Building construction robot control method and system

Abstract

The invention relates to the technical field of robot control, in particular to a building construction robot control method and system, and the method comprises the steps: receiving operation state parameters; dynamically updating a preset global construction environment model according to the operation state parameters and physical parameters of the corresponding construction robot; when a construction robot needs to re-plan a path and an emergency task exists, based on the updated global construction environment model, identifying potential time-space conflicts, forming a plurality of task execution schemes, obtaining corresponding cost evaluation values and income evaluation values, and calculating a total utility score; and selecting the scheme with the highest total utility score as an optimal scheduling decision, and issuing a corresponding control instruction to a related construction robot. The problems of operation conflict, efficiency reduction, resource waste and the like caused by the fact that a traditional robot control system is difficult to deal with the complex situation of a building construction site in real time and effectively are solved.

Description

Technical Field

[0001] This application relates to the field of robot control technology, and more specifically, to a control method and system for a construction robot. Background Technology

[0002] In modern construction, multiple intelligent construction robots are centrally connected to a central control system for unified path planning, task allocation, and status synchronization to improve efficiency and safety. However, construction sites are highly dynamic and uncertain: road conditions can deteriorate due to sudden weather changes, and high-priority emergency tasks may be interrupted at any time. Traditional control systems struggle to utilize the operational status and location information reported by individual robots in a timely manner, and cannot dynamically update the global construction environment model. They still rely on outdated data for scheduling and path planning, resulting in an inability to accurately identify temporal-spatial conflicts. A typical problem is that when a robot is blocked by mud and needs to detour, while a high-priority task occurs, the system often simply yields the way based on simple priority rules, ignoring the cascading effects of different scheduling schemes on the delay of other tasks, increased energy consumption, and overall progress.

[0003] To address the aforementioned issues, existing technologies urgently need improvement. Summary of the Invention

[0004] This application discloses a control method and system for construction robots, which aims to solve the problems that traditional robot control systems cannot respond in real time and effectively in complex situations such as construction sites, resulting in operational conflicts, reduced efficiency and waste of resources.

[0005] The technical solution of this application is as follows:

[0006] In a first aspect, this application discloses a control method for a construction robot, comprising:

[0007] Receive real-time operating status parameters of the construction robot that exceed the preset normal range. The operating status parameters include at least one or more of the following: drive motor load, deviation between actual travel speed and expected travel speed, and changes in vehicle posture.

[0008] Based on the operating status parameters and the physical parameters of the corresponding construction robot, the local environmental physical characteristics of the area where the construction robot is located are inverted and quantified. Based on the local environmental physical characteristics, the preset global construction environment model is dynamically updated, and the area where the construction robot is located is marked as a passage risk area.

[0009] Continuously monitor the task status of all construction robots and external emergency tasks, and when it is found that a construction robot needs to replan its path and there is an emergency task, predict the replanned path of the construction robot and the path corresponding to the emergency task based on the updated global construction environment model, and identify potential time-space conflicts.

[0010] Based on the updated global construction environment model, multiple alternative paths are generated for construction robots involved in potential time-space conflicts. These alternative paths are then combined with different task execution sequences, waiting strategies, and / or passage priority rules to form multiple task execution schemes.

[0011] For each task execution plan, based on the updated global construction environment model and the current task status of each construction robot, the multi-dimensional costs and multi-dimensional benefits corresponding to the task execution plan are quantitatively evaluated to obtain the corresponding cost evaluation value and benefit evaluation value.

[0012] Based on preset weight parameters and utility functions, the cost assessment value and benefit assessment value are uniformly converted into the total utility score of the corresponding task execution plan;

[0013] Compare the total utility scores of each task execution plan and select the plan with the highest total utility score as the optimal scheduling decision;

[0014] Based on the optimal scheduling decision, corresponding control commands are issued to the relevant construction robots.

[0015] Furthermore, based on preset weighting parameters and utility functions, the cost assessment value and benefit assessment value are uniformly converted into the total utility score of the corresponding task execution plan, including:

[0016] Obtain the structural integrity sensitivity profile of the component currently being transported by the construction robot. The structural integrity sensitivity profile includes at least the maximum differential settlement threshold and torsional deformation threshold that the component can withstand.

[0017] Based on the pavement deformation sensitivity index of the alternative paths corresponding to the task execution plan in the updated global construction environment model and the percentage decrease in instantaneous bearing capacity caused by external instantaneous disturbance, the differential settlement and torsional deformation of the components under the alternative paths are predicted.

[0018] When the differential settlement exceeds the maximum differential settlement threshold or the torsional deformation exceeds the torsional deformation threshold, calculate the component damage risk cost corresponding to the task execution plan;

[0019] The component damage risk cost, along with other quantitative assessment results of multi-dimensional costs and multi-dimensional benefits, are used as inputs to participate in the calculation of the total utility score corresponding to the task execution plan.

[0020] Furthermore, the system continuously monitors the task status of all construction robots and externally issued emergency tasks. When it identifies a construction robot that needs to replan its path and an emergency task, it predicts the replanned path of the construction robot and the path corresponding to the emergency task based on the updated global construction environment model, identifying potential time-space conflicts, including:

[0021] It receives external environment perception data in real time, which includes obstacle information, road surface information, and other dynamic environmental elements in the area where the construction robot is located.

[0022] Based on external environment perception data, the local area information in the updated global construction environment model is dynamically corrected, including updating obstacle locations, road traffic costs, and temporary work areas.

[0023] Based on the dynamically corrected global construction environment model, as well as the kinematic and dynamic parameters of the construction robot, multiple probabilistic predicted paths are generated for construction robots that need to replan their paths and construction robots corresponding to emergency tasks within a preset future time window.

[0024] A temporal-spatial overlap analysis is performed on multiple probabilistic predicted paths. Based on the probability distribution of the overlapping areas and a preset safe distance threshold, potential temporal-spatial conflicts are identified between construction robots that replan their paths and construction robots corresponding to emergency tasks.

[0025] Furthermore, based on the updated global construction environment model, multiple alternative paths are generated for construction robots involving potential time-space conflicts. These alternative paths are then combined with different task execution sequences, waiting strategies, and / or passage priority rules to form multiple task execution schemes, including:

[0026] Obtain the structural integrity sensitivity profile of the component currently being transported by the construction robot. The structural integrity sensitivity profile includes at least the maximum differential settlement threshold and torsional deformation threshold that the component can withstand.

[0027] Obtain the efficiency drift index and vibration characteristic changes of the construction robot drive system;

[0028] Obtain the percentage decrease in instantaneous bearing capacity caused by external instantaneous disturbance;

[0029] In the path planning algorithm, the maximum differential settlement threshold, torsional deformation threshold, efficiency drift index, vibration characteristic change, and instantaneous bearing capacity decrease percentage are used as dynamic constraints and optimization objectives. The path cost of each candidate path is adjusted and / or candidate paths that do not meet the preset safety constraints are eliminated to generate multiple alternative paths and their corresponding task execution schemes.

[0030] Furthermore, based on preset weighting parameters and utility functions, the cost assessment value and benefit assessment value are uniformly converted into the total utility score of the corresponding task execution plan, including:

[0031] Obtain the priority of the current task, the real-time environmental risk level of the construction site, and the structural integrity sensitivity of the components carried by the construction robot;

[0032] Based on the priority of the current task, the real-time environmental risk level of the construction site, and the sensitivity to structural integrity, the weight parameters are dynamically adjusted to obtain the quantitative weights corresponding to the multi-dimensional costs and multi-dimensional benefits.

[0033] Quantitative weights are applied to the cost assessment value and the benefit assessment value, and the cost assessment value and the benefit assessment value are weighted according to the utility function to obtain the total utility score of the corresponding task execution plan.

[0034] Furthermore, quantitative weights are applied to the cost assessment value and the benefit assessment value, and the cost assessment value and the benefit assessment value are weighted according to the utility function to obtain the total utility score of the corresponding task execution plan, including:

[0035] It receives multi-source data from construction robots, external environment sensors, task management systems, and component BIM models;

[0036] Timestamp alignment and data format conversion are performed on multi-source data to obtain aligned multi-source data.

[0037] The missing data in the aligned multi-source data is imputed to obtain the imputed multi-source data.

[0038] Abnormal data are removed from the imputed multi-source data to obtain the removed multi-source data.

[0039] Based on the removed multi-source data, the quantitative results of multi-dimensional costs and multi-dimensional benefits are calculated to obtain cost assessment values ​​and benefit assessment values.

[0040] Quantitative weights are applied to the cost assessment value and the benefit assessment value, and the cost assessment value and the benefit assessment value are weighted according to the utility function to obtain the total utility score of the corresponding task execution plan.

