A mechanical arm path planning method and system for construction work

Abstract

The application discloses a kind of mechanical arm path planning method and system for construction, it is related to mechanical arm movement planning technical field, including acquisition multi-source original data is registered with structured preprocessing in space, generates standardization observation data;The absolute space pose of each target operation point determined with the standardization observation data is fused and associated, and operation state data is output;According to the operation state data, the candidate path of end effector in target operation point is calculated using kinematic model of mechanical arm;The candidate path is corrected with safety constraint, and the final path instruction of feasible region constraint is generated.The method of the application solves multiple candidate paths using kinematic model of mechanical arm, and generates final path instruction under safety constraint and feasible region limitation, realizes high precision, strong robustness and safety executable of construction operation mechanical arm path planning, significantly improves operation efficiency and stability under complex environment.

Description

Technical Field

[0001] This invention relates to the field of robotic arm motion planning technology, specifically to a path planning method and system for a robotic arm used in construction operations. Background Technology

[0002] With the continuous development of building automation and intelligent construction technologies, robotic arms are widely used in high-risk, repetitive, and precision-critical construction operations. In recent years, construction robots have gradually shifted from fixed work scenarios to complex environmental scenarios, making path planning technology a crucial element for achieving high-precision operations. Traditional path planning often relies on preset trajectories or offline calculations based on CAD models, making it difficult to respond in real-time to dynamic changes at the construction site. With advancements in multi-source sensing and spatial perception technologies, path planning methods integrating visual, depth, and posture information have emerged, providing new technological directions for construction robotic arms to achieve dynamic obstacle avoidance, autonomous planning, and safe execution. However, the data redundancy and high noise levels in the construction environment mean that multi-source information fusion and path accuracy control remain critical technical challenges that urgently need to be addressed.

[0003] Existing path planning methods for construction robotic arms generally suffer from low data fusion accuracy, unstable path calculation, and insufficient safety constraints. Firstly, traditional solutions typically rely on data input from a single sensor or a static model, lacking spatial registration and structured processing of multi-source raw data. This leads to positional errors and temporal mismatches in the input information for path planning, making it difficult to generate a unified, standardized data support environment. Secondly, while some methods incorporate kinematic modeling, they fail to combine it with dynamic operational state data for real-time solution. This results in poor reachability of the inverse kinematic solution set, frequently leading to path discontinuities, posture jumps, and singular configurations, affecting the smoothness and stability of the operation. Thirdly, existing safety constraints are generally set offline or controlled by static thresholds, failing to provide real-time correction for the dynamic positions of obstacles, equipment, and personnel in complex construction scenarios. This results in collision risks and reduced operational efficiency during the path execution phase. Summary of the Invention

[0004] In view of the above-mentioned problems, the present invention is proposed.

[0005] Therefore, the technical problem solved by this invention is that existing construction operation robotic arm path planning methods suffer from low accuracy of multi-source data fusion, poor continuity of path solving, and sluggish response to safety constraints, as well as the problem of how to achieve high-precision path planning and safety command generation based on multi-source perception in complex construction environments.

[0006] To address the aforementioned technical problems, this invention provides the following technical solution: a path planning method for a robotic arm used in construction operations, comprising: collecting multi-source raw data for spatial registration and structured preprocessing to generate standardized observation data; fusing and associating the standardized observation data with the absolute spatial pose of each determined target work point to output work status data; calculating candidate paths for the end effector at the target work point using a robotic arm kinematic model based on the work status data; and performing safety constraint corrections on the candidate paths to generate a final path instruction with feasible region constraints.

[0007] As a preferred embodiment of the path planning method for a robotic arm used in construction operations according to the present invention, the acquisition of multi-source raw data includes acquiring data on the robotic arm's working environment and its own operating status using a vision acquisition device, a depth measurement device, a posture detection device, and a displacement measurement device.

[0008] The visual acquisition device is used to acquire two-dimensional image information of the construction site, continuously capture the scene of the work area and generate time-series image frames;

[0009] The depth measurement device is used to collect depth information at the construction site and output corresponding depth matrix data to reflect the spatial distance distribution between the object and the robotic arm.

