Robot Communication Method and Device Based on Multi-Network Communication

Abstract

This application discloses a robot communication method and apparatus based on multi-network communication, comprising: constructing a multi-network communication architecture; when the robot has a data transmission requirement, acquiring real-time status information of each network communication method in the multi-network communication architecture; allocating the optimal network communication method for the data to be transmitted according to the real-time status information and preset communication priority rules; determining neighboring robots through a neighbor discovery mechanism; dynamically adjusting and optimizing the data transmission path according to the current status information of the neighboring robots through a path planning strategy based on opposition learning and the black widow algorithm to determine the optimal data transmission path; and transmitting the data to be transmitted to the target robot through the optimal data transmission path and the optimal network communication method. This application effectively improves network resource utilization by setting multi-network communication priorities, and effectively improves the robot's communication efficiency and stability by introducing a path planning strategy based on opposition learning and the black widow algorithm.

Description

Technical Field

[0001] This application relates to the field of robot communication technology, and in particular to robot communication methods and devices based on multi-network communication. Background Technology

[0002] With the rapid development of robotics technology, robots have been widely used in various fields such as industry, service, and home. Communication, as a key link in enabling robots to achieve remote control, data transmission, and collaborative work, directly affects the robot's performance.

[0003] Currently, commonly used communication methods for robots include WiFi, 5G, 4G, LoRa, and Bluetooth. However, in practical applications, a single communication method is often insufficient to meet the communication needs of robots in complex environments. For example, WiFi networks are widely used in indoor environments, but they suffer from limited signal coverage and susceptibility to interference; 5G / 4G cellular networks have wide coverage, but the signal may be weak in some remote areas or areas with signal blockage, and the cost is high when transmitting large amounts of data; LoRa networks have long transmission distances and low power consumption, but the transmission rate is relatively slow, making them suitable for small data transmissions; Bluetooth and USB-Net networks have short transmission distances and are usually used as temporary communication methods.

[0004] Furthermore, existing robot communication systems often suffer from problems such as untimely switching and data transmission interruptions during network switching, affecting the normal operation of robots. At the same time, different types of data, such as video, high-volume data, and critical operational data, are not classified and transmitted, leading to inefficient use of network resources. Summary of the Invention

[0005] The main purpose of this application is to provide a robot communication method and device based on multi-network communication, which aims to solve the technical problems of poor communication stability and unreasonable utilization of network resources in the existing robot communication technology.

[0006] To achieve the above objectives, this application proposes a robot communication method based on multi-network communication, which includes:

[0007] Construct a multi-network communication architecture, which is used to integrate at least two different types of network communication methods, including at least wired network communication, wireless network communication, Bluetooth communication, or mobile data network communication.

[0008] When the robot has a data transmission requirement, the real-time status information of each network communication method in the multi-network communication architecture is obtained. The real-time status information includes at least one of signal strength, connection status, network bandwidth, and network latency.

[0009] Based on the real-time status information and the preset communication priority rules, the optimal network communication method is allocated for the data to be transmitted, wherein the preset communication priority rules are set according to the data type, the urgency of the data, and the quality of network communication.

[0010] The neighbor discovery mechanism identifies neighboring robots that are in the same communication environment as the current robot and can interact with it.

[0011] The current state information of the neighboring robots is obtained, and the data transmission path is dynamically adjusted and optimized according to the current state information of the neighboring robots through a path planning strategy to determine the optimal data transmission path. The path planning strategy is based on opposition learning and the Black Widow algorithm.

[0012] The data to be transmitted is transmitted to the target robot through the optimal data transmission path and the optimal network communication method.

[0013] In one embodiment, allocating the optimal network communication method for the data to be transmitted based on the real-time status information and preset communication priority rules includes:

[0014] The network quality feature vectors for each network communication method are determined based on the real-time status information.

[0015] Obtain the data type feature vector and the urgency feature vector of the data to be transmitted. The data type feature vector consists of semantic level, data volume coefficient and preset tolerable packet loss rate. The urgency feature vector consists of data timeliness requirements, queued delay and data urgency.

[0016] The static priority components of each network communication method are determined based on the data type feature vector and the urgency feature vector.

[0017] The real-time score of each network communication method is determined based on the network quality feature vector.

[0018] The dynamic priority components of each network communication method are determined based on the static priority components and the real-time scores.

[0019] The priority difference between any two network communication methods is determined based on the dynamic priority components of each network communication method.

[0020] If all priority differences are greater than or equal to a preset threshold, the network communication method with the highest dynamic priority component is selected as the optimal network communication method.

[0021] In one embodiment, after determining the priority difference between any two network communication methods based on the dynamic priority components of each network communication method, the method further includes:

[0022] If the priority difference is less than a preset threshold, a fragmentation dual-transmission strategy is executed, and both network communication methods corresponding to the priority difference are selected as candidate network communication methods.

[0023] The data to be transmitted is fragmented according to a preset fragmentation rule to obtain multiple data fragments;

[0024] The corresponding data fragments are transmitted simultaneously through the candidate network communication methods, and the transmission status of the two candidate network communication methods is monitored in real time during the transmission process.

[0025] If one of the candidate network communication methods experiences a transmission error, the remaining untransmitted data will be immediately fragmented and switched to another candidate network communication method that is transmitting normally for continued transmission.

[0026] Once all data fragments have been successfully transmitted to the target robot, the received data fragments are reassembled on the target robot to recover the complete data to be transmitted.

[0027] In one embodiment, the step of determining neighboring robots that are in the same communication environment and can interact with the current robot through the neighbor discovery mechanism includes:

[0028] The robot acquires feature data of its current communication environment and maps the feature data to a predefined environmental feature space to generate a communication environment feature vector for the current robot.

[0029] The communication environment similarity is obtained by calculating the similarity between the current robot's communication environment feature vector and the robot to be judged.

[0030] Based on the communication environment similarity, the current robot and the robot to be judged are determined to be in the same domain, and the same domain determination result is obtained.

[0031] If the same-domain determination result indicates that the current robot and the robot to be determined are in the same communication domain, the robot to be determined is determined to be a same-domain robot.

[0032] The interaction feasibility of the robots in the same domain is evaluated based on the robot interaction feasibility scoring model, and the interaction feasibility score of each robot in the same domain is obtained. The robot interaction feasibility scoring model is constructed based on the communication capability, remaining battery power, task load, historical interaction success rate and the stability of the current network environment of the robots in the same domain.

[0033] The robots in the same domain are sorted from high to low according to the interaction feasibility score, and the robots in the same domain with an interaction feasibility score higher than a preset score threshold are selected as neighbor robots that can perform data interaction.

[0034] In one embodiment, the step of obtaining the current state information of the neighboring robot and dynamically adjusting and optimizing the data transmission path based on the current state information of the neighboring robot using a path planning strategy to determine the optimal data transmission path includes:

[0035] Obtain the current status information of the neighboring robot, which includes location information, communication capability parameters, current task load, and network environment stability indicators;

[0036] Using the current robot as the starting node, the target robot as the target node, and the neighboring robots as intermediate nodes, an initial data transmission path set is constructed based on the current state information.

[0037] The opposition learning algorithm is used to evaluate each original path in the initial data transmission path set and generate an opposition path for each original path. The construction rule of the opposition path is to generate an alternative path with a different topology than the original path but connecting the starting node and the target node by changing the order of intermediate nodes or replacing some intermediate nodes.

[0038] The fitness values ​​of each original path and its corresponding opposing path are calculated based on path length, cumulative node communication quality, and overall path stability coefficient. The path length is obtained by weighted summation of physical distance or logical hop count between nodes. The cumulative node communication quality is the product of the communication capability parameters of each node in the path. The overall path stability coefficient is calculated by integrating the network environment stability indicators of each node and using a weighted average method.

[0039] If the fitness value of the opposing path is greater than the fitness value of the original path, then the original path is replaced according to the opposing path and the initial data transmission path set is updated to obtain the updated initial data transmission path set.

[0040] The Black Widow algorithm is used to optimize the updated initial data transmission path set to obtain the optimal data transmission path.

[0041] In one embodiment, optimizing the updated initial data transmission path set using the Black Widow algorithm to obtain the optimal data transmission path includes:

[0042] Treat each path in the updated initial data transmission path set as an individual spider to construct the initial spider population;

[0043] The fitness value of each spider in the initial spider population is calculated according to a preset fitness function, which is constructed based on path length, cumulative node communication quality, overall path stability coefficient, and data transmission timeliness requirements.

[0044] Based on the fitness value of each individual spider in the initial spider population, high-quality spider individuals with fitness values ​​higher than a preset threshold are selected to form a high-quality spider subpopulation.

[0045] The high-quality spider population is used for breeding operations to generate new spider individuals through crossover and mutation. The crossover operation adopts single-point crossover or multi-point crossover, and the mutation operation adopts the method of randomly changing some intermediate nodes in the path.

[0046] The new spider individuals are merged with the remaining unselected spider individuals to form a new spider population;

[0047] Determine whether the new spider population meets the preset termination conditions, including reaching the maximum number of iterations, fitness value convergence, or finding the optimal path that meets the data transmission timeliness requirements;

[0048] If the termination condition is met, the path corresponding to the spider individual with the highest fitness value from the new spider population is selected as the optimal data transmission path.

