Multi-robot high-throughput information transmission and real-time collaborative operation system and method

Abstract

The invention discloses a multi-robot high-throughput information transmission and real-time collaborative operation system and method. The method comprises the following steps: S1, establishing 5Gamp; information interaction between the robot and a 5G base station and information interaction between multiple robots are achieved through a WiFi networking architecture; s2, establishing a regional three-dimensional environment map based on data collected by binocular cameras and laser radars in the robots, and further obtaining regional situation awareness of the robots; s3, based on the area situation awareness obtained by each robot, combining with the position of each robot, carrying out multi-machine task allocation and multi-machine path planning in sequence, and obtaining an operation task allocated by each robot and a running path of the operation task; and S4, according to control signals generated by the operation tasks and the operation paths of the robots, the robots are controlled to complete the service tasks, and collaborative operation is achieved. According to the invention, aiming at a field complex ground environment, execution of a defense material transportation task is taken as an application scene, an operation system is formed by more than two robots, and reliable communication and real-time collaborative operation of a multi-node mobile ad hoc network are realized.

Description

Technical Field

[0001] The present invention belongs to the technical field of robot collaborative operation, and specifically relates to a multi-robot high-throughput information transmission and real-time collaborative operation system and method. Background Art

[0002] In recent years, the development of multi-sensor detection and information fusion technology has been paid more and more attention by countries. In 2001, Mahajan designed a model of intelligent multi-sensor synthesis using fuzzy logic methods, which fuses sensor data from different sensors and different solutions. In 2015, EthzASL Laboratory proposed a variety of fusion algorithms to achieve the mutual fusion of monocular vision and IMU inertial information and real-time estimation of posture information. The research on information fusion technology in China started relatively late. The main research work includes: THMR series mobile robot system developed by the State Key Laboratory of Intelligent Technology and System, Department of Computer Science, Tsinghua University, automatic navigation vehicle and service robot of Intelligent Design Automation and Manufacturing Center, City University of Hong Kong, AGV and explosion-proof robot of Shenyang Institute of Automation, Chinese Academy of Sciences, etc. The basis of robot intelligence is that various sensors can work in coordination. How to remove redundant information and leave useful information has become one of the important topics in the development of intelligent robots.

[0003] In terms of robot cluster communication, Inner Mongolia University of Technology and State Grid Shanghai Electric Power Company introduced ZigBee wireless communication modules in the robot master control system, realizing wireless multi-hop self-organizing network communication and positioning during the inspection process of cable tunnel inspection robots, solving the problems of insufficient wireless communication distance and difficult positioning of inspection robots in cable tunnels. However, the communication distance is too short, which is limited in the promotion of outdoor environments. In response to the development needs of unmanned ground operations in future battlefields, China Electronics Technology Group 28 and Northwestern Polytechnical University proposed a two-layer network system architecture of global network-task network collaboration, using the global network to ensure network connectivity, and building independent task networks for different tasks in different channels to provide multi-target and multi-task processing capabilities. Experiments have shown that the proposed robot network system can meet the task requirements of large-scale battlefield robot groups working in real-time multi-task parallel work, but the communication scheme has the problems of low speed and large delay, and cannot play an important role in today's high-bandwidth and real-time tasks. Shenyang Ligong University proposed an energy-efficient routing protocol EERP (Energy Efficiency Routing Protocol) based on the limited energy characteristics of distributed micro-robot networks, and used the NS2 simulator to simulate the network protocol, providing a reference for the selection and development of protocols during project implementation. This achievement mainly studies more efficient routing protocols, and does not build solutions with higher communication efficiency in PHY and MAC.

[0004] Robot collaborative work refers to meeting task requirements and environmental constraints, breaking down large and complex tasks into multiple subtasks, and assigning the subtasks to different robots to perform, thereby shortening the task time. Therefore, task decomposition and task assignment directly determine the performance of multi-robot systems in completing tasks. The swarm-bots project sponsored by the European Commission is dedicated to studying new methods to design and implement self-organizing and self-building robot systems, which can complete tasks such as cooperative transportation, path planning, formation control, and obstacle avoidance. In addition, there are the iRobot swarm project of the Massachusetts Institute of Technology, the Cooperative Robotics system of the Oak Ridge National Laboratory in the United States, and the CEBOT system of Nagoya University in Japan. In China, there are the Project SI project of the Embedded Laboratory of Shanghai Jiaotong University, the Sambot of the Beijing University of Aeronautics and Astronautics, and the MRCAS of the Shenyang Institute of Automation of the Chinese Academy of Sciences.

