A building construction intelligent operation method and system based on BIM-robot collaboration

Abstract

This invention relates to an intelligent construction operation method based on BIM-robot collaboration. The method includes the following steps: Step 1, BIM construction task parsing and conversion; Step 2, multi-source fusion high-precision positioning; Step 3, process constraint path planning; Step 4, real-time quality detection and closed-loop control. This invention achieves an automated process from digital design to robot execution through BIM model semantic parsing and automatic task generation technology. Without manual programming or teaching, the system can directly extract construction tasks from the BIM model and generate robot execution instructions, significantly reducing construction preparation time and labor costs. When design changes occur, only updating the BIM model is needed to automatically adjust the robot operation plan.

Description

Technical Field

[0001] This invention relates to an intelligent construction operation method based on BIM-robot collaboration, belonging to the field of intelligent construction and robot control technology. Background Technology

[0002] As the construction industry moves towards intelligent and automated operations, applying robotics to construction has become an important way to address labor shortages, improve construction quality, and ensure operational safety. However, existing construction robot systems face numerous technical challenges in practical applications, limiting their large-scale adoption.

[0003] Traditional construction robot systems mainly suffer from the following problems:

[0004] 1. Disconnect between BIM model and robot execution: Existing construction robot systems typically require manual programming or teaching methods to input work trajectories, and cannot directly extract construction tasks and parameters automatically from the BIM design model. Although the BIM model contains rich architectural geometry information and construction attributes, it lacks an effective parsing mechanism to convert it into motion commands that can be executed by the robot. As a result, the "design → construction" conversion process still relies heavily on manual intervention, which is inefficient and prone to errors.

[0005] 2. Insufficient positioning accuracy at construction sites: Construction sites are complex and ever-changing environments, with numerous temporary structures, equipment, and personnel activities. Traditional single positioning methods (such as relying solely on SLAM or UWB) struggle to maintain stable, high-precision positioning across all scenarios. Positioning errors directly impact construction accuracy, especially in operations requiring high positional precision, such as masonry and welding. Even centimeter-level or millimeter-level deviations can lead to quality issues.

[0006] 3. Lack of construction process constraints in path planning: General robot path planning algorithms mainly consider obstacle avoidance and motion efficiency, failing to fully incorporate the process characteristics of building construction. For example, masonry work requires following the "bottom to top, outside to inside" masonry sequence, and plastering work requires controlling the direction and pressure of application. These process constraints are difficult to reflect in traditional path planning, resulting in the robot's work quality failing to meet the specifications.

[0007] 4. Separation of Quality Inspection from Construction Process: Current construction quality inspection mainly adopts a post-construction acceptance method, with measurements and evaluations only conducted after construction is completed. Once quality problems are discovered, rework is costly and time-consuming. The lack of real-time quality inspection and dynamic correction mechanisms during construction makes it impossible to correct deviations in their early stages, hindering process quality control.

[0008] 5. Inconsistent coordinate systems between robots and BIM: Robots use a local coordinate system based on their own origin, while BIM models use a building project coordinate system or a measurement coordinate system. There is a lack of precise coordinate transformation and alignment mechanisms between the two. This inconsistency in coordinate systems can cause deviations between the robot's actual operating position and the design position, affecting construction accuracy and model consistency.

[0009] 6. Lack of adaptive operation capability: There are differences between the construction site environment and the ideal model, including material size deviations, uneven base surfaces, and temporary obstacles. Existing robot systems usually execute rigidly according to preset programs, lacking the ability to sense environmental changes and automatically adjust operation parameters, which can easily lead to operation failures or quality problems under non-ideal conditions.

[0010] Therefore, there is an urgent need for an intelligent construction operation system that can deeply integrate BIM model information, achieve high-precision autonomous positioning, meet construction process constraints in path planning, and possess real-time quality detection and dynamic correction capabilities. This invention solves the above-mentioned technical challenges by constructing a closed-loop collaborative mechanism of "BIM task analysis → high-precision positioning → intelligent path planning → real-time quality control," providing a complete and reliable technical solution for construction automation. Summary of the Invention

[0011] This invention addresses the technical problems existing in current technologies by providing a BIM-robot collaborative intelligent construction operation method. This system deeply integrates Building Information Modeling (BIM) with construction robots, achieving a seamless transition from digital design to automated construction. The system includes a BIM construction task parsing module, a multi-source fusion high-precision positioning module, a construction path intelligent planning module, and a real-time quality detection and feedback module. It can automatically extract construction parameters from the BIM model, plan robot motion trajectories, and dynamically correct quality deviations in real time during construction. This invention is applicable to various construction operation scenarios such as masonry, plastering, spraying, and welding.

[0012] This invention provides a BIM-robot collaborative intelligent construction operation system. It achieves a complete technical chain from BIM digital model to automated robotic construction by constructing four core modules: a BIM construction task parser, a multi-source fusion high-precision positioner, a process constraint path planner, and a real-time quality closed-loop controller. The four modules are closely linked in the order of "parsing → positioning → planning → execution and detection": First, the BIM task parsing module extracts construction information from the design model and generates a task sequence; second, the high-precision positioning module determines the robot's position and aligns it with the BIM coordinate system; then, the path planning module generates the operation trajectory based on the target location and process constraints; finally, the quality detection module monitors the construction process in real time and provides feedback for correction parameters. These four modules form a closed-loop collaborative mechanism, ensuring that the robot completes construction tasks accurately, efficiently, and with high quality.

[0013] To achieve the above objectives, the technical solution of the present invention is as follows: a BIM-robot collaborative intelligent construction operation method, the method comprising the following steps:

[0014] Step 1: BIM construction task analysis and conversion.

[0015] Step 2: Multi-source fusion high-precision positioning.

[0016] Step 3: Process constraint path planning.

[0017] Step 4: Real-time quality detection and closed-loop control.

[0018] Step 1 is as follows:

[0019] The BIM construction task parser described in this invention can automatically extract the geometric information, material properties, and process parameters of construction objects from the building information model and convert them into a task sequence that can be executed by a robot. This parsing process consists of three steps: model semantic parsing, construction area division, and task sequence generation.

[0020] 1-1 Semantic analysis of BIM model

[0021] The system extracts semantic information of building components by parsing BIM models in IFC (Industry Foundation Classes) or Revit native format. For wall components to be constructed, the system extracts attributes such as their spatial bounding box, material type, thickness parameters, and surface treatment requirements. Assuming there are a total of [missing information - likely related to a BIM model or system] in the BIM model... The first wall component to be constructed, the first The geometric properties of a wall are represented as follows: ,in The length of the wall. The height of the wall. For wall thickness, and These are the three-dimensional coordinates of the wall's start and end points in the BIM coordinate system.