[0041] Furthermore, based on the dynamically corrected global construction environment model and the kinematic and dynamic parameters of the construction robot, multiple probabilistic predicted paths are generated within a preset future time window for construction robots that require path replanning and for construction robots corresponding to emergency tasks, including:

[0042] Using the current pose and velocity of the construction robot as the initial state, a stochastic motion model including control error, ground friction coefficient fluctuation and sensor noise is constructed. The motion process within a preset future time window is sampled and simulated multiple times to obtain a set of candidate trajectories. The probability of occurrence is assigned to each candidate trajectory to obtain multiple probabilistic prediction paths.

[0043] A temporal-spatial overlap analysis is performed on multiple probabilistic predicted paths. Based on the probability distribution of the overlapping areas and a preset safety distance threshold, potential temporal-spatial conflicts are identified between the construction robot replanning its path and the construction robot corresponding to the emergency task, including:

[0044] Within a preset future time window, the spatial distance and probability distribution of the corresponding overlapping area of ​​the construction robot that needs to replan its path and the construction robot corresponding to the emergency task are calculated at each discrete moment. When the spatial distance at a certain moment is less than the safe distance threshold and the probability distribution of the overlapping area is greater than the preset probability threshold, it is determined that there is a potential time-space conflict.

[0045] Furthermore, the safe distance threshold is a dynamic safe distance threshold, and the process of determining the dynamic safe distance threshold includes:

[0046] Obtain the priority information of the current task, which includes at least the urgency and importance of the task;

[0047] Identify the types of construction robots involved in potential time-space conflicts, including at least one of tracked, wheeled, or flying types;

[0048] Obtain collision sensitivity information of the components currently being transported by the construction robot. The collision sensitivity information includes at least the component's ability to withstand impact and vibration.

[0049] Calculate the safety distance adjustment factor based on priority information, type, and collision sensitivity information;

[0050] The safety distance adjustment factor is applied to the preset base safety distance to obtain the dynamic safety distance threshold. The dynamic safety distance threshold is then used as the safety distance threshold. Within a preset future time window, the spatial distances of construction robots that need to replan their paths and construction robots corresponding to emergency tasks are compared at each discrete moment to identify potential time-space conflicts.

[0051] Furthermore, based on priority information, type, and collision sensitivity information, a safety distance adjustment factor is calculated, including:

[0052] Obtain the risk appetite and environmental conditions at the construction site during the current construction phase. The risk appetite should include at least the degree of emphasis on schedule, cost and safety.

[0053] Based on priority information, type, collision sensitivity information, risk preference, and environmental conditions, calculate the adjustment weight of each piece of information on the safety distance adjustment;

[0054] The adjustment weights are applied to priority information, type, collision sensitivity information, risk preference, and environmental conditions. Through nonlinear function mapping, a preliminary safe distance adjustment factor is obtained.

[0055] Based on the real-time environmental risk level at the construction site, the initial safety distance adjustment factor is corrected to obtain the final safety distance adjustment factor.

[0056] Secondly, this application also discloses a control system for a construction robot, comprising:

[0057] The monitoring and reporting module is used to receive real-time operating status parameters of the construction robot that exceed the preset normal range. The operating status parameters include at least one or more of the following: drive motor load, deviation between actual travel speed and expected travel speed, and changes in vehicle posture.

[0058] The environment model update module is used to inversely quantify the local environmental physical characteristics of the area where the construction robot is located based on the operating status parameters and the physical parameters of the corresponding construction robot, and dynamically update the preset global construction environment model based on the local environmental physical characteristics, and mark the area where the construction robot is located as a passage risk area.

[0059] The conflict prediction module is used to continuously monitor the task status of all construction robots and external emergency tasks. When it is found that a construction robot needs to replan its path and there is an emergency task, it predicts the replanned path of the construction robot and the path corresponding to the emergency task based on the updated global construction environment model, and identifies potential time-space conflicts.

[0060] The solution generation module is used to generate multiple alternative paths for construction robots involving potential time-space conflicts based on the updated global construction environment model, and to combine the alternative paths with different task execution orders, waiting strategies and / or passage priority rules to form multiple task execution solutions.

[0061] The cost-benefit calculation module is used to quantitatively evaluate the multi-dimensional costs and benefits corresponding to each task execution plan based on the updated global construction environment model and the current task status of each construction robot, and obtain the corresponding cost evaluation value and benefit evaluation value.

[0062] The utility evaluation module is used to convert the cost evaluation value and the benefit evaluation value into the total utility score of the corresponding task execution plan based on the preset weight parameters and utility function.

[0063] The decision selection module is used to compare the total utility scores of each task execution plan and select the plan with the highest total utility score as the optimal scheduling decision.

[0064] The instruction issuance module is used to issue corresponding control instructions to relevant construction robots based on the optimal scheduling decision.

[0065] Beneficial Effects: The construction robot control method disclosed in this application receives robot operating status parameters in real time and dynamically updates the global construction environment model, ensuring the timeliness and accuracy of environmental information and avoiding the drawbacks of making decisions based on outdated information. It can not only predict potential time-space conflicts but also generate multiple alternative paths and task execution schemes, providing a rich selection space for decision-making. Through multi-dimensional cost-benefit evaluation and total utility score calculation, it comprehensively considers various constraints such as task priority, time window, robot energy consumption, component structural integrity sensitivity, and task dependencies, avoiding the limitations of traditional systems that make decisions based solely on simple priority rules, thereby selecting the globally optimal scheduling decision. Through collaborative rescheduling, it effectively avoids the chain reaction of delays or operational conflicts in the entire construction process caused by local environmental problems, ensuring construction accuracy and safe collaboration, and significantly improving overall construction efficiency and safety. Attached Figure Description

[0066] Figure 1 This is a flowchart illustrating a control method for a construction robot provided in this application.

[0067] Figure 2 A flowchart of a construction robot control system provided in this application.

[0068] The diagram shows: 1. Monitoring and reporting module; 2. Environmental model update module; 3. Conflict prediction module; 4. Solution generation module; 5. Cost-benefit calculation module; 6. Utility evaluation module; 7. Decision selection module; 8. Instruction issuance module. Detailed Implementation

[0069] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments. The components of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0070] In modern construction, multiple intelligent construction robots are widely used to collaboratively complete complex tasks, improving efficiency and safety. These robots are typically managed and scheduled by a central control system responsible for path planning, task allocation, and state synchronization between robots. However, construction site environments are highly dynamic and uncertain; for example, road conditions may suddenly change due to weather conditions, or unexpected high-priority tasks may require urgent handling. In such complex and changing environments, traditional robot control systems often struggle to respond effectively and in real time, leading to problems such as task conflicts, decreased efficiency, and resource waste.

[0071] Reference Figure 1 This application proposes a control method for a construction robot, comprising:

[0072] S1000: Receives real-time operating status parameters of the construction robot that exceed the preset normal range. The operating status parameters include at least one or more of the following: drive motor load, deviation between actual travel speed and expected travel speed, and changes in vehicle posture.

[0073] S2000: Based on the operating status parameters and the physical parameters of the corresponding construction robot, the local environmental physical characteristics of the area where the construction robot is located are inverted and quantified, and the preset global construction environment model is dynamically updated based on the local environmental physical characteristics, and the area where the construction robot is located is marked as a passage risk area.

[0074] S3000: Continuously monitors the task status of all construction robots and emergency tasks issued from the outside. When it is found that there are construction robots that need to replan their paths and there are emergency tasks, it predicts the replanned paths of the construction robots and the paths corresponding to the emergency tasks based on the updated global construction environment model, and identifies potential time-space conflicts.

[0075] S4000: Based on the updated global construction environment model, it generates multiple alternative paths for construction robots involved in potential time-space conflicts, and combines the alternative paths with different task execution sequences, waiting strategies and / or passage priority rules to form multiple task execution schemes;

[0076] S5000: For each task execution plan, based on the updated global construction environment model and the current task status of each construction robot, the multi-dimensional cost and multi-dimensional benefit corresponding to the task execution plan are quantitatively evaluated to obtain the corresponding cost evaluation value and benefit evaluation value.

[0077] S6000: Based on preset weight parameters and utility functions, the cost assessment value and benefit assessment value are uniformly converted into the total utility score of the corresponding task execution plan;

[0078] S7000: Compare the total utility scores of each task execution plan and select the plan with the highest total utility score as the optimal scheduling decision;

[0079] S8000: Based on the optimal scheduling decision, it issues corresponding control commands to the relevant construction robots.

