Modular soft robot communication control system and method based on 5G communication

Abstract

The application discloses a kind of modular soft robot communication control system and method based on 5G communication, single modular soft robot is perceived to peripheral environment by reinforcement learning algorithm, act on each modular soft robot;Modular soft robot is identified to the feature point of robot by deep learning algorithm, data is transmitted by 5G industrial gateway, realizes the real-time acquisition of robot feature;Control by cluster algorithm, in the case where 5G industrial gateway is used as host computer, send control instruction to each soft robot slave, each modular soft robot receives the cluster instruction sent by 5G industrial gateway and can execute corresponding cooperative action.The application uses 5G industrial module to carry out the communication of modular soft robot, ensures the real-time performance, timeliness of soft robot communication, 5G industrial gateway can also be used as host computer to realize the control of slave, with multiple use functions.

Description

TECHNICAL FIELD

[0001] The application relates to a robot control method, in particular to a communication control system and method of a modular soft robot based on 5G communication. BACKGROUND

[0002] Modularity is the development trend of intelligent robots. By using basic modules with certain functions, robots suitable for specific tasks can be quickly assembled to adapt to different working conditions and task requirements. For example, modular soft robots can cope with various complex environments due to their flexible and variable structures, and have very wide application prospects. However, when intelligent modular soft robots work cooperatively, the communication of the robots often affects their control effect, especially in some unstructured environments such as disaster rescue scenarios. Due to their own characteristics, modular soft robots can drill into deeper and more complex working environments, and at this time the modular soft robots are required to communicate with the control system without delay, which puts high requirements on the communication quality.

[0003] Currently, with the continuous progress of science and technology, the rise of AI and the rise of autonomous driving, the traditional 4G network has been unable to meet the application requirements, and 5G communication has emerged as the times require. The low delay, large bandwidth and high reliability of 5G communication provide great help for the application of intelligent robots. 5G communication can realize nearly real-time communication between robots and people, robots and the cloud, and robots and control systems. Many tasks that robots could not complete in the past can be realized. At present, 5G communication technology has not been widely commercialized, especially for use on robots. SUMMARY

[0004] In order to solve the above problems, the application provides a modular soft robot communication control system and method based on 5G communication, which can realize data information collection and data processing of the soft robot end, cloud convergence, and can meet remote system monitoring and system rapid response.

[0005] Therefore, the technical scheme of the application is: a modular soft robot communication control method based on 5G communication, comprising the following steps:

[0006] 1) defining the state space of the modular soft robot, including the current configuration, position, speed, angle and direction of the robot, and the position of the obstacles and the target object in the environment;

[0007] 2) a single modular soft robot perceives the surrounding environment through a reinforcement learning algorithm, transmits information through a 5G industrial gateway, and acts on each modular soft robot, so that each modular soft robot has environmental perception capability;

[0008] 3) The modular soft robot uses deep learning algorithms to collect, train, and identify the robot's feature points, and transmits data through a 5G industrial gateway to achieve real-time acquisition of the robot's position information;

[0009] 4) The modular soft robots that obtain posture information are controlled by a cluster algorithm. With the 5G industrial gateway as the host, control instructions are sent to each modular soft robot. After receiving the cluster instructions sent by the 5G industrial gateway, each modular soft robot can perform the corresponding collaborative action.

[0010] On the basis of the above solution and as a preferred solution of the above solution: the reinforcement learning algorithm in step 2) includes the following steps:

[0011] a1) Define the state space of the modular soft robot, clarify the set of actions that the modular soft robot can acquire and the learning objectives. The soft robot obtains the optimal behavior strategy through structural deformation and angle adjustment.

[0012] a2) Selecting a neural network algorithm, collecting and generating a dataset of interactions with the modular soft robot, and training the modular soft robot;

[0013] a3) During the interaction between a single modular soft robot and the environment, the strategy parameters are updated using the soft robot's state information, and the optimal parameters of the model are obtained through continuous algorithm updates and iterations;

[0014] a4) The acquired environment and parameters of the modular soft robot are uploaded to the cloud via the 5G industrial gateway. The 5G industrial gateway is used to store and upload the soft robot information. After the cloud receives the soft robot's environmental information, it integrates the information through the control end and then feeds it back to each soft robot end, realizing the modular soft robot's perception of the environment.