[0022] The system also extracts the construction attributes of the wall. ,in Indicate the type of material (e.g., standard bricks, aerated concrete blocks, plastering mortar, etc.). Indicates the type of construction process (such as masonry, plastering, spraying). These attributes indicate the required surface quality level, and they determine the specific methods and parameter settings for subsequent robot operations.

[0023] 1-2 Intelligent division of construction areas,

[0024] For large walls, the system divides the wall into several working units based on the robot's operating range and efficiency. The core basis for this division is the effective working radius of the robot's robotic arm. When the robot is at a certain stationary position, the robotic arm can cover a circle centered on that point. For an area with a radius of [radius value], to ensure full wall coverage and no omissions in adjacent areas, it is necessary to reasonably set the spacing between anchor points and the overlap width.

[0025] For length of Height is The walls are divided horizontally into The area is divided vertically into several regions. Areas:

[0026]

[0027] In the above formula, the effective width of each region in the horizontal direction is (The robotic arm covers both sides) (subtract the overlapping parts) The overlap width between adjacent areas (usually 10-20cm) is used to ensure the continuity of work at the area boundaries and avoid missed work due to positioning errors; in the vertical direction, since the robotic arm usually works from bottom to top, the height of each area is taken as... ,symbol This indicates rounding up to the nearest whole number, ensuring the area completely covers the wall, and the center position of each work unit. As the target station coordinates of the robot mobile platform, For horizontal indexing, Using a vertical indexing method, the system breaks down the construction task of the entire wall into a sequence of sub-tasks that the robot can execute one by one. Task sequences 1-3 are generated.

[0028] According to construction process specifications, the system generates a sequence of tasks in a specific order. For masonry work, the principle of "from bottom to top, from one end to the other" is followed; for plastering work, the principle of "from top to bottom, preventing contamination of finished surfaces" is followed. Defined as an ordered set of tasks, where The total number of tasks, each task contains objectives. Mark the coordinates of the stationary point Based on the job type, process parameters, and quality requirements, the BIM task parsing module will output a task sequence. The location information of each station point in the BIM coordinate system will be transmitted to the high-precision positioning module for coordinate transformation. Since the robot uses its own local coordinate system, while the task location is defined in the BIM coordinate system, a coordinate alignment mechanism is needed to convert the BIM coordinates into target coordinates that the robot can execute, ensuring that the robot accurately reaches the designed location for construction.

[0029] Step 2 is as follows:

[0030] Obtain the task sequence output by the BIM task parsing module After obtaining the target location information, the system needs to solve two key problems: first, to determine the robot's precise location on the construction site in real time; and second, to establish the transformation relationship between the robot's coordinate system and the BIM coordinate system. Construction site environments are complex and ever-changing, and single positioning technologies are insufficient to meet the high-precision positioning requirements across all scenarios. This invention designs a multi-source fusion positioning method, comprehensively utilizing laser SLAM, visual odometry, UWB positioning, and total station calibration information to achieve centimeter-level positioning of the robot on the construction site and accurately align the robot's coordinate system with the BIM coordinate system.

[0031] 2-1 Acquisition of multi-sensor positioning information

[0032] The system is equipped with three positioning sensors: LiDAR, binocular vision camera, and UWB tags. The LiDAR SLAM module constructs a point cloud map of the construction site using two-dimensional or three-dimensional LiDAR and estimates the robot's pose. ,in , For planar coordinates, For the heading angle, visual odometry calculates the relative displacement by matching feature points in consecutive frames of images. The UWB positioning system measures the distance to the robot using multiple base stations deployed at the construction site and estimates its position using a trilateration algorithm. ,

[0033] 2-2 Extended Kalman filter fusion,

[0034] The Extended Kalman Filter (EKF) algorithm is used to fuse multi-source localization information to improve localization accuracy and robustness. The reason for choosing EKF is that both the robot motion model and the sensor observation model are nonlinear systems. EKF can effectively handle such nonlinear state estimation problems by linearizing the nonlinear function through a first-order Taylor expansion. At the same time, EKF automatically adjusts the fusion weights of each sensor through the covariance matrix. When the noise of a certain sensor increases, its weight automatically decreases, which has good adaptive ability.

[0035] The system state vector is defined as follows: ,in , For position coordinates, For heading angle, Linear velocity, Angular velocity,

[0036] The state prediction equation is based on a kinematic model:

[0037]

[0038] in For time step, To control the inputs (including linear velocity and angular velocity), the covariance prediction is: ,in Let be the state covariance matrix of the previous time step. To predict the state covariance matrix, The Jacobian matrix of the state transition matrix (for...) Regarding the status (Find the partial derivative) The process noise covariance matrix is...

[0039] When sensor observation data is received, a measurement update is performed, and the observation vector is updated. Including measurements from laser SLAM, visual odometry, and UWB, the Kalman gain is calculated as follows: ,in Let Jacobian matrix be the observation matrix. To observe the noise covariance matrix, the state is updated as follows:

[0040]

[0041] in Let be the observation function.

[0042] Through the EKF fusion algorithm, the system dynamically adjusts the weights of each sensor. When the signal quality of a sensor deteriorates, its contribution is automatically reduced, ensuring stable positioning accuracy even under complex conditions such as occlusion and interference; 2-3, BIM coordinate system alignment.

[0043] Before construction begins, the system uses a total station to measure the BIM coordinates and corresponding robot coordinates of several control points (no fewer than four) at the construction site. Let the first... The coordinates of each control point in the BIM coordinate system are: The coordinates in the robot coordinate system are The system uses the least squares method to solve for the coordinate transformation parameters. Its core idea is to find the optimal rotation matrix. Translation vector To minimize the sum of squared transformation errors for all control points, specifically by minimizing the objective function... The rotation matrix is ​​solved using singular value decomposition (SVD), and the translation vector is calculated based on the centroid alignment principle.