[0080] Specifically, "operational status parameters" refer to measurable data of the construction robot during task execution, such as drive motor load, deviation between actual and expected travel speed, and changes in vehicle posture. These are acquired by encoders, inertial measurement units (IMUs), force sensors, current or torque sensors, wheel encoders, or GPS modules, and are used to reflect the real-time status of the robot and its environment. "Local environmental physical characteristics" refer to the road surface bearing capacity, friction coefficient, and smoothness of the area where the robot is located. "Global construction environment model" is a digital model describing the geographic information, obstacle distribution, road surface conditions, temporary work areas, and other dynamic environmental elements of the entire construction site, and is the basis for path planning and task scheduling. "Time-space conflict" refers to two or more robots attempting to occupy the same spatial area within the same time window, which may lead to collisions, stalls, or task delays. "Multi-dimensional costs" include negative impacts such as time costs, energy costs, and safety costs. "Multi-dimensional benefits" include positive effects such as task completion efficiency, resource utilization, and improved construction progress. "Utility function" is used to map multi-dimensional costs and benefits into a single "total utility score" under given weight parameters.

[0081] The process is as follows: First, when the operating parameters exceed the preset normal range, such as abnormally high drive motor load, excessive deviation between actual and expected travel speed, or drastic changes in vehicle posture, the system determines that the robot may have encountered abnormalities such as slippery road surface, obstacles, or local collapse, and reports the operating parameters and high-precision position information to the central control system in real time. The central control system, combining physical parameters such as robot mass, size, and driving force, infers the local environmental physical characteristics of the area (e.g., excessive friction coefficient, uneven road surface), and dynamically updates the global construction environment model accordingly, marking the corresponding area as a passage risk zone.

[0082] The control system continuously monitors the task status of all construction robots and externally issued emergency tasks. When a robot needs to replan its path due to entering a high-risk area and there is a high-priority emergency task, the system predicts the replanned path of the robot and the path of the robot performing the emergency task based on the latest global construction environment model. It then performs a time-space overlap analysis on the two within a preset future time window to identify potential time-space conflicts.

[0083] After identifying potential conflicts, the system generates multiple alternative paths for the robots involved in the conflict based on an updated global construction environment model. It then combines these alternative paths with different task execution sequences, waiting strategies, and / or passage priority rules (such as priority passage for robots performing urgent tasks), forming multiple task execution plans. Subsequently, the system quantitatively evaluates the multi-dimensional costs and benefits of each task execution plan, considering constraints such as the current task status of each robot, time windows, robot energy consumption, and task dependencies, obtaining cost and benefit assessment values. Based on preset weight parameters and a utility function, these assessment values ​​are then uniformly converted into the total utility score for the corresponding plan.

[0084] For example, when a transport robot is blocked by muddy roads and needs to replan its route, and a high-priority urgent task arises at the same time, traditional systems often simply give way unconditionally according to simple priority rules. This application, however, dynamically updates the global construction environment model and generates multiple task execution plans for both robots. It comprehensively considers the delays of other tasks, increased energy consumption, and the impact on overall construction progress, calculates the total utility score through a utility function, and selects the optimal scheduling decision that ensures both the requirements of the urgent task and overall efficiency and safety.

[0085] Ultimately, the central control system compares the total utility scores of each task execution plan, selects the highest one as the optimal scheduling decision, and issues the corresponding path instructions, waiting instructions, or priority adjustment instructions to the relevant construction robots to guide them to execute tasks according to the optimal plan, thereby avoiding global chain delays or operational conflicts caused by local environmental anomalies.

[0086] In another embodiment of this application, S6000 specifically includes, based on preset weight parameters and a utility function:

[0087] S6100: Obtain the structural integrity sensitivity profile of the component currently being transported by the construction robot. The structural integrity sensitivity profile includes at least the maximum differential settlement threshold and torsional deformation threshold that the component can withstand.

[0088] S6200: Based on the pavement deformation sensitivity index of the alternative paths corresponding to the task execution plan in the updated global construction environment model and the percentage decrease in instantaneous bearing capacity caused by external instantaneous disturbances, predict the differential settlement and torsional deformation of the components under the alternative paths.

[0089] S6300: When the differential settlement exceeds the maximum differential settlement threshold or the torsional deformation exceeds the torsional deformation threshold, calculate the component damage risk cost corresponding to the task execution plan;

[0090] S6400: The component damage risk cost, along with other quantitative assessment results of multi-dimensional costs and multi-dimensional benefits, are used as inputs to participate in the calculation of the total utility score corresponding to the task execution plan.

[0091] Specifically, a structural integrity sensitivity profile for a component refers to a pre-established dataset describing the sensitivity of a particular component to changes in the external environment (such as settlement and torsion). In this profile, the maximum differential settlement threshold and torsional deformation threshold are key parameters, defining the maximum limits that a component can withstand in terms of differential settlement and torsional deformation, respectively. Once the actual differential settlement or torsional deformation exceeds either threshold, the structural integrity of the component is considered threatened. These thresholds can be determined based on the component's material properties, geometry, design specifications, and its importance in the building structure.

[0092] The pavement deformation sensitivity index characterizes the tendency or degree of pavement deformation under different loads and environmental conditions on alternative routes. For example, soft soil roads or temporary paved roads may have a higher pavement deformation sensitivity index. The percentage decrease in instantaneous bearing capacity caused by external transient disturbances reflects the degree of instantaneous weakening of pavement bearing capacity caused by sudden events such as local collapses or falling heavy objects. Combining the pavement deformation sensitivity index and the percentage decrease in instantaneous bearing capacity caused by external transient disturbances allows for a more accurate prediction of the differential settlement and torsional deformation that components may experience under various alternative routes.

[0093] In practical applications, component damage risk cost refers to the potential loss incurred when a component is damaged due to predicted differential settlement or torsional deformation exceeding its maximum differential settlement or torsional deformation threshold. This cost comprehensively considers factors such as the component's intrinsic value, repair costs, replacement costs, impact on construction schedule, and potential safety risks. Incorporating component damage risk cost into the quantitative assessment of multi-dimensional costs and benefits allows for a more complete and precise calculation of the total utility score.

[0094] This application proposes a quantitative assessment of the structural integrity risk of components, building upon the traditional multi-dimensional cost and benefit evaluation.

[0095] In some preferred embodiments, suppose a construction robot is carrying a large precast concrete wall panel that is highly sensitive to uneven settlement and torsional deformation. The system first obtains a structural integrity sensitivity profile for the wall panel, which records a maximum differential settlement threshold of 5 mm and a torsional deformation threshold of 0.5 degrees. When evaluating multiple alternative paths, the system analyzes the pavement deformation sensitivity index of each path (e.g., a section of the path has a higher pavement deformation sensitivity index due to recent rainfall causing soil softening) and potential external instantaneous disturbances (e.g., heavy machinery operation nearby may cause a 10% decrease in instantaneous bearing capacity) based on the updated global construction environment model, and predicts the differential settlement and torsional deformation of the wall panel under each alternative path. If a candidate path is predicted to cause differential settlement of the wall panel to reach 8 mm, exceeding the maximum differential settlement threshold of 5 mm, the system calculates the corresponding component damage risk cost, such as wall panel repair cost, wall panel replacement or rework cost, and penalties for construction delays. This component damage risk cost is integrated into the multi-dimensional cost of the task execution plan corresponding to the candidate path, and participates in the calculation of the total utility score along with other multi-dimensional costs and benefits. This identifies paths that may have advantages in time or energy consumption but have a higher risk of component damage, and tends to select task execution plans with lower component damage risk and higher total utility scores. Even if the plan has slight disadvantages in some conventional indicators, it can prioritize the integrity of construction components and construction safety.

[0096] In another embodiment of this application, step S3000 is further proposed to include:

[0097] S3100: Receives external environment perception data in real time, including obstacle information, road surface information, and other dynamic environmental elements in the area where the construction robot is located.

[0098] S3200: Based on external environment perception data, dynamically correct and update the local area information in the updated global construction environment model. Dynamic corrections include updating obstacle locations, road traffic costs, and temporary work areas.

[0099] S3300: Based on the dynamically corrected global construction environment model, as well as the kinematic and dynamic parameters of the construction robot, it generates multiple probabilistic predicted paths for construction robots that need to replan their paths and construction robots corresponding to emergency tasks within a preset future time window.

[0100] S3400: Performs temporal-spatial overlap analysis on multiple probabilistic predicted paths, and identifies potential temporal-spatial conflicts between construction robots that replan their paths and construction robots corresponding to emergency tasks based on the probability distribution of the overlapping areas and preset safety distance thresholds.

[0101] Specifically, real-time reception of external environmental perception data refers to continuously acquiring real-time information about the construction robot's current location and its surrounding environment through various sensors deployed at the construction site, such as LiDAR, cameras, and ultrasonic sensors, as well as the sensing devices onboard the construction robot itself. This information includes dynamic environmental elements such as the location and shape of newly appearing obstacles, the slipperiness or pothole condition of the road surface, and temporarily set work areas or restricted areas. The purpose is to provide the latest and more comprehensive environmental input for subsequent environmental model correction.