[0015] On the basis of the above solution and as a preferred solution of the above solution: the deep learning algorithm in step 3) includes the following steps:

[0016] b1) Using a depth camera to collect data on the feature points of a single modular soft robot, the omnidirectional collection of feature points is achieved by changing the angle and direction during data collection;

[0017] b2) Using a deep neural network algorithm to train the features of a single modular soft robot, obtaining weights for the robot's feature points, and then optimizing the modular soft robot. The optimized parameters are then used for feature recognition of the modular soft robot to obtain the robot's real-time pose parameters.

[0018] b3) the acquired pose information is uploaded to the cloud through the 5G industrial gateway, and then transmitted to the control end through the cloud to form a closed loop, thereby realizing the control of the modular soft robot.

[0019] As a preferred scheme of the above scheme and on the basis of the above scheme: the cluster algorithm in the step 4) comprises the following steps:

[0020] c1) acquiring the pose and state information of each modular soft robot through a deep learning algorithm to obtain a robot cluster;

[0021] c2) integrating and decomposing the cluster task to each modular soft robot, ensuring that each robot reasonably utilizes resources through task allocation and scheduling, and realizing information sharing between robots through communication and protocol;

[0022] c3) performing path planning on each modular soft robot to ensure that the robot can move according to the predetermined plan and path through a cooperative strategy; the cooperative strategy comprises the following steps:

[0023] c3.1) taking the 5G industrial gateway as a host and setting each modular soft robot as a slave, each modular soft robot being assigned a slave address, and the 5G industrial gateway having multiple data distribution interfaces and being capable of simultaneously controlling multiple slaves;

[0024] c3.2) the 5G industrial gateway integrates and processes the information of each robot sent from the cloud, and then simultaneously and in parallel sends the information to each robot slave, thereby ensuring the fairness and timeliness of information received by each robot slave;

[0025] c3.3) realizing the cooperative operation of the robot slave through the transmission of instructions to ensure the efficient cooperative cluster control of the robot through the interaction and cooperation of the robot.

[0026] Another technical scheme of the present application is a modular soft robot communication control system based on 5G communication, comprising: a robot end, a 5G industrial gateway, a 5G base station end, a cloud end and a remote control end;

[0027] The robot end is divided into a single robot control unit and multiple robot control units, the single robot control unit is used for acquiring the parameter value, pose and state information of a single modular soft robot, and sending the information to the cloud through the 5G industrial gateway, and then sending the information to the remote control end through the cloud;

[0028] The multiple robot control units are used for cooperative motion control of multiple modular soft robots;

[0029] The 5G industrial gateway can be used as a transmission medium between the robot end and the cloud end, and can play a role in information transmission.

[0030] The remote control end is used for information integration of the PC end and instruction control of the host computer.

[0031] As a preferred scheme of the above scheme, a plurality of data distribution interfaces are arranged on the 5G industrial gateway, each robot slave is assigned a slave address by a master control chip, communicates with the 5G industrial gateway, and controls a plurality of robot slaves simultaneously.

[0032] As a preferred scheme of the above scheme, the 5G base station end is used for transmitting robot information in the 5G industrial gateway to the cloud end, acquiring state information and environment information of the robot in the cloud end, and transmitting corresponding information to the PC end to control the robot by combining a PC control algorithm.

[0033] As a preferred scheme of the above scheme, the remote control end includes a man-machine interface, the man-machine interface can select the robot by the number of the robot, and control the robot by sending instructions.

[0034] Compared with the prior art, the present application has the following advantages:

[0035] 1. The 5G industrial module is used for communication of the modular soft robot, which ensures the real-time and timeliness of the soft robot communication and realizes non-differential transmission of information between the soft robot end and the cloud end.

[0036] 2. The 5G industrial gateway can be used as an information transmission medium and a host computer to control the slave, has multiple functions, and is used as a host computer for soft robot control, sends instructions from the host to the slave, and realizes rapid cooperative control of each soft robot.

[0037] 3. The control algorithm is modularly designed and transmitted to the master control chip of each modular soft robot, and when a certain algorithm needs to be improved, the algorithm can be directly called, which is efficient and convenient.

[0038] 4. The soft robot can quickly perceive the environment, quickly identify the characteristics of the robot individual, quickly move the robot cluster, quickly optimize the robot motion gait, and has higher intelligence.