[0044] Finally, the translation vector is obtained. and rotation matrix (Orthogonal matrix) such that any point Coordinates in the robot coordinate system can be converted to coordinates in the BIM coordinate system: ,because Since it is an orthogonal matrix, its inverse is equal to its transpose. Therefore, BIM coordinates can also be converted to robot coordinates through inverse transformation. ,

[0045] Through the aforementioned coordinate alignment mechanism, the high-precision positioning module achieves bidirectional conversion between the BIM coordinate system and the robot coordinate system: on the one hand, it converts the robot's real-time pose... The position components in the image are transformed to the BIM coordinate system through a positive transformation, i.e., the position... heading angle (in Rotation matrix The corresponding rotation angle around the z-axis, and They are respectively The elements in the second row and first column, and the first row and first column, are used for progress tracking and quality recording; on the other hand, the target stationary point coordinates output by the BIM task parsing module are... These points are converted into navigation target points in the robot coordinate system. These target points are then passed to the path planning module to generate the operation trajectory from the current position to each target point.

[0046] Step 3 is as follows:

[0047] Based on the robot's real-time pose and the converted target stationary coordinates provided by the high-precision positioning module, the path planning module needs to calculate the motion trajectory from the current position to each target point. Traditional robot path planning mainly focuses on obstacle avoidance and motion efficiency, while construction operations also need to meet specific process constraints. This invention designs a process-constrained path planner that incorporates construction process requirements into motion planning, generating work trajectories that are both safe and efficient and comply with construction specifications.

[0048] 3-1 Construction process constraint modeling,

[0049] Different types of construction work have different technological constraints. For masonry work, the main constraints include: the effective time window after mortar laying. Horizontal and vertical accuracy of brick placement and Distance between staggered joints of adjacent layers For plastering operations, constraints include: plaster thickness range. Single application coverage width Consistent calendering direction

[0050] The system formalizes process constraints into optimization objectives and constraints for path planning, and defines a process compliance function. Evaluation of the degree to which the evaluation path meets the process requirements:

[0051]

[0052] in For task sequence The total number of tasks in the process. For the first One task, For the task Time constraint penalty (when the operation time exceeds the material's expiration date) (Punishment will be imposed at that time) A penalty term for accuracy constraints (when the positioning deviation exceeds the allowable accuracy). or (Punishment will be imposed at that time) This is a penalty item for process sequence constraints (a penalty is incurred when the process sequence such as "bottom first, top last" is violated). , , For the corresponding weight coefficients, satisfying By minimizing The path planner generates the optimal job sequence that meets the process requirements.

[0053] 3-2 Multi-objective path optimization,

[0054] Path planning requires simultaneous consideration of multiple optimization objectives: minimizing operation time, minimizing energy consumption, maximizing construction quality, and avoiding collisions with obstacles. This invention employs an improved A* algorithm combined with process constraints for path search. While the traditional A* algorithm only considers the shortest geometric path, the improvement of this invention lies in introducing a process constraint penalty term into the cost function, ensuring that path search not only pursues the shortest distance but also meets construction process specifications.

[0055] Define the comprehensive cost function ,in This represents the actual travel cost from the starting point to the current node (considering distance and energy consumption). The heuristic cost estimate (using Euclidean distance) from the current node to the target point is calculated. The process constraint cost is calculated based on whether the current path meets the construction process requirements: if the path crosses an unfinished construction area, causes the mortar to expire or fails, or violates the work sequence, a higher cost is assigned; the cost of a path that meets the process requirements is zero. In this way, the algorithm automatically avoids paths that do not meet the process requirements during the search process and outputs the optimal path that meets both obstacle avoidance requirements and construction specifications.

[0056] For the precise working trajectory of the robotic arm, the system performs trajectory interpolation in Cartesian space to ensure smooth movement of the end effector along the construction surface. This invention employs a fifth-order polynomial for trajectory smoothing. The reason for choosing a fifth-order polynomial is that it can simultaneously constrain six boundary conditions—position, velocity, and acceleration—at both the starting and ending points, which correspond precisely to the six coefficients of the fifth-order polynomial, thus generating a smooth curve where both velocity and acceleration are zero at the starting and ending points. This trajectory planning method ensures shock-free start-up and shutdown of the robotic arm, and continuous changes in velocity and acceleration during movement, effectively reducing the impact of vibration on construction quality. It is particularly suitable for precision operations requiring smooth movement, such as bricklaying, plastering, and other similar tasks.

[0057] The work trajectory output by the path planning module will be executed by the robot controller, driving the mobile platform and robotic arm to complete specific construction actions. During construction, the system needs to monitor the construction quality in real time and adjust subsequent construction parameters based on the detection results. This function is achieved by the real-time quality detection and closed-loop control module.

[0058] Step 4 is as follows: Unlike traditional post-construction quality acceptance, this invention designs a real-time quality inspection and closed-loop control module, forming a closed loop with the path planning module and the execution controller. During the robot's construction actions according to the planned trajectory, the quality inspection module synchronously collects the geometric parameters of the completed work surface. Once a deviation is detected, a correction command is immediately fed back to the controller to adjust subsequent construction parameters, thus achieving process control of quality.

[0059] 4-1 Real-time monitoring of construction quality.

[0060] The system is equipped with a structured light 3D scanner and a laser rangefinder to monitor the geometric parameters of the work surface in real time during construction. For masonry work, the monitoring includes the verticality of the wall. Flatness mortar joint thickness For plastering operations, the thickness of the plaster layer should be checked. and surface flatness ,

[0061] Structured light scanning acquires 3D point cloud data of the construction surface. ,in To determine the number of point clouds, the system employs the Random Sample Consensus (RANSAC) algorithm for robust plane fitting. This algorithm fits a planar model by randomly sampling a subset of the point cloud and iteratively filters interior points, effectively eliminating noise and outlier interference to obtain stable planar parameters.

[0062] Wall verticality The flatness is obtained by calculating the angle deviation between the normal vector of the fitted plane and the direction of gravity (vertically upward); The surface undulation is characterized by the maximum deviation or standard deviation of the distance from all points to the fitted plane; grout joint thickness. This is obtained by identifying the edges of adjacent blocks and measuring the spacing.

[0063] Let the detected quality index vector be... The quality requirements defined in the BIM model Compare and calculate the quality deviation:

[0064]

[0065] When the deviation exceeds the allowable tolerance, the system triggers a correction mechanism.

[0066] 4-2 Dynamic correction of construction parameters

[0067] Based on the detected quality deviations, the system dynamically adjusts the construction parameters using a PID control strategy. PID control is chosen because it is the most mature and reliable feedback control method in industrial control, with a simple structure, easily adjustable parameters, and low requirements for the system model, making it very suitable for closed-loop control of continuously adjustable variables such as construction parameters. Taking plastering as an example, when the plaster thickness is detected to be too thin, the amount of subsequent plastering is increased; when local bulges are detected, the amount of plastering in that area is reduced or the troweling intensity is increased.