[0049] If the termination condition is not met, return to the step of calculating the fitness value of each spider individual in the new spider population according to the preset fitness function, and continue iterative optimization until the preset termination condition is met, and obtain the optimal data transmission path.

[0050] In one embodiment, transmitting the data to be transmitted to the target robot via the optimal data transmission path and the optimal network communication method includes:

[0051] Obtain the list of available channels corresponding to the optimal network communication method, monitor the channels in the list of available channels in real time, and obtain the current occupancy status and signal quality parameters of each channel;

[0052] Based on the current occupancy status and signal quality parameters of each channel, a dynamic channel allocation strategy is adopted to allocate the optimal transmission channel for the data to be transmitted. The dynamic channel allocation strategy is set according to the channel idle time, signal interference intensity and historical transmission success rate.

[0053] The data to be transmitted is encapsulated according to a preset data encapsulation format. Channel allocation information and target robot identification information are added to the header of the encapsulated data to generate a data packet to be sent.

[0054] The data packet to be sent is transmitted to the target robot through the optimal transmission channel along the optimal data transmission path.

[0055] In one embodiment, after transmitting the data packet to be sent to the target robot via the optimal transmission channel along the optimal data transmission path, the method further includes:

[0056] If, during transmission, it is detected that the optimal transmission channel experiences increased signal interference or is occupied by other devices, the channel contention backoff mechanism will be activated immediately.

[0057] Based on the contention level of the optimal transmission channel and the preset backoff algorithm, the backoff time window is dynamically adjusted, data transmission is paused within the backoff time window, and the state changes of the optimal transmission channel are continuously monitored.

[0058] If the optimal transmission channel becomes available again when the backoff time window ends, data transmission is restarted.

[0059] If the optimal transmission channel is still unavailable, the dynamic channel allocation strategy is re-executed, and the second-best transmission channel is selected from the list of available channels to continue transmitting the data packet to be sent.

[0060] If there are no other available channels in the available channel list, data transmission is paused, and the channel list expiration repair mechanism is triggered to search for and update the available channel list again, resulting in an updated available channel list.

[0061] Select the new optimal transmission channel from the updated list of available channels to transmit the data packet to be sent.

[0062] In one embodiment, after transmitting the data packet to be sent to the target robot via the optimal transmission channel along the optimal data transmission path, the method further includes:

[0063] During data transmission, if a fault or reduced communication capability is detected in an intermediate node on the optimal data transmission path, the intermediate node will be designated as the faulty intermediate node, and a multi-channel virtual monitoring mechanism will be immediately activated.

[0064] Monitor the channel status information on other available paths around the faulty intermediate node, and determine the backup transmission channel based on the channel status information;

[0065] The data stream passing through the faulty intermediate node is switched to the backup transmission channel, and the optimal data transmission path is updated to bypass the faulty intermediate node and continue to transmit data to the target robot.

[0066] Furthermore, to achieve the above objectives, this application also proposes a robot communication device based on multi-network communication, which includes:

[0067] A building module is used to build a multi-network communication architecture, which integrates at least two different types of network communication methods, including at least wired network communication, wireless network communication, Bluetooth communication, or mobile data network communication.

[0068] The acquisition module is used to acquire real-time status information of each network communication method in the multi-network communication architecture when the robot has a data transmission requirement. The real-time status information includes at least one of signal strength, connection status, network bandwidth, and network latency.

[0069] The allocation module is used to allocate the optimal network communication method to the data to be transmitted based on the real-time status information and the preset communication priority rules, wherein the preset communication priority rules are set according to the data type, the urgency of the data, and the quality of network communication.

[0070] The determination module is used to determine neighboring robots that are in the same communication environment as the current robot and can interact with data, through a neighbor discovery mechanism.

[0071] The acquisition module is also used to acquire the current state information of the neighboring robots, and dynamically adjust and optimize the data transmission path according to the current state information of the neighboring robots through a path planning strategy to determine the optimal data transmission path. The path planning strategy is based on opposition learning and the Black Widow algorithm.

[0072] The transmission module is used to transmit the data to be transmitted to the target robot through the optimal data transmission path and the optimal network communication method.

[0073] This application proposes one or more technical solutions to construct a multi-network communication architecture, which integrates at least two different types of network communication methods, including wired network communication, wireless network communication, Bluetooth communication, or mobile data network communication. When the current robot has a data transmission requirement, the real-time status information of each network communication method in the multi-network communication architecture is obtained. The real-time status information includes at least one of signal strength, connection status, network bandwidth, and network latency. Based on the real-time status information and preset communication priority rules, the optimal network communication method is allocated for the data to be transmitted. The preset communication priority rules are set according to data type, data urgency, and network communication quality. Through a neighbor discovery mechanism, neighboring robots that are in the same communication environment as the current robot and can interact with data are identified. The current status information of the neighboring robots is obtained, and the data transmission path is dynamically adjusted and optimized based on the current status information of the neighboring robots using a path planning strategy to determine the optimal data transmission path. The path planning strategy is based on adversarial learning and the Black Widow algorithm. The data to be transmitted is transmitted to the target robot through the optimal data transmission path and the optimal network communication method. By setting multiple network communication priorities, the utilization efficiency of network resources is effectively improved. Furthermore, by introducing a path planning strategy based on adversarial learning and the Black Widow algorithm, the communication efficiency and stability of the robot are effectively enhanced. Attached Figure Description

[0074] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0075] To more clearly illustrate the technical solutions in the embodiments of this application 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.

[0076] Figure 1 This is a flowchart illustrating an embodiment of the robot communication method based on multi-network communication in this application.

[0077] Figure 2 This is a schematic diagram of the module structure of a robot communication device based on multi-network communication according to an embodiment of this application.

[0078] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0079] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application.

[0080] To better understand the technical solution of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.

[0081] This application provides a solution that effectively improves the utilization efficiency of network resources by setting multiple network communication priorities, and effectively improves the communication efficiency and stability of the robot by introducing a path planning strategy based on opposition learning and the Black Widow algorithm.

[0082] It should be noted that the executing entity in this embodiment can be a computing service device with data processing, network communication, and program execution functions, such as a tablet computer, personal computer, or mobile phone, or an electronic device capable of performing the above functions, a robot communication device based on multi-network communication, etc. The following description uses a robot communication device based on multi-network communication as an example to illustrate this embodiment and the subsequent embodiments.

[0083] Based on this, embodiments of this application provide a robot communication method based on multi-network communication, referring to... Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of the robot communication method based on multi-network communication of this application.

[0084] In this embodiment, the robot communication method based on multi-network communication includes steps S10~S60:

[0085] Step S10: Construct a multi-network communication architecture, which is used to integrate at least two different types of network communication methods, including at least wired network communication, wireless network communication, Bluetooth communication, or mobile data network communication.

[0086] It should be noted that a multi-network communication architecture refers to an intelligent system that integrates multiple network communication methods to automatically select the optimal communication path and method based on environmental changes and communication needs. This architecture not only enhances the flexibility and reliability of communication but also enables rapid switching to other available networks when a single network fails or experiences performance degradation, ensuring the continuity and stability of data transmission.

[0087] In this embodiment, the multi-network communication architecture is used to integrate at least two different types of network communication methods, including any combination of two or more of wired network communication, wireless network communication, Bluetooth communication, or mobile data network communication. Among them, wired network communication is such as USB-Net network, wireless network communication is such as WiFi network or LoRa network, Bluetooth communication is communication via Bluetooth network, and mobile data network communication is such as 5G / 4G cellular network. In this embodiment, different network communication methods have different priorities, and the communication priorities from high to low are WiFi network, 5G / 4G cellular network, LoRa network, Bluetooth or USB-Net network.

[0088] In practical implementation, building a multi-network communication architecture ensures that the robot has stable and reliable communication capabilities in different environments. For example, in an indoor environment, the robot can use both wired and wireless network communication simultaneously. When the wired network fails, it can seamlessly switch to the wireless network to ensure communication continuity. In an outdoor environment, the robot can use mobile data network communication and Bluetooth communication. When the mobile data network signal is weak, it can use Bluetooth communication to interact with other nearby devices, thereby maintaining normal communication functions.

[0089] Step S20: When the robot has a data transmission requirement, obtain the real-time status information of each network communication method in the multi-network communication architecture. The real-time status information includes at least one of signal strength, connection status, network bandwidth, and network latency.

[0090] It should be noted that when the robot has data transmission needs, a status monitoring mechanism is quickly activated to perform a comprehensive and real-time status assessment of each network communication method in the multi-network communication architecture, including but not limited to signal strength, connection status, network bandwidth, and network latency. By monitoring the signal strength of each network communication method in real time, the quality of the current network coverage can be determined, avoiding data transmission interruptions caused by weak signals; by monitoring the connection status, network disconnections or abnormal connections can be detected in a timely manner, and recovery measures can be taken quickly; monitoring network bandwidth helps to rationally allocate data transmission volume and prevent transmission delays caused by insufficient bandwidth; and monitoring network latency ensures the real-time nature of data transmission, meeting the robot's need for instant communication.

[0091] Understandably, real-time status information such as signal strength, connection status, network bandwidth, and network latency can be obtained through the built-in network monitoring module or third-party network monitoring tools and fed back to the robot communication system in real time so that the system can make optimal communication decisions based on this information.

[0092] Step S30: Based on the real-time status information and the preset communication priority rules, allocate the optimal network communication method for the data to be transmitted, wherein the preset communication priority rules are set according to the data type, the urgency of the data, and the quality of network communication.

