A multi-machine linkage-based intelligent matching and scheduling method and system for construction robots

Abstract

The application provides a kind of based on multi-machine linkage's building robot intelligent matching scheduling method and system, including multiple robots and cloud;Including the following steps: input building drawing set to cloud, three-dimensional construction scene set and total coordinate system are updated to the whole robot in this construction site;Robot scans environment and identifies the fixed obstacle of space and sends to cloud, and sends the updated three-dimensional construction scene to other robots;Robot updates process stage label by process stage model, and cloud selects robot to go to construction coordinate and carries out construction task.Using the above method, three-dimensional construction scene set is constructed through cloud and a unified total coordinate system is defined, to ensure that all robots work under the same coordinate system. Robots can share the position of fixed obstacles in real time. Cloud can flexibly schedule robots according to the current construction progress and demand, and intelligently select the most suitable robot for operation.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of construction robots, in particular to a construction robot intelligent matching and scheduling method and system based on multi-machine linkage. BACKGROUND

[0002] With the development of the construction industry towards intelligence and digitization, construction robots are increasingly widely used in construction sites. At present, construction robots have been gradually applied to masonry, plastering, spraying, handling and other construction operations, which has improved construction efficiency and reduced labor intensity to a certain extent. However, there are still many problems in multi-robot collaborative operation. Due to the complex construction site environment and variable process flow, there is a lack of unified task scheduling platform among different types of robots, making it difficult to realize information interconnection and efficient cooperation, resulting in resource waste, task conflict and even frequent safety hazards.

[0003] In actual engineering, the task allocation of construction robots usually relies on manual experience for on-site scheduling, which is not only inefficient, but also difficult to adapt to the changing process stages and site conditions in the construction process. Some systems try to remotely manage robots through centralized control platforms, but when facing simultaneous operation of multiple robots, they often cannot establish a unified spatial coordinate system according to the construction drawings and the site environment, resulting in frequent problems such as robot positioning deviation and path conflict.

[0004] In addition, most current construction robot systems only rely on static path planning according to architectural drawings before construction, lacking real-time obstacle recognition and feedback mechanisms, which leads to frequent failures in obstacle avoidance or task interruption during construction. At the same time, for dynamic updates in the construction phase, such as process flow adjustment and construction sequence change, existing systems also lack effective automatic identification and response capabilities, still requiring a large amount of manual intervention, limiting the intelligent level and adaptability of construction robot systems.

[0005] Therefore, how to realize task collaboration and intelligent scheduling among construction robots in complex construction environments has become a key technical problem to be solved in the field of construction robots. SUMMARY

[0006] To overcome the existing technical problems, the present application provides a construction robot intelligent matching and scheduling method and system based on multi-machine linkage.

[0007] The present application adopts the following technical solutions.

[0008] A construction robot intelligent matching and scheduling method based on multi-machine linkage, comprising a plurality of robots with a camera module, and a cloud connected to the plurality of robots;

[0009] The method comprises the following steps:

[0010] S1, input the building drawing set to the cloud, the cloud constructs a three-dimensional construction scene set according to the building drawing set and defines a total coordinate system, and updates the three-dimensional construction scene set and the total coordinate system to all robots in the construction site;

[0011] S2, the robot scans the environment to identify the fixed obstacles in the space and sends them to the cloud, the cloud updates and integrates them into the three-dimensional construction scene set, and updates the three-dimensional construction scene to other robots;

[0012] S3, manually update the process stage label or the robot updates the process stage label through the process stage model, and the cloud selects the robot to go to the construction coordinate according to the process stage label and the robot data to perform the construction task.

[0013] As a further improvement of the application, the robot has a state label synchronized with the cloud, which includes a construction state, a charging state and an idle state;

[0014] If the robot has a construction task, it is marked as a construction state, and if there is no construction task and charging task, it is marked as an idle state;

[0015] When the power of the robot marked as an idle state is lower than the preset charging threshold, perform the charging task, and switch the idle state to the charging state;

[0016] The robot is also provided with a limit power threshold lower than the charging threshold, and when the power of the robot marked as a construction state is lower than the limit power threshold, the charging task is performed, and the construction state is marked as a charging state.

[0017] As a further improvement of the application, each process stage label includes a plurality of construction steps, and the robot is provided with a construction step set including at least one construction step;

[0018] The step S3 further includes a step S4:

[0019] The robot has a display module, and the display module displays the current construction step of the robot;

[0020] When there are at least two robots performing the same construction step, each robot scans the environment through a camera module and corrects the posture to obtain an environment scan map, and calculates the similarity between the environment scan map and the step scene map corresponding to the last construction step in the three-dimensional construction scene set.

[0021] If the similarity is greater than the construction scene threshold, the robot is selected, and all selected robots are integrated to form a pre-selected election set;

[0022] If the similarity is less than the construction scene threshold value, each robot moves around a circle with itself as the origin and a preset election scene value as the radius, stops multiple times during the process of surrounding to perform environmental scanning and corrects the posture to obtain a three-dimensional nearby environment scanning map, calculates the similarity between the three-dimensional nearby environment scanning map and the step scene map corresponding to the previous construction step, selects the robot with the highest similarity value as the master robot, and the master robot performs real-time environmental scanning to construct a three-dimensional nearby environment scanning map;

[0023] If there is a pre-selected election set, the robots therein are calculated respectively The robot with the highest cooperative election score is selected as the master robot, and the coordinates of the robot when performing environmental scanning are recorded as the supervision coordinates;

[0024] ,

[0025] ,

[0026] Among them, is the percentage of the power of the i th robot, is the power weight coefficient, is the sum of the distances between the i th robot and the j th robot in the same construction task, is the transmission average distance weight coefficient, is the construction efficiency of the i th robot, is the construction efficiency weight coefficient, is the average construction amount of the i th robot, is the average construction amount of the same type of robot as the i th robot, is the construction efficiency decay of extreme weather, is the weight coefficient of , is the weight coefficient of , is the average construction amount of the e th extreme weather type;

[0027] If the current construction step requires at least two robots to perform together, the master robot sends an action synchronization instruction to the co-performing robot.