[0068] This invention employs incremental PID control to calculate the correction amount for construction parameters. Compared to positional PID, incremental PID only outputs the increment of the control quantity, facilitating amplitude limiting and preventing overshoot due to integral saturation. The formula for calculating the correction amount is:

[0069]

[0070] in For the first Quality deviation at any given time (the difference between the detected value and the target value). and This is the deviation between the first two moments. It is a proportionality coefficient (which determines the speed of response to changes in the current deviation). The integral coefficient (to eliminate steady-state error). The differential coefficient (to suppress the trend of deviation change). This refers to the adjustment amount for construction parameters (such as plastering amount, pressure, and speed).

[0071] The system sends the corrected parameters to the robot controller in real time, forming a closed-loop control loop of "detection → analysis → correction → execution". Through continuous quality monitoring and dynamic adjustment, the system ensures that the quality of the entire construction process is under control and that the final result meets the design requirements.

[0072] Intelligent construction operation system based on BIM-robot collaboration

[0073] The system includes a BIM task parsing module, a high-precision positioning module, a path planning module, and a quality inspection module.

[0074] Data flow between modules: After the BIM task parsing module parses the BIM model, it outputs a task sequence. and coordinates of each station The coordinates are then passed to the high-precision positioning module. After the high-precision positioning module completes the coordinate system alignment, it converts the stationary point coordinates in the BIM coordinate system into the target navigation point in the robot coordinate system, while also providing the robot's real-time pose. The data is then passed to the path planning module; the path planning module calculates the process compliance function based on the target point, real-time pose, and process constraints. The optimal work trajectory is output to the execution controller; the quality inspection module collects quality indicators in real time during construction. Calculate the deviation The correction amount is then calculated using a PID controller. This feedback is sent to the execution controller to adjust the construction parameters.

[0075] The system adopts a layered architecture: the decision-making layer is responsible for task scheduling and exception handling, the planning layer is responsible for path calculation and parameter optimization, and the execution layer is responsible for motion control and sensor data acquisition. The layers communicate with each other through standard interfaces, using a publish / subscribe pattern to achieve real-time data sharing.

[0076] The system deploys edge computing servers for real-time data processing and synchronizes construction progress and quality data with the BIM cloud platform via a 5G network. After construction is completed, the system writes the measured data back into the BIM model to form a finished model, providing data support for subsequent operation and maintenance management.

[0077] Through the above technical solution, this invention achieves deep collaboration between BIM digital models and construction robots. The four core modules work closely together and are seamlessly connected to construct an automated closed-loop process from design to construction. This solves the problems of model and execution disconnect, insufficient positioning accuracy, lack of process constraints and quality feedback in traditional construction robot systems, and provides a complete technical solution for intelligent building construction.

[0078] Compared with the prior art, the present invention has the following advantages:

[0079] 1. This invention utilizes BIM model semantic parsing and automatic task generation technology to automate the process from digital design to robot execution. Without manual programming or teaching, the system can directly extract construction tasks from the BIM model and generate robot execution instructions, significantly reducing construction preparation time and labor costs. When design changes occur, simply updating the BIM model automatically adjusts the robot's operational plan.

[0080] 2. High-precision positioning ensures construction accuracy. Through multi-source fusion positioning technology and precise coordinate system alignment, the system achieves centimeter-level positioning accuracy on the construction site. The robot's working position precisely corresponds to the BIM design position, ensuring the construction quality of tasks requiring high positional accuracy, such as masonry and welding. Compared to traditional manual surveying and layout methods, this is more accurate and efficient.

[0081] 3. Process-constrained path planning ensures construction quality. This invention integrates construction process requirements into the path planning algorithm. The generated work trajectory is not only highly efficient in obstacle avoidance, but more importantly, it conforms to building construction specifications. Through process compliance evaluation and multi-objective optimization, it ensures that the robot performs construction in the correct sequence and manner, guaranteeing construction quality from the source.

[0082] 4. Real-time closed-loop quality control enables process control. Unlike traditional post-construction quality acceptance, this invention performs quality inspection and parameter correction simultaneously during construction. Any deviations are immediately adjusted, eliminating quality problems at their inception, avoiding rework and waste, and achieving a quality management upgrade from "post-construction inspection" to "process control."

[0083] 5. Improved construction efficiency and safety: Robots can operate continuously 24 hours a day, unaffected by fatigue, weather, or other factors, increasing construction efficiency by more than 30% compared to manual labor. Simultaneously, robots handle high-altitude, repetitive, and highly dangerous tasks, improving the working environment for workers and reducing the risk of construction safety accidents.

[0084] 6. Promoting the digital transformation of the construction industry, this invention organically integrates BIM models, construction robots, and quality data, with construction process data being written back to the BIM platform in real time, forming a digital management system covering the entire lifecycle from design to construction to completion. This provides technical support for the construction industry's transformation from labor-intensive to technology-intensive, and promotes the development of intelligent construction.

[0085] 7. Excellent scalability and adaptability: The system architecture of this invention adopts a modular design, which can be easily expanded to various construction operation types (masonry, plastering, spraying, welding, 3D printing, etc.). By changing the end effector and adjusting the process parameters, the same system can adapt to different construction tasks. The positioning fusion algorithm can flexibly configure sensor combinations according to site conditions, exhibiting good scene adaptability.

[0086] 8. Significant economic benefits: Although the initial investment in the robotic system is high, long-term operation can achieve significant economic benefits by improving construction efficiency, reducing labor costs, and minimizing quality problems and rework losses. It is particularly suitable for construction projects with high standardization and repetitive tasks, enabling large-scale application and cost amortization. Attached Figure Description

[0087] Figure 1 This is a schematic diagram of the process of the present invention;

[0088] Figure 2 This is a schematic diagram of the architecture of the present invention. Detailed Implementation

[0089] To enhance understanding of the present invention, the embodiments will be described in detail below with reference to the accompanying drawings.

[0090] Example 1: See Figure 1 , Figure 2 This embodiment details the implementation process of a BIM-robot collaborative intelligent construction operation system, using interior masonry and plastering operations of a residential building project as an application scenario. The implementation process strictly follows the technical solution described in the invention, sequentially completing five stages: hardware deployment, BIM task analysis, high-precision positioning, path planning, and quality closed-loop control. The data flow and module integration at each stage are completely consistent with the invention. The following sections will explain in detail the implementation process of each module and their interrelationships, using specific numerical values ​​and examples.