[0102] The dynamic correction and updating of local area information in the global construction environment model based on external environment perception data refers to the system using the aforementioned real-time environmental perception data to finely update local areas related to the current position of the construction robot. For example, when a new obstacle is detected, the system immediately adds or updates the obstacle's location information in the global construction environment model; when water accumulation or mud appears on the road surface, the system adjusts the passage cost of the corresponding road segment; when temporary work areas or restricted areas are designated or removed, the system updates the area markings in a timely manner. Through this dynamic correction process, the global construction environment model can reflect the real-time dynamics of the construction site as realistically as possible, providing a reliable environmental foundation for subsequent path prediction.

[0103] In practical applications, based on the dynamically corrected global construction environment model and the kinematic and dynamic parameters of the construction robot, multiple probabilistic prediction paths are generated. This means that the system, considering the kinematic parameters such as the maximum speed, acceleration, and turning radius of the construction robot, as well as the dynamic parameters such as friction and inertia, and combining the real-time updated environment model, uses probabilistic prediction algorithms such as Monte Carlo simulation and Kalman filtering to simulate the movement process of the construction robot within a preset future time window multiple times, obtains multiple possible movement trajectories, and assigns a probability of occurrence to each trajectory to quantify its likelihood of occurrence. This captures the uncertainty in the movement process of the construction robot and forms a more comprehensive prediction of future movement trends.

[0104] Furthermore, the system performs temporal-spatial overlap analysis on multiple probabilistic predicted paths and identifies potential temporal-spatial conflicts based on the probability distribution of overlapping areas and a preset safety distance threshold. This means that within a preset future time window, the system calculates the predicted positions of each construction robot at each discrete moment, determines whether the predicted positions of different robots overlap spatially or are less than the safety distance threshold, and comprehensively assesses the conflict risk by combining the probability of occurrence corresponding to the overlapping area. For example, if at a certain moment the minimum distance between the predicted paths of two construction robots is less than the safety distance threshold and the probability of overlapping areas is greater than a preset probability threshold, a potential temporal-spatial conflict is determined to exist. Thus, by introducing probabilistic factors, the system more accurately and robustly identifies risks that may lead to collisions or task delays.

[0105] In some preferred embodiments, assuming a construction site is being monitored, a tracked construction robot A is transporting heavy components along a predetermined path, while a wheeled construction robot B is performing an emergency material transport task, and their paths may potentially intersect. The system receives real-time environmental perception data collected by sensors such as LiDAR and cameras. For example, it detects temporary piles of building materials ahead of robot A (obstacle information) or slippery road conditions ahead of robot B (road surface condition information). Based on this, the system dynamically adjusts the global construction environment model, updates obstacle positions, and increases the travel cost on slippery sections. Subsequently, based on the adjusted global construction environment model and the kinematic and dynamic parameters of robots A and B, the system generates multiple probabilistic predicted paths for both robots within the next 30 seconds. For example, robot A may deviate from its predetermined trajectory due to slippery conditions, while robot B may accelerate due to an emergency task. The system performs time-space overlap analysis on these probabilistic predicted paths, calculates the spatial distance between the predicted positions of robot A and robot B at different future times, and combines the probability distribution of the overlapping area. If the spatial distance between the two is less than the preset dynamic safety distance threshold at a certain time (for example, a larger safety distance threshold is set to consider robot A carrying heavy objects), and the probability of the overlapping area occurring exceeds 80%, then it is determined that there is a potential time-space conflict, providing constraints and decision-making basis for the subsequent generation of alternative task execution plans.

[0106] In another embodiment of this application, S4000 is further proposed to include:

[0107] S4100: Obtain the structural integrity sensitivity profile of the component currently being transported by the construction robot. The structural integrity sensitivity profile includes at least the maximum differential settlement threshold and torsional deformation threshold that the component can withstand.

[0108] S4200 acquires the efficiency drift index and vibration characteristic changes of the construction robot drive system;

[0109] The S4300 obtains the percentage decrease in instantaneous bearing capacity caused by external instantaneous disturbances;

[0110] In its path planning algorithm, S4400 uses the maximum differential settlement threshold, torsional deformation threshold, efficiency drift index, vibration characteristic changes, and instantaneous bearing capacity decrease percentage as dynamic constraints and optimization objectives to adjust the path cost of each candidate path and / or eliminate candidate paths that do not meet the preset safety constraints, thereby generating multiple alternative paths and their corresponding task execution schemes.

[0111] Specifically, obtaining the structural integrity sensitivity profile of components currently being transported by construction robots refers to acquiring information about the component's ability to withstand external stress and deformation during transportation from the component's BIM model, design specifications, or a pre-set database before the task begins or during component loading. The structural integrity sensitivity profile describes the component's structural response characteristics under different stress conditions. Specifically, the maximum differential settlement threshold limits the maximum permissible vertical displacement difference between different support points during transportation to avoid excessive stress caused by uneven settlement; the torsional deformation threshold limits the maximum permissible torsional angle or amount of twist to prevent structural damage caused by torsion. This provides physical constraints on the component itself for path planning, ensuring component safety during transportation.

[0112] Acquiring the efficiency drift index and vibration characteristic changes of the construction robot's drive system refers to real-time monitoring of the operating status of key components such as the drive motor and transmission mechanism. The efficiency drift index represents the deviation between the actual operating efficiency and the design efficiency of the drive system, reflecting trends of wear, aging, or failure. Vibration characteristic changes indicate the deviation of the drive system's vibration frequency, amplitude, and other characteristics from the normal baseline, typically indicating abnormal wear or loosening of mechanical components. These parameters can be collected and analyzed in real-time by internal sensors such as current sensors and accelerometers to assess the health status and operational stability of the construction robot, reducing safety risks caused by its own malfunctions during task execution.

[0113] In practical applications, obtaining the percentage decrease in instantaneous bearing capacity caused by external instantaneous disturbances refers to assessing the degree of temporary decrease in the bearing capacity of the road surface or supporting structure at a construction site at a specific moment due to external factors such as local water accumulation, temporary excavation, or the passage of heavy equipment, through external environmental sensors such as ground-penetrating radar and pressure sensors, and / or combined with information such as weather forecasts and construction logs. This decrease is quantified as a percentage to characterize the vulnerability of the road surface or supporting structure under instantaneous disturbances and to provide real-time environmental risk information for route planning.

[0114] Furthermore, in the path planning algorithm, the maximum differential settlement threshold, torsional deformation threshold, efficiency drift index, vibration characteristic change, and instantaneous bearing capacity reduction percentage are used as dynamic constraints and optimization objectives. As dynamic constraints, candidate paths predicted to cause component differential settlement or torsional deformation to exceed the corresponding threshold, drive system efficiency drift index or vibration characteristic change to exceed the safe range, or paths passing through sections with insufficient bearing capacity to support the construction robot are directly eliminated. As optimization objectives, under the premise of satisfying the above safety constraints, the path planning algorithm adjusts the path costs of each candidate path, prioritizing paths that reduce component stress, lower drive system load, and avoid high instantaneous risk areas. For example, although a path may be the shortest in geometric distance, if it is predicted to cause significant stress to the component or a significant increase in drive system load, the algorithm will significantly increase the path cost of that path, or even mark it as infeasible. This generates multiple candidate paths that are comprehensively optimized in terms of safety, efficiency, and robustness. These candidate paths are then combined with different task execution sequences, waiting strategies, and / or passage priority rules to form multiple task execution schemes.

[0115] In some preferred embodiments, suppose a construction robot is transporting a large precast concrete beam. The structural integrity sensitivity profile of the beam shows a maximum differential settlement threshold of 5 mm and a maximum torsional deformation threshold of 0.1 degrees. Simultaneously, a slight increase in the efficiency drift index of the robot's drive motor is detected, and vibration characteristics indicate a minor anomaly. Furthermore, due to localized rainfall the previous night, the instantaneous load-bearing capacity of a certain section of the construction site has decreased by 20%. When generating alternative paths, the path planning algorithm uses the 5 mm differential settlement threshold, the 0.1 degree torsional deformation threshold, the drive motor efficiency drift index and vibration characteristics, and the 20% instantaneous load-bearing capacity decrease as dynamic constraints and optimization objectives: first, any paths predicted to cause differential settlement exceeding 5 mm or torsional deformation exceeding 0.1 degrees in the concrete beam are eliminated; for the remaining candidate paths, their impact on drive motor efficiency and vibration is evaluated, and paths passing through sections with decreased load-bearing capacity are subject to higher path cost penalties. For example, even if a path directly through a flooded section of road has the shortest geometric distance, it may be subject to extremely high costs due to its high risk and potential negative impact on the concrete beam and construction robot, and may even be eliminated altogether. The final set of alternative paths may not be the shortest in geometric distance, but it can simultaneously ensure the structural integrity of the concrete beam, the stable operation of the construction robot, and avoid high-risk sections, thus forming a safer and more reliable task execution plan.