[0093] It should be noted that the preset communication priority rules are set based on data type, data urgency, and network communication quality. Data to be transmitted is categorized into at least one of the following based on data type: streaming media data, mission-critical data, or network self-healing instruction data. This allows for categorized data transmission. Streaming media data refers to video and high-volume data; mission-critical data refers to critical operational data; and network self-healing instruction data refers to operational instructions that restore network data. Selecting the appropriate network based on the characteristics and needs of different data types can improve network resource utilization efficiency and reduce communication costs. Data to be transmitted is categorized into at least one of the following based on urgency: urgent data, important data, ordinary data, or data that can be delayed. Network communication quality is categorized into at least one of the following based on network communication quality: high-quality communication, good communication, general communication, or poor communication. A comprehensive judgment is made based on the real-time status information of each network communication method, combined with the preset communication priority rules considering data type, data urgency, and network communication quality, to accurately allocate the optimal network communication method for the data to be transmitted.

[0094] Understandably, after determining the optimal network communication method, the robot needs to access the corresponding network communication channel. When selecting a WiFi network, it automatically adapts and polls for access to pre-stored service set identifiers (SSIDs), prioritizing access to WiFi networks with signal strength meeting a preset threshold. In this embodiment, the robot pre-stores more than 20 WiFi SSIDs and their corresponding passwords. When accessing a WiFi network, the robot sequentially tries these pre-stored WiFi SSIDs and automatically detects the signal strength of each network. When a WiFi network's signal strength meets the preset threshold, the robot automatically connects to that network to transmit video and high-bandwidth data, ensuring stable and high-speed data transmission. If the signal strength of the currently accessed WiFi network is lower than the preset threshold, it automatically polls for access to the next WiFi network with a signal strength meeting the preset threshold. By setting multiple network communication priorities and combining automatic adaptation and polling access mechanisms, the robot can prioritize WiFi networks and automatically switch between multiple WiFi networks, ensuring stable transmission of video and high-bandwidth data while improving the utilization rate of WiFi networks.

[0095] If all pre-stored WiFi networks fail to meet signal strength requirements, the robot will select other available network communication methods according to preset communication priority rules. For example, it may switch to a 5G or 4G cellular network to transmit critical operational data. When neither cellular nor WiFi networks can guarantee communication, it will switch to a LoRa network to transmit operational commands to restore network data. When none of the above networks can meet communication needs, it will initiate Bluetooth or USB-Net communication, temporarily borrowing a network-connected mobile phone to ensure robot communication. This multi-layered communication guarantee mechanism significantly improves the reliability and adaptability of robot communication.

[0096] In one feasible implementation, step S30 may include: determining network quality feature vectors for each network communication method based on the real-time status information; obtaining data type feature vectors and urgency feature vectors for the data to be transmitted, wherein the data type feature vectors consist of semantic level, data volume coefficient, and a preset tolerable packet loss rate, and the urgency feature vectors consist of data timeliness requirements, queued delay, and data urgency; determining static priority components for each network communication method based on the data type feature vectors and urgency feature vectors; determining real-time scores for each network communication method based on the network quality feature vectors; determining dynamic priority components for each network communication method based on the static priority components and the real-time scores; determining the priority difference between any two network communication methods based on the dynamic priority components of each network communication method; and selecting the network communication method with the highest dynamic priority component as the optimal network communication method when all priority differences are greater than or equal to a preset threshold.

[0097] It should be noted that signal strength, connection status, network bandwidth, and network latency are determined based on real-time status information, and then a network quality feature vector is constructed. This feature vector consists of signal strength, network bandwidth, round-trip time, latency jitter, packet loss rate, and network tariff weight. These parameters comprehensively reflect the real-time performance and quality of network communication. The network tariff weight C is greater than 0 in cellular mobile data networks and equal to 0 in Wi-Fi, Bluetooth, or wired networks.

[0098] The data type feature vector of the data to be transmitted consists of a semantic level, a data volume coefficient, and a preset tolerable packet loss rate. The semantic level α is determined as follows: α = 2 indicates operable commands to recover network data, α = 1 indicates critical operational data, and α = 0 indicates large data such as video or files. The data volume coefficient β is determined based on the data packet size: β is 0.1 when the data packet is less than 1KB, β is 0.5 when the data packet is between 1KB and 10KB, and β is 1 when the data packet is greater than 10KB. The preset tolerable packet loss rate γ is set according to the data type and application scenario. If the actual transmission packet loss rate is greater than 2γ, the corresponding network dynamic priority component is immediately set to zero, achieving circuit breaking.

[0099] The urgency feature vector of the data to be transmitted consists of data timeliness requirements, queued delay, and data urgency. Data timeliness requirement δ reflects the time window from data generation to when it must be processed. For data with high real-time requirements, such as robot control commands, the δ value is small; while for non-real-time data, such as log records, the δ value is large. Queued delay ε represents the time data waits in the transmission queue; the larger this value, the higher the urgency of the data transmission. Data urgency ζ is determined based on the business importance of the data; for example, security monitoring data is usually considered highly urgent.

[0100] The determination of static priority components comprehensively considers data type feature vectors and urgency feature vectors, and is obtained through weighted summation. The weights of each feature are updated online using the entropy weight method according to actual application scenarios and needs, with an update cycle of 1 minute to 24 hours. The real-time score is calculated based on the network quality feature vector. That is, based on the specific values ​​of parameters such as signal strength, network bandwidth, round-trip time, latency jitter, packet loss rate, and network cost weight, a specific scoring algorithm, such as linear weighting or analytic hierarchy process, is used to calculate a comprehensive score for each network communication method. This score represents the communication quality of the network at the current moment.

[0101] The calculation of dynamic priority components involves fusing static priority components with real-time scores, as shown in the following formula:

[0102] Pdynamic,k=Pstatic·(1+Slink,k)

[0103] Where Pdynamic,k is the dynamic priority component of communication network mode k, Pstatic is the static priority component, and Slink,k is the real-time score of communication network mode k.

[0104] After calculating the dynamic priority components, the priority ranking of each network communication method can be determined by comparing the magnitudes of their dynamic priority components. Based on this ranking, when selecting the optimal network communication method for data transmission, the method with the highest dynamic priority component is chosen first. However, to ensure accuracy, the priority difference between any two network communication methods needs to be further evaluated. Only when the priority differences are both greater than or equal to a preset threshold is the network communication method with the highest dynamic priority component selected as the optimal method. This approach effectively avoids frequent network switching due to small priority differences, reduces unnecessary network switching overhead, and improves the stability and efficiency of network communication.

[0105] In one feasible implementation, after determining the priority difference between any two network communication methods based on the dynamic priority components of each network communication method, the method further includes: if the priority difference is less than a preset threshold, executing a fragmented dual-transmission strategy and treating both network communication methods corresponding to the priority difference as candidate network communication methods; fragmenting the data to be transmitted according to a preset fragmentation rule to obtain multiple data fragments; simultaneously transmitting the corresponding data fragments through the candidate network communication methods, and monitoring the transmission status of the two candidate network communication methods in real time during transmission; if one of the candidate network communication methods experiences a transmission anomaly, immediately switching the remaining untransmitted data fragments to another candidate network communication method that is transmitting normally for continued transmission; and after all data fragments have been successfully transmitted to the target robot, reassembling the received data fragments at the target robot end to recover the complete data to be transmitted.

[0106] It should be noted that if the priority difference is less than the preset threshold, it indicates that the priorities of the various network communication methods are relatively close. In this case, the network with the highest dynamic priority component cannot be simply selected. Directly selecting it may lead to transmission interruptions or reduced efficiency due to minor fluctuations in network status. Therefore, implementing a fragmented dual-transmission strategy becomes a more prudent choice. This strategy first meticulously fragments the data to be transmitted according to preset fragmentation rules, ensuring that each data fragment contains a complete portion of the information, while facilitating independent transmission and reassembly. Subsequently, the corresponding data fragments are transmitted simultaneously through two candidate network communication methods with similar priorities.

[0107] During transmission, the system monitors the transmission status of both networks in real time, including but not limited to key indicators such as signal strength, transmission rate, and packet loss rate. If an anomaly is detected in one network communication method, such as a sudden weakening of the signal or a significant drop in transmission rate, the system immediately responds by rapidly switching the remaining untransmitted data fragments to the other network communication method that is still transmitting normally, ensuring the continuity and integrity of data transmission. Finally, once all data fragments have successfully arrived at the target robot, the target robot will accurately reassemble the received data fragments according to preset reassembly rules, thereby recovering complete information that is completely consistent with the original data to be transmitted. This process not only improves the reliability of data transmission but also effectively utilizes network resources, providing strong support for robot communication in complex and ever-changing network environments.

[0108] Understandably, preset fragmentation rules can be flexibly set according to actual needs. For example, data can be fragmented equally according to its size, dividing the data to be transmitted into multiple data fragments of similar size. This ensures that the load on each fragment during transmission is relatively balanced. Alternatively, fragmentation can be based on the logical structure of the data. For instance, data with a specific hierarchical or modular structure can be split into different parts as data fragments according to its inherent logic, so that the original data can be more accurately recovered during reassembly. At the same time, preset fragmentation rules must also consider the security and integrity requirements of the data, ensuring that no information is lost or erroneous during the transmission and reassembly of the fragmented data.