[0010] The attitude detection device is used to measure the angle changes of each joint of the robotic arm and the spatial attitude of the end effector, and generate attitude sequence data;

[0011] The displacement measuring device is used to measure the displacement changes of the robotic arm base and key components in real time during operation and output displacement trajectory data.

[0012] As a preferred embodiment of the path planning method for a robotic arm used in construction operations according to the present invention, the spatial registration includes mapping visual images, depth matrices, posture sequences and displacement trajectories in the same coordinate system, so that the geometric and temporal relationships between the multi-source raw data remain consistent.

[0013] After spatial registration is completed, the multi-source raw data undergoes structured preprocessing, including format unification, field parsing, and structure reorganization of the registered data.

[0014] Extract key fields from each data source; delete redundant or incomplete records, repair sequence data with abnormal time intervals, and reorganize them into data tables according to a unified field standard;

[0015] The numerical precision, value range, and unit system are standardized during the structuring process.

[0016] As a preferred embodiment of the path planning method for a robotic arm used in construction operations according to the present invention, the fusion association includes retrieving a local region corresponding to the absolute spatial pose of the target work point based on the pixel points, depth values, attitude angles and displacement trajectory information in the standardized observation data.

[0017] Calculate the position deviation and attitude error in the local area corresponding to the absolute spatial pose of the target work point to determine the degree of matching between the standardized observation data and the target work point.

[0018] As a preferred embodiment of the path planning method for a robotic arm used in construction operations according to the present invention, the determination of the matching degree between standardized observation data and target operation points includes determining that the correspondence between the operation point and the observation data is valid when both the position deviation and the posture error are lower than a preset threshold, and marking the matching result as an effective fusion state.

[0019] If either the position deviation or the attitude error exceeds a preset threshold, a local adjustment process will be automatically triggered to dynamically correct the local spatial mapping of the work point.

[0020] Dynamically correcting the local spatial mapping of the work point includes recalculating the position deviation, attitude error, and depth matching degree in the neighborhood of the current position, and updating the spatial pose description of the target point based on the correction results;

[0021] When either the corrected position deviation or the attitude error meets the preset threshold condition again, the system automatically performs a matching confirmation operation and locks the current result;

[0022] If the preset threshold requirement is still not met after continuous adjustments, the weight will be reduced and the next round of fusion iteration will begin.

[0023] The preset threshold is determined based on the characteristics of the collected and processed data and the motion accuracy of the robotic arm itself.

[0024] The position deviation threshold is determined based on standardized observation data.

[0025] The attitude error threshold is set based on the attitude deviation of the robotic arm's end effector.

[0026] As a preferred embodiment of the path planning method for a robotic arm used in construction operations according to the present invention, the kinematic model of the robotic arm includes a forward kinematic model and an inverse kinematic model;

[0027] The forward kinematics model is used to calculate the position and orientation of the end effector based on the angles of each joint;

[0028] The inverse kinematics model is used to solve for the combination of values ​​of the joint variables of the robotic arm given the target pose of the end effector;

[0029] The kinematic model of the robotic arm is based on the link length, joint type, rotation axis direction and pose transformation matrix of the robotic arm, and expresses the motion characteristics of the robotic arm through joint parameterization.

[0030] As a preferred embodiment of the path planning method for a robotic arm used in construction operations according to the present invention, the calculation of the candidate path of the end effector at the target work point includes calculating the feasibility index of each candidate path, including reading the absolute spatial pose of each target work point and inputting it into the kinematic model of the robotic arm.

[0031] Based on the inverse kinematics model, the feasible solution set of each joint variable is solved, and the output is the achievable end effector posture combination;

[0032] For each feasible solution, invalid solutions caused by joint constraints, singular configurations, and mechanism interference are filtered out by a forward kinematics model, and candidate posture solutions that satisfy the kinematic constraints are output.

[0033] A continuous motion trajectory is generated between each candidate attitude solution to form a candidate path.

[0034] As a preferred embodiment of the path planning method for a robotic arm used in construction operations according to the present invention, the step of correcting the safety constraints of the candidate path includes establishing a system that performs constraint detection on each trajectory segment of the candidate path to determine whether there is any boundary crossing, collision, or speed exceeding the limit.