[0028] As a further improvement of the present application, the specific steps of updating the process stage label through the process stage model include:

[0029] According to the environmental scanning map or the three-dimensional nearby environment scanning map, and the step scene map corresponding to the construction step in the three-dimensional construction scene set, the progress of the construction step is analyzed, and when the progress of the construction step is greater than the preset step threshold value, the master robot broadcasts to the robot of the next construction step;

[0030] If the current construction step is the last construction step of the process stage, the progress of the construction step is compared with the step threshold value, and when the progress of the construction step is greater than the preset stage threshold value, the host robot sends a process stage label marked as a preparation state to the cloud, and the cloud selects a robot marked as an idle state according to the process stage label marked as the preparation state and the robot data, and switches the idle state of the robot to the preparation state;

[0031] When the power of the robot in the preparation state is lower than the preset preparation power threshold value, the preparation state is maintained while being marked as a charging state to perform a charging task;

[0032] When the progress of the last construction step of the process stage is greater than the preset stage completion threshold value, the host robot sends an environment scan map or a three-dimensional adjacent environment scan map and a next process stage label to the cloud, and the cloud judges whether the progress of the step is greater than the stage completion threshold value according to the environment scan map or the three-dimensional adjacent environment scan map and the step scene map corresponding to the construction step in the three-dimensional construction scene, and updates the process stage label if yes;

[0033] According to the environment scan map or the three-dimensional adjacent environment scan map and the step scene map corresponding to the construction step in the three-dimensional construction scene, the specific step of the progress of the construction step is analyzed and obtained, including:

[0034] The host robot goes to a supervision coordinate according to a preset supervision interval to perform environment scanning and obtains an environment scan map through pose correction, or the host robot performs real-time environment scanning to construct a three-dimensional adjacent environment scan map, the environment scan map and the three-dimensional adjacent environment scan map are both point cloud data, the step scene map corresponding to the construction step has a construction component in the form of point cloud data, and the progress of the construction step ;

[0035] ,

[0036] Among them, is the nth three-dimensional coordinate of the jth construction component of the ith process stage, is an error range of the mth three-dimensional coordinate in the environment scan map or the three-dimensional adjacent environment scan map, and the unit is millimeter, is the number of three-dimensional coordinates of the construction component falling into the error range of the three-dimensional coordinate of the environment scan map or the three-dimensional adjacent environment scan map, is the total amount of point cloud data of the jth construction component of the ith process stage.

[0037] As a further improvement of the present application, the specific step of selecting a robot by the cloud according to the process stage label and the robot data includes:

[0038] If the construction steps of the robot are the same as the construction steps in the process stage label, and the robot is not in the charging state, then the robot is selected, and a plurality of preselected construction sets are integrated according to each construction step in the process stage label;

[0039] Each construction step is pre-set with a standard construction machine quantity. If the robot in the preparation state exists in the construction step, the robot is selected, and the preparation state is switched to the construction state. If the robot is also marked as the charging state, the construction state is switched after the charging task is completed. The required robot quantity of each construction step ;

[0040] ,

[0041] wherein, is the required robot quantity per thousand square, is the construction size of the step, is the number of robots in the preparation state, and j is the number of construction steps;

[0042] If the required robot quantity of the construction step is greater than the robot quantity of the corresponding preselected construction set, all robots in the preselected construction set are marked as the construction state.

[0043] If the required robot quantity of the construction step is less than the robot quantity of the corresponding preselected construction set, the selected construction score of each robot in the preselected construction set is calculated ;

[0044] ,

[0045] wherein, is the standard moving length, is the path length of the i th robot to the construction area, is the power percentage of the i th robot, is the total number of construction steps in the process stage, is the number of i th robots selected by the preselected construction set, , and is the selected score weight factor;

[0046] According to the selected construction score from large to small, the robot is selected and marked as the construction state.

[0047] As a further improvement of the present application, the specific steps of the robot going to the construction coordinates for the construction task include: the three-dimensional construction scene set includes step scene graphs of each construction step of each process stage and real-time three-dimensional scene graphs, and the step scene graph has a construction area.

[0048] Randomly select multiple coordinates in the step scene graph as construction coordinates, and send the construction coordinates to the robot one by one, and the robot carries out Monte Carlo path planning according to the real-time three-dimensional scene graph and the construction coordinates, and in the process of going to the construction coordinates, the robot scans the environment in real time to carry out Monte Carlo path planning to avoid obstacles;

[0049] If there are at least two robots in the same construction step, and all robots establish signal connection with each other when they all reach the construction coordinates, each robot scans the environment to identify whether there is a construction component of the construction step in the environment scan graph, if yes, the construction component area is set as a material area and is synchronized to other robots under the same construction step, if not, all robots randomly move and re-scan the environment, when there is a material area and a construction area, each robot carries out the construction task according to the preset program.

[0050] As a further improvement of the application, the step of constructing a three-dimensional construction scene set and defining a total coordinate system according to the building drawing set by the cloud end comprises:

[0051] The names of the drawings in the building drawing set are identified and classified by NLP natural language processing technology to obtain a plot master drawing, a single building design drawing set and a building construction drawing set, the drawings in the single building design drawing set and the building construction drawing set are arranged according to the drawing numbers and imported into the BIM software, and a three-dimensional construction scene set of point cloud data is generated, and the southwest corner of the plot master drawing is selected as the origin to generate a total coordinate system;

[0052] The three-dimensional construction scene set includes step scene graphs of each construction step in each process stage and real-time three-dimensional scene graphs.

[0053] As a further improvement of the application, the robot scans the environment to identify the fixed obstacles in the space and sends them to the cloud end, and the cloud end updates and integrates them into the three-dimensional construction scene set, and the three-dimensional construction scene is updated to other robots. The specific steps include:

[0054] The robot is pre-set with an obstacle detection interval, when the robot scans the environment and identifies that there is an obstacle in a certain space, it continuously scans the environment, if the obstacle moves within the obstacle detection interval, it is considered that the obstacle is a dynamic obstacle, and the continuous scanning of the environment is stopped, if the obstacle remains stationary beyond the obstacle detection interval, it is considered that the obstacle is a fixed obstacle, the robot takes multiple views of the fixed obstacle and sends it to the cloud end, the cloud end generates three-dimensional reconstruction point cloud data through COLMAP and integrates and updates it into the three-dimensional construction scene set.

[0055] As a further improvement of the application, when the power of the robot marked as the construction state is lower than the limit power threshold, the construction state is switched to the charging state and is synchronized to the master robot;

[0056] The host calculates the power required for the remaining task ;

[0057] ,

[0058] wherein, is the average power of each robot of the same type as the robot switched to the charging state at the beginning of the construction step, is the average power of each robot of the same type as the robot switched to the charging state when the robot exited the construction task, is the number of robots of the same type as the robot switched to the charging state at the beginning of the construction step;

[0059] The host calculates the total available power of the robots currently in the construction state and of the same type as the robot switched to the charging state ;

[0060] ,

[0061] wherein, is the current power of the tth robot of the same type, is the threshold value of the limit power, is the power surplus value, if the total available power is less than the power required for the remaining task , the robot in the charging state and not in the preparation state of the same type is selected and sent to the construction coordinates, and after the robot completes the charging task, the robot goes to the construction coordinates to perform the construction task.