[0091] 1 System Hardware Configuration and Deployment

[0092] The construction robot system consists of four main hardware components: a mobile platform, a robotic arm, an end effector, and a sensor system.

[0093] Mobile platform configuration: - Chassis type: Mecanum wheel omnidirectional mobile chassis, maximum load capacity 200kg - Moving speed: maximum 1.0m / s, 0.3m / s during operation - Positioning sensors: 16-line LiDAR (detection range 100m), binocular vision camera (resolution 1280×720), UWB positioning tag (positioning accuracy ±10cm)

[0094] Robotic arm configuration: - Model: Six-axis industrial robotic arm, load capacity 10kg - Working radius: m - Repeatability: ±0.1mm - Maximum joint speed: 180° / s

[0095] End effector configuration: - Masonry work: equipped with a clamp and mortar spray head - Plastering work: equipped with a trowel and hopper

[0096] Quality inspection equipment: - Structured light 3D scanner: accuracy ±0.5mm, scanning range 1m×1m - Laser rangefinder: accuracy ±1mm, measuring range 0.05-30m

[0097] Before construction, four UWB base stations were deployed on the construction floor to form a positioning coverage network. Total station targets were set up at key control points for coordinate system alignment and calibration. Edge computing servers were deployed in temporary workshops and connected to the BIM cloud platform via a 5G network.

[0098] 2. BIM Task Analysis and Implementation

[0099] After the hardware system is deployed, the first step is to perform BIM task parsing, converting the design model into a sequence of tasks that the robot can execute. This stage strictly follows the method described in the BIM construction task parsing and conversion section of the invention.

[0100] 2.1 Model Import and Semantic Parsing

[0101] Import the IFC format BIM model of this residential building project and extract wall component information using the IFC parsing engine. In this embodiment, a standard floor includes... A wall to be constructed. Taking wall number W-12 as an example, according to the geometric attribute format defined in the invention description. The analysis yielded the following:

[0102]

[0103] Simultaneously, in accordance with the construction attribute format defined in the invention content. The construction attributes were obtained through analysis. ,in For material type, This refers to the type of construction procedure. This refers to the surface quality requirement level.

[0104] 2.2 Construction Area Division

[0105] Based on the area division formula in the intelligent division section of the invention, combined with the working radius of the robotic arm... m and region overlap width m, for wall W-12 (length) m, height m) Divide the region:

[0106]

[0107]

[0108] That is, the wall is divided into Each work unit corresponds to 6 robot docking points. Each work unit has a horizontal dimension of 2.0m and a vertical dimension of 1.4m. Since the wall extends along the x-axis (y-coordinate is constant at 5.0m), the robot moving platform needs to dock sequentially in front of the wall (offset approximately 0.8m in the y-direction to allow for working space). The planar positions of each docking point are as follows. The corresponding operating height ranges are as follows (wherein) For horizontal indexing, (for vertical indexing)

[0109]

[0110] 2.3 Task Sequence Generation

[0111] Based on the masonry construction principle of "bottom to top, left to right," a sequence of tasks is generated. Task Sequence Organize according to the format defined in the invention description:

[0112]

[0113] Total number of tasks in this embodiment Each task Includes: target station coordinates Parameters such as the range of masonry layers, block type, and mortar mix ratio. Task Corresponding element (1,1), stationary point coordinates m, when the robot is in this position, the robotic arm can cover an area of ​​3.0m to 5.0m in the x direction of the wall and 0 to 1.4m in height, requiring a total of 7 layers of blocks (block height 200mm).

[0114] After BIM task analysis is completed, the task sequence is... The coordinate information of each station point is transmitted to the high-precision positioning module for coordinate system transformation and positioning guidance.

[0115] 3. Implementation of high-precision positioning

[0116] After the BIM task parsing is completed, the system obtains the task sequence. And the coordinates of each station in the BIM coordinate system. Next, the high-precision positioning module needs to complete two tasks: first, to determine the precise position of the robot on the construction site in real time; and second, to establish the transformation relationship between the BIM coordinate system and the robot coordinate system.

[0117] 3.1 Multi-sensor data acquisition

[0118] After the robot starts, it simultaneously collects data from three positioning sensors. Laser SLAM outputs pose estimates at a frequency of 10Hz. The positioning accuracy is approximately ±3cm. The visual odometry outputs relative displacement at a frequency of 30Hz. While offering high accuracy over short distances, the UWB system suffers from cumulative drift. It outputs absolute position at a frequency of 5Hz. The accuracy is approximately ±10cm.

[0119] 3.2 EKF Fusion Positioning

[0120] Following the method described in the Extended Kalman Filter (EKF) fusion section of the invention, the EKF algorithm is used to fuse multi-source localization information. EKF combines motion model predictions with sensor observations through a two-step "prediction-update" iterative process, outputting the optimal state estimate. (System state vector) The initial state is obtained by total station calibration.

[0121] Prediction Phase: Calculate the predicted state based on the robot's kinematic model. Let the time step be... s, control input Obtained from the motor encoder:

[0122]

[0123] According to the definition in the invention, the process noise covariance matrix Set as The diagonal elements correspond to the positions respectively. ,Location Heading angle linear velocity angular velocity The process noise variance reflects the uncertainty of the motion model.

[0124] Update Phase: Upon receiving sensor observation data, the predicted value and the observed value are fused using Kalman gain weighting. Observation Vector Includes the position and heading of the laser SLAM and the position of the UWB. Observation noise covariance matrix. Based on the accuracy settings of each sensor The first three terms correspond to laser SLAM measurement noise (accuracy ±3cm), and the last two terms correspond to UWB measurement noise (accuracy ±10cm). Sensors with smaller noise variance receive greater weight during fusion.

[0125] Based on the state update formula in the invention, the updated state estimate is as follows:

[0126]

[0127] After EKF fusion, the positioning accuracy is improved to ±2cm, which meets the construction accuracy requirements.

[0128] 3.3 Coordinate System Alignment

[0129] Before construction, a total station was used to measure the coordinates of four control points in both the BIM coordinate system and the robot coordinate system:

[0130]

[0131] According to the coordinate transformation formula in the invention The translation vector is obtained by solving for the coordinate transformation parameters using the least squares method. Rotation matrix It approaches the identity matrix (with only a tiny angular deviation, approximately 0.1°). Subsequently, the robot's real-time position can be precisely mapped to the BIM coordinate system using this transformation relationship.