[0035] If the detection results meet the safety constraints, the trajectory segment is retained;

[0036] If the conditions are not met, a local correction is initiated to adjust the position or orientation of the trajectory points and return them to the safe and feasible region.

[0037] Another objective of this invention is to provide a path planning system for a robotic arm used in construction operations, which can generate a final path instruction with feasible domain constraints by correcting the candidate path for safety constraints. This solves the problem that current path planning methods for robotic arms used in construction operations cannot achieve high-precision path planning and safety instruction generation based on multi-source perception in complex construction environments.

[0038] As a preferred embodiment of the construction operation robotic arm path planning system of the present invention, it includes: a multi-source data acquisition and spatial registration preprocessing module, a work point pose fusion and state generation module, a kinematics solution and candidate path generation module, and a safety constraint correction and path command output module; the multi-source data acquisition and spatial registration preprocessing module is used to acquire multi-source raw data from the construction site, and generate standardized observation data in a unified format through spatial registration and structured preprocessing; the work point pose fusion and state generation module is used to fuse and correlate the standardized observation data with the absolute spatial pose of the target work point to generate work state data describing the relationship between the work environment and the target position; the kinematics solution and candidate path generation module is used to solve multiple sets of feasible trajectories of the end effector at the target work point based on the work state data and using the robotic arm kinematic model to form a candidate path set; the safety constraint correction and path command output module is used to apply safety and feasible domain constraints to the candidate paths, and generate the final path command after correcting the trajectory.

[0039] The beneficial effects of this invention are as follows: The standardized observation data provided by this invention for the path planning of a robotic arm in construction operations eliminates spatiotemporal and format differences within a unified coordinate and field system, significantly reducing input uncertainty. By fusing standardized observation data with absolute spatial pose, operational status data containing environmental features and quality markers is constructed, providing precise basis for target constraints and boundary conditions. Candidate paths are calculated based on the robotic arm's kinematic model, avoiding singular configurations and limit conflicts, obtaining continuous, reachable, and diverse solution sets, enhancing adaptability to dynamic working conditions. Safety constraints and feasible domain limitations are introduced to perform minimum disturbance correction and time parameterization on the path, transforming it into an executable final path instruction, suppressing collisions, velocity mutations, and boundary violations from the source. In summary, this invention achieves synergistic improvements in positioning accuracy, path smoothness, execution safety, and online stability compared to existing technologies, significantly reducing the frequency of replanning and on-site intervention costs, and improving the reliability and efficiency of operations in complex construction scenarios. Attached Figure Description

[0040] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0041] Figure 1 The first embodiment of the present invention provides an overall flowchart of a path planning method for a robotic arm used in construction operations. Detailed Implementation

[0042] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0043] Example 1, referring to Figure 1 As an embodiment of the present invention, a path planning method for a robotic arm used in construction operations is provided, comprising:

[0044] S1: Collect raw data from multiple sources, perform spatial registration and structured preprocessing, and generate standardized observation data.

[0045] Furthermore, data acquisition units are deployed at the construction site to collect multi-source raw data on the robotic arm's working environment and its own operating status.

[0046] The acquisition unit includes a vision acquisition device, a depth measurement device, a posture detection device, and a displacement measurement device; the devices work together to acquire multi-source raw information within the work space.

[0047] The vision acquisition device is used to acquire two-dimensional image information of the construction site, continuously capture the scene of the work area and generate time-series image frames; the depth measurement device is used to acquire depth information of the construction site and output the corresponding depth matrix data to reflect the spatial distance distribution between the object and the robotic arm; the posture detection device is used to measure the angle changes of each joint of the robotic arm and the spatial posture of the end effector, and generate posture sequence data; the displacement measurement device is used to measure the displacement changes of the robotic arm base and key components in real time during the operation and output displacement trajectory data.

[0048] Spatial registration of multi-source raw data maps visual images, depth matrices, pose sequences, and displacement trajectories to the same coordinate system, ensuring that the geometric and temporal relationships among the multi-source raw data remain consistent.