[0109] Step S40: Through the neighbor discovery mechanism, identify neighboring robots that are in the same communication environment as the current robot and can interact with data.

[0110] It's important to note that the neighbor discovery mechanism is a key technology ensuring that robots can accurately identify and connect to interactive neighbor robots in complex and ever-changing network environments. The neighbor discovery mechanism comprises two crucial stages: same-domain determination and interaction feasibility assessment. Same-domain determination primarily involves analyzing the network domain information of the robot to determine whether other nearby robots are in the same network domain. For example, it can compare key network parameters such as the network identifier and IP address range the robot accesses. When these parameters meet preset same-domain conditions, it can be preliminarily determined that the robot in question is in the same communication environment as the current robot.

[0111] The interaction feasibility assessment, building upon the existing domain identification, further evaluates the feasibility of data interaction with other robots in the same domain. This assessment process comprehensively considers multiple factors, such as whether network bandwidth is sufficient, whether network latency is within acceptable limits, and whether the communication protocols of both robots are compatible. Only when all these factors meet the preset interaction conditions will the robot be identified as a neighboring robot with whom data interaction can occur.

[0112] After accurately identifying neighboring robots through the neighbor discovery mechanism, the current robot can establish an effective data interaction channel with these neighboring robots. This not only helps improve the overall working efficiency of the robot group, but also enhances the stability and reliability of the robot system in complex network environments.

[0113] In one feasible implementation, step S40 may include: acquiring feature data of the current robot's communication environment and mapping the feature data to a predefined environmental feature space to generate a communication environment feature vector for the current robot; calculating the similarity between the current robot's communication environment feature vector and the communication environment feature vector of the robot to be judged to obtain a communication environment similarity; determining whether the current robot and the robot to be judged are in the same domain based on the communication environment similarity to obtain a same domain determination result; determining the robot to be judged as a same domain robot if the same domain determination result indicates that the current robot and the robot to be judged are in the same communication domain; evaluating the interaction feasibility of the same domain robots based on a robot interaction feasibility scoring model to obtain an interaction feasibility score for each same domain robot, wherein the robot interaction feasibility scoring model is constructed based on the communication capability, remaining battery power, task load, historical interaction success rate, and current network environment stability of the same domain robots; sorting the same domain robots from high to low according to the interaction feasibility scores, and selecting the same domain robots with interaction feasibility scores higher than a preset scoring threshold as neighbor robots that can perform data interaction.

[0114] It's important to note that the characteristic data of the robot's current communication environment refers to various data information that reflects the characteristics of the network environment in which the robot is currently located. For example, network signal strength reflects the strength of the current network signal, directly affecting the stability and speed of data transmission; network bandwidth reflects the amount of data the network can transmit per unit time, which is crucial for large-scale data exchange; network latency is the time required for data to travel from the sender to the receiver; excessive latency can lead to poor real-time communication; and network packet loss rate indicates the proportion of data packets lost during network transmission; a high packet loss rate will seriously affect the integrity and accuracy of the data. These characteristic data, taken together, can comprehensively and accurately depict the current communication environment of the robot.

[0115] Mapping feature data to a predefined environmental feature space transforms complex real-world environmental information into numerical vectors that can be mathematically calculated and analyzed, facilitating subsequent operations such as similarity calculations. The predefined environmental feature space is a mathematical model space pre-constructed based on the actual communication environment and requirements, specifying the representation and dimensional structure of the feature data within it.

[0116] After generating the communication environment feature vector of the current robot, the similarity between the two robots' communication environments can be calculated by comparing it with the communication environment feature vector of the robot to be judged. The similarity calculation can use weighted cosine similarity, as shown in the following formula:

[0117]

[0118] in, Communication environment feature vector and Weighted cosine similarity between them , These are the feature vectors of the communication environment. and The i-th component, Let be the weights of the i features.

[0119] Based on the calculated communication environment similarity, the current robot and the robot to be judged are determined to be in the same communication domain. When the similarity reaches a certain level and meets the preset same-domain conditions, it can be determined that the two are in the same communication domain. If the same-domain determination result indicates that they are in the same communication domain, the robot to be judged is further confirmed as a same-domain robot.

[0120] Subsequently, the feasibility of interaction between robots in the same domain was assessed based on a robot interaction feasibility scoring model. This model comprehensively considers the robot's communication capabilities, such as data transmission rate and maximum transmission distance; remaining battery power, which relates to the robot's ability to continuously engage in data interaction; task load, as excessive task load may affect the robot's interaction performance; historical interaction success rate, reflecting the stability and reliability of the robot's past interactions with other robots; and the stability of the current network environment, including network fluctuations and signal interference. By constructing a scoring model that integrates these factors, the feasibility of data interaction between robots in the same domain can be assessed more comprehensively and accurately.

[0121] Based on the interaction feasibility scoring model, an interaction feasibility score is calculated for each robot in the same domain, as shown in the following formula:

[0122]

[0123] in, Score the feasibility of interaction between robot i and the same domain at time t. For the dynamic weights associated with the j-th factor, For the rating between the i-th robot and the j-th factor, N is the time decay factor. i Let i be the set of robots in the neighborhood of robot i. and Robot's communication environment feature vector The environmental distance between robots, This is an evaluation function for emergency response capabilities, used based on communication capabilities. Remaining battery power Historical interaction success rate A comprehensive emergency response score is calculated.

[0124] Robots within the same domain are ranked from highest to lowest based on their interaction feasibility scores. Robots with interaction feasibility scores above a preset threshold are selected as neighboring robots capable of data interaction. This selection method ensures that the current robot prioritizes establishing connections with neighboring robots that offer better interaction conditions, thereby improving the efficiency and reliability of data transmission. The preset threshold can be flexibly set according to actual communication needs and environmental conditions. When the environment is complex or high communication quality requirements are high, the threshold can be appropriately increased to ensure that the selected neighboring robots have sufficient interaction capabilities.

[0125] Step S50: Obtain the current state information of the neighboring robots, and dynamically adjust and optimize the data transmission path according to the current state information of the neighboring robots through a path planning strategy to determine the optimal data transmission path. The path planning strategy is based on adversarial learning and the Black Widow algorithm.

[0126] It should be noted that by combining opposition learning and the Black Widow algorithm, the convergence speed, optimization accuracy, and global search capability can be improved, thereby constructing an efficient and adaptive data transmission path planning strategy, which effectively improves the efficiency and accuracy of determining the optimal data transmission path.

[0127] Oppositional learning algorithms can expand the search space during path planning by generating solutions that are opposite to the current solution, thus avoiding getting trapped in local optima and improving global search capabilities. This method of generating opposing solutions allows the algorithm to consider multiple directions simultaneously during exploration, increasing the likelihood of finding a better path.

[0128] The Black Widow algorithm simulates the predation and survival strategies of the black widow spider, seeking optimal solutions through swarm intelligence and adaptive adjustment. In path planning, the Black Widow algorithm dynamically adjusts the weights and directions of the path based on the current path status and the state information of neighboring robots, guiding data transmission towards a better path. Simultaneously, the algorithm exhibits good robustness, maintaining stable performance in complex and ever-changing network environments. Combining the opposition learning algorithm and the Black Widow algorithm fully leverages the advantages of both. The opposition learning algorithm expands the search space, avoiding the problem of local optima; while the Black Widow algorithm further optimizes the path planning results through swarm intelligence and adaptive adjustment. This combined approach makes the path planning strategy more efficient and adaptive, dynamically adjusting the data transmission path according to the actual situation to ensure that data is transmitted to the target robot in the optimal way.

[0129] Step S60: Transmit the data to be transmitted to the target robot through the optimal data transmission path and the optimal network communication method.

[0130] It should be noted that different network communication methods involve different channels. After determining the optimal network communication method, it is necessary to further select the optimal transmission channel as the specific channel for data transmission. Then, the data to be transmitted is encapsulated into data packets suitable for transmission over the network. The encapsulated data packets are then transmitted according to the optimal data transmission path and the channel specified by the optimal network communication method.

[0131] During transmission, the data transmission status is monitored in real time, including data transmission rate and data packet integrity. If any abnormalities are detected, such as a sudden and significant drop in data transmission rate or data packet loss, the corresponding response mechanism will be triggered promptly.

[0132] Once data is successfully transmitted to the target robot, the target robot verifies the received data. Verification includes data integrity checks, comparing the received data with key characteristics of the original data, such as data length and checksum, to determine if the data was transmitted intact. It also performs data accuracy checks, verifying that the received data conforms to the expected content and structure according to preset data formats and logical rules, thus ensuring data accuracy. If the verification finds a problem, the target robot will send an error message to the sending robot. Upon receiving the feedback, the sending robot will take appropriate action based on the error type. For example, if the data is incomplete due to lost data packets, the sending robot will resend the lost packets; if the data has an accuracy issue, the sending robot will check the data generation and processing stages, correct the errors, and resend. If the verification passes, the target robot will further process and apply the received data to complete the data interaction task between the robots.

