Dual-frequency adaptive transmission method and device for industrial mobile platform

Abstract

The application provides a dual-frequency adaptive transmission method and device of an industrial mobile platform, relates to the technical field of industrial wireless communication, and comprises the following steps: identifying service of a to-be-transmitted data packet and obtaining a service level; performing mode matching on the basis of the service level and current channel quality data, and obtaining an initial transmission mode; when the initial transmission mode is a normal single-transmission mode or a fast retransmission redundancy mode, inputting historical motion state data and historical channel quality data into a time sequence prediction model to perform channel switching prediction, and obtaining a switching prediction result; when the switching prediction result does not satisfy a preset trigger condition or the initial transmission mode is a full dual-transmission mode, taking the initial transmission mode as a final transmission mode; when the switching prediction result satisfies the preset trigger condition, forcibly switching the final transmission mode to the full dual-transmission mode through a hysteresis switching mechanism; and sending a to-be-transmitted data packet to the industrial mobile platform according to the final transmission mode.

Description

Technical Field

[0001] This application relates to the field of industrial wireless communication technology, and in particular to a dual-frequency adaptive transmission method and apparatus for an industrial mobile platform. Background Technology

[0002] In Industrial Internet of Things (IIoT) scenarios, industrial mobile platforms (such as Automated Guided Vehicles (AGVs), unmanned forklifts, and industrial inspection robots) rely on wireless networks to transmit critical business data such as motion control commands, real-time status reporting, and video monitoring. These services are extremely sensitive to the reliability and latency of wireless networks; packet loss or delays in control commands can lead to equipment malfunctions, collisions, or even safety incidents.

[0003] Existing data transmission methods lack sophisticated strategies for dynamically adjusting redundancy levels, making it difficult to achieve the optimal balance between critical data reliability and spectrum utilization efficiency; they also lack the ability to predict channel quality, making it impossible to switch transmission modes in advance to avoid roaming packet loss; and the transmission modes cannot accurately match the operating environment, resulting in low data transmission accuracy and affecting the normal operation of industrial mobile platforms. Summary of the Invention

[0004] To address the aforementioned problems, in a first aspect, the present invention provides a dual-frequency adaptive transmission method for an industrial mobile platform, comprising: The system acquires the data packet to be transmitted and the current channel quality data from the sending end, identifies the service in the data packet to be transmitted, and obtains the service level; it then performs pattern matching based on the service level and the current channel quality data to obtain the initial transmission mode, which includes normal single transmission mode, fast retransmission redundancy mode, and full dual transmission mode. When the initial transmission mode is normal single-transmission mode or fast retransmission redundancy mode, acquire the historical motion status data of the industrial mobile platform and the historical channel quality data of the transmitting end; input the historical motion status data and historical channel quality data into the time series prediction model to perform channel handover prediction and obtain the handover prediction result. When the handover prediction result does not meet the preset triggering conditions, or the initial transmission mode is full dual-transmission mode, the initial transmission mode is used as the final transmission mode; when the handover prediction result meets the preset triggering conditions, the final transmission mode is forcibly switched to full dual-transmission mode through a hysteresis handover mechanism. The sending end sends the data packet to be transmitted to the industrial mobile platform according to the final transmission mode, and the industrial mobile platform executes the corresponding operation instructions according to the data packet to be transmitted.

[0005] In one embodiment, service identification is performed on the data packet to be transmitted to obtain the service level, including: Obtain the socket identifier corresponding to the data packet to be transmitted; Execute the first identification path, read the service quality mark value associated with the socket identifier, and when the service quality mark value is a preset priority value, identify the data packet to be transmitted as critical business data; The second identification path is executed, and deep packet inspection is performed on the payload part of the data packet to be transmitted. The protocol feature fields of the payload part are matched with the preset industrial control protocol feature library. When the match is successful, the data packet to be transmitted is identified as key business data. When either the first identification path or the second identification path identifies the data packet to be transmitted as critical business data, the business level corresponding to the data packet to be transmitted is determined to be a high-priority business level; when neither the first identification path nor the second identification path identifies the data packet to be transmitted as critical business data, the business level corresponding to the data packet to be transmitted is determined to be a normal business level.