[0116] In another embodiment of this application, S6000 is further proposed to include:

[0117] S6100: Obtain the priority of the current task, the real-time environmental risk level of the construction site, and the structural integrity sensitivity of the components carried by the construction robot;

[0118] S6200: Based on the priority of the current task, the real-time environmental risk level of the construction site, and the sensitivity to structural integrity, the weight parameters are dynamically adjusted to obtain the quantitative weights corresponding to the multi-dimensional costs and multi-dimensional benefits.

[0119] S6300: Apply quantitative weights to the cost assessment value and the benefit assessment value, and perform weighted processing on the cost assessment value and the benefit assessment value according to the utility function to obtain the total utility score of the corresponding task execution plan.

[0120] Specifically, prioritizing current tasks refers to assessing the importance and urgency of tasks currently being executed or about to be executed within the overall construction project, categorized as "urgent," "high," "medium," or "low." Real-time environmental risk levels at the construction site assess the potential hazards of the current construction area or the entire site, based on sensor data, weather forecasts, personnel density, and other factors, and categorized as "high-risk," "medium-risk," or "low-risk." The structural integrity sensitivity of components transported by construction robots refers to the component's ability to withstand external impacts, vibrations, deformations, etc. For example, some precision components are highly sensitive to vibration, while some large structural components are highly sensitive to differential settlement. This information can be obtained from data sources such as task management systems, environmental monitoring systems, and component BIM models.

[0121] The system dynamically adjusts preset weight parameters based on the acquired task priorities, the real-time environmental risk level at the construction site, and the structural integrity sensitivity of the components. This adjustment can be implemented using predefined rule sets, adaptive algorithms, or machine learning models. For example, when the task priority is high, the weight of benefits related to task completion time is increased, while the weight of costs related to resource consumption is appropriately reduced. When the real-time environmental risk level is high, the weight of safety-related costs is significantly increased to prioritize construction safety. When the structural integrity sensitivity of the transported components is high, the weight of costs related to component damage risk is amplified to prevent irreversible damage to the components. Through these adjustments, quantitative weights corresponding to multi-dimensional costs and benefits that match the current scenario can be obtained.

[0122] In practical applications, dynamically adjusted quantitative weights are applied to the calculated cost and benefit assessment values. The weighted cost and benefit assessment values ​​are then comprehensively processed according to a preset utility function to obtain the total utility score for each task execution plan. The utility function can employ linear weighted summation, nonlinear mapping, or other multi-objective decision-making methods. Its purpose is to unify multi-dimensional costs and benefits onto a single, comparable numerical scale, facilitating subsequent plan optimization.

[0123] This application's solution addresses the problem of decision-making biases arising from traditional fixed weights in complex and variable construction environments by introducing a dynamic adjustment mechanism for weight parameters. In some preferred embodiments, suppose a construction robot is transporting a precision component highly sensitive to vibration and impact, whose structural integrity sensitivity profile indicates a low maximum vibration threshold; simultaneously, a strong wind warning is issued at the construction site, and the real-time environmental risk level is assessed as "high risk." In this case, the control method first obtains the task priority (e.g., transporting the precision component is a higher priority), the real-time environmental risk level at the construction site (high risk), and the component's structural integrity sensitivity (high sensitivity), and dynamically adjusts the weight parameters accordingly: the weight of costs related to component damage risk is significantly increased, the weight of costs related to safety (e.g., obstacle avoidance, slow passage) is simultaneously increased, while the weight of benefits related to time efficiency is appropriately decreased. For example, when calculating the total utility score, the weight of component damage risk cost can be adjusted from 0.2 to 0.5, the weight of safety cost from 0.3 to 0.4, and the weight of time benefits from 0.4 to 0.1. In this way, when evaluating alternative paths and task execution plans, the system will prioritize the option that minimizes the risk of component damage and ensures construction safety, even if the option requires longer transportation time or higher energy consumption. Ultimately, the system selects the option with the highest total utility score, which may be a smoother, slower, but safer path, thereby effectively protecting precision components and ensuring construction safety in high-risk environments.

[0124] In another embodiment of this application, S6300 further includes:

[0125] S6310: Receives multi-source data from construction robots, external environment sensors, task management systems, and component BIM models;

[0126] S6320: Perform timestamp alignment and data format conversion on multi-source data to obtain aligned multi-source data;

[0127] S6330: Imput missing data in the aligned multi-source data to obtain imputed multi-source data;

[0128] S6340: Remove outliers from the imputed multi-source data to obtain the removed multi-source data;

[0129] S6350: Based on the removed multi-source data, calculate the quantitative results of multi-dimensional costs and multi-dimensional benefits to obtain cost assessment values ​​and benefit assessment values;

[0130] S6360: Apply quantitative weights to the cost assessment value and the benefit assessment value, and perform weighted processing on the cost assessment value and the benefit assessment value according to the utility function to obtain the total utility score of the corresponding task execution plan.

[0131] Specifically, receiving multi-source data from construction robots, external environment sensors, task management systems, and component BIM models refers to collecting various types of data related to the task execution of construction robots from different information sources. Among these, the construction robot provides real-time information such as its operating status, position, and sensor readings; external environment sensors provide environmental perception data such as obstacles, road conditions, and weather at the construction site; the task management system provides high-level instructions such as task priority, schedule requirements, and resource allocation; and the component BIM model provides detailed structural information and sensitivity profiles of the components being transported. This multi-source data forms the basis for comprehensively evaluating the multi-dimensional costs and benefits of task execution plans.

[0132] Furthermore, timestamp alignment and data format conversion are performed on the multi-source data to obtain aligned multi-source data. This ensures the synchronization of data from different sources in the time dimension and unifies their representation format. Since different sensors and systems have different data acquisition frequencies and data formats, timestamp alignment ensures the comparability of various types of data at the same point in time, avoiding evaluation bias caused by asynchrony. Data format conversion transforms heterogeneous data into a unified structured data format, facilitating subsequent unified processing.

[0133] Based on this, missing data in the aligned multi-source data is imputed to obtain imputed multi-source data. This is to address the problem of accidental missing data during data collection and to avoid inaccurate evaluation results or algorithm failure due to missing data. Imputation can be performed using linear imputation, mean imputation, or predictive imputation based on machine learning to reasonably estimate missing values, thereby improving data integrity and usability.

[0134] Simultaneously, outlier data is removed from the interpolated multi-source data to obtain the removed multi-source data. This process eliminates abnormal data points caused by sensor malfunctions, environmental interference, or transmission errors. Failure to remove outlier data can severely distort the evaluation results of multi-dimensional costs and benefits. Identifying and removing outlier data using statistical methods such as Z-score and IQR, or machine learning-based methods, ensures that the data entering the evaluation stage is authentic and reliable.

[0135] Therefore, based on the removed multi-source data, the quantitative results of multi-dimensional costs and multi-dimensional benefits are calculated to obtain cost assessment values ​​and benefit assessment values. That is, on the basis of high-quality data, the multi-dimensional costs involved in each task execution plan, such as time cost, energy consumption, potential risks, component damage risks, etc., as well as the multi-dimensional benefits such as task completion, efficiency improvement, and safety assurance, are accurately quantified to form corresponding cost assessment values ​​and benefit assessment values.

[0136] Finally, the quantified weights are applied to the cost and benefit assessment values, and the cost and benefit assessment values ​​are weighted according to the utility function to obtain the total utility score of the corresponding task execution plan. That is, using the aforementioned dynamically adjusted quantified weights, the cost and benefit assessment values ​​of each dimension are weighted, and a preset utility function (which can be linear or nonlinear) is used for comprehensive calculation, unifying multi-dimensional costs and benefits onto a single comparable numerical scale to reflect the overall merits of the task execution plan.

[0137] The solution proposed in this application introduces a systematic data preprocessing process to ensure that the multi-source data used to calculate cost and revenue assessments meet the requirements in terms of time synchronization, format consistency, completeness, and accuracy.

[0138] In another embodiment of this application, S3300 further includes:

[0139] S3310: Using the pose and velocity of the construction robot at the current moment as the initial state, a stochastic motion model including control error, ground friction coefficient fluctuation and sensor noise is constructed. The motion process within a preset future time window is sampled and simulated multiple times to obtain a set of candidate trajectories. The probability of occurrence is assigned to each candidate trajectory to obtain multiple probabilistic prediction paths.

[0140] The S3400 includes:

[0141] S3410: Within a preset future time window, calculate the spatial distance and probability distribution of the overlapping area of ​​the construction robot that needs to replan its path and the construction robot corresponding to the emergency task at each discrete moment. When the spatial distance at a certain moment is less than the safe distance threshold and the probability distribution of the overlapping area is greater than the preset probability threshold, it is determined that there is a potential time-space conflict.