[0049] After spatial registration, the multi-source raw data undergoes structured preprocessing. This process includes: standardizing the format, parsing fields, and reorganizing the structure of the registered multi-source raw data; extracting each key field, such as position coordinates, attitude angle, depth pixel values, and time labels; deleting redundant or incomplete records, repairing multi-source raw data with abnormal time intervals, and reorganizing the data into a multi-source raw data table according to unified field standards; and standardizing numerical precision, value range, and unit system during the structuring process to ensure that the data can be directly accessed and analyzed under the same format.

[0050] Through the above spatial registration and structured preprocessing, standardized observation data conforming to a unified coordinate system and data specifications are generated.

[0051] S2: The standardized observation data is fused and correlated with the absolute spatial pose of each target operation point to output the operation status data.

[0052] Furthermore, in construction operation scenarios, multiple target operation points are pre-determined based on construction design drawings and task planning results. These target operation points refer to the specific spatial locations where the robotic arm needs to perform operations during construction tasks, including assembly points, welding points, grinding points, coating points, or inspection points. Each target operation point is defined by a unified spatial reference system, and its absolute spatial pose consists of three-dimensional position coordinates and attitude angle information, uniquely representing the target position and orientation of the robotic arm's end effector in the world coordinate system. Absolute spatial pose refers to the combination of the target operation point's three-dimensional position and attitude in the unified world coordinate system, represented by a position vector and attitude angles.

[0053] The standardized observation data is spatially fused with the target operation points in a one-to-one correspondence. By reading the absolute spatial pose of the target operation points, spatial mapping and feature matching are performed between them and the standardized observation data in a unified coordinate system.

[0054] During the fusion and association process, a local region corresponding to the absolute spatial pose of the target work point is retrieved based on the standardized observation data; then, the position deviation and attitude error are calculated within this region to determine the degree of matching between the standardized observation data and the target work point.

[0055] When both position deviation and attitude error are below preset thresholds, the correspondence between the target work point and the standardized observation data is determined to be valid, and the matching result is marked as a valid fusion state for subsequent work status data generation. If either error exceeds the preset threshold range, the control terminal automatically triggers a local adjustment process to dynamically correct the local spatial mapping corresponding to the target work point. Local adjustment includes recalculating the position deviation, attitude error, and depth matching degree within the current location neighborhood, and updating the spatial pose description of the target work point based on the correction results. When the corrected error meets the preset threshold condition again, the system automatically performs a matching confirmation operation and locks the current result. If continuous correction still fails to meet the threshold requirement, the system reduces the fusion weight of the target work point and enters the next round of fusion iteration. Through the above judgment and local adjustment mechanism, the matching process achieves self-checking, self-correction, and self-convergence, ensuring the accuracy and consistency of the fusion results in complex construction work environments.

[0056] The preset thresholds are determined based on the characteristics of the data collected and processed by the system and the motion accuracy of the robotic arm itself. Specifically, the preset thresholds include a position deviation threshold and a posture error threshold. The position deviation threshold is determined comprehensively based on standardized observation data; the posture error threshold is set based on the posture deviation of the robotic arm's end effector. In the initial stage of operation, the preset thresholds are obtained through empirical calibration and can be dynamically corrected according to the on-site noise level, sensor stability, and environmental changes. This ensures that the matching judgment maintains sufficient sensitivity while avoiding misjudgments caused by minor disturbances, thereby ensuring the reliability and stability of the fusion and association process. After fusion is completed, operational status data is generated.

[0057] S3: Calculate the candidate path of the end effector at the target work point using the robotic arm kinematic model based on the work status data.

[0058] Furthermore, after generating the work status data, the candidate path of the end effector at the target work point is calculated based on the kinematic model of the robotic arm.

[0059] A robotic arm kinematic model is a mathematical model used to describe the mapping relationship between the motion of each joint of a robotic arm and the spatial pose of the end effector. This model consists of two parts: a forward kinematic model and an inverse kinematic model. The forward kinematic model calculates the position and orientation of the end effector based on the joint angles; the inverse kinematic model, given the target pose of the end effector, solves for the combination of values ​​for each joint variable of the robotic arm. Based on the link lengths, joint types, rotation axis directions, and pose transformation matrices of the robotic arm, this model expresses the motion characteristics of the robotic arm through joint parameterization.

