The data acquisition module is configured to: obtain surveying and mapping data of the construction site, as well as geometric dimension data and geographic coordinate system data of relevant mechanical equipment; The mixed reality model building module is configured to: build a mixed reality model using the acquired surveying and mapping data and geometric dimension data; A coordinate system data mapping module is configured to: map geographic coordinate system data to a spatial coordinate system of a mixed reality model; wherein, during the mapping process, the longitude and latitude of the geographic coordinate system data are converted into plane coordinates, and the plane coordinates are converted into a spatial mapping grid local coordinate system formed by simultaneous positioning and mapping technology; The collision detection module is configured to: assign an initial weight value to each grid cell in the space mapping grid, dynamically adjust the weight to change the grid size, and dynamically manage the space mapping grid data; perform collision detection based on the dynamically managed space mapping grid data to obtain a collision detection result; The safety warning module is configured to: perform risk assessment and safety warning based on the collision detection results.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the construction safety early warning method based on mixed reality as described in any one of claims 1 to 6 are implemented.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and capable of running on the processor, characterized in that: When the processor executes the program, the steps of the construction safety early warning method based on mixed reality are implemented as described in any one of claims 1-6.

10. A computer program product, characterized in that The computer program product includes a computer program, and when the computer program is executed by a processor, the steps of the construction safety early warning method based on mixed reality are implemented as described in any one of claims 1 to 6.

Images