[0142] The construction of a stochastic motion model incorporating control errors, fluctuations in ground friction coefficients, and sensor noise involves modeling various uncertainties of the construction robot in the actual environment during path prediction. Control errors refer to deviations in the robot's control system when executing commands; fluctuations in ground friction coefficients refer to the unevenness of road conditions at the construction site, leading to variations in friction force in different areas or at different times; sensor noise refers to random interference introduced by the robot's onboard sensors when collecting data. These uncertainties are integrated into the stochastic motion model, which can be achieved through methods such as Monte Carlo simulation, Kalman filtering, or particle filtering. For a preset future time window, the motion process is sampled and simulated multiple times, generating a large number of candidate trajectories. Each candidate trajectory is assigned a probability of occurrence, thus forming a set of candidate trajectories with a probability distribution, i.e., multiple probabilistic predicted paths.

[0143] Furthermore, performing temporal-spatial overlap analysis on multiple probabilistic predicted paths involves calculating the spatial distance between the construction robots requiring path replanning and those corresponding to emergency tasks at discrete moments within a preset future time window, and then determining the probability distribution of the corresponding overlapping areas. Spatial distance refers to the geometric distance between the two robots at a given moment; the probability distribution of the overlapping area quantifies the likelihood of the two robots appearing simultaneously within a given spatial region. When the spatial distance at a given moment is less than a preset safe distance threshold, and the probability distribution of the overlapping area at that moment is greater than the preset probability threshold, the system determines that a potential temporal-spatial conflict exists. The safe distance threshold ensures sufficient physical separation between robots to avoid collisions, while the preset probability threshold filters out extremely low-probability, accidental overlaps, focusing on conflict events with higher actual risks.

[0144] By introducing a stochastic motion model and performing multiple sampling simulations of the motion process, this scheme simultaneously considers uncertainties such as control error, fluctuations in ground friction coefficient, and sensor noise during path prediction. This makes the generated probabilistic predicted path closer to the actual motion trajectory of the construction robot in complex and variable construction environments. Compared with traditional deterministic path prediction, it significantly reduces the risk of potential conflict omissions caused by deviations between the predicted and actual paths. Combining the dual judgment of spatial distance and probability distribution of overlapping areas in temporal-spatial overlap analysis, a potential conflict is only identified when robots are close to each other and have a high probability of occurring simultaneously. This effectively avoids false alarms or missed alarms caused by single-condition judgments, making conflict identification more accurate and reliable, and providing a more solid foundation for subsequent scheduling decisions.

[0145] In some preferred embodiments, suppose that at a construction site, a tracked construction robot A is carrying a large precast component and needs to traverse an area with unstable road conditions. Another wheeled construction robot B is performing high-precision welding nearby, and its work area is highly sensitive to vibration and impact. To avoid collisions between robot A and robot B due to swaying or path deviation when traversing the unstable area, the system constructs a stochastic motion model using robot A's current pose and speed as the initial state. This model incorporates potential minor control errors in robot A's control system, fluctuations in the ground friction coefficient caused by uneven road surfaces, and noise from positioning sensors. Then, the system performs multiple (e.g., 1000) sampling simulations of robot A's motion over the next 30 seconds, generating 1000 candidate trajectories. Each candidate trajectory is assigned a probability of occurrence, thus forming a probabilistic predicted path set for robot A over the next 30 seconds.

[0146] Subsequently, the system performs a temporal-spatial overlap analysis on the probabilistic predicted path of robot A and the predicted path of robot B. Within a preset future time window, for example, every second, the system calculates the spatial distance between robot A and robot B at each discrete moment, and calculates the probability distribution of robot A appearing near robot B's work area based on robot A's probabilistic predicted path. If, at a certain moment, the spatial distance between the two robots is less than a preset safe distance threshold (e.g., 2 meters), and the probability distribution of robot A appearing in robot B's work area is greater than a preset probability threshold (e.g., 0.1), the system determines that a potential temporal-spatial conflict exists. For example, simulation results show that at the 15th second, robot A enters a range less than 2 meters from robot B with a 15% probability. The system identifies this situation as a potential conflict and triggers subsequent scheduling processes to generate alternative solutions to avoid the conflict.

[0147] In another embodiment of this application, it is further proposed that the safety distance threshold is a dynamic safety distance threshold, and the process of determining the dynamic safety distance threshold includes:

[0148] A1: Obtain the priority information of the current task. The priority information should include at least the urgency and importance of the task.

[0149] A2: Identify the types of construction robots involved in potential time-space conflicts, including at least one of tracked, wheeled, or flying robots;

[0150] A3: Obtain the collision sensitivity information of the components currently being transported by the construction robot. The collision sensitivity information includes at least the ability of the transported components to withstand impact and vibration.

[0151] A4: Calculate the safety distance adjustment factor based on priority information, type, and collision sensitivity information;

[0152] A5: Apply the safety distance adjustment factor to the preset base safety distance to obtain the dynamic safety distance threshold. Use the dynamic safety distance threshold as the safety distance threshold to compare the spatial distance of construction robots that need to replan their paths and construction robots corresponding to emergency tasks at each discrete moment within a preset future time window, in order to identify potential time-space conflicts.

[0153] Specifically, dynamic safety distance thresholds refer to safety distances that are adaptively adjusted based on real-time changes in the construction environment, task characteristics, and robot and component attributes. This replaces fixed, one-size-fits-all thresholds, making safety distances more closely reflect actual risks. Priority information indicates the importance and time urgency of a task within the construction plan; for example, urgency can be categorized as "high," "medium," or "low," and importance as "critical," "secondary critical," or "general." This information is typically provided by the task management system to guide scheduling priorities. The type of construction robot refers to its motion characteristics and physical dimensions. For example, tracked robots have strong off-road capabilities but large turning radii, wheeled robots are fast but have high requirements for road surfaces, and aircraft possess three-dimensional motion capabilities. Different types have different braking distances, obstacle avoidance capabilities, and post-collision consequences, resulting in different safety distance requirements. Collision sensitivity information refers to the probability and severity of damage to components carried by the construction robot when subjected to impact or vibration. For example, precast concrete components are more sensitive to differential settlement and torsional deformation, while steel structure components have higher impact resistance. This information can be obtained from component BIM models or structural integrity sensitivity files. The safety distance adjustment factor is a dimensionless multiplier or additive / subtractive factor used to correct for a pre-set basic safety distance. The pre-set basic safety distance is the initial safety distance determined based on experience or specifications under standard or average construction conditions.

[0154] This application's solution, when identifying potential temporal-spatial conflicts, does not directly use a fixed safety distance threshold. Instead, it first obtains the priority information of the current task: urgent and important tasks typically require a higher safety margin, leading to an increased safety distance. Next, it identifies the type of construction robot involved in the conflict and adjusts the safety distance based on its speed, braking performance, size, and other characteristics. For example, high-speed wheeled robots or large tracked robots require a larger safety distance. Simultaneously, it integrates the collision sensitivity information of the components currently being transported by the construction robot, significantly amplifying the safety distance requirement for situations involving fragile or high-value components. The system integrates priority information, construction robot type, and collision sensitivity information to calculate a safety distance adjustment factor and applies it to a preset base safety distance to obtain a dynamic safety distance threshold matching the current scenario. Subsequently, within a preset future time window, it compares the spatial distances of construction robots requiring path replanning and those corresponding to urgent tasks at each discrete moment. Combining this with the probability distribution of overlapping areas, when the spatial distance is less than the dynamic safety distance threshold and the overlap probability exceeds a preset probability threshold, a potential temporal-spatial conflict is determined, thus achieving more reasonable conflict identification.

[0155] In some preferred embodiments, it is assumed that there are two construction robots at a construction site that may have temporal-spatial conflicts. Robot A performs high-priority emergency tasks, transporting precision prefabricated components that are extremely sensitive to vibration and impact, and is a high-speed wheeled robot; Robot B performs routine material transportation tasks, transporting ordinary steel bars, and is a low-speed tracked robot. According to the scheme of this application, when determining the safety distance threshold:

[0156] 1. Obtain priority information: Robot A's task priority is "high urgency" and "high importance", while Robot B's priority is "medium urgency" and "general importance".

[0157] 2. Determine the type of construction robot: Robot A is "wheeled", and Robot B is "tracked";

[0158] 3. Obtain collision sensitivity information: The collision sensitivity of robot A carrying precision prefabricated components is "extremely high", while the collision sensitivity of robot B carrying steel bars is "low".

[0159] 4. Calculate the safety distance adjustment factor: The system uses a preset algorithm (such as a weighted function or fuzzy logic combining priority, type and sensitivity) to integrate the above information and give a larger safety distance adjustment factor for the conflict scenario between robot A and robot B, so that the final safety distance is significantly increased, while the adjustment range for the low priority and low sensitivity side is smaller.

[0160] 5. Obtaining the dynamic safety distance threshold: Given a preset basic safety distance of 2 meters, the system may calculate a dynamic safety distance threshold of 3.5 meters for this scenario. Subsequently, within a preset future time window, the system uses 3.5 meters as the safety distance threshold to compare the spatial distance between robot A and robot B at various discrete moments. If the spatial distance between the two is less than 3.5 meters at a certain moment and the probability distribution of the corresponding overlapping area is greater than the preset probability threshold, a potential time-space conflict is determined, and the subsequent scheduling decision process is triggered.