[0039] 5. Through the transmission of the cloud, various information values of the robot can be obtained in real time through the PC end, and various information can be processed and controlled in a timely manner, and through the combination of the 5G industrial gateway and the modular soft robot, guidance can be provided for the future more intelligent robot control. BRIEF DESCRIPTION OF DRAWINGS

[0040] Figure 1 It is a schematic diagram of the robot end communication of the application.

[0041] Figure 2 It is a schematic diagram of the robot end communication of the application.

[0042] Figure 3 It is a schematic diagram of the robot end communication of the application.

[0043] Figure 4 It is a schematic diagram of the robot end communication of the application.

[0044] Figure 5 It is a schematic diagram of the robot end communication of the application. DETAILED DESCRIPTION

[0045] I. Robot cooperative communication platform

[0046] As shown in Figure 1 , the robot cooperative communication platform mainly consists of three parts, which are the soft robot end, the information interaction end, and the remote control end, and in detail:

[0047] The soft robot end is divided into two parts, which are single robot control and multiple robot control. When single robot control is performed, the straight line motion of single module robot can be controlled, such as forward movement, backward movement, and the bending motion of single module robot, such as bending, deflection, etc. When multiple robot control is performed, the cooperative motion control of multiple robots is mainly performed, such as assembly control, obstacle avoidance control, positioning control, obstacle crossing control, etc.

[0048] The information interaction end mainly performs information interaction and reading between robots, which includes interaction of state information and interaction of environmental information. The state information mainly includes battery capacity information display of the robot, motion speed display of the robot, bending angle display, deflection angle display, motor running state monitoring, and position feedback of the robot, etc. The environmental information interaction includes obstacle information, terrain environmental information, position information, and other environmental parameter information, such as temperature and humidity, air pressure information, etc.

[0049] The remote control end mainly carries out information integration of the PC end and instruction control of the upper computer, which is specifically embodied as follows: in the PC end integration, on the one hand, detailed information of the robot transmitted by the cloud server is received, such as battery power information, motor state, robot running state and the like, to realize data monitoring of the robot.

[0050] On the other hand, the robot is controlled by intelligent algorithms, and the algorithms mainly used in the robot control are deep learning algorithm, strength learning algorithm and cluster algorithm. The deep learning algorithm is used to extract and fit the features of the robot to obtain coordinate parameter values. The strength learning algorithm is used to obtain environmental information and make intelligent decisions according to the changes of the environment. The cluster algorithm is used for cluster collaborative motion of the robot.

[0051] When the upper computer is controlled, the running state control of the system and the control of the man-machine interface are mainly carried out. In the system running state control, the running state of each robot system is monitored, and the fault condition of the robot is judged and analyzed. In the man-machine interface, the accurate control of the robot can be realized, such as single robot control, mixed robot control, selection of the robot through the number of the robot, and control of the robot through sending instructions.

[0052] II. Communication process of modular soft robot end

[0053] As shown in Figure 2 When the robot communicates, it is mainly composed of several parts, which are external environment action, robot end information processing, 5G device end information transmission, 5G base station end information sending, cloud end data monitoring and PC end data analysis.

[0054] Firstly, the external environment end mainly responds to the environment in which the soft robot is located, senses the changes of the environment and adjusts the running state according to the feedback of the environment, such as the robot in the obstacle environment, the robot in different temperature and humidity scenes and the like. The intelligent soft robot can adjust the posture in time according to the environment, action, reward and other factors, and achieve the adaptation to the environment under the guidance of reinforcement learning.

[0055] Secondly, at the robot end, each robot is an independent individual. In this embodiment, ESP32 is used as the main control chip. The ESP32 chip has the functions of 485 communication and WiFi communication, and can receive and send robot data signals. Each robot module can have corresponding learning ability through intelligent algorithms such as deep learning, machine learning, reinforcement learning, ant colony or particle swarm algorithm, genetic algorithm and the like, so as to obtain depth value, characteristic value, state value and pose information of the robot. Then the information is sent to the cloud platform through the 5G channel, and the cloud end can monitor the information of the robot, so as to form a closed loop.

[0056] Then, at the 5G device end, the transmission of soft robot information is mainly achieved through the 5G gateway. The 5G gateway used in this embodiment is the industrial-grade product FCU2303. The core board supports the 5G modules of Huawei MH5000-31 and Quectel RM500Q-GL. The 5G core component used in this embodiment is the Quectel RM500Q-GL module. The FCU2303 industrial module supports 5G communication and 4G full network access. It has multiple RS485 and RS232 communication modules, 8-way Gigabit Ethernet, and a standardized E-KEY interface, which can realize WiFi communication and realize the transmission of 5G signals by inserting a 5G Nano card.