[0133] In one feasible implementation, step S60 may include: obtaining a list of available channels corresponding to the optimal network communication method; monitoring the channels in the list in real time to obtain the current occupancy status and signal quality parameters of each channel; allocating the optimal transmission channel for the data to be transmitted using a dynamic channel allocation strategy based on the current occupancy status and signal quality parameters of each channel, wherein the dynamic channel allocation strategy is set according to the channel idle time, signal interference intensity, and historical transmission success rate; encapsulating the data to be transmitted according to a preset data encapsulation format, adding channel allocation information and target robot identification information to the header of the encapsulated data to generate a data packet to be sent; and transmitting the data packet to be sent to the target robot through the optimal transmission channel along the optimal data transmission path.

[0134] It's important to note that obtaining the list of available channels corresponding to the optimal network communication method is crucial for comprehensively understanding all channel resources available for data transmission in the current network environment. Real-time monitoring of the available channels allows for timely understanding of their current occupancy status, such as whether the channel is being used by other data transmission tasks and the extent of occupancy. Simultaneously, signal quality parameters, such as signal strength and signal-to-noise ratio, can be obtained, directly reflecting the reliability and stability of data transmission through the channel.

[0135] Based on the current occupancy status and signal quality parameters of each channel, it is crucial to employ a dynamic channel allocation strategy to assign the optimal transmission channel for the data to be transmitted. The dynamic channel allocation strategy is set based on channel idle time, signal interference intensity, and historical transmission success rate. A longer channel idle time indicates a lower probability that the channel is currently occupied, making it more suitable for data transmission; lower signal interference intensity reduces the likelihood of external interference during data transmission, ensuring accurate data transmission; and a higher historical transmission success rate indicates that the channel has performed well in previous data transmission tasks and can reliably complete the data transmission mission. By comprehensively considering these factors, the optimal transmission channel best suited for the current data transmission can be selected.

[0136] Encapsulating the data to be transmitted according to a preset data encapsulation format is to ensure that the data can be transmitted in a standardized and efficient manner over the network. Channel allocation information and target robot identification information are added to the header of the encapsulated data to generate a data packet to be sent. The channel allocation information allows the target robot to accurately know which channel the data was transmitted through after receiving it, enabling subsequent processing; the target robot identification information ensures that the data packet is accurately delivered to the target robot, avoiding data transmission errors.

[0137] Finally, the data packets to be sent are transmitted to the target robot through the optimal transmission channel along the optimal data transmission path. This ensures that the data reaches the target robot in the most efficient and reliable way, completing the entire data interaction process.

[0138] In one feasible implementation, after transmitting the data packet to be sent to the target robot via the optimal transmission channel along the optimal data transmission path, the method further includes: during transmission, if signal interference is detected to be enhanced or the optimal transmission channel is occupied by other devices, a channel contention backoff mechanism is immediately activated; the backoff time window is dynamically adjusted according to the contention level of the optimal transmission channel and a preset backoff algorithm, data transmission is suspended within the backoff time window, and the status change of the optimal transmission channel is continuously monitored; when the backoff time window ends, if the optimal transmission channel becomes available again, data transmission is restarted; if the optimal transmission channel is still unavailable, the dynamic channel allocation strategy is re-executed, and a suboptimal transmission channel is selected from the available channel list to continue transmitting the data packet to be sent; if there are no other available channels in the available channel list, data transmission is suspended, and a channel list expiration repair mechanism is triggered to search and update the available channel list again, resulting in an updated available channel list; a new optimal transmission channel is selected from the updated available channel list to transmit the data packet to be sent.

[0139] It should be noted that when increased signal interference or channel contention is detected, the activation of the channel contention backoff mechanism can effectively prevent continuous data transmission on unreliable channels, reducing the risk of data loss and errors.

[0140] The pre-defined backoff algorithm dynamically adjusts the backoff time window based on the channel contention level. This design considers both the real-time channel status and the flexibility of data transmission. Data transmission is paused within the backoff time window, and the channel status is continuously monitored, ensuring that data can be immediately retransmitted when the channel becomes available again, thereby improving the efficiency and reliability of data transmission.

[0141] If the channel has not been restored by the end of the backoff time window, the dynamic channel allocation strategy is re-executed to select the suboptimal transmission channel, demonstrating the system's adaptive capability. When there are no other available channels in the available channel list, data transmission is paused and the channel list expiration repair mechanism is triggered. The available channel list is then searched and updated again, ensuring the system's stability and recovery capability in extreme situations. Finally, a new optimal transmission channel is selected from the updated available channel list to transmit data, ensuring the successful completion of the data exchange task.

[0142] In one feasible implementation, after transmitting the data packet to be sent to the target robot via the optimal transmission channel along the optimal data transmission path, the method further includes: during data transmission, if a fault or reduced communication capability is detected in an intermediate node on the optimal data transmission path, the intermediate node is designated as a faulty intermediate node, and a multi-channel virtual monitoring mechanism is immediately activated; channel status information on other available paths around the faulty intermediate node is monitored, and a backup transmission channel is determined based on the channel status information; the data stream passing through the faulty intermediate node is switched to the backup transmission channel, and the optimal data transmission path is updated to bypass the faulty intermediate node and continue transmitting data to the target robot.

[0143] It's important to note that during data transmission, intermediate node failures or decreased communication capabilities are common problems. If not addressed promptly, these can lead to data transmission interruptions or reduced efficiency. The activation of a multi-channel virtual monitoring mechanism allows for rapid detection of channel status on other available paths around the failed intermediate node. By monitoring this channel status information, backup transmission channels can be accurately identified, ensuring a quick alternative solution is found when the original path encounters problems. Switching the data stream passing through the failed intermediate node to the backup transmission channel and updating the optimal data transmission path, bypassing the failed intermediate node to continue data transmission, effectively avoids the failure of the entire data transmission task due to a single node failure. This improves the stability and reliability of data transmission, ensuring the smooth operation of data interaction tasks between robots.

[0144] As is understandable, virtual monitoring refers to simulating real-time monitoring of multiple channels without actually occupying or connecting to all possible channels. This mechanism enables rapid response to changes in the network environment without significantly increasing hardware costs. When the channel status around a faulty intermediate node changes, the virtual monitoring mechanism can instantly capture these changes. For example, a channel with poor signal quality may suddenly become available due to the disappearance of surrounding interference sources, or a channel may be temporarily congested due to other low-priority tasks but is expected to be released soon.

[0145] Based on this real-time monitored channel status information, the most suitable option for backup transmission channels can be quickly evaluated and determined. This backup channel may not be spatially adjacent to the faulty intermediate node, but it can bypass the fault point with the fewest hops and the highest efficiency through the network topology. Once the backup transmission channel is determined, a data stream switching operation is immediately performed. This process needs to ensure that the integrity and order of the data are not compromised, which is usually achieved by adding sequence numbers and checksum information to the data packets.

[0146] Meanwhile, updating the optimal data transmission path involves not only replacing nodes along the path but also recalculating key performance indicators such as total latency and bandwidth utilization to ensure that the updated path still meets the real-time and reliability requirements of data transmission. In this way, even in the face of uncertainties and dynamic changes in the network, data interaction between robots can maintain a high degree of stability and efficiency.

[0147] This embodiment provides a robot communication method based on multi-network communication. A multi-network communication architecture is constructed to integrate at least two different types of network communication methods, including wired network communication, wireless network communication, Bluetooth communication, or mobile data network communication. When the robot has a data transmission requirement, the real-time status information of each network communication method in the multi-network communication architecture is obtained. This real-time status information includes at least one of signal strength, connection status, network bandwidth, and network latency. Based on the real-time status information and preset communication priority rules, the optimal network communication method is allocated to the data to be transmitted. The preset communication priority rules are set according to data type, data urgency, and network communication quality. A neighbor discovery mechanism is used to identify neighboring robots that are in the same communication environment and can interact with the current robot. The current status information of the neighboring robots is obtained, and the data transmission path is dynamically adjusted and optimized based on the current status information using a path planning strategy to determine the optimal data transmission path. This path planning strategy is based on adversarial learning and the Black Widow algorithm. The data to be transmitted is then transmitted to the target robot through the optimal data transmission path and the optimal network communication method. By setting multiple network communication priorities, the utilization efficiency of network resources is effectively improved. Furthermore, by introducing a path planning strategy based on adversarial learning and the Black Widow algorithm, the communication efficiency and stability of the robot are effectively enhanced.

[0148] Based on the first embodiment of this application, in the second embodiment of this application, the content that is the same as or similar to that in the first embodiment described above can be referred to the above description and will not be repeated hereafter. Based on this, step S50 includes steps S501 to S506:

[0149] Step S501: Obtain the current status information of the neighboring robot, which includes location information, communication capability parameters, current task load, and network environment stability indicators.

[0150] It should be noted that current status information can be obtained through multiple means. Firstly, it can be obtained through periodic status information messages exchanged between robots. These messages contain data such as the location and communication capabilities proactively reported by neighboring robots. For example, a neighboring robot can, at preset time intervals, encapsulate its current geographical coordinates, maximum supported data transmission rate, number of tasks currently being executed, and signal strength fluctuations in its network environment into specific status information messages and send them to the current robot. Secondly, the current robot can also proactively initiate query requests to obtain the current status information of neighboring robots. Based on its own data transmission needs and its perception of the surrounding environment, the current robot sends a query command to a specific neighboring robot, requesting it to provide relevant status information. Upon receiving the query command, the neighboring robot quickly collects its own status data and replies to the current robot.