[0006] In one embodiment, pattern matching is performed based on the service level and current channel quality data to obtain an initial transmission mode, including: When the service level is normal service level, the initial transmission mode is determined to be normal single transmission mode; When the service level is high priority, obtain the current signal strength, current signal-to-noise ratio, current packet loss rate, and current frequency domain interference from the current channel quality data; The hierarchical reinforcement learning agent model generates preset quality conditions corresponding to the current signal strength, current signal-to-noise ratio, current packet loss rate, and current frequency domain interference. If the current signal strength, current signal-to-noise ratio, current packet loss rate, and current frequency domain interference all meet their respective preset quality conditions, then the initial transmission mode will be determined as the fast retransmission redundancy mode; otherwise, the initial transmission mode will be determined as the full dual-transmission mode.

[0007] In one embodiment, the temporal prediction model includes a feature extraction layer, an attention weighting layer, a feature fusion layer, and a mapping layer. The process of obtaining the switching prediction result includes: Historical motion state data and historical channel quality data are obtained through historical time windows. Historical motion state data includes historical motion speed and historical motion direction angle, while historical channel quality data includes historical signal strength, historical signal-to-noise ratio, historical packet loss rate, and historical frequency domain interference. Based on the combination of historical motion state data and historical channel quality data into a multi-dimensional feature sequence, the temporal dependency relationship of each feature dimension in the multi-dimensional feature sequence is extracted by the feature extraction layer to obtain the temporal feature vector; The attention-weighted layer assigns attention weights to temporal feature vectors through an attention mechanism to obtain weighted feature vectors. The last temporal feature and the weighted feature vector in the temporal feature vector are input into the feature fusion layer and concatenated to obtain the fused feature vector. The fused feature vector is input into the mapping layer for mapping calculation to obtain the handover prediction result, which includes the channel handover probability and the prediction confidence score.

[0008] In one embodiment, when the channel switching probability is greater than a first preset threshold and the prediction confidence score is greater than a second preset threshold, the switching prediction result is determined to meet the preset triggering condition. When the channel switching probability is less than or equal to the first preset threshold, or the prediction confidence score is less than or equal to the second preset threshold, it is determined that the switching prediction result does not meet the preset triggering condition.

[0009] In one embodiment, the sending end sends the data packet to be transmitted to the industrial mobile platform according to the final transmission mode, including: Fault detection is performed on the main communication link and the secondary communication link. When the fault detection fails, the main communication link and the secondary communication link are degraded and self-healed. When the final transmission mode is the normal single-transmission mode, the sending end sends the data packet to be transmitted to the main access point of the industrial mobile platform through the main communication link, and the main access point sends the data packet to be transmitted to the control unit of the industrial mobile platform. When the final transmission mode is fast retransmission redundancy mode, the sending end sends the data packet to be transmitted to the main access point through the main communication link and continuously monitors the transmission feedback status of the main communication link; when no acknowledgment frame is received from the main access point within the preset timeout period, it is determined that the main communication link triggers retransmission, and the sending end sends the retransmission data packet of the data packet to be transmitted to the secondary access point of the industrial mobile platform through the secondary communication link; the main access point or the secondary access point sends the data packet to be transmitted to the control unit of the industrial mobile platform; When the final transmission mode is full dual-transmission mode, the sending end copies the data packet to be transmitted to obtain the first copy data packet and the second copy data packet; the first copy data packet is sent to the primary access point through the primary communication link, and the second copy data packet is sent to the secondary access point through the secondary communication link; The central controller verifies the validity of the first and second copy data packets. If neither the first nor the second copy data packets pass the validity verification, the sending end is triggered to perform a fast retransmission operation on the data packet to be transmitted. Otherwise, the quality of the first and second copy data packets is compared to determine the final copy data packet. The central controller then sends the final copy data packet to the control unit of the industrial mobile platform.