[0062] The application also provides a building robot intelligent selection and scheduling system based on multi-machine linkage, which uses the building robot intelligent selection and scheduling method based on multi-machine linkage as described above, and comprises the following modules:

[0063] A plurality of robots, which are provided with a construction program, are used for scanning the environment and performing the construction task;

[0064] A cloud, which is in signal connection with the plurality of robots, is used for receiving the building drawing set, defining the total coordinate system, issuing the construction coordinates and the construction task to the robots;

[0065] A three-dimensional construction scene construction module, which is arranged in the cloud, is used for constructing the three-dimensional construction scene set according to the building drawing set or updating the three-dimensional construction scene set according to the fixed obstacles recognized by the robots;

[0066] A process stage label updating module, which is used for updating the process stage label by manual updating or by a process stage model, is used for the cloud to select the robots for construction according to the updated process stage label;

[0067] The path planning module, located on the robot, plans the path to the construction coordinates based on the robot's overall coordinate system and the construction coordinates.

[0068] The task selection module, located in the cloud, selects the robot to perform the construction task based on the process stage label and robot data.

[0069] The status monitoring module is used to detect the robot's operating status and can change the robot's status label.

[0070] The process stage feedback module is used to send the robot's construction results to the cloud and trigger the allocation of tasks for the next process stage.

[0071] The beneficial effects of this invention are as follows: By constructing a 3D construction scene set in the cloud and defining a unified overall coordinate system, all robots are ensured to work under the same coordinate system, reducing construction deviations caused by human measurement and positioning errors. Robots can share the positions of fixed obstacles in real time, and these positions are updated and integrated into the 3D construction scene set via the cloud, allowing each robot to obtain the latest construction site conditions. Process stage labels are updated manually or automatically through the process stage model, enabling the cloud to flexibly schedule robots according to the current construction progress and needs, intelligently selecting the most suitable robot to work at specific construction coordinates, maximizing resource utilization and reducing idle time. Attached Figure Description

[0072] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0073] Figure 1 This is a flowchart illustrating the present invention;

[0074] Figure 2 This is an example diagram of the robot of the present invention reaching the construction coordinates. Detailed Implementation

[0075] The accompanying drawings are for illustrative purposes only and should not be construed as limiting the scope of this patent. To better illustrate this embodiment, some parts in the drawings may be omitted, enlarged, or reduced, and do not represent the actual dimensions of the product.

[0076] It will be understood by those skilled in the art that certain well-known structures and their descriptions may be omitted in the accompanying drawings. The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0077] Reference Figure 1 and Figure 2The application discloses a multi-machine linkage-based intelligent matching and scheduling method for construction robots, which comprises multiple robots provided with camera modules and a cloud end connected with the multiple robots.

[0078] The method comprises the following steps:

[0079] S1, inputting a set of construction drawings to the cloud end, constructing a set of three-dimensional construction scenes and defining a total coordinate system according to the set of construction drawings, and updating the set of three-dimensional construction scenes and the total coordinate system to all robots of the construction site;

[0080] As a further improvement of the application, the step of constructing a set of three-dimensional construction scenes and defining a total coordinate system according to the set of construction drawings specifically comprises:

[0081] The names of the drawings in the set of construction drawings are recognized and classified through NLP natural language processing technology, a general plot plan, a single building design drawing set and a construction drawing set are obtained, the drawings in the single building design drawing set and the construction drawing set are arranged according to drawing numbers and imported into BIM software, a set of three-dimensional construction scenes in the form of point cloud data is generated, and a total coordinate system is generated by taking the southwest corner of the general plot plan as an origin;

[0082] Taking the southwest corner as the origin to generate the total coordinate system can realize that the coordinates of the robots moving in the entire construction site are all positive, reduce the burden of data processing, and provide intuitive position information for the operator. Since there are strict drawing specifications in China, for some larger construction projects, the construction site will have a general plot plan containing the positions of all single building designs, such as "Plot A-Overall Planning Plan". If the project is small, there may be no general plot plan, but only a single building design drawing. If there are multiple single building design drawings but no general plot plan for some construction projects, the construction cannot be approved, so at this time, if there is no general plot plan, the single building design drawing is also regarded as the general plot plan, and a total coordinate system is generated by taking the southwest corner of the single building design drawing as an origin. The coordinate system of the entire construction site can be established.

[0083] Since each drawing has strict naming specifications, such as "Building B-1st Floor Plan", "Building B-Standard Floor Plan" and "Building B-Structural Construction Drawing- Foundation Detail", the drawing names can be directly automatically classified by NLP natural language processing technology. Moreover, the drawings have numbers, for example, "Building B-1st Floor Plan" may be B01, which can be directly imported into the BIM software according to the order of the numbers.

[0084] The three-dimensional construction scene set generated as point cloud data in the BIM software can be three-dimensional modeling in the BIM software according to drawing information by manual modeling, which needs to be modeled gradually according to the process stage and the construction step, and finally point cloud data is directly generated by the software. Through the point cloud data, subsequent Monte Carlo path planning, building component comparison, and real-time three-dimensional scene graph updating in the three-dimensional construction scene set can be facilitated.

[0085] It should be pointed out that although it is necessary to model gradually according to the order of each process stage and construction step by manual modeling, the modeling can be directly used to guide the subsequent automatic construction of the robot after the modeling is completed, and the robot can operate uninterruptedly for 24 hours, greatly increasing the efficiency and safety of building construction. The physical labor in high-temperature environment is transferred to the mental activity in the office, and the modeling is in the early stage, which will not delay the progress of the later construction. From the effect, the construction time can be greatly reduced, so it has obvious practicability. Since the current process stage and construction step have a clear sequence specification, each construction step in each process stage can be individually modeled by manual modeling to obtain a step scene graph and label the name of the construction step, and the step scene graph is automatically sorted according to the construction step sequence preset in the system. Of course, since the construction of the whole building is almost incremental construction except the construction step of removing the template, the modeling basis of the next construction step is the step scene graph of the previous construction step when modeling manually, so the modeling will be performed according to the order of the process stage and the construction step.

[0086] The three-dimensional construction scene set includes the step scene graph of each process stage and each construction step and the real-time three-dimensional scene graph.

[0087] S2, the robot scans the fixed obstacles in the space and sends them to the cloud, the cloud updates and integrates them into the three-dimensional construction scene set, and the three-dimensional construction scene is updated to other robots;

[0088] As a further improvement of the present application, the specific steps of the robot scanning the fixed obstacles in the space and sending them to the cloud, the cloud updating and integrating them into the three-dimensional construction scene set, and the three-dimensional construction scene updating to other robots include:

[0089] The robot is preset with an obstacle detection interval. When the robot scans the environment and identifies that a space has an obstacle, it continuously scans the environment. If the obstacle moves within the obstacle detection interval, it is considered a dynamic obstacle, and the continuous scanning of the environment is stopped. If the obstacle remains stationary beyond the obstacle detection interval, it is considered a fixed obstacle, and the robot takes a multi-view photograph of the fixed obstacle and sends it to the cloud. The cloud generates three-dimensional reconstruction point cloud data using COLMAP and integrates and updates it to the three-dimensional construction scene set.