[0132] After the high-precision positioning module completes coordinate alignment, bidirectional conversion between the BIM coordinate system and the robot coordinate system can be achieved. For task sequences... BIM coordinates of each station Through inverse transformation Convert to target navigation point in robot coordinate system (due to (It is an orthogonal matrix, whose inverse is equal to its transpose), and is passed to the path planning module for trajectory calculation.

[0133] 4. Implementation of Path Planning

[0134] After receiving the target navigation point output by the high-precision positioning module, the path planning module begins to calculate the operation trajectory that meets the process constraints.

[0135] 4.1 Definition of process constraints: For masonry operations, based on the process constraint modeling method in the invention, the following specific parameters are defined: Effective time window for mortar: s, Block placement horizontal accuracy: mm, vertical accuracy of block placement: mm, staggered joint distance between adjacent layers: mm,

[0136] Based on the process compliance function defined in the invention, the weighting coefficient in this embodiment is set to... , , (satisfy The process compliance function is specified as follows:

[0137]

[0138] in For task sequence The first in One task. Time constraint penalty. The work time exceeds the effective time of the mortar. A penalty is incurred at time s; precision constraint penalty term. When the positioning deviation exceeds the allowable accuracy (horizontal) mm or vertical Penalties are incurred when the time exceeds mm; process sequence penalty items Penalties will be imposed for violating the "bottom-up" principle.

[0139] 4.2 Path Search and Optimization

[0140] Following the method described in the multi-objective path optimization section of the invention, an improved A* algorithm is used to plan the travel path of the mobile platform between various docking points. The core improvement lies in adding a process constraint penalty term to the cost function of the traditional A* algorithm. Taking path planning from stationary point (1,1) to stationary point (2,1) as an example, the comprehensive cost function is:

[0141] ,

[0142] in The cost of the actual distance traveled. The heuristic value is the Euclidean distance. Based on whether the path traverses an unfinished construction area and whether it causes the mortar to fail due to timeout (exceeding...) Calculations based on factors such as s). If the path meets the process requirements, then... Otherwise, a larger penalty value is assigned to guide the algorithm to avoid obstacles. Through this improvement, the path output by the algorithm avoids obstacles while also satisfying construction process constraints.

[0143] For the trajectory of the robotic arm's end effector, a fifth-order polynomial is used for trajectory interpolation according to the method described in the invention. The fifth-order polynomial can constrain six boundary conditions, including the position, velocity, and acceleration of the start and end points, ensuring no impact vibration during start-up and shutdown. Taking the trajectory of placing a single block as an example, the motion time from the gripping point to the placement point is set to 2 seconds, and the velocities and accelerations at the start and end points are both set to zero. This determines the six coefficients of the fifth-order polynomial, generating a smooth Cartesian space trajectory.

[0144] 5. Implementation of quality inspection and control.

[0145] While the robot performs masonry work according to the trajectory generated by the path planning module, the quality inspection module simultaneously monitors the quality of the work surface, forming a closed-loop control loop of "detection → analysis → correction → execution" as described in the invention.

[0146] 5.1 Real-time quality inspection,

[0147] After each layer of masonry is completed, a structured light scanner scans the constructed wall surface to acquire 3D point cloud data. Following the method described in the real-time construction quality monitoring section of the invention, the RANSAC algorithm is used for robust plane fitting. After eliminating noise points, the best-fitting plane for the wall surface is obtained. Based on the fitted plane, the quality index, wall verticality, is calculated. The flatness is calculated by the angle deviation between the plane normal vector and the vertical direction. The mortar joint thickness is calculated by the maximum distance deviation from the point to the fitted plane. Measurements are taken by identifying the distance between the edges of the building blocks.

[0148] Assuming the detected quality indicator is: wall verticality: mm / m (measured deviation value), wall surface flatness: mm (actual deviation value), mortar joint thickness: mm (actual thickness value)

[0149] The quality objectives defined in the BIM model are: The target values ​​for verticality and flatness are 0 (indicating no deviation), and the target value for mortar joint thickness is 10mm. Allowable tolerances are defined as verticality ≤ 3mm / m, flatness ≤ 4mm, and mortar joint thickness 10±1mm.

[0150] Calculate the quality deviation (the difference between the measured value and the target value):

[0151]

[0152] The verticality deviation of 3mm / m and the flatness deviation of 4mm are both within the allowable tolerance range, but the grout joint thickness deviation of 1mm (actual measurement 11mm, target 10mm) exceeds the allowable range (±1mm), triggering the correction mechanism.

[0153] 5.2 Dynamic parameter correction,

[0154] To address the issue of excessive mortar joint thickness, construction parameters are adjusted according to the incremental PID control formula in the dynamic correction section of the invention. Assuming the target mortar joint thickness is 10mm, the current deviation... mm (th (Time-based detection value 11mm minus target value 10mm), previous deviation mm, deviation from the previous period mm.

[0155] According to the PID control formula in the invention Set the scaling factor Integral coefficient Differential coefficients Calculate the mortar volume adjustment:

[0156]

[0157] The system reduces the amount of mortar sprayed by 0.35 relative units, making the thickness of the mortar joints in subsequent masonry work approach the target value. This correction instruction is sent to the execution controller in real time through a closed-loop control circuit, reflecting the complete closed-loop mechanism of "detection → analysis → correction → execution" in the invention.

[0158] 6. Construction process and results

[0159] By integrating the collaborative work of the four modules—BIM task analysis, high-precision positioning, path planning, and quality closed-loop control—the robot follows the planned task sequence. Complete the masonry work on wall W-12 sequentially. The entire wall construction will take approximately 4.5 hours and includes 6 checkpoints (corresponding to 6 tasks). to A total of approximately 420 blocks were placed.

[0160] Construction quality statistics: Wall verticality: average 2.1mm / m, maximum 3.2mm / m, meets Class I standard (≤4mm / m); Wall flatness: average 2.8mm, maximum 4.1mm, meets Class I standard (≤5mm); Grout thickness: average 9.8mm, standard deviation 0.8mm, meets specification requirements (8-12mm).

[0161] Compared to manual bricklaying, robotic bricklaying is about 30% more efficient (about 6 hours for manual work) and significantly improves quality stability (the standard deviation of mortar joint thickness for manual work is about 2mm).

[0162] After construction is completed, the system uploads the measured point cloud data and quality inspection report to the BIM cloud platform and updates the as-built model. Project managers can view the construction progress, quality indicators, and deviation distribution on the BIM platform, realizing digital management of the construction process.