[0151] In addition, external positioning systems and management platforms can be used to assist in obtaining the current status information of neighboring robots. For example, the GPS system can be used to obtain the location information of neighboring robots more accurately; through a centralized management platform, the task load and network environment stability indicators of each robot can be monitored and managed in real time, and the current robot can obtain the required status information of neighboring robots from the management platform.

[0152] After obtaining the current state information of the neighboring robots, the current robot can use this information, combined with a path planning strategy based on adversarial learning and the Black Widow algorithm, to more accurately and efficiently adjust and optimize the data transmission path, thereby determining the optimal data transmission path and further improving the efficiency and stability of data transmission between robots.

[0153] Specifically, location information helps the current robot accurately determine the spatial location of neighboring robots, thereby planning the optimal data transmission path and reducing latency and packet loss during data transmission.

[0154] Communication capability parameters reflect the neighboring robot's ability to receive and send data, including maximum transmission rate and supported communication protocols. This information helps the current robot choose a communication method that matches the neighboring robot, thereby improving data transmission efficiency.

[0155] The current task load is a crucial indicator of whether the neighboring robot can process received data in a timely manner. If the neighboring robot's current task load is too heavy, it may lead to data transmission delays or untimely processing, affecting the communication performance of the entire robot system. Therefore, the current robot needs to dynamically adjust the timing and priority of data transmission based on the neighboring robot's task load to ensure that data is processed at the appropriate time.

[0156] Network environment stability metrics reflect the reliability and stability of the network environment in which neighboring robots operate. An unstable network environment can lead to data transmission interruptions or errors, affecting the quality of communication between robots. By obtaining network environment stability metrics, the current robot can choose neighboring robots with more stable network environments for data transmission, thereby improving the reliability and stability of data transmission.

[0157] Step S502: Using the current robot as the starting node, the target robot as the target node, and the neighboring robots as intermediate nodes, construct an initial data transmission path set based on the current state information.

[0158] It's important to note that when constructing the initial data transmission path set, the starting node is the current robot, which acts as the initiator of the data transmission, possessing the data to be sent and a clearly defined target robot. The target node is the robot that ultimately receives the data; the entire data transmission process aims to deliver the data accurately to this node. Neighboring robots, acting as intermediate nodes, serve as bridges and relays in the data transmission process.

[0159] Based on the acquired current status information of neighboring robots, including location information, communication capability parameters, current task load, and network environment stability indicators, the current robot can comprehensively evaluate the suitability of each neighboring robot as an intermediate node. Location information determines the spatial distribution of intermediate nodes, thus affecting the length and hop count of data transmission paths; communication capability parameters reflect the intermediate node's data processing capabilities, including transmission rate and protocol support; current task load relates to whether the intermediate node can process and forward data in a timely manner; and network environment stability indicators reflect the reliability of the network in which the intermediate node resides.

[0160] Based on these assessments, the robot can construct multiple possible initial data transmission paths, each containing a starting node, a series of intermediate nodes, and a target node. This set of initial paths provides the foundation for subsequent path optimization and selection, ensuring that the optimal data transmission path can be chosen, thereby improving the efficiency and stability of data transmission between robots.

[0161] Step S503: Use the opposition learning algorithm to evaluate each original path in the initial data transmission path set and generate an opposition path for each original path. The construction rule of the opposition path is to generate an alternative path with a different topology than the original path but connecting the starting node and the target node by changing the order of intermediate nodes or replacing some intermediate nodes.

[0162] It's important to note that the core idea of ​​the opposition learning algorithm lies in discovering potentially better solutions by exploring the solution space that contrasts with the initial solution. In the context of robot communication path planning, this means not only considering existing path combinations but also actively seeking new paths formed by adjusting the order of intermediate nodes or replacing some nodes. This strategy effectively avoids getting trapped in local optima and increases the likelihood of finding the globally optimal path.

[0163] Specifically, for each path in the initial set of data transmission paths, the algorithm analyzes its topology, identifying key intermediate nodes and their connection order. Then, according to preset rules, such as randomly swapping the positions of two non-adjacent intermediate nodes or replacing an intermediate node in the current path with another suitable neighboring robot, a corresponding opposing path is generated. These opposing paths, while maintaining the connection between the starting and target nodes, offer different transmission options than the original path, potentially resulting in lower latency, higher bandwidth utilization, or better network stability.

[0164] After generating opposing paths, the algorithm further evaluates the performance metrics of these new paths, including but not limited to total transmission delay, path reliability, energy consumption, and compliance with preset communication priority rules. During the evaluation process, the current state information of neighboring robots, such as location, communication capabilities, task load, and network environment stability, is comprehensively considered to ensure that the selected paths are both efficient and reliable in practical applications.

[0165] Through this process, the opposition learning algorithm not only expands the scope of path search, but also provides more comprehensive and accurate data support for subsequent path selection by comparing the performance of the initial path and the opposition path. Finally, combined with the black widow algorithm, an optimization algorithm that simulates the natural selection process, iteratively eliminates inferior solutions, retains and optimizes superior solutions, and further filters and optimizes to determine the optimal data transmission path, thereby significantly improving the efficiency and stability of data transmission between robots.

[0166] Step S504: Calculate the fitness value of each original path and its corresponding opposing path based on the path length, the cumulative value of node communication quality, and the overall stability coefficient of the path. The path length is obtained by weighted summation of the physical distance or logical hop count between nodes. The cumulative value of node communication quality is the product of the communication capability parameters of each node in the path. The overall stability coefficient of the path is calculated by integrating the network environment stability indicators of each node and using a weighted average method.

[0167] It should be noted that path length, calculated by weighted summation of physical distances or logical hop counts between nodes, accurately quantifies the total distance or number of hops required for data to travel from the originating node to the destination node. Shorter path lengths typically imply lower transmission latency and higher transmission efficiency, thus playing a crucial role in fitness evaluation.

[0168] The cumulative node communication quality value reflects the overall communication capability level of each node in the path. By multiplying the communication capability parameters of each node in the path, a cumulative value can be obtained. The larger this value, the stronger the communication capability of each node in the path, and the lower the risk of communication failures or data loss during data transmission. Therefore, the cumulative node communication quality value is also an important reference indicator in fitness evaluation.

[0169] The overall path stability coefficient comprehensively considers the network environment stability of each node in the path. By integrating the network environment stability indicators of each node and calculating using a weighted average method, a coefficient reflecting the overall stability of the path can be obtained. The higher this coefficient, the more stable the network environment of each node in the path, and the lower the probability of interruptions or errors during data transmission. Therefore, the overall path stability coefficient is also an indispensable factor in fitness evaluation.

[0170] Based on the three indicators mentioned above, the fitness value of each original path and its corresponding opposing path can be calculated. A higher fitness value indicates better performance in data transmission efficiency, communication quality, and stability, and makes it more likely to be the optimal data transmission path. This process allows for a more scientific and objective evaluation and selection of data transmission paths, thereby significantly improving the efficiency and stability of data transmission between robots. The formula for calculating the fitness value is:

[0171]

[0172] in, Let N be the fitness value of the path, and N be the total number of nodes in the path. Let be the physical distance from the i-th node to the (i+1)-th node in the path. Let be the stability weight of the i-th node. It is the communication capability parameter of the i-th node in the path. The stability index of the i-th node.

[0173] Step S505: If the fitness value of the opposing path is greater than the fitness value of the original path, then the original path is replaced according to the opposing path and the initial data transmission path set is updated to obtain the updated initial data transmission path set.

[0174] It should be noted that when the fitness value of the opposing path is greater than that of the original path, it means that the new path generated by changing the order of intermediate nodes or replacing some intermediate nodes performs better in terms of data transmission efficiency, communication quality, and stability. Therefore, replacing the original path with the opposing path is a reasonable choice, as it helps improve the overall performance of data transmission.

[0175] During the replacement process, the algorithm accurately identifies the original path that needs to be replaced and adds the corresponding opposing path to the initial data transmission path set, while removing the original path being replaced. This operation ensures that the initial data transmission path set always contains the currently known optimal path options, providing a better foundation for subsequent path selection.

[0176] After obtaining the updated initial data transmission path set, the algorithm continues with the next round of evaluation and optimization. This includes generating new opposing paths again using the opposition learning algorithm, calculating the fitness value of each path, and replacing and updating paths based on the fitness value. Through this iterative approach, the algorithm can continuously approach the global optimum, thereby significantly improving the efficiency and stability of data transmission between robots. The algorithm continues until it reaches a preset iteration limit, or until no better path solution is found in several consecutive iterations. At this point, the algorithm can be considered to have converged to a relatively stable state. In this relatively stable state, the updated initial data transmission path set obtained by the algorithm already contains the better data transmission paths that can be found under the current network environment and communication conditions.

[0177] Step S506: Optimize the updated initial data transmission path set using the Black Widow algorithm to obtain the optimal data transmission path.

[0178] It should be noted that the Black Widow algorithm, as an optimization algorithm that simulates the natural selection process, exhibits unique advantages in robot communication path planning. The core mechanism of this algorithm lies in iteratively eliminating path solutions with lower fitness values ​​while retaining and optimizing those with higher fitness values, thereby gradually approaching the global optimum.

[0179] When applying the Black Widow algorithm to optimize the updated initial set of data transmission paths, the algorithm first evaluates the fitness of each path in the set. Then, based on the fitness values, the algorithm divides the path set into a set of high-quality solutions and a set of low-quality solutions.