[0090] COLMAP is an open-source computer vision library mainly used to generate three-dimensional reconstruction point clouds from multi-view images. Specifically, to facilitate the acquisition of obstacle depth information and subsequent acquisition of environment scan maps or three-dimensional nearby environment scan maps, the robot's camera module uses an RGB-D depth camera, reducing the computational burden of generating point cloud data and extending the service life of the construction robot. By presetting the obstacle detection interval, it can effectively determine whether the obstacle is temporary or always present. If it is temporary, it will not affect the pre-planning of the path of other robots and does not need to be uploaded to the cloud, such as when a worker or other robot passes by. However, if it is always present, it may have a significant impact on the actual path of the robot when it receives the construction coordinates for path pre-planning, such as a pile of construction components placed directly in the middle of the road causing the robot to take a detour, affecting the authenticity of the construction score selection Therefore, the point cloud data of the fixed obstacle needs to be reported to the cloud and updated to other robots. Specifically, except for workers who are resting and will stay for a long time without obvious body movement, workers at other times will not remain motionless. Robots will move when in operation, or like the front fork of a forklift, part of the structure will move up and down. Therefore, setting the obstacle detection interval to twenty seconds can effectively detect fixed obstacles, such as sleeping workers, placed construction components or mechanical devices, etc. Of course, if the fixed obstacle in the area disappears when a robot passes by and performs an environment scan, it can be updated to the cloud. Multi-view photography specifically involves taking at least four direction RGB images by circling the obstacle. Of course, due to the complexity of the construction scene, it is not always possible to circle the obstacle. At this time, a telescopic device can be added to the robot's camera module to enable multi-angle photography at different heights.

[0091] S3, manually update the process phase label or the robot updates the process phase label through the process phase model, and the cloud selects a robot to go to the construction coordinates for a construction task based on the process phase label and robot data.

[0092] The process stage label includes a foundation engineering stage, a main structure construction stage, an enclosure construction stage, a decoration stage, and a device installation stage.

[0093] The foundation engineering stage includes construction steps such as foundation excavation, piling, pouring of concrete cushion and foundation structure, backfilling and compaction treatment. The main structure construction stage includes construction steps such as erecting scaffolding and formwork, binding reinforcement frames, pouring concrete, and removing formwork and scaffolding. The enclosure construction stage includes construction steps such as external wall masonry or installation of prefabricated external wall panels, roof waterproofing, insulation layer construction, window installation, and door frame installation. The decoration stage includes construction steps such as internal and external wall plastering, painting, floor laying, ceiling installation, and partition installation. The device installation stage includes construction steps such as installation of drainage pipes, electrical lines, and heating, ventilation, and air conditioning systems.

[0094] As a further improvement of the present application, each robot has a state label synchronized with the cloud, which includes a construction state, a charging state, and an idle state.

[0095] If the robot has a construction task, it is marked as the construction state, and if there is no construction task and charging task, it is marked as the idle state.

[0096] When the power of the robot marked as the idle state is lower than a preset charging threshold, a charging task is performed to switch the idle state to the charging state.

[0097] The robot is also preset with a limit power threshold lower than the charging threshold, and when the power of the robot marked as the construction state is lower than the limit power threshold, a charging task is performed to mark the construction state as the charging state.

[0098] As a specific embodiment of the present application, the limit power threshold is 5% to 15%, which needs to be controlled according to the distance from the position where the robot is in the construction state to the charging position. If there is a situation that the robot reaches the limit power threshold to charge and cannot reach the charging position smoothly, the value of the limit power threshold needs to be adjusted. The charging threshold is 70~90%, which is selected according to the size of the construction project, for the convenience of subsequent selection of robots, and there is enough power to perform the construction task. At this time, it needs to be noted that the robot in the idle state is not directly executed to perform the charging task, and there are two reasons. First, not every robot is suitable for every construction step, so theoretically there will be many idle robots. If the idle robot directly performs the charging task, it will cause a large number of robots to go to charge after each construction step, and the number of charging piles is insufficient, and it will cause great congestion on the construction site, and frequent charging may affect the battery of the robot. Second, the idle robot can monitor the surrounding environment in real time, and can perform safety monitoring. Once a serious safety accident occurs, the camera can record and view, and the environment can be scanned to update the three-dimensional construction scene set.

[0099] Synchronizing the state label to the cloud can make the robot in the construction state perform the charging task, and let the cloud assess whether the remaining robots can complete the current construction step.

[0100] As a further improvement of the present application, each process stage label includes a plurality of construction steps, and each robot is preconfigured with a construction step set including at least one construction step;

[0101] The step S3 further includes a step S4:

[0102] The robot has a display module, and the display module displays the current construction step of the robot;

[0103] When there are at least two robots performing the same construction step, each robot scans the environment through the camera module and corrects the posture to obtain an environment scan map, and calculates the similarity between the environment scan map and the step scene map corresponding to the last construction step in the three-dimensional construction scene set;

[0104] At this time, it should be pointed out that the step scene diagram refers to the effect diagram after the completion of the construction step, so at the beginning of the construction, similarity calculation is needed with the step scene diagram of the previous construction step, and then it is judged whether the current robot is in the correct construction area. The environment scanning is carried out by the RGB-D depth camera of the robot. The depth camera can provide depth information after scanning, which is convenient for generating point cloud data later, and the robot will mark the coordinates in the global coordinate system during scanning, so the data after environment scanning can be directly compared with the three-dimensional construction scene set. Of course, the density of the point cloud data after environment scanning and the point cloud data in the three-dimensional construction scene set is similar.

[0105] The attitude correction is carried out by the first IMU inertial measurement system on the robot base and the second IMU inertial measurement system arranged on the camera module. Since the attitude correction is a common technical means, as described in CN110321754B, it will not be described again.

[0106] If the similarity is greater than the construction scene threshold, the robot is selected, and all selected robots are integrated to form a pre-selected election set;

[0107] If the similarity is less than the construction scene threshold, each robot moves around the circumference with itself as the origin and a preset election scene value as the radius, stops multiple times during the process of surrounding to perform environment scanning and attitude correction, to obtain a three-dimensional nearby environment scanning diagram. The three-dimensional nearby environment scanning diagram is compared with the step scene diagram of the previous construction step to select the robot with the highest similarity value as the master robot, and the master robot performs real-time environment scanning to construct a three-dimensional nearby environment scanning diagram.

[0108] If the similarity is less than the construction scene threshold, it means that all robots in the current construction scene are blocked by obstacles to block the construction target, referring to Figure 2 The positions of robot b and robot c in the robot b and robot c are blocked at this time. Common construction steps such as ground laying may have a large number of building components (tiles) or columns blocking the field of view. Therefore, by moving around the circumference, the degree of obstruction of the environment around the current robot can be detected, and the robot with the smallest degree of obstruction is selected as the master robot. If the master robot switches to the charging state due to the low battery level below the limit battery level threshold, the second robot with the highest similarity value is directly selected as the master robot.