[0010] In one embodiment, sending a first replica data packet to the primary access point via the primary communication link and sending a second replica data packet to the secondary access point via the secondary communication link includes: Obtain the first estimated completion time of the first replica data packet on the primary communication link, and the second estimated completion time of the second replica data packet on the secondary communication link; Based on the difference between the first estimated transmission completion time and the second estimated transmission completion time, the first initial transmission time of the main communication link and the second initial transmission time of the secondary communication link are offset and adjusted to obtain the first optimized transmission time and the second optimized transmission time. The primary communication link sends the first replica data packet to the primary access point according to the first optimized sending time, and the secondary communication link sends the second replica data packet to the secondary access point according to the second optimized sending time.

[0011] Secondly, the present invention provides a dual-frequency adaptive transmission device for an industrial mobile platform, used to implement the dual-frequency adaptive transmission method for the industrial mobile platform, the device comprising: The initial transmission mode acquisition module is used to acquire the data packet to be transmitted and the current channel quality data from the sending end, identify the service of the data packet to be transmitted, and obtain the service level; and perform pattern matching based on the service level and the current channel quality data to obtain the initial transmission mode, which includes normal single transmission mode, fast retransmission redundancy mode and full dual transmission mode. The handover prediction result acquisition module is used to acquire historical motion status data of the industrial mobile platform and historical channel quality data of the transmitting end when the initial transmission mode is normal single transmission mode or fast retransmission redundancy mode; input the historical motion status data and historical channel quality data into the time series prediction model to perform channel handover prediction and obtain the handover prediction result. The transmission mode switching module is used to use the initial transmission mode as the final transmission mode when the switching prediction result does not meet the preset triggering conditions, or when the initial transmission mode is a full dual-transmission mode; when the switching prediction result meets the preset triggering conditions, the final transmission mode is forcibly switched to the full dual-transmission mode through a hysteresis switching mechanism. The data transmission module is used by the sending end to send the data packet to be transmitted to the industrial mobile platform according to the final transmission mode, and the industrial mobile platform executes the corresponding operation instructions according to the data packet to be transmitted.

[0012] Thirdly, the present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the dual-frequency adaptive transmission method of the industrial mobile platform.

[0013] Fourthly, the present invention provides a non-transitory computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the dual-frequency adaptive transmission method of the industrial mobile platform.

[0014] The present invention has the following beneficial effects: 1. By matching the service level with the current channel quality data, the redundancy strategy is applied in a differentiated manner, which greatly saves air interface resources while ensuring the reliability of critical data. By inputting historical motion status and historical channel quality data into the time series prediction model, the channel switching trend can be predicted in advance. When the switching prediction result meets the preset trigger conditions, the transmission mode can be forcibly switched to full dual transmission mode, which effectively avoids frequent oscillations of the mode at the threshold edge and ensures the stability of the switching decision. The data packets to be transmitted are sent to the industrial mobile platform according to the final transmission mode, so that the transmission mode matches the current working environment and improves the data transmission accuracy of the industrial mobile platform.

[0015] 2. By constructing a time-series prediction model that includes a feature extraction layer, an attention weighting layer, a feature fusion layer, and a mapping layer, the model integrates historical motion state data and historical channel quality data of industrial mobile platforms in multiple dimensions. An attention mechanism automatically assigns weights to different times and features. Combined with inference optimizations such as sliding window updates and quantization compression, high-precision prediction of future channel switching trends is achieved. The channel switching probability and prediction confidence score in the prediction results are used together as triggering conditions. Only when the switching probability is higher than a first preset threshold and the confidence score is higher than a second preset threshold is the preset triggering condition met and a mode switch is executed. This avoids false triggering and invalid switching caused by the susceptibility of single-threshold judgment to instantaneous fluctuations, significantly improving the triggering accuracy of predictive roaming.