[0060] When calculating candidate paths, the absolute spatial pose of each target work point is read and input into the robotic arm's kinematics model. The feasible solution set for each joint variable is solved using the inverse kinematics model, yielding multiple reachable end effector posture combinations. For each feasible solution, its reachability and posture continuity in a unified spatial coordinate system are further verified using the forward kinematics model. Invalid solutions caused by joint limitations, singular configurations, or mechanism interference are eliminated, thus obtaining several candidate posture solutions that satisfy the kinematic constraints.

[0061] Subsequently, continuous motion trajectories are generated among the candidate attitude solutions to form candidate paths. A candidate path refers to a set of multiple feasible trajectories from the current state of the robotic arm's end effector to the target work point. During trajectory generation, environmental constraints in the work state data are considered, and the attitude change rate, joint angular velocity, and acceleration between path segments are smoothed to avoid path discontinuities or abrupt motion changes.

[0062] A constraint filtering mechanism is introduced to prioritize the selection of work points with high-quality fusion matching, and to calculate their corresponding candidate paths first. For each path, its feasibility indicators are calculated, including path length, joint motion range, attitude change range, and minimum distance to obstacle boundaries. Using these feasibility indicators as evaluation criteria, a set of candidate paths that meet both mechanical motion characteristics and operational safety requirements is selected, providing input for the next step of safety constraint correction.

[0063] S4: Correct the safety constraints of the candidate path to generate the final path instruction with feasible domain constraints.

[0064] Furthermore, after generating candidate paths, safety constraints are modified on the candidate paths to generate final path instructions that satisfy the feasible domain constraints that meet the operating environment limitations and the operating characteristics of the robotic arm.

[0065] Safety constraints refer to the safety boundary conditions and operational constraints that must be followed during the movement of a robotic arm, in order to avoid interference, collisions, or overload operation between the robotic arm and its environment. Safety constraints include, but are not limited to: rotational limit constraints of each joint of the robotic arm, upper limit constraints of movement speed and acceleration, minimum safe distance constraints between the end effector and surrounding obstacles, and avoidance constraints for dynamic objects (such as personnel or equipment) in the working environment.

[0066] During the safety constraint correction phase, each trajectory segment of the candidate path is evaluated point-by-point. First, constraint checks are performed on each trajectory segment to determine if it exceeds limits, collides, or exceeds speed limits. If the checks meet the safety constraints, the trajectory segment is retained; otherwise, local correction is initiated, fine-tuning the position or attitude of each trajectory segment to bring it back within the safe and feasible region. Through iterative calculations and continuous checks, it is ensured that the corrected trajectory remains within the safety boundaries throughout the entire motion process.

[0067] The feasible region constraint refers to the comprehensive constraint conditions used to ensure the continuous feasibility of a path in terms of mechanics and kinematics, based on the satisfaction of safety constraints. By calculating the feasible region boundary of each trajectory segment of the modified candidate path, the candidate paths are screened through multi-dimensional constraints, eliminating trajectories that do not meet the continuous feasibility condition, and retaining only candidate paths that meet the constraint boundaries.

[0068] After completing safety constraints and feasible region corrections, the final path command is generated. This final path command includes the spatial position command, attitude angle command, joint angle sequence command, and time interpolation parameters of the end effector. The final path command is then sent to the robotic arm control system. Upon receiving the final path command, the robotic arm control system smoothly moves the end effector along the corrected path within the workspace. To improve the stability and real-time performance of path execution, consistency verification is performed after the final path command is generated. This verifies the continuity and accuracy of the final path command, ensuring that there are no speed jumps, abrupt attitude changes, or out-of-bounds operations during execution. Once the verification is successful, the final path command is stored and then sent out.

[0069] By modifying safety constraints and limiting feasible domains, the final generated path instructions can ensure the safety and executability of the robotic arm in complex construction operation environments, achieve smooth transitions and spatial avoidance of operation actions, and provide a reliable path execution foundation for the automated operation of the robotic arm.