[0109] Specifically, since there are errors in similarity calculation and environment scanning, the construction scene threshold can be set to 92%.

[0110] The method has the advantages that different types of robots perform different construction tasks in the same construction step, such as a carrying robot, a tiling robot, and a caulking robot, and the cloud sends the construction coordinates in a random selection manner, so that each robot is blocked by obstacles to different degrees, and the method can select the robot with the lowest influence degree, thereby facilitating the calculation of the progress of the subsequent construction step and the progress of the process stage.

[0111] If there is a pre-election set, the collaborative election score of each robot in the set is calculated , the robot with the highest collaborative election score is selected as the master robot, and the coordinates of the robot when performing environment scanning are recorded as the monitoring coordinates;

[0112] ,

[0113] ,

[0114] Among them, is the percentage of the power of the i-th robot, is the power weight coefficient, is the sum of the distances between the i-th robot and the j robots in the same construction task, is the transmission average distance weight coefficient, is the construction efficiency of the i-th robot, is the construction efficiency weight coefficient, is the average construction amount of the i-th robot, is the average construction amount of the same type of robot as the i-th robot, is the construction efficiency decay of extreme weather, is the weight coefficient of , is the weight coefficient of , is the average construction amount of the e-th extreme weather type;

[0115] Since the robots in the pre-election set can almost observe the entire construction object without obstacles, the collaborative election score is mainly determined by the power, the transmission distance with other robots, and the construction efficiency of the master robot, especially the master robot that tries to complete the entire construction step, The unit of distance is meters. The extreme weather types include various levels of high temperature warning, various levels of heavy rain warning, various levels of strong wind warning, etc. The construction efficiency of the robot under each extreme weather is calculated. If there is no construction amount value of the extreme weather, the is considered equivalent to to make the difference between them 0. , if there is no data of N extreme weather, then . Wherein, since the operation steps of each robot can be simply regarded as four steps: grabbing, carrying, placing, and fixing, the calculation of construction amount can be confirmed by how many times the robot carries the building component.

[0116] If the current construction step needs at least two robots to jointly execute, the master robot sends an action synchronization instruction to the jointly executed robot. For example, when a part of the large building component needs two carrying robots to synchronously execute the carrying task, in order to avoid the two robots walking out of sync, the master robot can send an action instruction synchronously to make them walk synchronously, so as to avoid the building component from falling off due to walking out of sync, and to avoid damaging the building component, the robot, and even threatening the safety of nearby construction workers.

[0117] As a further improvement of the present application, the specific steps of updating the process stage label through the process stage model include:

[0118] According to the environmental scan map or the three-dimensional adjacent environmental scan map, and the three-dimensional construction scene corresponding to the step scene map of the construction step, the progress of the construction step is analyzed, and when the progress of the construction step is greater than the preset step threshold, the master robot broadcasts to the robot of the next construction step;

[0119] At this time, it needs to be noted that since the robot will go to the construction coordinate when entering a new process stage, but as a specific embodiment of the present application, after the robot goes to the construction coordinate and the master robot is elected for the current construction step, the master robot will establish a signal connection with all the robots of the process stage, and will send standby coordinates and standby instructions to the robots not in the current construction step, thereby increasing the activity range of the construction step and saving the power of the subsequent construction step. Specifically, the step threshold is 90%, which can be modified for different construction steps. For example, the next construction step of window installation is door frame installation, and the degree of mutual influence of the two construction steps is low, so the robot of the next construction step can be awakened to enter the construction coordinate at a lower step threshold. If the construction step is like tiling, if the robot of the next construction step is awakened in advance, the ground that has not been tiled will be easily stepped on by the robot.

[0120] ​If the current construction step is the last construction step of the process stage, the progress of the construction step is compared with the step threshold value, and when the progress of the construction step is greater than the preset stage threshold value, the host robot sends a process stage label marked as a preparation state to the cloud, and the cloud selects a robot marked as an idle state according to the process stage label marked as the preparation state and the robot data, and switches the idle state of the robot to the preparation state;

[0121] When the power of the robot in the preparation state is lower than the preset preparation power threshold value, the preparation state is maintained while being marked as a charging state to perform a charging task;

[0122] Since the robot is selected in advance and charged at this step, the stage threshold value can be determined according to the charging time and the construction time, that is, the robot in the preparation state can start to construct in the next process stage after being fully charged.

[0123] When the progress of the last construction step of the process stage is greater than the preset stage completion threshold value, the host robot sends an environment scan map or a three-dimensional adjacent environment scan map and a next process stage label to the cloud, and the cloud determines whether the progress of the step is greater than the stage completion threshold value according to the environment scan map or the three-dimensional adjacent environment scan map and the step scene graph corresponding to the construction step in the three-dimensional construction scene, and updates the process stage label if so.

[0124] This step is used for secondary determination to avoid the host robot from making a wrong judgment or someone from attacking the cloud and sending an incorrect process stage label instruction to the cloud, which leads to early updating to the next stage label and further leads to the disorder of the entire process.

[0125] According to the environment scan map or the three-dimensional adjacent environment scan map and the step scene graph corresponding to the construction step in the three-dimensional construction scene, the specific steps of the progress of the construction step are analyzed, including:

[0126] The host robot goes to a supervision coordinate to perform environment scanning and obtains an environment scan map through pose correction according to a preset supervision interval, or the host robot constructs a three-dimensional adjacent environment scan map in real time, and the environment scan map and the three-dimensional adjacent environment scan map are both point cloud data, the step scene graph corresponding to the construction step has construction components in the form of point cloud data, and the progress of the construction step .

[0127] ,

[0128] Among them, is the nth three-dimensional coordinate of the jth construction component of the ith process stage, is the error range of the mth three-dimensional coordinate in the environmental scan or the three-dimensional adjacent environmental scan, in millimeters, is the number of three-dimensional coordinates of the construction component that fall within the error range of the three-dimensional coordinates in the environmental scan or the three-dimensional adjacent environmental scan, is the total amount of point cloud data of the construction component in the jth construction step in the ith process stage.

[0129] Since the monitoring coordinate is actually a random construction coordinate sent by the cloud to the robot, and the robot starts to perform the preset construction task based on this construction coordinate, the robot will not be too far away from the monitoring coordinate when performing the construction task. Therefore, scanning the monitoring coordinate at regular intervals will not cause the robot to detour too much. At this time, it needs to be reminded that, with reference to Figure 2 , the cloud randomly sends construction coordinates to robots not only to avoid robots from gathering and blocking after reaching the construction site, but also to enable robots to work together around the entire construction object, rather than all robots gathering in one place to work together, which can easily cause construction conflicts. For example Figure 2 , robots a and b in the construction task perform different paths to the building component and do not affect each other, and can be located on opposite sides of the construction object when performing the construction task.