[0005] In the existing multi-robot collaborative operation system, different collaborative operation technologies are used for different operation purposes. One is to use WIFI local area network for real-time transmission of image data, video data and navigation data, and the map constructed by the robot is a two-dimensional map, which cannot be applied to complex environments such as mountains. Another is to build maps by designing a special main control mechanism, but a single robot does not have the ability to build maps, only simple obstacle avoidance function. Summary of the invention

[0006] In view of the above-mentioned deficiencies in the prior art, the present invention provides a multi-robot high-throughput information transmission and real-time collaborative operation system and method, which is aimed at complex outdoor ground environments and takes the execution of defense material transportation tasks as the application scenario. The operation system is composed of more than two robots to achieve reliable communication and real-time collaborative operation in a multi-node mobile ad hoc network.

[0007] In order to achieve the above-mentioned invention object, the technical solution adopted by the present invention is: a multi-robot high-throughput information transmission and real-time collaborative operation system, including a plurality of robots and an operation coordination terminal, each of the robots including an intelligent unit, and a perception unit, a communication unit and an electromechanical unit connected to the intelligent unit; The communication unit is used to realize information interaction between the robot and the 5G base station and other robots; The sensing unit is used to collect surrounding environment data and transmit it to the intelligent unit; The intelligent unit is used to process the received surrounding environment data, build regional situation awareness of the current robot, and transmit it to the operation coordination end; The operation coordination end is used to run the collaborative operation algorithm to perform operation task allocation and local path planning based on the constructed regional situation awareness, generate control signals and send them to the corresponding robots; The electromechanical unit controls the robot to complete the task according to the received control signal.

[0008] Furthermore, the communication unit is a mobile communication gateway device equipped with a 5G communication module and a WiFi6 communication module; wherein the 5G communication module is connected to a 5G base station, and the WiFi6 communication module is connected to WiFi6 communication modules carried by other robots to form a Mesh network.

[0009] Furthermore, when the communication quality of the Mesh network between the robots is lower than a preset threshold, the communication mode between the robots is switched to: using the 5G base station as a communication transfer station, and exchanging information between the robots through the 5G communication module; When the 5G communication quality is lower than the preset threshold, it switches back to the Mesh network between robots for communication.

[0010] Furthermore, the operation coordination end is deployed with a serial multi-machine task allocation model and a multi-machine path planning model; The multi-machine path planning model performs path planning according to the task allocation result of the multi-machine task allocation model.

[0011] A collaborative working method comprises the following steps: S1. Establish 5G&WiFi6 networking architecture to realize information interaction between robots and 5G base stations and multiple robots; S2, based on the data collected by the binocular camera and laser radar in each robot, establish a three-dimensional environmental map of the area, and then obtain the regional situation awareness of each robot; S3, based on the regional situation awareness obtained by each robot and combined with the position of each robot, multi-machine task allocation and multi-machine path planning are performed in sequence to obtain the operation tasks assigned to each robot and their operation paths; S4. Based on the control signals generated by the operation tasks and operation paths of each robot, each robot is controlled to complete the business tasks and realize collaborative operation.

[0012] Furthermore, in the step S1, in the 5G&WiFi6 networking architecture, the robots form a Mesh network through the communication connection between the WiFi6 communication modules, and the robots are connected to the 5G base station through the 5G communication module; When the communication quality of the Mesh network is lower than a preset threshold, the communication mode between the robots is switched to: using the 5G base station as a communication transfer station, and information exchange between the robots through the 5G communication module; When the 5G communication quality is lower than the preset threshold, it switches back to the Mesh network between robots for communication.

[0013] Furthermore, the step S2 is specifically as follows: S21. Determine the relative sensing height of each robot according to the number of robots and the safe obstacle avoidance height range; S22. For each robot, decompose the three-dimensional environment into multiple two-dimensional planes at different heights according to the data collected by its binocular camera and laser radar; S23, converting the two-dimensional scanning data of each two-dimensional plane into local three-dimensional point cloud data of a fixed height; S24, according to the relative perception height of each robot, fusing the local three-dimensional point cloud data at different heights into global three-dimensional point cloud data through coordinate transformation; S25. Run the 3D mapping algorithm LeGO-LOAM based on the global 3D point cloud data to build a regional 3D environment map.