[0070] Example 2, one embodiment of the present invention, provides a path planning method for a robotic arm used in construction operations. To verify the beneficial effects of the present invention, scientific demonstration is carried out through economic benefit calculations and simulation experiments.

[0071] First, to verify the effectiveness of the path planning method for the robotic arm used in construction operations, a steel beam welding operation scenario was selected as the experimental subject. A multi-source data acquisition system consisting of a vision acquisition device, a depth measurement device, a posture detection device, and a displacement measurement device was deployed at the experimental site to collect raw data on the robotic arm's working area and its own motion state. The robotic arm used in the experiment was a six-degree-of-freedom industrial-grade robotic arm with a length of 1.8 meters. The end effector was an arc welding gun, and the number of welding points was set to eight, located at different heights and angles on the building components.

[0072] At the beginning of the experiment, two-dimensional image information of the welding operation area was first acquired through a vision acquisition device, sampling 12 frames per second to record the spatial texture features of the site. Simultaneously, depth matrix data was acquired through a depth measurement device with a resolution of 1280×720 to reflect the spatial distance distribution between the welding point and the end effector of the robotic arm. An attitude detection device was installed at each joint of the robotic arm to collect joint angles and end effector attitude changes in real time; a displacement measurement device was used to monitor the displacement trajectory of the base and end effector, with a sampling frequency of 50Hz. All sensor data were spatially registered and unified to the robotic arm's base coordinate system to ensure consistent spatial correspondence between vision, depth, and attitude information. After structured preprocessing, standardized observation data containing fields such as timestamp, spatial position, attitude angle, depth value, and displacement parameters were generated, providing standard input for subsequent processing.

[0073] Subsequently, the absolute spatial pose of the target work point is defined according to the architectural construction design drawings. The system reads standardized observation data and the spatial pose information of the target work point, and matches the two through a fusion and correlation process. When the position deviation is less than ±4.0mm and the attitude error is less than ±0.25°, the correspondence is considered valid; if it exceeds the range, it is automatically readjusted. After multiple rounds of fusion, work status data containing work point identification, spatial coordinates, attitude angle, and environmental characteristics is generated.

[0074] During the path planning phase, the system solves the inverse solution set for each target work point based on the robotic arm's kinematic model, filtering out reachable end-effector posture combinations. Combining work status data, the system generates multiple candidate paths, each containing approximately 200 consecutive pose points. A safety constraint correction module then dynamically detects and corrects the paths, ensuring the end-effector maintains a minimum safe distance of 100mm from obstacles during movement. The system further performs feasible region constraint calculations, comprehensively filtering based on posture change rate, joint acceleration, and path smoothness. Finally, a final path instruction file is generated, including joint angle sequences, posture angle commands, interpolation parameters, and execution cycles, and the instructions are sent to the robotic arm control system for execution. The experiment was repeated three times under the same working conditions to verify the stability and repeatability of the path planning.

[0075] Table 1: Experimental Data Table

[0076]

[0077] As can be seen from the experimental data in Table 1, the path planning method for construction robotic arms proposed in this invention exhibits significant advantages in positioning accuracy, attitude control, path smoothness, and operational safety. In Experiment A, after adopting the method of this invention, the average positioning error was 3.7 mm and the attitude error was 0.21°, which are approximately 56.5% and 59.6% lower than the 8.5 mm and 0.52° of traditional path planning (Experiment B), respectively. This demonstrates that the fusion of correlation and kinematic model calculations effectively improves the spatial positioning accuracy and attitude stability of the end effector.

[0078] Regarding path smoothness, the smoothness index of the method of this invention reaches 0.93, while that of the traditional method is only 0.71, indicating that by applying feasible region constraints and safety constraints to the candidate path, the continuity and dynamic stability of the robotic arm's motion trajectory are significantly enhanced. In terms of safety distance, the method of this invention maintains a minimum distance of 126mm on average, which is nearly double that of experiment D without safety constraints, significantly reducing the potential collision risk.