[0130] Since the environmental scan and the three-dimensional adjacent environmental scan are based on scanning results, errors can easily occur. If error ranges are not set for direct comparison, comparison failure can easily occur. For example, in the tile laying construction step, the tile as a building component is compared to determine whether the height of the ground has changed, and the change range is approximately the thickness of the tile. In the door frame installation construction step, the door frame as a building component is compared to determine whether the door area has an object approximately the size of the door frame, thereby determining the progress of the construction step. The error value , , can be defined according to different construction steps. For example, the main error of the tile is the thickness, so can be adjusted accordingly, but this value cannot be approximately the thickness of the tile. For example, the door frame, due to its special shape, has a low probability of misjudgment, so the error value can be set to a large degree.

[0131] The preset monitoring interval can be determined according to the time required for the construction step, or it can be uniformly set to a fixed value, such as ten minutes or half an hour.

[0132] The method for calculating how many three-dimensional coordinates of the construction component fall into the error range of the three-dimensional coordinates of the environmental scanning map or the three-dimensional nearby environmental scanning map can adopt a combination elimination method, that is, two groups of point cloud data are combined, if falling into the range, the point cloud data of the two groups are reserved, if not falling into the range, the point cloud data of the two groups are deleted at the same time, and the number of the remaining point cloud data is calculated.

[0133] As a further improvement of the present application, the specific steps of selecting the robot according to the process phase label and the robot data by the cloud server include:

[0134] If the construction steps of the robot exist the same construction steps as those in the process phase label, and the robot is not in the charging state, the robot is selected, and a plurality of preselected construction sets are integrated according to each construction step in the process phase label;

[0135] If the robots of a certain type are all in the charging state, the selection is performed again after a period of time to avoid the robot just entering the charging state being selected, and then entering the construction state again when not having enough power, and going to charge again after entering the construction state and being below the limit power threshold, which seriously affects the construction efficiency.

[0136] Each construction step is preset with a standard construction machine quantity, if the robot in the preparation state exists in the construction step, the robot is selected, and the preparation state is switched to the construction state, if the robot is also marked as the charging state, the construction state is switched after the charging task is completed, the number of robots required for each construction step ;

[0137] ,

[0138] wherein, is the number of robots required per thousand square, is the construction size of the step, is the number of robots in the preparation state, and j is the number of construction steps;

[0139] If the number of robots required for the construction step is greater than the number of robots of the corresponding preselected construction set, all the robots in the preselected construction set are marked as the construction state;

[0140] At this time, it is necessary to remind that if there is at least one type of robot entering the preselected construction set, the construction process is below the limit power threshold to charge, at this time the same type of robot of the construction step is 0. Since the state label is synchronized with the cloud, the cloud will reselect the robot after receiving the information, and the robot that has just performed the charging task may be idle, which can directly make it perform the construction task. In addition, if the robot of this type is still in the charging state, it will be selected again after a period of time. The purpose is to ensure efficient charging and avoid multiple plug and unplug, which affects the overall construction efficiency.

[0141] If the number of robots required for the construction step is less than the number of robots in the corresponding preselected construction set, calculate the selected construction score of each robot in the preselected construction set ;

[0142] ,

[0143] wherein, is the standard moving length, is the path length of the ith robot to the construction area, is the power percentage of the ith robot, is the total number of construction steps in the process stage, is the number of the ith robot selected by the preselected construction set, , and is the selected construction score weight factor;

[0144] According to the selected construction score , the robots are selected in descending order and marked as construction state.

[0145] By calculating the selected construction score , the robots with short distance, high power and less number of selections by the preselected construction set can be selected first. For the number of selections by the preselected construction set, the principle is that if there are multiple construction steps in the process stage that can be constructed by a type of robot, and some of the steps can be constructed by b type of robot, if these steps are all constructed by a type of robot, the power of a type of robot is obviously insufficient to complete so many construction steps, therefore, for each robot in the same preselected construction set, the robot with less construction steps should be selected first. By log processing and combining the total number of construction steps for calculation, the result of this item can be more stable and reliable.

[0146] The path length of the robot to the construction area is planned according to the current position of the robot, the construction coordinate position, and the real-time three-dimensional scene graph, specifically, a Monte Carlo path planning is adopted, if the path exists obstacles in the real-time three-dimensional scene graph, the selection probability of the path is reduced, the path with the maximum probability is selected, and a path planning method of CN116519005B can be referred to.

[0147] As a further improvement of the application, the specific steps of the robot going to the construction coordinate for the construction task include: the three-dimensional construction scene set includes step scene graphs of each construction step of each process stage and a real-time three-dimensional scene graph, and the step scene graph has a construction area;

[0148] A plurality of coordinates in the step scene graph are randomly selected as construction coordinates, and the construction coordinates are sent to the robot one by one, the robot performs Monte Carlo path planning according to the real-time three-dimensional scene graph and the construction coordinates, and the robot performs Monte Carlo path planning by real-time scanning the environment during going to the construction coordinate to avoid obstacles;

[0149] Randomly selecting the construction coordinate is not only to avoid the robot from gathering and blocking after reaching the construction site, but also to enable the robots for construction to jointly construct around the entire construction object, rather than all the robots gathering in one place for joint construction, which is easy to cause construction conflicts.

[0150] If there are at least two robots in the same construction step, and all the robots establish signal connection with each other when reaching the construction coordinate, each robot performs environment scanning to identify whether there is a construction component of the construction step in the environment scanning graph, if yes, the construction component area is set as a material area and is synchronized to other robots under the same construction step, if not, all the robots randomly move and re-perform environment scanning, and when there is a material area and a construction area, each robot performs the construction task according to a preset program.

[0151] Establishing signal connection after reaching the construction coordinate can reduce the burden of communication establishment, specifically, after establishing signal connection, the robots transmit their own data such as power, time, state, position, etc. to each other. Since the robots reach the site through random construction coordinates, some robots may not be able to identify the position of the material, by scanning and communicating with each other, the position of the construction component can be quickly found, thereby helping the robots to smoothly perform the construction task.

[0152] As a further improvement of the application, when the power of the robot marked as the construction state is lower than the limit power threshold, the construction state is switched to the charging state and is synchronized to the master robot;

[0153] The master robot calculates the power required for the remaining task ;

[0154] ,

[0155] wherein, is the average power of each robot of the same type as the robot switched to the charging state at the beginning of the construction step, is the average power of each robot of the same type as the robot switched to the charging state when the robot exited the construction task, is the number of robots of the same type as the robot switched to the charging state at the beginning of the construction step;

[0156] calculating the total available power of the robots currently in the construction state and of the same type as the robot switched to the charging state ;

[0157] ,

[0158] wherein, is the current power of the tth robot of the same type, is the threshold value of the limit power, is the power surplus value, if the total available power is less than the required power of the remaining task , the robot in the charging state and not in the standby state of the same type is selected and sent to the construction coordinates, and after the robot completes the charging task, the robot goes to the construction coordinates to perform the construction task. Thus, the entire construction step can be completed, and the construction task can be avoided.