[0163] 7. Multi-process collaboration,

[0164] After the masonry work is completed, the same wall enters the plastering process, demonstrating the multi-process collaborative scheduling capability of this invention. The system replaces the end effector of the robotic arm with a plastering trowel and a hopper. The BIM task parsing module regenerates the task sequence according to the plastering process specifications, the high-precision positioning module continues to provide real-time pose, and the path planning module and quality inspection module work collaboratively according to the method described in the invention.

[0165] Plastering is done from top to bottom, the opposite of masonry, to prevent the plaster from contaminating the finished surface below. The BIM task parsing module regenerates the task sequence according to the plastering process specifications:

[0166]

[0167] This task sequence is related to the bricklaying task sequence. The order is exactly reversed, reflecting the differences in process between different steps. During the plastering process, the quality inspection module simultaneously detects the thickness of the plaster layer. and surface flatness The target plaster thickness is 15mm, with an allowable deviation of ±3mm. The system dynamically adjusts the plastering trowel pressure and movement speed based on the detection results, according to the PID control strategy described in the invention, to ensure uniform plaster layer thickness.

[0168] By coordinating the masonry and plastering processes, the system achieved a fully automated construction process from BIM design to finished wall structure. Throughout the implementation, the four core modules—BIM task analysis, high-precision positioning, path planning, and quality inspection—worked closely together and seamlessly connected according to the data flow relationships defined in the invention, verifying the feasibility and effectiveness of the technical solution.

[0169] It should be noted that the above embodiments are not intended to limit the scope of protection of the present invention. Equivalent transformations or substitutions made based on the above technical solutions all fall within the scope of protection of the claims of the present invention.

Claims

1. A BIM-robot collaborative intelligent construction operation method, characterized in that, The method includes the following steps: Step 1: BIM construction task analysis and conversion. Step 2: Multi-source fusion high-precision positioning. Step 3: Process constraint path planning. Step 4: Real-time quality detection and closed-loop control.

2. The intelligent construction operation method based on BIM-robot collaboration according to claim 1, characterized in that, Step 1 is as follows: 1-1 Semantic analysis of BIM model The system extracts semantic information of building components by parsing BIM models in IFC (Industry Foundation Classes) or Revit native format. For wall components to be constructed, the system extracts their spatial bounding box, material type, thickness parameters, and surface treatment requirements. Assuming the BIM model contains a total of... The first wall component to be constructed, the first The geometric properties of a wall are represented as follows: ,in The length of the wall. The height of the wall. For wall thickness, and These are the three-dimensional coordinates of the wall's start and end points in the BIM coordinate system. The system also extracts the construction attributes of the wall. ,in Indicates the material type. Indicates the type of construction procedure. Indicates the surface quality requirement level. 1-2 Intelligent division of construction area: For large-area walls, the system divides the wall into several working units based on the robot's operating range and efficiency. The core basis for this division is the effective working radius of the robot's robotic arm. When the robot is at a certain stationary position, the robotic arm can cover a circle centered on that point. For an area with a radius of [radius value], to ensure full wall coverage and no omissions in adjacent areas, it is necessary to reasonably set the spacing between anchor points and the overlap width. For length of Height is The walls are divided horizontally into The area is divided vertically into several regions. Areas: In the above formula, the effective width of each region in the horizontal direction is (The robotic arm covers both sides) (minus the overlapping parts) The overlap width between adjacent regions; In the vertical direction, since robotic arms typically operate from bottom to top, the height of each area is taken as follows: ,symbol This indicates rounding up to the nearest whole number, ensuring the area completely covers the wall, and the center position of each work unit. As the target station coordinates of the robot mobile platform, For horizontal indexing, Using vertical indexing, this partitioning method breaks down the construction task of the entire wall into a sequence of sub-tasks that the robot can execute one by one. 1-3 Task sequence generation, According to construction process specifications, the system generates a sequence of tasks in a specific order. For masonry work, the principle of "from bottom to top, from one end to the other" is followed; for plastering work, the principle of "from top to bottom, preventing contamination of finished surfaces" is followed. Defined as an ordered set of tasks, where The total number of tasks, each task contains objectives. Mark the coordinates of the stationary point Based on the job type, process parameters, and quality requirements, the BIM task parsing module will output a task sequence. The location information of each station in the BIM coordinate system will be transmitted to the high-precision positioning module for coordinate transformation.

3. The intelligent construction operation method based on BIM-robot collaboration according to claim 2, characterized in that, Step 2 is as follows: 2-1 Multi-sensor localization information acquisition: The system is equipped with three types of localization sensors: LiDAR, binocular vision camera, and UWB tag. The LiDAR SLAM module constructs a point cloud map of the construction site using two-dimensional or three-dimensional LiDAR and estimates the robot pose. ,in , For planar coordinates, For the heading angle, visual odometry calculates the relative displacement by matching feature points in consecutive frames of images. The UWB positioning system measures the distance to the robot using multiple base stations deployed at the construction site and estimates its position using a trilateration algorithm. , 2-2 Extended Kalman Filter Fusion: The Extended Kalman Filter (EKF) algorithm is used to fuse multi-source localization information. The system state vector is defined as follows: ,in , For position coordinates, For heading angle, Linear velocity, Angular velocity, The state prediction equation is based on a kinematic model: in For time step, To control the inputs (including linear velocity and angular velocity), the covariance prediction is: ,in Let be the state covariance matrix of the previous time step. To predict the state covariance matrix, The Jacobian matrix of the state transition matrix (for...) Regarding the status (Find the partial derivative) The process noise covariance matrix is... When sensor observation data is received, a measurement update is performed, and the observation vector is updated. Including measurements from laser SLAM, visual odometry, and UWB, the Kalman gain is calculated as follows: ,in Let Jacobian matrix be the observation matrix. To observe the noise covariance matrix, the state is updated as follows: in As the observation function, the system dynamically adjusts the weights of each sensor through the EKF fusion algorithm. When the signal quality of a sensor deteriorates, its contribution is automatically reduced to ensure stable positioning accuracy under complex conditions such as occlusion and interference. 2-3. BIM Coordinate System Alignment: Before construction begins, the system uses a total station to measure the BIM coordinates and corresponding robot coordinates of several control points (no fewer than 4) at the construction site. Let the first... The coordinates of each control point in the BIM coordinate system are: The coordinates in the robot coordinate system are The system uses the least squares method to solve for the coordinate transformation parameters. Its core idea is to find the optimal rotation matrix. Translation vector To minimize the sum of squared transformation errors for all control points, specifically by minimizing the objective function... The rotation matrix is ​​solved using singular value decomposition (SVD), and the translation vector is calculated based on the centroid alignment principle. Finally, the translation vector is obtained. and rotation matrix (Orthogonal matrix) such that any point Coordinates in the robot coordinate system can be converted to coordinates in the BIM coordinate system: ,because Since it is an orthogonal matrix, its inverse is equal to its transpose. Therefore, BIM coordinates can also be converted to robot coordinates through inverse transformation. , Through the aforementioned coordinate alignment mechanism, the high-precision positioning module achieves bidirectional conversion between the BIM coordinate system and the robot coordinate system: on the one hand, it converts the robot's real-time pose... The position components in the image are transformed to the BIM coordinate system through a positive transformation, i.e., the position... heading angle (in Rotation matrix The corresponding rotation angle around the z-axis, and They are respectively The elements in the second row and first column, and the first row and first column, are used for progress tracking and quality recording; on the other hand, the target stationary point coordinates output by the BIM task parsing module are... These points are converted into navigation target points in the robot coordinate system. These target points are then passed to the path planning module to generate the operation trajectory from the current position to each target point.