[0057] Then, at the 5G base station end, the robot information in the 5G industrial gateway is transmitted to the cloud through the 5G base station, and the robot's status information, environmental information, etc. are acquired in the cloud. The corresponding information is then transmitted to the PC end, and the robot is controlled in conjunction with the PC control algorithm.

[0058] Finally, the PC side mainly analyzes and controls the soft robot data, including algorithm regulation, data monitoring, system operation and maintenance, and human-computer interaction, and modifies and transmits instructions, which are then sent to each robot side through the cloud, realizing closed-loop control operations from the robot individual to the cloud and then to the device side.

[0059] The upper computer operation interface of the modular soft robot mainly serves as the user's operation interface. It is connected to the host through the serial port and can perform functions such as single module control, multi-module control, multi-modular mixed control, and slave status display.

[0060] 3. Modular soft robot communication control method based on 5G communication:

[0061] like Figure 4 、 Figure 5 As shown in the figure, the algorithms that can be used for modular soft robot cluster control include reinforcement learning algorithms, deep learning algorithms, and cluster algorithms.

[0062] (1) Reinforcement Learning Algorithm

[0063] The implementation of reinforcement learning algorithms helps modular soft robots quickly adapt to environmental changes, gaining rewards from environmental perception, and thus guiding their movements. A single modular soft robot uses reinforcement learning algorithms to perceive its surroundings. This information is transmitted via 5G industrial gateways and applied to each modular soft robot, giving each modular soft robot environmental perception capabilities.

[0064] The specific implementation plan is:

[0065] a1) Define the state space of the modular soft robot, including information such as the robot's current configuration, position, speed, angle, and direction, as well as the location of obstacles and targets in the environment. Identify the set of actions that the modular soft robot can capture and design appropriate reward rules to guide the robot's behavior, such as giving positive rewards when the robot approaches an obstacle and negative rewards when it collides with an obstacle.

[0066] a2) Adopt an appropriate reinforcement learning implementation scheme. For example, if the environment model is known or can be learned, use model-based reinforcement learning. If the environment model is unknown or difficult to learn, use a model-free reinforcement learning algorithm. Collect and generate robot interaction datasets for training modular soft robots.

[0067] a3) Through simple control of a single robot, the strategy parameters are updated through the state information of the soft robot during its interaction with the environment, and the optimal parameters of the model are obtained through continuous updating and iteration of the algorithm.

[0068] a4) The acquired environment and robot parameters and other related information are uploaded to the cloud through the 5G industrial gateway. At this time, the 5G industrial gateway is equivalent to the transmission medium, which is used to store and upload the robot information. After the cloud receives the robot's environmental information, it integrates the information through the control end and then feeds it back to each robot end, realizing the robot's perception of the environment.

[0069] (2) Deep Learning Algorithms

[0070] Deep learning algorithms can help modular soft robots obtain their own status and posture information in real time, without being affected by the model configuration or the environment. Once the modular soft robots controlled by reinforcement learning have a certain degree of environmental perception, deep learning algorithms can be implemented. Deep learning algorithms collect, train, and identify the robot's feature points and transmit data through 5G industrial gateways to achieve real-time acquisition of robot features.

[0071] The specific implementation plan is:

[0072] b1) Use a depth camera to collect data on the feature points of a single modular soft robot. During data collection, data enhancement processing must be performed first, such as adjusting the brightness, angle, image rotation, and noise removal of the collected soft robot image, to achieve all-round data collection and help improve recognition accuracy.

[0073] b2) Use a suitable deep neural network algorithm to train the features of a single modular soft robot, obtain the mask values ​​and weight values ​​of the robot's feature points, and then optimize the modular soft robot. The optimized parameters are used for feature recognition of the modular soft robot to obtain the robot's real-time posture parameter values; such as the robot's three-dimensional coordinate values, angle joint values, etc., thereby providing preparation for further realizing the motion control of the modular soft robot.

[0074] b3) After acquiring the characteristic pose of the modular soft robot, the pose information is uploaded to the cloud via the 5G industrial gateway, and then transmitted to the control end via the cloud, thus forming a closed loop to realize the control of the modular soft robot. At this time, the 5G industrial gateway still acts as a transmission medium, mainly transmitting the robot's pose and status information.