[0079] In terms of efficiency, the path execution time of the method in this invention is 12.4 seconds, which is about 20.5% shorter than that of traditional path planning. This is mainly due to the introduction of feasibility index calculation in the candidate path selection stage, which reduces the number of invalid trajectory traversals. At the same time, the number of dynamic adjustments is reduced from 5 times in the traditional scheme to 1 time, indicating that the fused and matched job status data improves the stability and adaptability of path planning.

[0080] In terms of task completion rate, the method of this invention achieved 99.2%, while random path generation only achieved 83.7%, and manually preset paths achieved 89.3%, demonstrating that the method has higher reliability and execution consistency in complex building environments. Multiple sets of experimental comparisons confirm that traditional solutions suffer from problems such as large path jitter, untimely obstacle avoidance, and lagging attitude adjustment. In contrast, this invention achieves closed-loop optimization of the entire path planning process through standardized observation data fusion, precise kinematic model calculation, and safety constraint correction.

[0081] In summary, this embodiment verifies the feasibility and innovation of the method of the present invention in construction operation scenarios. Its innovation lies in achieving integrated and adaptive optimization of the entire process from multi-source data acquisition to path instruction generation, significantly improving the operational safety and path execution accuracy of the robotic arm in dynamic environments. It overcomes problems such as large positioning deviations, uneven trajectories, and insufficient safety margins in traditional methods, demonstrating significant engineering application value and promising prospects for widespread adoption.

[0082] Example 3, one embodiment of the present invention, provides a path planning system for a robotic arm used in construction operations, including a multi-source data acquisition and spatial registration preprocessing module, a work point pose fusion and state generation module, a kinematics solution and candidate path generation module, and a safety constraint correction and path command output module.

[0083] The multi-source data acquisition and spatial registration preprocessing module is used to collect multi-source raw data from the construction site and generate standardized observation data in a unified format through spatial registration and structured preprocessing. The operation point pose fusion and state generation module is used to fuse and correlate the standardized observation data with the absolute spatial pose of the target operation point to generate operation state data describing the relationship between the operation environment and the target position. The kinematics solution and candidate path generation module is used to solve multiple feasible trajectories of the end effector at the target operation point based on the operation state data and the kinematic model of the robotic arm to form a set of candidate paths. The safety constraint correction and path command output module is used to apply safety and feasible domain constraints to the candidate paths, correct the trajectories, and generate the final path command.

[0084] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0085] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-including system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0086] More specific examples (a non-exhaustive list) of computer-readable media include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which programs can be printed, because programs can be obtained electronically, for example, by optically scanning the paper or other media, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0087] It should be understood that various parts of the present invention can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc. It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

[0088] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A path planning method for a robotic arm used in construction operations, characterized in that, include: Collect raw data from multiple sources, perform spatial registration and structured preprocessing, and generate standardized observation data; The standardized observation data is fused and correlated with the absolute spatial pose of each target operation point to output operation status data; Based on the work status data, the candidate path of the end effector at the target work point is calculated using the kinematic model of the robotic arm; The candidate paths are modified to meet safety constraints, and a final path instruction with feasible domain constraints is generated. The spatial registration includes mapping visual images, depth matrices, pose sequences, and displacement trajectories in the same coordinate system to maintain consistency in the geometric and temporal relationships among the multi-source raw data. After spatial registration is completed, structured preprocessing is performed on the multi-source raw data, including: The registered data undergoes format standardization, field parsing, and structure reorganization. Extract key fields from each data source; delete redundant or incomplete records, repair sequence data with abnormal time intervals, and reorganize them into data tables according to a unified field standard; During the structuring process, numerical precision, value range, and unit system are standardized and uniformly defined. The fusion and association process includes: In construction operation scenarios, multiple target operation points are pre-determined based on construction design drawings and task planning results; Based on the pixel points, depth values, attitude angles and displacement trajectory information in the standardized observation data, the local region corresponding to the absolute spatial pose of the target work point is retrieved. Calculate the position deviation and attitude error in the local area corresponding to the absolute spatial pose of the target work point to determine the degree of matching between the standardized observation data and the target work point; The determination of the matching degree between the standardized observation data and the target operation point includes: When both the position deviation and attitude error are below the preset threshold, the correspondence between the operation point and the observation data is determined to be valid, and the matching result is marked as an effective fusion state. If either the position deviation or the attitude error exceeds a preset threshold, a local adjustment process will be automatically triggered to dynamically correct the local spatial mapping of the work point. Dynamically correcting the local spatial mapping of the work point includes recalculating the position deviation, attitude error, and depth matching degree in the neighborhood of the current position, and updating the spatial pose description of the target point based on the correction results; When either the corrected position deviation or the attitude error meets the preset threshold condition again, the system automatically performs a matching confirmation operation and locks the current result; If the preset threshold requirement is still not met after continuous adjustments, the weight will be reduced and the next round of fusion iteration will begin. The preset threshold is determined based on the characteristics of the collected and processed data and the motion accuracy of the robotic arm itself. The position deviation threshold is determined based on standardized observation data. The attitude error threshold is set based on the attitude deviation of the robotic arm's end effector; The candidate paths for the computational end effector at the target job point include: The feasibility index of each candidate path is calculated by reading the absolute spatial pose of each target work point and inputting it into the kinematic model of the robotic arm. Based on the inverse kinematics model, the feasible solution set of each joint variable is solved, and the output is the achievable end effector posture combination; For each feasible solution, invalid solutions caused by joint constraints, singular configurations, and mechanism interference are filtered out by a forward kinematics model, and candidate posture solutions that satisfy the kinematic constraints are output. A continuous motion trajectory is generated between each candidate attitude solution to form a candidate path.