[0014] Furthermore, in step S3, multi-machine task allocation is performed according to a multi-machine task allocation model, and the multi-machine task allocation model is expressed as:

[0015]

[0016]

[0017]

[0018]

[0019] In the formula, represents the multi-machine task allocation objective function, , and represents the weight coefficient, Represents the total number of robots, Represents the robot index, Indicates Robots The length of the path to complete the material distribution of all mission points, represents the time cost function for all robots to complete the task, Represents the energy cost function of the i-th robot completing all task points; Indicates i Robots Ri The time required to complete the material distribution of all mission points. virepresents the movement speed of the i-th robot Ri; Pi Indicates i Robots Ri The power consumed to complete the material distribution of all mission points; The constraint condition of the multi-machine task allocation model is the load constraint, which is expressed as:

[0020] In the formula, LDi Indicates the load of the robot. Lm Indicates the maximum load of the robot.

[0021] Furthermore, the improved dung beetle algorithm is used to solve the multi-machine task allocation model and obtain the multi-machine task allocation plan. The specific method is as follows: S31. Set the number of mission points that require materials and the weight of materials required for each mission point; S32, initializing the dung beetle algorithm parameters, including the maximum number of iterations, the population size, the proportion of rolling cockroaches, the proportion of foraging dung beetles, and the proportion of thief dung beetles; S33, calculating the fitness value of each dung beetle in the initialized population, defining the global optimal fitness and its position, the global worst fitness and its position, and selecting the rolling dung beetle, the hatching dung beetle, the foraging dung beetle and the thief dung beetle in proportional order; S34, setting the current number of iterations t=0; S35, using the multi-machine task allocation model as the fitness function of the dung beetle algorithm, executing the dung beetle algorithm, sorting the calculated fitness values, finding the dung beetle individual with the best global fitness and the dung beetle individual with the worst global fitness, and using the Cubic chaotic mapping method to modify parameters related to the number of iterations; S36, determining the type of each dung beetle in turn, and using t-distribution variation to update the current position of the dung beetle and calculate the next position information; S37, performing boundary determination on each dung beetle after the position information is updated, calculating its fitness value and selecting the optimal fitness value and its position; S38, determining whether the maximum number of iterations has been reached; If so, the optimal multi-machine task allocation solution is obtained; If not, return to step S35.

[0022] Furthermore, in step S3, path planning is performed for each robot according to a multi-machine path planning model. The multi-machine path planning model aims to optimize the time for each robot to complete the assigned task. The multi-machine path planning model is solved by using an improved dung beetle algorithm and an artificial potential field fusion method to plan its working path, including a task starting point, a task end point, and a robot walking path.

[0023] The beneficial effects of the present invention are: (1) The present invention uses multi-source sensing detection and information fusion technology to effectively expand the perception space, reduce the time required for three-dimensional mapping, improve the real-time performance of three-dimensional mapping, solve the perception blind spot problem and information loss problem caused by the limited field of view and coverage of a single sensor, improve the system's dynamic environment adaptability and response speed, and realize multi-node collaborative perception.

[0024] (2) The method of the present invention solves the difficulties of high latency, limited bandwidth and difficulty in real-time processing of information in multi-node large-volume data transmission through a multi-modal communication mechanism and large-volume data transmission technology based on 5G & WiFi technology, and realizes the construction of a reliable multi-node mobile self-organizing communication network.

[0025] (3) The present invention solves the problem of adapting to complex terrain in the wild environment through a robot path planning method based on the fusion of the improved dung beetle algorithm and the artificial potential field method, better performs dynamic obstacle avoidance, and realizes multi-robot task planning, task allocation and path planning in the wild environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 This is a block diagram of the collaborative operation coordination structure in the present invention.

[0027] Figure 2 It is a structural block diagram of the robot in the present invention.

[0028] Figure 3 It is a flow chart of the collaborative operation method in the present invention.

[0029] Figure 4 This is a schematic diagram of communication between robots in the present invention via WiFi6.

[0030] Figure 5 This is a schematic diagram of communication between robots in the present invention using a 5G base station as a transfer station.

[0031] Figure 6 A schematic diagram of multiple two-dimensional nodes in the present invention cooperating to realize three-dimensional environment perception.