[0159] The application also provides a building robot intelligent selection and scheduling system based on multi-machine linkage, which uses the building robot intelligent selection and scheduling method based on multi-machine linkage as described above, and comprises the following modules.

[0160] A plurality of robots, which are provided with a construction program, are used for scanning the environment and performing the construction task;

[0161] A cloud end, which is signal connected with the plurality of robots, is used for receiving the building drawing set, defining the total coordinate system, issuing the construction coordinates and the construction task to the robots;

[0162] A three-dimensional construction scene construction module, which is arranged in the cloud end, is used for constructing the three-dimensional construction scene set according to the building drawing set or updating the three-dimensional construction scene set according to the fixed obstacles recognized by the robots;

[0163] A process stage label updating module, which is used for updating the process stage label by manual updating or by the process stage model, is used for the cloud end to select the robots for construction according to the updated process stage label;

[0164] A path planning module is arranged in the robot, and a path to the construction coordinate is planned according to a total coordinate system of the robot and the construction coordinate;

[0165] A task matching module is arranged in the cloud, and a robot performing a construction task is selected according to a process stage label and robot data;

[0166] A state monitoring module is used for detecting a running state of the robot and capable of changing a state label of the robot;

[0167] A process stage feedback module is used for feeding back a construction result of the robot to the cloud and triggering distribution of a next process stage task.

[0168] Obviously, the above embodiments of the present application are merely exemplary and are not intended to limit the embodiments of the present application. Based on the above description, other different forms of changes or variations can be made by those skilled in the art. Here, all the embodiments are not required to be exhausted. Any modification, equivalent replacement and improvement within the spirit and principle of the present application shall be included in the protection scope of the claims of the present application.

Claims

1. A multi-machine linkage-based building robot intelligent matching and scheduling method, characterized in that, The application relates to a system comprising a plurality of robots with camera modules and a cloud connected to the plurality of robots. The application comprises the following steps: S1, inputting a set of construction drawings into the cloud, the cloud constructing a set of three-dimensional construction scenes and defining a total coordinate system according to the set of construction drawings, and updating the set of three-dimensional construction scenes and the total coordinate system into all robots at the construction site; S2, the robots scanning the environment to identify fixed obstacles in the space and sending the fixed obstacles to the cloud, the cloud updating and integrating the fixed obstacles into the set of three-dimensional construction scenes, and updating the three-dimensional construction scenes to other robots; S3, manually updating a process stage label or the robots updating the process stage label through a process stage model, and the cloud selecting robots to go to construction coordinates to perform construction tasks according to the process stage label and robot data; Each process stage label comprises a plurality of construction steps, and each robot is preconfigured with a construction step set comprising at least one construction step; The step S3 further comprises a step S4: The robot has a display module, and the display module displays the current construction step of the robot; When there are at least two robots performing the same construction step, each robot scans the environment through a camera module and corrects the posture to obtain an environment scanning map, and similarity calculation is performed between the environment scanning map and a step scene map corresponding to the previous construction step in the set of three-dimensional construction scenes; If the similarity is greater than a construction scene threshold, the robot is selected, and all selected robots form a preselected election set; If the similarity is less than the construction scene threshold, each robot moves around a circle with itself as the origin and a preset election scene value as the radius, stops multiple times during the movement to scan the environment and correct the posture, obtains a three-dimensional adjacent environment scanning map, and performs similarity calculation between the three-dimensional adjacent environment scanning map and the step scene map corresponding to the previous construction step, selects the robot with the highest similarity value as the master robot, and the master robot performs real-time environment scanning to construct a three-dimensional adjacent environment scanning map; If there is a pre-selected election set, the collaborative election score of each robot in the set is calculated respectively The robot with the highest collaborative election score is selected as the host robot, and the coordinates of the robot when performing environmental scanning are recorded as the supervision coordinates ， ， in, It represents the battery percentage of the i-th robot. It is the power weighting coefficient. It is the sum of the distances between the i-th robot and the j-th robots within the same construction task. It is the average transmission distance weighting coefficient. It is the construction efficiency of the i-th robot. It is the construction efficiency weighting coefficient. It is the average construction volume of the i-th robot. It is the average construction volume of robots of the same type as the i-th robot. The construction efficiency is reduced due to extreme weather conditions. yes The weighting coefficients, yes The weighting coefficients, This is the average construction volume for the e-th extreme weather type; If the current construction step needs to be performed by at least two robots, the master robot sends a motion synchronization instruction to the robots.

2. The method of claim 1, wherein, Each robot has a state label synchronized with the cloud, and the state label comprises a construction state, a charging state and an idle state; If the robot has a construction task, the robot is marked as the construction state, and if there is no construction task and charging task, the robot is marked as the idle state; When the power of the robot marked as the idle state is lower than a preset charging threshold, the robot performs a charging task, and the idle state is switched to the charging state; The robot is further preconfigured with a limit power threshold lower than the charging threshold, and when the power of the robot marked as the construction state is lower than the limit power threshold, the robot performs a charging task, and the construction state is marked as the charging state. 3.The method of claim 2, wherein, The specific steps of updating the process stage label through the process stage model comprise: According to the environment scanning map or the three-dimensional adjacent environment scanning map and the step scene map corresponding to the construction step in the set of three-dimensional construction scenes, the progress of the construction step is analyzed, and when the progress of the construction step is greater than a preset step threshold, the master robot broadcasts to the robots of the next construction step; If the current construction step is the last construction step of the process stage, the progress of the construction step is compared with the step threshold value, and when the progress of the construction step is greater than the preset stage threshold value, the host robot sends a process stage label marked as a preparation state to the cloud, and the cloud selects a robot marked as an idle state according to the process stage label marked as the preparation state and the robot data, and switches the idle state of the robot to the preparation state; When the power of the robot in the preparation state is lower than the preset preparation power threshold value, the preparation state is maintained while being marked as a charging state to perform a charging task; When the progress of the last construction step of the process stage is greater than the preset stage completion threshold value, the host robot sends an environment scan map or a three-dimensional adjacent environment scan map and a next process stage label to the cloud, and the cloud judges whether the progress of the step is greater than the stage completion threshold value according to the environment scan map or the three-dimensional adjacent environment scan map and the step scene graph corresponding to the construction step in the three-dimensional construction scene, and if so, updates the process stage label; According to the environment scan map or the three-dimensional adjacent environment scan map and the step scene graph corresponding to the construction step in the three-dimensional construction scene set, the specific steps of the progress of the construction step are analyzed and obtained, and the specific steps of the progress of the construction step include: The host robot goes to the supervision coordinate according to the preset supervision interval to perform environment scanning and obtains an environment scanning map through pose correction, or the host robot performs real-time environment scanning to construct a three-dimensional nearby environment scanning map. Both the environment scanning map and the three-dimensional nearby environment scanning map are point cloud data. The step scene map corresponding to the construction step has a construction component in the form of point cloud data, and the progress of the construction step ; ， wherein, is the n-th three-dimensional coordinate of the construction component of the j-th construction step in the i-th process stage, is the error range of the m-th three-dimensional coordinate in the environmental scan or the three-dimensional neighboring environmental scan, in millimeters, is the number of three-dimensional coordinates of the construction component falling within the error range of the three-dimensional coordinates of the environmental scan or the three-dimensional neighboring environmental scan, is the total amount of point cloud data of the construction component of the j-th construction step in the i-th process stage.