[0075] (3) Clustering algorithm

[0076] The swarming algorithm is highly adaptable and robust, enabling efficient collaboration between modular soft robots and enabling rapid changes through swarm response. Under the control of the swarming algorithm, the modular soft robots, with a 5G industrial gateway acting as the host, send control commands to each individual robot. Upon receiving the swarming commands from the host, each individual robot executes the corresponding coordinated action.

[0077] The specific implementation plan is

[0078] c1) The position and state information of each robot has been obtained through the deep learning algorithm, resulting in a group of characteristic robot clusters;

[0079] c2) Integrate and decompose cluster tasks to each modular soft robot, ensure that each robot can make rational use of resources through reasonable task allocation and scheduling, and realize information sharing between robots through communication and protocols;

[0080] c3) Path planning: Each robot needs to plan its own path to complete the task, and collaborative strategies must be used to ensure that the robots can move according to the predetermined plan and path;

[0081] The collaborative strategy includes the following steps:

[0082] c3.1) During collaborative robot motion, the host computer controls the motion of each modular soft robot slave. Each slave robot is assigned a slave address via the ESP32 master control chip, while the 5G industrial gateway FCU2303 acts as the host. This industrial gateway has multiple data distribution interfaces and can control multiple slaves simultaneously.

[0083] c3.2) the host integrates the information of each aspect of the robot sent by the cloud, and then simultaneously and in parallel sends the information to each robot slave, to ensure the fairness and timeliness of information received by each slave;

[0084] c3.3) then the slaves are coordinated through the transmission of instructions, such as obstacle avoidance, climbing, etc., to ensure the interaction and cooperation of the robots, and finally the algorithm is improved according to the problems encountered during the movement of the robots, so as to realize the efficient cooperative cluster control of the robots.

[0085] As shown in Figure 3 The working principle is: the host computer sends instructions to the robot, and the communication between the host computer and the modular software robot can be carried out through private protocol data transmission through UART. The information can be transmitted to the host computer first, and the 5G industrial gateway is used as the host computer to distribute the information. After the 5G industrial gateway is used as the host computer, the control information is sent to each robot slave through multiple WiFi antennas, such as modular software robot slave 1, modular software robot slave 2,..., modular software robot slave n, etc. Then each robot slave transmits and receives information through the RS485 data transmission chip, so as to complete the transmission control of the host computer control instruction to the single robot.

[0086] During the execution of the control algorithm of the modular software robot, each algorithm has a high execution rate. In the reinforcement learning execution stage, the learning efficiency of the reinforcement learning can reach 90% or more. In the deep learning algorithm stage, the recognition accuracy of the feature points can reach 98% or more, and multiple soft robots with feature points can be recognized. In the cluster algorithm execution stage, the average time from the sending of the instruction by the 5G industrial module as the host computer to the receiving of the instruction by the slave is within 5ms, and each algorithm has a high execution accuracy.

[0087] The above is only the preferred embodiment of the present application, and the protection scope of the present application is not limited to the above-mentioned embodiments. Any technical solution falling within the scope of the present application shall be considered as falling within the protection scope of the present application. It should be noted that for ordinary skilled persons in the art, some improvements and refinements without departing from the principles of the present application shall also be considered as falling within the protection scope of the present application.

Claims

1. A modular soft robot communication control method based on 5G communication, characterized by: The following steps are involved: 1) Define the state space of the modular soft robot, including the robot's current configuration, position, speed, angle and direction, the location of obstacles in the environment, and the location of the target object; 2) A single modular soft robot uses a reinforcement learning algorithm to perceive its surrounding environment. This information is transmitted via a 5G industrial gateway and acts on each modular soft robot, giving each modular soft robot environmental perception capabilities. 3) The modular soft robot uses deep learning algorithms to collect, train, and identify the robot's feature points, and transmits data through a 5G industrial gateway to achieve real-time acquisition of the robot's position information; 4) The modular soft robots that obtain pose information are controlled by a cluster algorithm. With the 5G industrial gateway as the host, control instructions are sent to each modular soft robot. After receiving the cluster instructions sent by the 5G industrial gateway, each modular soft robot can perform the corresponding collaborative action. The clustering algorithm in step 4) includes the following steps: c1) Obtain the position and state information of each modular soft robot through a deep learning algorithm to obtain a robot cluster; c2) Integrate and decompose cluster tasks to each modular soft robot, ensure that each robot makes rational use of resources through task allocation and scheduling, and enable information sharing between robots through communication and protocols; c3) Plan paths for each modular soft robot and use collaborative strategies to ensure the robots can move according to the predetermined plan and path. The collaborative strategy includes the following steps: c3.1) Use the 5G industrial gateway as the host and each modular soft robot as a slave. Each modular soft robot is assigned a slave address. The 5G industrial gateway has multiple data distribution interfaces and can control multiple slaves simultaneously. c3.2) The 5G industrial gateway integrates and processes all aspects of the robot information sent from the cloud, and then sends it to each slave robot simultaneously, ensuring fairness and timeliness in receiving information from each slave robot. c3.3) Realize collaborative operation of robot slaves through the transmission of instructions, ensure interactive collaboration of robots and realize efficient collaborative cluster control of robots.