4. The intelligent construction operation method based on BIM-robot collaboration according to claim 3, characterized in that, Step 3 is as follows: 3-1 Construction process constraint modeling, Different types of construction work have different technological constraints. For masonry work, the main constraints include: the effective time window after mortar laying. Horizontal and vertical accuracy of brick placement and Distance between staggered joints of adjacent layers For plastering operations, constraints include: plaster thickness range Single application coverage width Consistent calendering direction The system formalizes process constraints into optimization objectives and constraints for path planning, and defines a process compliance function. Evaluation of the degree to which the evaluation path meets the process requirements: in For task sequence The total number of tasks in the process. For the first One task, For the task Time constraint penalty (when the operation time exceeds the material's expiration date) (Punishment will be imposed at that time) A penalty term for accuracy constraints (when the positioning deviation exceeds the allowable accuracy). or (Punishment will be imposed at that time) This is a penalty item for process sequence constraints (a penalty is incurred when the process sequence such as "bottom first, top last" is violated). , , For the corresponding weight coefficients, satisfying By minimizing The path planner generates the optimal job sequence that meets the process requirements; 3-2 Multi-objective path optimization, An improved A* algorithm combined with process constraints is used for path search. Define the comprehensive cost function ,in This represents the actual travel cost from the starting point to the current node (considering distance and energy consumption). The heuristic cost estimate (using Euclidean distance) from the current node to the target point is calculated. The process constraint cost is calculated based on whether the current path meets the construction process requirements: if the path traverses an unfinished construction area, causes mortar to expire or fails, or violates the work sequence, a higher cost is assigned; the cost of a path that meets the process requirements is zero. In this way, the algorithm automatically avoids paths that do not meet the process requirements during the search process and outputs the optimal path that meets both obstacle avoidance requirements and construction specifications. The work trajectory output by the path planning module will be executed by the robot controller, driving the mobile platform and robotic arm to complete specific construction actions. During the construction process, the system needs to monitor the construction quality in real time and adjust subsequent construction parameters based on the detection results. This function is realized by the real-time quality detection and closed-loop control module.

5. The intelligent construction operation method based on BIM-robot collaboration according to claim 2, characterized in that, Step 4 is as follows: 4-1 Real-time monitoring of construction quality. The system is equipped with a structured light 3D scanner and a laser rangefinder to monitor the geometric parameters of the work surface in real time during construction. For masonry work, the monitoring includes the verticality of the wall. Flatness mortar joint thickness For plastering operations, the thickness of the plaster layer should be checked. and surface flatness , Structured light scanning acquires 3D point cloud data of the construction surface. ,in To determine the number of point clouds, the system employs the Random Sample Consensus (RANSAC) algorithm for robust plane fitting. This algorithm fits a planar model by randomly sampling a subset of the point cloud and iteratively filters interior points, effectively eliminating noise and outlier interference to obtain stable planar parameters. Wall verticality The flatness is obtained by calculating the angle deviation between the normal vector of the fitted plane and the direction of gravity (vertically upward); The surface undulation is characterized by the maximum deviation or standard deviation of the distance from all points to the fitted plane; grout joint thickness. This is obtained by identifying the edges of adjacent blocks and measuring the spacing. Let the detected quality index vector be... The quality requirements defined in the BIM model Compare and calculate the quality deviation: When the deviation exceeds the allowable tolerance, the system triggers a correction mechanism. 4-2 Dynamic correction of construction parameters Based on the detected quality deviation, the system uses a PID control strategy to dynamically adjust the construction parameters. Incremental PID control is used to calculate the correction amount of the construction parameters, and the formula for calculating the correction amount is as follows: in For the first Quality deviation at any moment and This is the deviation between the first two moments. This is the proportionality coefficient. The integral coefficient is... The differential coefficients are... This refers to the adjustment amount of construction parameters. The system sends the corrected parameters to the robot controller in real time, forming a closed-loop control loop of "detection → analysis → correction → execution". Through continuous quality monitoring and dynamic adjustment, the system ensures that the quality of the entire construction process is under control and that the final result meets the design requirements.

6. A BIM-robot collaborative intelligent construction operation system, characterized in that, The system includes a BIM task parsing module, a high-precision positioning module, a path planning module, and a quality inspection module. Data flow between modules: After the BIM task parsing module parses the BIM model, it outputs a task sequence. and coordinates of each station This information is then transmitted to the high-precision positioning module. After the high-precision positioning module completes the coordinate system alignment, it converts the stationary point coordinates in the BIM coordinate system into the target navigation point in the robot coordinate system, while providing the robot's real-time pose. The data is then passed to the path planning module; the path planning module calculates the process compliance function based on the target point, real-time pose, and process constraints. The optimal work trajectory is output to the execution controller; The quality inspection module collects quality indicators in real time during construction. Calculate the deviation The correction amount is then calculated using a PID controller. This feedback is sent to the execution controller to adjust the construction parameters.

7. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the intelligent construction operation method based on BIM-robot collaboration as described in any one of claims 1 to 5.

8. A computer-readable storage medium storing computer instructions thereon, characterized in that, When the computer instructions are executed by the processor, they implement the intelligent construction operation method based on BIM-robot collaboration as described in any one of claims 1-5.

Images