[0180] During the iteration process, the Black Widow algorithm employs specific selection strategies, such as roulette wheel selection or tournament selection, to choose a subset of paths from the high-quality solution set as parents. New child paths are then generated through crossover and mutation operations. Crossover aims to generate paths with new topologies by swapping some nodes or their order in the parent paths; mutation introduces new genetic information by randomly changing a node or its order in the path, thereby increasing the diversity of path solutions. After generating child paths, the algorithm re-evaluates their fitness, comparing it with the fitness values ​​of the parent paths and paths in the low-quality solution set. If the fitness value of a child path is higher than that of the parent path or some paths in the low-quality solution set, it is added to the high-quality solution set, and the eliminated paths are removed. This process is repeated until the algorithm reaches a preset maximum number of iterations or meets other convergence conditions.

[0181] Through optimization using the Black Widow algorithm, the path solutions in the updated initial data transmission path set are continuously optimized and improved. The final optimal data transmission path will have lower transmission latency, higher bandwidth utilization, and better network stability. Determining this optimal path provides a more reliable and efficient communication guarantee for data transmission between robots.

[0182] In one feasible implementation, step S506 may include: treating each path in the updated initial data transmission path set as a spider individual to construct an initial spider population; calculating the fitness value of each spider individual in the initial spider population according to a preset fitness function, wherein the fitness function is constructed based on path length, cumulative node communication quality, overall path stability coefficient, and data transmission timeliness requirements; selecting high-quality spider individuals with fitness values ​​higher than a preset threshold based on the fitness values ​​of each spider individual in the initial spider population to form a high-quality spider subpopulation; and performing a breeding operation on the high-quality spider subpopulation to generate new spider individuals through crossover and mutation, wherein the crossover operation adopts single-point crossover or multi-point crossover. The mutation operation randomly changes some intermediate nodes in the path; the new spider individuals are merged with the remaining unselected spider individuals to form a new spider population; it is determined whether the new spider population meets the preset termination conditions, which include reaching the maximum number of iterations, fitness value convergence, or finding the optimal path that meets the timeliness requirements of data transmission; if the termination conditions are met, the path corresponding to the spider individual with the highest fitness value in the new spider population is selected as the optimal data transmission path; if the termination conditions are not met, the process returns to the step of calculating the fitness value of each spider individual in the new spider population according to the preset fitness function, and continues iterative optimization until the preset termination conditions are met, thus obtaining the optimal data transmission path.

[0183] It's worth noting that treating each path as an individual spider and constructing an initial spider population provides an intuitive and effective model framework for the Black Widow algorithm. This modeling approach allows the path optimization process to draw upon the natural selection mechanism in biological evolution, thus approximating the optimal solution more efficiently.

[0184] The preset fitness function comprehensively considers multiple dimensions, including path length, cumulative node communication quality, overall path stability coefficient, and data transmission timeliness requirements, ensuring the comprehensiveness and accuracy of fitness evaluation. The formula for the preset fitness function is:

[0185]

[0186] in, The fitness value, T, for the path corresponding to each individual spider. a T represents the actual transmission time of the path. r This refers to the timeliness requirements for data transmission along the path.

[0187] In the process of selecting high-quality spider individuals, a preset threshold ensures that only those paths that perform well in fitness assessment can be included in the high-quality spider subpopulation, thereby ensuring the efficiency and effectiveness of the optimization process, avoiding the algorithm from getting stuck in local optima, and improving the ability of global search.

[0188] Crossover and mutation in the reproductive operations are the core techniques used by the Black Widow algorithm to achieve path optimization. The crossover operation generates paths with new topological structures by exchanging some nodes or the order of nodes in the parent path, increasing the diversity of path solutions. The mutation operation, on the other hand, introduces new genetic information by randomly changing a node or the order of nodes in the path, helping the algorithm to escape local optima and explore a wider solution space.

[0189] The new spider individuals are merged with the remaining unselected spider individuals to form a new spider population. This step ensures that the algorithm can make full use of the high-quality path information found during the iteration process, while retaining a certain degree of diversity, providing more possibilities for subsequent optimization.

[0190] Determining whether a new spider population meets the preset termination condition is crucial for ensuring the algorithm's convergence. The algorithm terminates its iteration when it reaches the maximum number of iterations, the fitness value converges, or it finds the optimal path that meets the timeliness requirements of data transmission. It then selects the path corresponding to the spider with the highest fitness value from the new spider population as the optimal data transmission path. This termination condition ensures both the algorithm's efficiency and the quality of the final result.

[0191] In this embodiment, by introducing the opposition learning algorithm and the black widow algorithm, a comprehensive optimization of the robot's data transmission path is achieved. The opposition learning algorithm generates opposition paths and compares their fitness with the original paths, effectively improving the quality of the initial data transmission path set. Then, the black widow algorithm simulates the natural selection process, iteratively optimizing the path set, gradually approaching the global optimal solution, accurately determining the optimal data transmission path, and effectively improving the efficiency and stability of data transmission between robots.

[0192] It should be noted that the above examples are only for understanding this application and do not constitute a limitation on the robot communication method based on multi-network communication in this application. Any simple modifications based on this technical concept are within the protection scope of this application.

[0193] This application also provides a robot communication device based on multi-network communication, please refer to... Figure 2 Robot communication devices based on multi-network communication include:

[0194] Module 10 is used to build a multi-network communication architecture, which integrates at least two different types of network communication methods, including at least wired network communication, wireless network communication, Bluetooth communication, or mobile data network communication.

[0195] The acquisition module 20 is used to acquire real-time status information of each network communication method in the multi-network communication architecture when the robot has a data transmission requirement. The real-time status information includes at least one of signal strength, connection status, network bandwidth, and network latency.

[0196] The allocation module 30 is used to allocate the optimal network communication method for the data to be transmitted according to the real-time status information and the preset communication priority rules, wherein the preset communication priority rules are set according to the data type, the urgency of the data, and the quality of network communication.

[0197] The determination module 40 is used to determine neighboring robots that are in the same communication environment as the current robot and can interact with data through a neighbor discovery mechanism.

[0198] The acquisition module 20 is also used to acquire the current state information of the neighboring robots, and dynamically adjust and optimize the data transmission path according to the current state information of the neighboring robots through a path planning strategy to determine the optimal data transmission path. The path planning strategy is based on opposition learning and the Black Widow algorithm.

[0199] The transmission module 50 is used to transmit the data to be transmitted to the target robot through the optimal data transmission path and the optimal network communication method.

[0200] The robot communication device based on multi-network communication provided in this application, employing the robot communication method based on multi-network communication in the above embodiments, can solve the technical problems of poor communication stability and unreasonable network resource utilization in existing robot communication technologies. Compared with the prior art, the beneficial effects of the robot communication device based on multi-network communication provided in this application are the same as those of the robot communication method based on multi-network communication provided in the above embodiments, and other technical features in the robot communication device based on multi-network communication are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0201] The above are only some embodiments of this application and do not limit the patent scope of this application. All equivalent structural transformations made under the technical concept of this application and using the contents of the specification and drawings of this application, or direct / indirect applications in other related technical fields, are included in the patent protection scope of this application.

Claims

1. A robot communication method based on multi-network communication, characterized in that, The method includes: Construct a multi-network communication architecture, which is used to integrate at least two different types of network communication methods, including at least wired network communication, wireless network communication, Bluetooth communication, or mobile data network communication. When the robot has a data transmission requirement, the real-time status information of each network communication method in the multi-network communication architecture is obtained. The real-time status information includes at least one of signal strength, connection status, network bandwidth, and network latency. Based on the real-time status information and the preset communication priority rules, the optimal network communication method is allocated for the data to be transmitted, wherein the preset communication priority rules are set according to the data type, the urgency of the data, and the quality of network communication. The neighbor discovery mechanism identifies neighboring robots that are in the same communication environment as the current robot and can interact with it. The current state information of the neighboring robots is obtained, and the data transmission path is dynamically adjusted and optimized according to the current state information of the neighboring robots through a path planning strategy to determine the optimal data transmission path. The path planning strategy is based on opposition learning and the Black Widow algorithm. The data to be transmitted is transmitted to the target robot through the optimal data transmission path and the optimal network communication method; The step of obtaining the current state information of neighboring robots and dynamically adjusting and optimizing the data transmission path based on the current state information of neighboring robots through a path planning strategy to determine the optimal data transmission path includes: Obtain the current status information of the neighboring robot, which includes location information, communication capability parameters, current task load, and network environment stability indicators; Using the current robot as the starting node, the target robot as the target node, and the neighboring robots as intermediate nodes, an initial data transmission path set is constructed based on the current state information. The opposition learning algorithm is used to evaluate each original path in the initial data transmission path set and generate an opposition path for each original path. The construction rule of the opposition path is to generate an alternative path with a different topology than the original path but connecting the starting node and the target node by changing the order of intermediate nodes or replacing some intermediate nodes. The fitness values ​​of each original path and its corresponding opposing path are calculated based on path length, cumulative node communication quality, and overall path stability coefficient. The path length is obtained by weighted summation of physical distance or logical hop count between nodes. The cumulative node communication quality is the product of the communication capability parameters of each node in the path. The overall path stability coefficient is calculated by integrating the network environment stability indicators of each node and using a weighted average method. If the fitness value of the opposing path is greater than the fitness value of the original path, then the original path is replaced according to the opposing path and the initial data transmission path set is updated to obtain the updated initial data transmission path set. The Black Widow algorithm is used to optimize the updated initial data transmission path set to obtain the optimal data transmission path.