[0032] Figure 7 It is a schematic diagram of a two-dimensional decomposition of a three-dimensional environment in the present invention. DETAILED DESCRIPTION

[0033] The specific implementation modes of the present invention are described below so that those skilled in the art can understand the present invention. However, it should be clear that the present invention is not limited to the scope of the specific implementation modes. For those of ordinary skill in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the attached claims, these changes are obvious, and all inventions and creations utilizing the concept of the present invention are protected.

[0034] The embodiment of the present invention provides a multi-robot high-throughput information transmission and real-time collaborative operation system, such as Figure 1 As shown, it includes several robots and an operation coordination terminal, each of the robots includes an intelligent unit, and a sensing unit, a communication unit and an electromechanical unit connected to the intelligent unit; The communication unit is used to realize information interaction between the robot and the 5G base station and other robots; The sensing unit is used to collect surrounding environment data and transmit it to the intelligent unit; The intelligent unit is used to process the received surrounding environment data, build regional situation awareness of the current robot, and transmit it to the operation coordination end; The operation coordination end is used to run the collaborative operation algorithm to perform operation task allocation and local path planning based on the constructed regional situation awareness, generate control signals and send them to the corresponding robots; The electromechanical unit controls the robot to complete the task according to the received control signal.

[0035] In the embodiment of the present invention, the structure of each robot is as follows: Figure 2 As shown, the perception unit includes a laser radar, a binocular camera, an inertial navigation and a GNSS, which are used to collect environmental data around the robot and transmit it to the intelligent unit through a LAN cable; the intelligent unit includes a ROS main control board and an STM32 chip, wherein the STM32 chip is connected to the motor and the servo in the electromechanical unit to control its running actions to complete the specified tasks; the 5G communication module in the communication unit is connected to the ROS main control board through a LAN cable and connected to a 5G antenna, and the WiFi6 communication module is connected to the ROS main control board and connected to a WiFi antenna.

[0036] In this embodiment, Figure 2 The communication unit is a mobile communication gateway device equipped with a 5G communication module and a WiFi6 communication module; wherein the 5G communication module is connected to a 5G base station, and the WiFi6 communication module is connected to the WiFi6 communication modules carried by other robots to form a Mesh network. Among them, the robots form a point-to-point Mesh network through the WiFi6 module for short-range high-passband communication, and the 5G communication module is used for long-range wide-area communication.

[0037] In this embodiment, when the communication quality of the Mesh network between robots is lower than a preset threshold, the communication mode between robots is switched to: using the 5G base station as a communication transfer station, and exchanging information between robots through the 5G communication module; when the 5G communication quality is lower than the preset threshold, it is switched back to the Mesh network between robots for communication. For example, when the Mesh network formed by the interconnection of robots through WiFi6 is affected by the mountainous environment and the communication quality is seriously disturbed, the car can switch the communication mode.

[0038] In this embodiment, in the intelligent unit, based on the data collected by the binocular cameras and lidars of other robots received by a single robot and the communication unit, a multi-robot collaborative SLAM algorithm is run, and combined with GNSS and inertial navigation, a regional three-dimensional environmental map of the robot coverage area is established, and then a regional situation awareness map of multiple robots is obtained, providing perception information for the commanders at the on-site command and the back-end coordination operation end.

[0039] In the embodiment of the present invention, the operation coordination end is deployed with a serial multi-machine task allocation model and a multi-machine path planning model; the multi-machine path planning model performs path planning according to the task allocation result of the multi-machine task allocation model. Furthermore, the command robot sends the above task allocation scheme and path planning structure to each robot through the communication unit, controls each robot to reach the specified position, and completes the operation task.

[0040] In the embodiment of the present invention, based on the above multi-robot high-throughput information transmission and real-time collaborative operation system, a corresponding collaborative operation method is provided, such as Figure 3 As shown, the following steps are included: S1. Establish 5G&WiFi6 networking architecture to realize information interaction between robots and 5G base stations and multiple robots; S2, based on the data collected by the binocular camera and laser radar in each robot, establish a three-dimensional environmental map of the area, and then obtain the regional situation awareness of each robot; S3, based on the regional situation awareness obtained by each robot and combined with the position of each robot, multi-machine task allocation and multi-machine path planning are performed in sequence to obtain the operation tasks assigned to each robot and their operation paths; S4. Based on the control signals generated by the operation tasks and operation paths of each robot, each robot is controlled to complete the business tasks and realize collaborative operation.