[0161] In another embodiment of this application, A4 is further proposed to include:

[0162] A41: Obtain the risk appetite and environmental conditions of the construction site at the current construction stage. The risk appetite should include at least the degree of emphasis on schedule, cost and safety.

[0163] A42: Calculate the adjustment weight of each piece of information on the safety distance adjustment based on priority information, type, collision sensitivity information, risk preference, and environmental conditions;

[0164] A43: The adjustment weights are applied to priority information, type, collision sensitivity information, risk preference, and environmental conditions. A preliminary safe distance adjustment factor is obtained through nonlinear function mapping.

[0165] A44: Based on the real-time environmental risk level of the construction site, the preliminary safety distance adjustment factor is corrected to obtain the final safety distance adjustment factor.

[0166] Specifically, obtaining the risk preference at the current construction stage refers to understanding the relative emphasis placed on construction progress, cost control, and safety by the project management or on-site command at this stage. For example, the initial stage of the project may prioritize progress, while the stage of high-altitude operations or precision installation may prioritize safety. Environmental conditions at the construction site refer to external factors affecting the operation of construction robots, such as weather conditions (rain, snow, strong winds), lighting conditions, ground slipperiness, visibility, and the presence of temporary obstacles or areas with high personnel activity. This information can be obtained through environmental sensors, meteorological data interfaces, or manual input.

[0167] The adjustment weights for each piece of information in relation to safety distance adjustments are calculated based on priority information, type, collision sensitivity information, risk preference, and environmental conditions. This involves assigning different importance coefficients to each input parameter. For example, under extreme weather conditions, the weight of environmental conditions increases significantly; when transporting highly sensitive components, the weight of collision sensitivity information is higher; and during the safety-focused construction phase, the weight of the "safety" dimension in risk preference increases. These adjustment weights can be determined based on expert experience, historical data analysis, or machine learning models.

[0168] In practical applications, the adjustment weights are applied to priority information, type, collision sensitivity information, risk preference, and environmental conditions. A preliminary safety distance adjustment factor is obtained through nonlinear function mapping. This means that the weighted information is input into a predefined nonlinear function (such as the Sigmoid function, ReLU function, or polynomial function) to characterize the complex interactions and nonlinear cumulative effects between the factors, thereby generating a preliminary adjustment factor that reflects the overall risk level.

[0169] Furthermore, based on the real-time environmental risk level of the construction site, the initial safety distance adjustment factor is revised to obtain the final safety distance adjustment factor. This refers to a secondary correction after the initial calculation, combined with the overall real-time risk assessment results of the current construction site. For example, when the site is determined to be at a high risk level (such as a sudden fire, structural abnormalities, or personnel accidentally entering a dangerous area), even if the initial adjustment factor tends to reduce the safety distance, it is forcibly corrected to a larger safety distance to increase the safety margin. The real-time environmental risk level can be comprehensively assessed through multi-source sensor data, monitoring images, and manual reporting.

[0170] In some preferred embodiments, assuming that during the steel structure hoisting phase of a high-rise building, a tracked construction robot is carrying a precision component highly sensitive to vibration and impact, and needs to traverse narrow passages, while external environmental sensors detect strong winds and slightly slippery ground in the area. According to the solution of this application, the risk preference for the current construction phase is first determined. Since it is a high-altitude precision hoisting phase, the project management prioritizes "safety" the most, followed by "schedule." Simultaneously, the environmental conditions at the construction site are determined to be "strong winds and slightly slippery ground." Next, the system calculates the adjustment weights of each piece of information on the safety distance adjustment based on task priority (precision component hoisting has a higher priority), robot type (tracked), component collision sensitivity (high sensitivity), risk preference (high safety focus), and environmental conditions (strong wind, slippery ground). Among these, the weights of component sensitivity and environmental conditions are significantly increased. Subsequently, the weighted information is input into a preset nonlinear function to obtain a larger initial safety distance adjustment factor. Finally, if the real-time environmental risk level assessment system detects a sudden high-risk event in the area (such as a malfunction of other nearby equipment), the initial adjustment factor is further amplified to obtain the final safety distance adjustment factor, thereby reserving a more generous dynamic safety distance between the construction robot and the surrounding environment and other robots in this scenario.

[0171] Reference Figure 2 The specific embodiments of this application also disclose a construction robot control system, including:

[0172] The monitoring and reporting module 1 is used to receive real-time operating status parameters of the construction robot that exceed the preset normal range. The operating status parameters include at least one or more of the following: drive motor load, deviation between actual travel speed and expected travel speed, and changes in vehicle posture.

[0173] The environment model update module 2 is used to inversely quantify the local environmental physical characteristics of the area where the construction robot is located based on the operating status parameters and the physical parameters of the corresponding construction robot, and dynamically update the preset global construction environment model based on the local environmental physical characteristics, and mark the area where the construction robot is located as a passage risk area.

[0174] The conflict prediction module 3 is used to continuously monitor the task status of all construction robots and external emergency tasks. When it is found that there are construction robots that need to replan their paths and there are emergency tasks, it predicts the replanned paths of the construction robots and the paths corresponding to the emergency tasks based on the updated global construction environment model, and identifies potential time-space conflicts.

[0175] The scheme generation module 4 is used to generate multiple alternative paths for construction robots involving potential time-space conflicts based on the updated global construction environment model, and to combine the alternative paths with different task execution orders, waiting strategies and / or passage priority rules to form multiple task execution schemes.

[0176] The cost-benefit calculation module 5 is used to quantitatively evaluate the multi-dimensional costs and benefits corresponding to each task execution plan based on the updated global construction environment model and the current task status of each construction robot, and obtain the corresponding cost evaluation value and benefit evaluation value.

[0177] The utility evaluation module 6 is used to convert the cost evaluation value and the benefit evaluation value into the total utility score of the corresponding task execution plan according to the preset weight parameters and utility function.

[0178] Decision selection module 7 is used to compare the total utility scores of each task execution plan and select the plan with the highest total utility score as the optimal scheduling decision;

[0179] The instruction issuing module 8 is used to issue corresponding control instructions to relevant construction robots based on the optimal scheduling decision.

[0180] The construction robot control system proposed in this application aims to solve the problems of traditional construction robot control systems, such as difficulty in responding to operational conflicts, efficiency decline, and resource waste in real time in dynamic and uncertain construction environments.

[0181] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A control method for a construction robot, characterized in that, include: The system receives real-time operational status parameters from the construction robot that exceed the preset normal range. These operational status parameters include at least one or more of the following: drive motor load, deviation between actual and expected travel speed, and vehicle body posture changes. Based on the operating status parameters and the corresponding physical parameters of the construction robot, the local environmental physical characteristics of the area where the construction robot is located are inverted and quantified, and the preset global construction environment model is dynamically updated based on the local environmental physical characteristics, and the area where the construction robot is located is marked as a passage risk area. The system continuously monitors the task status of all construction robots and externally issued emergency tasks. When it is identified that a construction robot needs to replan its path and there is an emergency task, it predicts the replanned path of the construction robot and the path corresponding to the emergency task based on the updated global construction environment model, and identifies potential time-space conflicts. Based on the updated global construction environment model, multiple alternative paths are generated for construction robots involved in the potential time-space conflict. These alternative paths are then combined with different task execution sequences, waiting strategies, and / or passage priority rules to form multiple task execution schemes. For each task execution plan, based on the updated global construction environment model and the current task status of each construction robot, the multi-dimensional costs and multi-dimensional benefits corresponding to the task execution plan are quantitatively evaluated to obtain the corresponding cost evaluation value and benefit evaluation value. Based on preset weight parameters and utility functions, the cost assessment value and the benefit assessment value are uniformly converted into the total utility score of the corresponding task execution plan; Compare the total utility scores of each task execution plan and select the plan with the highest total utility score as the optimal scheduling decision; Based on the optimal scheduling decision, corresponding control commands are issued to the relevant construction robots.

2. The construction robot control method according to claim 1, characterized in that, Based on preset weight parameters and utility functions, the cost assessment value and the benefit assessment value are uniformly converted into the corresponding total utility score of the task execution plan, including: Obtain the structural integrity sensitivity profile of the component currently being transported by the construction robot. The structural integrity sensitivity profile includes at least the maximum differential settlement threshold and torsional deformation threshold that the component can withstand. Based on the pavement deformation sensitivity index of the alternative paths corresponding to the task execution scheme in the updated global construction environment model and the percentage decrease in instantaneous bearing capacity caused by external instantaneous disturbance, the differential settlement and torsional deformation of the component under the alternative paths are predicted. When the differential settlement exceeds the maximum differential settlement threshold or the torsional deformation exceeds the torsional deformation threshold, calculate the component damage risk cost corresponding to the task execution plan; The component damage risk cost, along with other quantitative assessment results of the multi-dimensional costs and multi-dimensional benefits, are used as inputs to participate in the calculation of the total utility score corresponding to the task execution plan.