4. The intelligent scheduling method of claim 3, wherein, The specific steps of selecting the robot by the cloud according to the process stage label and the robot data include: If there is a construction step in the construction step set of the robot that is the same as the construction step in the process stage label, and the robot is not in the charging state, the robot is selected, and a plurality of preselected construction sets are integrated according to each construction step in the process stage label; Each construction step is preset with a standard number of construction machines, if a robot in a ready state exists for this construction step, this robot is selected and the ready state is switched to a construction state, if this robot is also marked as being in a charging state, the switch to the construction state is made when the charging task is completed, the number of robots required for each construction step ; ， wherein, is the number of robots per thousand square meters required, is the construction size of the step, is the number of robots in the ready state, j is the number of construction steps; If the number of robots required for the construction step is greater than the number of robots corresponding to the preselected construction set, all the robots under the preselected construction set are marked as a construction state; If the number of robots required for the construction steps is less than the number of robots of the corresponding preselected construction set, calculate the selected construction score of each robot under the preselected construction set ; ， wherein, is the standard movement length, is the path length of the ith robot to the construction area, is the percentage of power of the ith robot, is the total number of construction steps at this process stage, is the number of the ith robot selected by the preselected construction set, , and is the fitting score weight factor; According to the fit-up construction score The robots are selected in descending order and marked as under construction.

5. The method of claim 1, wherein, The specific steps of the robot going to the construction coordinate to perform the construction task include that the three-dimensional construction scene set includes the step scene graph of each construction step of each process stage and the real-time three-dimensional scene graph, and the step scene graph has a construction area; A plurality of coordinates in the step scene graph are randomly selected as construction coordinates, and the construction coordinates are sent to the robot one by one, the robot performs Monte Carlo path planning according to the real-time three-dimensional scene graph and the construction coordinates, and the robot performs Monte Carlo path planning in real time during the process of going to the construction coordinate to avoid obstacles; If there are at least two robots in the same construction step, and all the robots establish signal connection with each other when they all arrive at the construction coordinate, each robot performs environment scanning, and identifies whether there is a construction component of the construction step in the environment scan map, if so, the construction component area is set as a material area and is synchronized to other robots under the same construction step, if not, all the robots randomly move and perform environment scanning again, and when there is a material area and a construction area, each robot performs the construction task according to the preset program.

6. The method of claim 1, wherein, The steps of constructing the three-dimensional construction scene set and defining the total coordinate system by the cloud according to the building drawing set specifically include: The names of the drawings in the set of architectural drawings are identified and classified by NLP natural language processing technology, and the total plot drawing, the single building design drawing set and the building construction drawing set are obtained. The drawings in the single building design drawing set and the building construction drawing set are arranged according to the drawing number and imported into the BIM software, and a three-dimensional construction scene set of point cloud data is generated. The southwest corner of the total plot drawing is selected as the origin to generate the total coordinate system; The three-dimensional construction scene set includes step scene graphs of each process stage and each construction step, and real-time three-dimensional scene graphs.

7. The method of claim 1, wherein, The robot scans the environment to identify fixed obstacles in the space and sends them to the cloud, which updates and integrates them into the three-dimensional construction scene set, and updates the three-dimensional construction scene to other robots. The specific steps include: The robot is pre-set with an obstacle detection interval. When the robot scans the environment and identifies that there is an obstacle in a certain space, it continuously scans the environment. If the obstacle moves within the obstacle detection interval, it is considered a dynamic obstacle, and continuous scanning of the environment is stopped. If the obstacle remains stationary beyond the obstacle detection interval, it is considered a fixed obstacle. The robot takes multiple views of the fixed obstacle and sends them to the cloud. The cloud generates three-dimensional reconstruction point cloud data through COLMAP and integrates and updates it into the three-dimensional construction scene set. 8.The method of claim 4, wherein, When the power of the robot marked as the construction state is lower than the limit power threshold, the construction state is switched to the charging state and is synchronized to the master robot; Host robot calculates power required for remaining tasks ; ， wherein is the average power of the robots of the same type as the robot which switches to the charging state at the beginning of the construction step, is the average power of the robots of the same type as the robot which switches to the charging state at the end of the construction step, is the number of robots of the same type as the robot which switches to the charging state at the beginning of the construction step; total available power of the robots of the same type as the robot that is switched to the charging state ; ， wherein, is the current power of the tth robot of the same type, is the limit power threshold, is the power surplus, if the total available power is less than the power required for the remaining tasks , a robot in the charging state and not in the preparation state of the same type is selected and sent to the construction coordinates, and after the robot completes the charging task, it goes to the construction coordinates to perform the construction task.

9. A multi-machine linkage-based intelligent matching and scheduling system for construction robots, which applies the multi-machine linkage-based intelligent matching and scheduling method for construction robots according to any one of claims 1-8. It includes the following modules: A plurality of robots, pre-set with a construction program, for scanning the environment and performing construction tasks; Cloud, signal connected with multiple robots, for receiving a set of architectural drawings, defining a total coordinate system, issuing construction coordinates and construction tasks to robots; Three-dimensional construction scene construction module, located in the cloud, for constructing a three-dimensional construction scene set according to the set of architectural drawings or updating the three-dimensional construction scene set according to the fixed obstacles identified by the robot; Process stage label update module, for updating process stage labels by manual update or by process stage model, for the cloud to select robots for construction according to the updated process stage labels; Path planning module, located in the robot, for planning a path to the construction coordinate according to the total coordinate system and the construction coordinate of the robot; Task selection module, located in the cloud, for selecting robots to perform construction tasks according to process stage labels and robot data; State monitoring module, for detecting the running state of the robot and being able to change the state label of the robot; Process stage feedback module, for feeding back the construction results of the robot to the cloud and triggering the distribution of the next process stage task.

Images