[0041] In step S1 of the embodiment of the present invention, in the 5G&WiFi6 networking architecture, the robots are connected to form a Mesh network through the communication between the WiFi6 communication modules, and the robots are connected to the 5G base station through the 5G communication module; When the communication quality of the Mesh network is lower than a preset threshold, the communication mode between the robots is switched to: using the 5G base station as a communication transfer station, and exchanging information between the robots through the 5G communication module; when the 5G communication quality is lower than the preset threshold, it is switched back to the Mesh network between the robots for communication.

[0042] For example, when the Mesh network formed by the WiFi6 interconnection between robots is affected by the mountain environment and the communication quality is seriously disturbed, the car can switch the communication mode. Figure 4 As shown in the figure, the schematic diagram of robots communicating with each other using 5G base stations as transfer stations is as follows Figure 5 shown.

[0043] In this embodiment, the use of 5G and WiFi6 hybrid networking can shield the robot from underlying communication differences, provide a unified communication interface, and continuously detect the communication quality of different links in the communication network. When a link fails, the robot can switch the communication link without perception. The robot uses the Overlay logical network for business communications, regardless of the physical communication pipeline, and the underlying 5G or WiFi network is used for actual connection. According to communication requirements, different services select different links to send data to achieve the best communication effect. When the optimal communication link deteriorates, it can dynamically sense the occurrence of degradation and select the suboptimal link for communication to ensure that the communication meets business requirements as much as possible.

[0044] In the embodiment of the present invention, Figure 6 As shown, step S2 is specifically as follows: S21. Determine the relative sensing height of each robot according to the number of robots and the safe obstacle avoidance height range; For example, the relative sensing height H of each robot is determined according to the number of robots N and the safe obstacle avoidance height range ∆H i for: H i = i * ( ∆H / N ), i= 1, 2, ..., N ; S22. For each robot, according to the data collected by its binocular camera and laser radar, decompose the three-dimensional environment into multiple two-dimensional planes at different heights; illustratively, Figure 7 It shows the decomposition of a three-dimensional environment into a two-dimensional plane; S23, converting the two-dimensional scanning data of each two-dimensional plane into local three-dimensional point cloud data of a fixed height; S24, according to the relative perception height of each robot, fusing the local three-dimensional point cloud data at different heights into global three-dimensional point cloud data through coordinate transformation; S25. Run the 3D mapping algorithm LeGO-LOAM based on the global 3D point cloud data to build a regional 3D environment map.

[0045] In step S3 of the embodiment of the present invention, multi-machine task allocation is performed according to a multi-machine task allocation model. The multi-machine task allocation model in this embodiment is expressed as:

[0046]

[0047]

[0048]

[0049]

[0050] In the formula, represents the multi-machine task allocation objective function, , and represents the weight coefficient, Represents the total number of robots, Represents the robot index, Indicates Robots The length of the path to complete the material distribution of all mission points, represents the time cost function for all robots to complete the task, Represents the energy cost function of the i-th robot completing all task points; Indicates the starting point coordinates, Indicates the coordinates of the first task point, Indicates the total number of tasks, that is, the number of task points that require supplies. Indicates the sequence number of the task. T Mt Indicates Robots The coordinates of the last mission point, represents the coordinates of the end point, Indicates the distance between two points; Indicates i Robots Ri The time required to complete the material distribution of all mission points. vi represents the movement speed of the i-th robot Ri; PiIndicates i Robots Ri The power consumed to complete the material distribution of all mission points.

[0051] In this embodiment, the robot is limited by its own performance during the distribution process and its load does not exceed its maximum load. Therefore, when allocating task points, it is necessary to obtain how much material is needed for the task point. Considering the robot's own maximum load, based on this, in this embodiment, the constraint condition of the multi-machine task allocation model is set to the load constraint, and its expression is:

[0052] In the formula, Indicates the load of the robot. is the maximum load of the robot.

[0053] In this embodiment, the improved dung beetle algorithm is used to solve the multi-machine task allocation model to obtain a multi-machine task allocation solution, and the specific method is as follows: S31. Set the number of mission points that require materials and the weight of materials required for each mission point; S32, initializing the dung beetle algorithm parameters, including the maximum number of iterations, the population size, the proportion of rolling cockroaches, the proportion of foraging dung beetles, and the proportion of thief dung beetles; S33, calculating the fitness value of each dung beetle in the initialized population, defining the global optimal fitness and its position, the global worst fitness and its position, and selecting the rolling dung beetle, the hatching dung beetle, the foraging dung beetle and the thief dung beetle in proportional order; S34, setting the current number of iterations t=0; S35, using the multi-machine task allocation model as the fitness function of the dung beetle algorithm, executing the dung beetle algorithm, sorting the calculated fitness values, finding the dung beetle individual with the best global fitness and the dung beetle individual with the worst global fitness, and using the Cubic chaotic mapping method to modify parameters related to the number of iterations; S36, determining the type of each dung beetle in turn, and using t-distribution variation to update the current position of the dung beetle and calculate the next position information; S37, performing boundary determination on each dung beetle after the position information is updated, calculating its fitness value and selecting the optimal fitness value and its position; S38, determining whether the maximum number of iterations has been reached; If so, the optimal multi-machine task allocation solution is obtained; If not, return to step S35.