2. The modular soft robot communication control method based on 5G communication according to claim 1, characterized in that: The reinforcement learning algorithm in step 2) includes the following steps: a1) Define the state space of the modular soft robot, clarify the set of actions that the modular soft robot can acquire and its learning objectives, and the soft robot acquires the optimal behavior strategy through structural deformation and angle adjustment; a2) Select a neural network algorithm, collect and generate a dataset of interactions with the modular soft robot, and train the modular soft robot; a3) During the interaction between a single modular soft robot and its environment, the strategy parameters are updated using the soft robot’s state information, and the optimal parameters of the model are obtained through continuous algorithm updates and iterations. a4) The acquired environment and parameters of the modular soft robot are uploaded to the cloud via the 5G industrial gateway. The 5G industrial gateway is used to store and upload the soft robot information. After the cloud receives the soft robot's environmental information, it integrates the information through the control end and then feeds it back to each soft robot end, realizing the modular soft robot's perception of the environment.

3. The modular soft robot communication control method based on 5G communication according to claim 1, characterized in that: The deep learning algorithm in step 3) includes the following steps: b1) Using a depth camera to collect data on the feature points of a single modular soft robot, the omnidirectional collection of feature points is achieved by changing the angle and direction during data collection; b2) Using a deep neural network algorithm to train the features of a single modular soft robot, the weights of the robot's feature points are obtained, and then the modular soft robot is optimized. The optimized parameters are used for feature recognition of the modular soft robot to obtain the robot's real-time pose parameter values; b3) The acquired posture information is uploaded to the cloud via the 5G industrial gateway, and then transmitted to the control end via the cloud, forming a closed loop to realize the control of the modular soft robot.

4. A modular soft robot communication control system based on the method according to any one of claims 1 to 3, characterized in that: include: Robot side, 5G industrial gateway, 5G base station side, cloud side, and remote control side; The robot end is divided into a single robot control unit and multiple robot control units. The single robot control unit is used to obtain the parameter value, posture and status information of a single modular soft robot, and send it to the cloud through the 5G industrial gateway, and then send it to the remote control end through the cloud; Multiple robot control units are used to perform coordinated motion control on multiple modular soft robots; The 5G industrial gateway can serve as a transmission medium between the robot and the cloud, playing a role in information transmission. At the same time, the 5G industrial gateway can serve as the host, and each soft robot as a slave. The host receives information sent from the cloud, integrates and processes it, and then sends it to each slave robot simultaneously. The transmission of instructions enables the collaborative operation of the slave robots. The remote control terminal is used for information integration of the PC terminal and command control of the host computer.

5. The modular soft robot communication control system based on 5G communication according to claim 4, characterized in that: The 5G industrial gateway is equipped with multiple data distribution interfaces. Each robot slave is assigned a slave address through the main control chip, communicates with the 5G industrial gateway, and then controls multiple robot slaves at the same time.

6. The modular soft robot communication control system based on 5G communication according to claim 4, characterized in that: The 5G base station is used to transmit the robot information in the 5G industrial gateway to the cloud, acquire the robot's status information and environmental information in the cloud, and transmit the corresponding information to the PC, and control the robot in conjunction with the PC control algorithm.

7. The modular soft robot communication control system based on 5G communication according to claim 4, characterized in that: The remote control terminal includes a human-machine interaction interface, which can realize the selection of robots by the number of robots and the control of robots by sending instructions.

Images