2. The path planning method for a robotic arm used in construction operations as described in claim 1, characterized in that, The collection of multi-source raw data includes: The robot arm collects data on its working environment and its own operating status using vision acquisition devices, depth measurement devices, posture detection devices, and displacement measurement devices. The visual acquisition device is used to acquire two-dimensional image information of the construction site, continuously capture the scene of the work area and generate time-series image frames; The depth measurement device is used to collect depth information at the construction site and output corresponding depth matrix data to reflect the spatial distance distribution between the object and the robotic arm. The attitude detection device is used to measure the angle changes of each joint of the robotic arm and the spatial attitude of the end effector, and generate attitude sequence data; The displacement measuring device is used to measure the displacement changes of the robotic arm base and key components in real time during operation and output displacement trajectory data.

3. The path planning method for a robotic arm used in construction operations as described in claim 1, characterized in that, The kinematic model of the robotic arm includes a forward kinematic model and an inverse kinematic model; The forward kinematics model is used to calculate the position and orientation of the end effector based on the angles of each joint; The inverse kinematics model is used to solve for the combination of values ​​of the joint variables of the robotic arm given the target pose of the end effector; The kinematic model of the robotic arm is based on the link length, joint type, rotation axis direction and pose transformation matrix of the robotic arm, and expresses the motion characteristics of the robotic arm through joint parameterization.

4. The path planning method for a robotic arm used in construction operations as described in claim 1, characterized in that, The step of modifying the candidate path by security constraints includes: Establish a constraint detection mechanism for each trajectory segment of the candidate path to determine whether it exceeds the limit, collides, or exceeds the speed limit. If the detection results meet the safety constraints, the trajectory segment is retained; If the conditions are not met, a local correction is initiated to adjust the position or orientation of the trajectory points and return them to the safe and feasible region.

5. A system employing the path planning method for a construction robotic arm as described in any one of claims 1 to 4, characterized in that, include: Multi-source data acquisition and spatial registration preprocessing module, operation point pose fusion and state generation module, kinematics solution and candidate path generation module, safety constraint correction and path command output module; The multi-source data acquisition and spatial registration preprocessing module is used to acquire multi-source raw data from the construction site and generate standardized observation data in a unified format through spatial registration and structured preprocessing. The work point pose fusion and state generation module is used to fuse and correlate standardized observation data with the absolute spatial pose of the target work point to generate work state data describing the relationship between the work environment and the target position. The kinematics solution and candidate path generation module is used to solve multiple feasible trajectories of the end effector at the target work point based on the work status data and the kinematic model of the robotic arm, forming a set of candidate paths. The safety constraint correction and path instruction output module is used to apply safety and feasible domain constraints to candidate paths, and generate the final path instruction after correcting the trajectory.

Images