2. The method as described in claim 1, characterized in that, The step of allocating the optimal network communication method for the data to be transmitted based on the real-time status information and preset communication priority rules includes: The network quality feature vectors for each network communication method are determined based on the real-time status information. Obtain the data type feature vector and the urgency feature vector of the data to be transmitted. The data type feature vector consists of semantic level, data volume coefficient and preset tolerable packet loss rate. The urgency feature vector consists of data timeliness requirements, queued delay and data urgency. The static priority components of each network communication method are determined based on the data type feature vector and the urgency feature vector. The real-time score of each network communication method is determined based on the network quality feature vector. The dynamic priority components of each network communication method are determined based on the static priority components and the real-time scores. The priority difference between any two network communication methods is determined based on the dynamic priority components of each network communication method. If all priority differences are greater than or equal to a preset threshold, the network communication method with the highest dynamic priority component is selected as the optimal network communication method.

3. The method as described in claim 2, characterized in that, After determining the priority difference between any two network communication methods based on the dynamic priority components of each network communication method, the method further includes: If the priority difference is less than a preset threshold, a fragmentation dual-transmission strategy is executed, and both network communication methods corresponding to the priority difference are selected as candidate network communication methods. The data to be transmitted is fragmented according to a preset fragmentation rule to obtain multiple data fragments; The corresponding data fragments are transmitted simultaneously through the candidate network communication methods, and the transmission status of the two candidate network communication methods is monitored in real time during the transmission process. If one of the candidate network communication methods experiences a transmission error, the remaining untransmitted data will be immediately fragmented and switched to another candidate network communication method that is transmitting normally for continued transmission. Once all data fragments have been successfully transmitted to the target robot, the received data fragments are reassembled on the target robot to recover the complete data to be transmitted.

4. The method as described in claim 1, characterized in that, The process of identifying neighboring robots that are in the same communication environment and can interact with the current robot through a neighbor discovery mechanism includes: The robot acquires feature data of its current communication environment and maps the feature data to a predefined environmental feature space to generate a communication environment feature vector for the current robot. The communication environment similarity is obtained by calculating the similarity between the current robot's communication environment feature vector and the robot to be judged. Based on the communication environment similarity, the current robot and the robot to be judged are determined to be in the same domain, and the same domain determination result is obtained. If the same-domain determination result indicates that the current robot and the robot to be determined are in the same communication domain, the robot to be determined is determined to be a same-domain robot. The interaction feasibility of the robots in the same domain is evaluated based on the robot interaction feasibility scoring model, and the interaction feasibility score of each robot in the same domain is obtained. The robot interaction feasibility scoring model is constructed based on the communication capability, remaining battery power, task load, historical interaction success rate and the stability of the current network environment of the robots in the same domain. The robots in the same domain are sorted from high to low according to the interaction feasibility score, and the robots in the same domain with an interaction feasibility score higher than a preset score threshold are selected as neighbor robots that can perform data interaction.

5. The method as described in claim 1, characterized in that, The step of optimizing the updated initial data transmission path set using the Black Widow algorithm to obtain the optimal data transmission path includes: Treat each path in the updated initial data transmission path set as an individual spider to construct the initial spider population; The fitness value of each spider in the initial spider population is calculated according to a preset fitness function, which is constructed based on path length, cumulative node communication quality, overall path stability coefficient, and data transmission timeliness requirements. Based on the fitness value of each individual spider in the initial spider population, high-quality spider individuals with fitness values ​​higher than a preset threshold are selected to form a high-quality spider subpopulation. The high-quality spider population is used for breeding operations to generate new spider individuals through crossover and mutation. The crossover operation adopts single-point crossover or multi-point crossover, and the mutation operation adopts the method of randomly changing some intermediate nodes in the path. The new spider individuals are merged with the remaining unselected spider individuals to form a new spider population; Determine whether the new spider population meets the preset termination conditions, including reaching the maximum number of iterations, fitness value convergence, or finding the optimal path that meets the timeliness requirements of data transmission; If the termination condition is met, the path corresponding to the spider individual with the highest fitness value from the new spider population is selected as the optimal data transmission path. If the termination condition is not met, return to the step of calculating the fitness value of each spider individual in the new spider population according to the preset fitness function, and continue iterative optimization until the preset termination condition is met, and obtain the optimal data transmission path.

6. The method as described in claim 1, characterized in that, The step of transmitting the data to be transmitted to the target robot through the optimal data transmission path and the optimal network communication method includes: Obtain the list of available channels corresponding to the optimal network communication method, monitor the channels in the list of available channels in real time, and obtain the current occupancy status and signal quality parameters of each channel; Based on the current occupancy status and signal quality parameters of each channel, a dynamic channel allocation strategy is adopted to allocate the optimal transmission channel for the data to be transmitted. The dynamic channel allocation strategy is set according to the channel idle time, signal interference intensity and historical transmission success rate. The data to be transmitted is encapsulated according to a preset data encapsulation format. Channel allocation information and target robot identification information are added to the header of the encapsulated data to generate a data packet to be sent. The data packet to be sent is transmitted to the target robot through the optimal transmission channel along the optimal data transmission path.

7. The method as described in claim 6, characterized in that, After transmitting the data packet to be sent to the target robot via the optimal transmission channel and along the optimal data transmission path, the method further includes: If, during transmission, it is detected that the optimal transmission channel experiences increased signal interference or is occupied by other devices, the channel contention backoff mechanism will be activated immediately. Based on the contention level of the optimal transmission channel and the preset backoff algorithm, the backoff time window is dynamically adjusted, data transmission is paused within the backoff time window, and the state changes of the optimal transmission channel are continuously monitored. If the optimal transmission channel becomes available again when the backoff time window ends, data transmission is restarted. If the optimal transmission channel is still unavailable, the dynamic channel allocation strategy is re-executed, and the second-best transmission channel is selected from the list of available channels to continue transmitting the data packet to be sent. If there are no other available channels in the available channel list, data transmission is paused, and the channel list expiration repair mechanism is triggered to search for and update the available channel list again, resulting in an updated available channel list. Select a new optimal transmission channel from the updated list of available channels to transmit the data packet to be sent.

8. The method as described in claim 6, characterized in that, After transmitting the data packet to be sent to the target robot via the optimal transmission channel and along the optimal data transmission path, the process further includes: During data transmission, if a fault or reduced communication capability is detected in an intermediate node on the optimal data transmission path, the intermediate node will be designated as the faulty intermediate node, and a multi-channel virtual monitoring mechanism will be immediately activated. Monitor the channel status on other available paths around the faulty intermediate node, and determine the backup transmission channel based on the channel status information; The data stream passing through the faulty intermediate node is switched to the backup transmission channel, and the optimal data transmission path is updated to bypass the faulty intermediate node and continue to transmit data to the target robot.

9. A robot communication device based on multi-network communication, characterized in that, The robot communication device based on multi-network communication includes: A building module is used to build a multi-network communication architecture, which integrates at least two different types of network communication methods, including at least wired network communication, wireless network communication, Bluetooth communication, or mobile data network communication. The acquisition module is used to acquire real-time status information of each network communication method in the multi-network communication architecture when the robot has a data transmission requirement. The real-time status information includes at least one of signal strength, connection status, network bandwidth, and network latency. The allocation module is used to allocate the optimal network communication method to the data to be transmitted based on the real-time status information and the preset communication priority rules, wherein the preset communication priority rules are set according to the data type, the urgency of the data, and the quality of network communication. The determination module is used to determine neighboring robots that are in the same communication environment as the current robot and can interact with data, through a neighbor discovery mechanism. The acquisition module is also used to acquire the current state information of the neighboring robots, and dynamically adjust and optimize the data transmission path according to the current state information of the neighboring robots through a path planning strategy to determine the optimal data transmission path. The path planning strategy is based on opposition learning and the Black Widow algorithm. The transmission module is used to transmit the data to be transmitted to the target robot through the optimal data transmission path and the optimal network communication method; The acquisition module is further configured to acquire the current state information of neighboring robots, including location information, communication capability parameters, current task load, and network environment stability indicators; using the current robot as the starting node, the target robot as the target node, and neighboring robots as intermediate nodes, an initial data transmission path set is constructed based on the current state information; each original path in the initial data transmission path set is evaluated using an opposition learning algorithm to generate an opposition path for each original path. The construction rule for the opposition path is to generate an alternative path that connects the starting node and the target node but has a different topology than the original path by changing the order of intermediate nodes or replacing some intermediate nodes; based on path length, nodes... The fitness values ​​of each original path and its corresponding opposing path are calculated using the cumulative communication quality value and the overall path stability coefficient. The path length is obtained by weighted summation of physical distances or logical hop counts between nodes. The cumulative node communication quality value is the product of the communication capability parameters of each node in the path. The overall path stability coefficient is calculated by integrating the network environment stability indicators of each node and using a weighted average method. If the fitness value of the opposing path is greater than the fitness value of the original path, the original path is replaced by the opposing path, and the initial data transmission path set is updated to obtain an updated initial data transmission path set. The Black Widow algorithm is then used to optimize the updated initial data transmission path set to obtain the optimal data transmission path.

Images