3. The construction robot control method according to claim 1, characterized in that, Continuously monitor the task status of all construction robots and externally issued emergency tasks. When it is identified that a construction robot needs to replan its path and an emergency task exists, based on the updated global construction environment model, predict the replanned path of the construction robot and the path corresponding to the emergency task, and identify potential time-space conflicts, including: The robot receives external environment perception data in real time, including obstacle information, road surface condition information, and other dynamic environmental elements in the area where the construction robot is located. Based on the external environment perception data, the local area information in the updated global construction environment model is dynamically corrected, including updating the location of obstacles, road traffic costs, and temporary work areas. Based on the dynamically corrected global construction environment model, as well as the kinematic and dynamic parameters of the construction robot, multiple probabilistic predicted paths are generated for the construction robot that needs to replan its path and for the construction robot corresponding to the emergency task within a preset future time window. A time-space overlap analysis is performed on the multiple probabilistic predicted paths, and based on the probability distribution of the overlapping areas and a preset safe distance threshold, potential time-space conflicts are identified between the construction robot that replans its path and the construction robot corresponding to the emergency task.

4. The construction robot control method according to claim 1, characterized in that, Based on the updated global construction environment model, multiple alternative paths are generated for construction robots involved in the potential time-space conflict. These alternative paths are then combined with different task execution sequences, waiting strategies, and / or passage priority rules to form multiple task execution schemes, including: Obtain the structural integrity sensitivity profile of the component currently being transported by the construction robot. The structural integrity sensitivity profile includes at least the maximum differential settlement threshold and torsional deformation threshold that the component can withstand. Obtain the efficiency drift index and vibration characteristic changes of the construction robot drive system; Obtain the percentage decrease in instantaneous bearing capacity caused by external instantaneous disturbance; In the path planning algorithm, the maximum differential settlement threshold and torsional deformation threshold, the efficiency drift index and vibration characteristic change, and the instantaneous bearing capacity decrease percentage are used as dynamic constraints and optimization objectives to adjust the path cost of each candidate path and / or eliminate candidate paths that do not meet the preset safety constraints, so as to generate multiple alternative paths and their corresponding task execution schemes.

5. The construction robot control method according to claim 1, characterized in that, Based on preset weight parameters and utility functions, the cost assessment value and the benefit assessment value are uniformly converted into the corresponding total utility score of the task execution plan, including: Obtain the priority of the current task, the real-time environmental risk level of the construction site, and the structural integrity sensitivity of the components carried by the construction robot; Based on the priority of the current task, the real-time environmental risk level of the construction site, and the structural integrity sensitivity, the weight parameters are dynamically adjusted to obtain the quantitative weights corresponding to the multi-dimensional costs and multi-dimensional benefits. The quantitative weights are applied to the cost assessment value and the benefit assessment value, and the cost assessment value and the benefit assessment value are weighted according to the utility function to obtain the total utility score of the corresponding task execution plan.

6. The construction robot control method according to claim 5, characterized in that, The quantitative weights are applied to the cost assessment value and the benefit assessment value, and the cost assessment value and the benefit assessment value are weighted according to the utility function to obtain the total utility score of the corresponding task execution plan, including: It receives multi-source data from construction robots, external environment sensors, task management systems, and component BIM models; The multi-source data is timestamped and converted to the data format to obtain aligned multi-source data; The missing data in the aligned multi-source data is imputed to obtain the imputed multi-source data. Abnormal data are removed from the imputed multi-source data to obtain the removed multi-source data. Based on the removed multi-source data, the quantitative results of the multi-dimensional costs and multi-dimensional benefits are calculated to obtain the cost assessment value and the benefit assessment value. The quantitative weights are applied to the cost assessment value and the benefit assessment value, and the cost assessment value and the benefit assessment value are weighted according to the utility function to obtain the total utility score of the corresponding task execution plan.

7. The construction robot control method according to claim 3, characterized in that, Based on the dynamically corrected global construction environment model, and the kinematic and dynamic parameters of the construction robot, multiple probabilistic predicted paths are generated within a preset future time window for the construction robot requiring path replanning and the construction robot corresponding to the emergency task, including: Using the current pose and speed of the construction robot as the initial state, a stochastic motion model including control error, ground friction coefficient fluctuation and sensor noise is constructed. The motion process within a preset future time window is sampled and simulated multiple times to obtain a set of candidate trajectories. The probability of occurrence is assigned to each candidate trajectory to obtain multiple probabilistic prediction paths. A temporal-spatial overlap analysis is performed on the multiple probabilistic predicted paths. Based on the probability distribution of the overlapping areas and a preset safety distance threshold, potential temporal-spatial conflicts are identified between the construction robot that replans its path and the construction robot corresponding to the emergency task, including: Within the preset future time window, the spatial distance and probability distribution of the corresponding overlapping area of ​​the construction robot that needs to replan its path and the construction robot corresponding to the emergency task at each discrete moment are calculated. When the spatial distance at a certain moment is less than the safe distance threshold and the probability distribution of the overlapping area is greater than the preset probability threshold, it is determined that there is a potential time-space conflict.

8. The construction robot control method according to claim 7, characterized in that, The safety distance threshold is a dynamic safety distance threshold, and the process of determining the dynamic safety distance threshold includes: Obtain the priority information of the current task, wherein the priority information includes at least the urgency and importance of the task; Identify the types of construction robots involved in the potential time-space conflict, wherein the types include at least one of tracked, wheeled, or flying robots; Obtain the collision sensitivity information of the component currently being transported by the construction robot, wherein the collision sensitivity information includes at least the component's ability to withstand impact and vibration. Calculate the safety distance adjustment factor based on the priority information, the type, and the collision sensitivity information; The safety distance adjustment factor is applied to a preset base safety distance to obtain the dynamic safety distance threshold. The dynamic safety distance threshold is then used as the safety distance threshold. Within the preset future time window, the spatial distances of the construction robots that need to replan their paths and the construction robots corresponding to the emergency tasks are compared at each discrete moment to identify potential time-space conflicts.

9. The control method for a construction robot according to claim 8, characterized in that, Based on the priority information, the type, and the collision sensitivity information, a safety distance adjustment factor is calculated, including: Obtain the risk preference and environmental conditions of the construction site at the current construction stage, wherein the risk preference includes at least the degree of emphasis on schedule, cost and safety; Based on the priority information, the type, the collision sensitivity information, the risk preference, and the environmental conditions, calculate the adjustment weight of each piece of information on the safety distance adjustment; The adjustment weights are applied to the priority information, the type, the collision sensitivity information, the risk preference, and the environmental conditions, and a preliminary safe distance adjustment factor is obtained through nonlinear function mapping. Based on the real-time environmental risk level at the construction site, the preliminary safety distance adjustment factor is corrected to obtain the final safety distance adjustment factor.

10. A control system for a construction robot, characterized in that, include: The monitoring and reporting module is used to receive real-time operating status parameters of the construction robot that exceed the preset normal range. The operating status parameters include at least one or more of the following: drive motor load, deviation between actual travel speed and expected travel speed, and changes in vehicle posture. The environment model update module is used to inversely quantify the local environmental physical characteristics of the area where the construction robot is located based on the operating status parameters and the physical parameters of the corresponding construction robot, and dynamically update the preset global construction environment model based on the local environmental physical characteristics, and mark the area where the construction robot is located as a passage risk area. The conflict prediction module is used to continuously monitor the task status of all construction robots and external emergency tasks. When it is identified that a construction robot needs to replan its path and there is an emergency task, it predicts the replanned path of the construction robot and the path corresponding to the emergency task based on the updated global construction environment model, and identifies potential time-space conflicts. The scheme generation module is used to generate multiple alternative paths for construction robots involved in potential time-space conflicts based on the updated global construction environment model, and to combine the alternative paths with different task execution orders, waiting strategies and / or passage priority rules to form multiple task execution schemes. The cost-benefit calculation module is used to quantitatively evaluate the multi-dimensional costs and multi-dimensional benefits corresponding to each task execution plan based on the updated global construction environment model and the current task status of each construction robot, and obtain the corresponding cost evaluation value and benefit evaluation value. The utility evaluation module is used to convert the cost evaluation value and the benefit evaluation value into the total utility score of the corresponding task execution plan according to preset weight parameters and utility functions. The decision selection module is used to compare the total utility scores of each task execution plan and select the plan with the highest total utility score as the optimal scheduling decision. The instruction issuing module is used to issue corresponding control instructions to the relevant construction robots based on the optimal scheduling decision.

Images