[0054] In step S3 of the embodiment of the present invention, path planning is performed for each robot according to the multi-machine path planning model. The multi-machine path planning model aims to optimize the time for each robot to complete the assigned task. The multi-machine path planning model is solved by using the improved dung beetle algorithm and the artificial potential field fusion method to plan its operation path, including the task starting point, the task end point and the robot walking path.

[0055] In an embodiment of the present invention, a serial model of multi-machine task allocation and multi-machine path planning is established to solve the problem of material distribution in a complex environment with multiple robots, wherein the multi-machine task allocation model is used to solve the task allocation problem of multiple robots, that is, to distribute materials to various material storage points, establish a multi-robot task allocation objective function that meets the constraints, and use the improved dung beetle algorithm to output the optimal task allocation plan; the multi-machine path planning model in the second part is used to solve the task allocation results of multiple robots, and use the fusion algorithm of the improved dung beetle algorithm and the artificial potential field method to plan a multi-robot collision-free path from the starting point to the task point and from the task point to the task point.

[0056] The present invention uses specific embodiments to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

[0057] Those skilled in the art will appreciate that the embodiments described herein are intended to help readers understand the principles of the present invention, and should be understood that the protection scope of the present invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific variations and combinations that do not deviate from the essence of the present invention based on the technical revelations disclosed by the present invention, and these variations and combinations are still within the protection scope of the present invention.

Claims

1. A multi-robot high-throughput information transmission and real-time collaborative operation system, characterized in that: It includes several robots and an operation coordination terminal, each of the robots includes an intelligent unit, and a sensing unit, a communication unit and an electromechanical unit connected to the intelligent unit; The communication unit is used to realize information interaction between the robot and the 5G base station and other robots; The sensing unit is used to collect surrounding environment data and transmit it to the intelligent unit; The intelligent unit is used to process the received surrounding environment data, build regional situation awareness of the current robot, and transmit it to the operation coordination end; The operation coordination end is used to run the collaborative operation algorithm to perform operation task allocation and local path planning based on the constructed regional situation awareness, generate control signals and send them to the corresponding robots; The electromechanical unit controls the robot to complete the task according to the received control signal.

2. The multi-robot high-throughput information transmission and real-time collaborative operation system according to claim 1, characterized in that: The communication unit is a mobile communication gateway device equipped with a 5G communication module and a WiFi6 communication module; wherein the 5G communication module is connected to a 5G base station, and the WiFi6 communication module is connected to the WiFi6 communication modules carried by other robots to form a Mesh network.

3. The multi-robot high-throughput information transmission and real-time collaborative operation system according to claim 2, characterized in that: When the communication quality of the Mesh network between robots is lower than the preset threshold, the communication mode between the robots is switched to: using the 5G base station as a communication transfer station, and information exchange between the robots through the 5G communication module; When the 5G communication quality is lower than the preset threshold, it switches back to the Mesh network between robots for communication.

4. The multi-robot high-throughput information transmission and real-time collaborative operation system according to claim 1, characterized in that: The job coordination end is deployed with a serial multi-machine task allocation model and a multi-machine path planning model; The multi-machine path planning model performs path planning according to the task allocation result of the multi-machine task allocation model.

5. A collaborative operation method based on the multi-robot high-throughput information transmission and real-time collaborative operation system according to any one of claims 1 to 4, characterized in that: The following steps are involved: S1. Establish 5G&WiFi6 networking architecture to realize information interaction between robots and 5G base stations and multiple robots; S2, based on the data collected by the binocular camera and laser radar in each robot, establish a three-dimensional environmental map of the area, and then obtain the regional situation awareness of each robot; S3, based on the regional situation awareness obtained by each robot and combined with the position of each robot, multi-machine task allocation and multi-machine path planning are performed in sequence to obtain the operation tasks assigned to each robot and their operation paths; S4. Based on the control signals generated by the operation tasks and operation paths of each robot, each robot is controlled to complete the business tasks and realize collaborative operation.

6. The collaborative operation method according to claim 5, characterized in that: In the step S1, in the 5G&WiFi6 networking architecture, the robots form a Mesh network through the communication connection between the WiFi6 communication modules, and the robots are connected to the 5G base station through the 5G communication module; When the communication quality of the Mesh network is lower than a preset threshold, the communication mode between the robots is switched to: using the 5G base station as a communication transfer station, and information exchange between the robots through the 5G communication module; When the 5G communication quality is lower than the preset threshold, it switches back to the Mesh network between robots for communication.

7. The collaborative operation method according to claim 5, characterized in that: The step S2 is specifically as follows: S21. Determine the relative sensing height of each robot according to the number of robots and the safe obstacle avoidance height range; S22. For each robot, decompose the three-dimensional environment into multiple two-dimensional planes at different heights according to the data collected by its binocular camera and laser radar; S23, converting the two-dimensional scanning data of each two-dimensional plane into local three-dimensional point cloud data of a fixed height; S24, according to the relative perception height of each robot, fusing the local three-dimensional point cloud data at different heights into global three-dimensional point cloud data through coordinate transformation; S25. Run the 3D mapping algorithm LeGO-LOAM based on the global 3D point cloud data to build a regional 3D environment map.

8. The collaborative operation method according to claim 5, characterized in that: In step S3, multi-machine task allocation is performed according to a multi-machine task allocation model, and the multi-machine task allocation model is expressed as: In the formula, represents the multi-machine task allocation objective function, , and represents the weight coefficient, Represents the total number of robots, Represents the robot index, Indicates Robots The length of the path to complete the material distribution of all mission points, represents the time cost function for all robots to complete the task, Represents the energy cost function of the i-th robot completing all task points; Indicates the starting point coordinates, Indicates the coordinates of the first task point, Indicates the total number of tasks, that is, the number of task points that require supplies. Indicates the sequence number of the task. T Mt Indicates Robots The coordinates of the last mission point, represents the coordinates of the end point, Indicates the distance between two points; Indicates i Robots Ri The time required to complete the material distribution of all mission points. vi represents the movement speed of the i-th robot Ri; Pi Indicates i Robots Ri The power consumed to complete the material distribution of all mission points; The constraint condition of the multi-machine task allocation model is the load constraint, which is expressed as: In the formula, Indicates the load of the robot. Indicates the maximum load of the robot.

9. The collaborative operation method according to claim 8, characterized in that: The improved dung beetle algorithm is used to solve the multi-machine task allocation model and obtain the multi-machine task allocation plan. The specific method is as follows: S31. Set the number of mission points that require materials and the weight of materials required for each mission point; S32, initializing the dung beetle algorithm parameters, including the maximum number of iterations, the population size, the proportion of rolling cockroaches, the proportion of foraging dung beetles, and the proportion of thief dung beetles; S33, calculating the fitness value of each dung beetle in the initialized population, defining the global optimal fitness and its position, the global worst fitness and its position, and selecting the rolling dung beetle, the hatching dung beetle, the foraging dung beetle and the thief dung beetle in proportional order; S34, setting the current number of iterations t=0; S35, using the multi-machine task allocation model as the fitness function of the dung beetle algorithm, executing the dung beetle algorithm, sorting the calculated fitness values, finding the dung beetle individual with the best global fitness and the dung beetle individual with the worst global fitness, and using the Cubic chaotic mapping method to modify parameters related to the number of iterations; S36, determining the type of each dung beetle in turn, and using t-distribution variation to update the current position of the dung beetle and calculate the next position information; S37, performing boundary determination on each dung beetle after the position information is updated, calculating its fitness value and selecting the optimal fitness value and its position; S38, determining whether the maximum number of iterations has been reached; If so, the optimal multi-machine task allocation solution is obtained; If not, return to step S35.

10. The collaborative operation method according to claim 8, characterized in that: In step S3, the path of each robot is planned according to the multi-machine path planning model. The multi-machine path planning model aims to optimize the time for each robot to complete the assigned task. The multi-machine path planning model is solved by using the improved dung beetle algorithm and the artificial potential field fusion method to plan its working path, including the task starting point, the task end point and the robot walking path.

Images