[0016] 3. During the process of sending data packets to the industrial mobile platform, the introduction of fault detection and fault degradation self-healing mechanisms ensures that the system can still operate in a degraded manner and automatically recover under single-frequency band hardware failure or severe interference conditions, fundamentally improving the robustness and maintainability of the system. In the normal single-transmission mode, only the main communication link is used; in the fast retransmission redundancy mode, conditional retransmission is performed; and in the full dual-transmission mode, offset adjustment, validity verification, and quality comparison are performed. High-quality data transmission can be achieved in various environments, significantly improving the quality and stability of data transmission. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments or related technologies of this application, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart of a method according to an embodiment of the present invention; Figure 2 This is a structural diagram of the device according to an embodiment of the present invention. Detailed Implementation

[0019] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments.

[0020] The terminology used in the following embodiments of this application is for the purpose of describing particular embodiments only and is not intended to be limiting of this application. As used in the specification of this application, the singular expressions “a,” “an,” “the,” “the,” “the,” and “this” are intended to include the plural expressions as well, unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used in this application refers to and includes any or all possible combinations of one or more of the listed items.

[0021] Hereinafter, the terms "first" and "second" are used for descriptive purposes only and should not be construed as implying or suggesting relative importance or implicitly indicating the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature, and in the description of the embodiments of this application, unless otherwise stated, "multiple" means two or more.

[0022] To enable those skilled in the art to better understand the technical solution of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings.

[0023] Reference Figure 1 The present invention provides a dual-frequency adaptive transmission method for an industrial mobile platform, including steps S100-S400.

[0024] S100 acquires the data packet to be transmitted and the current channel quality data from the sending end, identifies the service in the data packet to be transmitted, and obtains the service level; it performs pattern matching based on the service level and the current channel quality data to obtain the initial transmission mode, which includes normal single transmission mode, fast retransmission redundancy mode, and full dual transmission mode.

[0025] In some embodiments, service identification is performed on the data packets to be transmitted to obtain service levels, including: Obtain the socket identifier corresponding to the data packet to be transmitted; Execute the first identification path, read the service quality mark value associated with the socket identifier, and when the service quality mark value is a preset priority value, identify the data packet to be transmitted as critical business data; The second identification path is executed, and deep packet inspection is performed on the payload part of the data packet to be transmitted. The protocol feature fields of the payload part are matched with the preset industrial control protocol feature library. When the match is successful, the data packet to be transmitted is identified as key business data. When either the first identification path or the second identification path identifies the data packet to be transmitted as critical business data, the business level corresponding to the data packet to be transmitted is determined to be a high-priority business level; when neither the first identification path nor the second identification path identifies the data packet to be transmitted as critical business data, the business level corresponding to the data packet to be transmitted is determined to be a normal business level.

[0026] In some embodiments, the socket identifier is used to uniquely identify the communication connection between the sending device and the receiving device.

[0027] In industrial control systems, when creating sockets, critical business applications can set the IP_TOS or SO_PRIORITY option via the setsockopt() interface provided by the operating system to set the Quality of Service (QoS) flag value to a specific priority value. When the sending device reads this QoS flag value as the preset priority value, it identifies the data packet to be transmitted as critical business data. The first identification path has the advantages of zero overhead and support for encrypted streams because the QoS flag value can be directly read by the driver layer before the data packet is encrypted.

[0028] Deep packet inspection is a service identification technology based on the content of data packet payloads. It can parse application layer protocol headers and extract key fields for pattern matching. The industrial control protocol feature library includes feature fields for at least Modbus, OPC UA, PROFINET, and EtherNet / IP protocols, which are widely used in industrial IoT scenarios. Taking the Modbus TCP protocol as an example, Modbus TCP packets begin with an MBAP header, followed by a function code field. When the sending device detects that the destination port of a data packet is 502 and the function code is 0x05 or 0x0F, it determines that the data packet is a Modbus write command, which is a critical control command. A successful match is achieved, and the data packet to be transmitted is identified as critical business data. The second identification path, as a supplement to the first identification path, can cover data packets that do not have a set QoS flag value but actually contain critical control commands, achieving precise protection for all industrial control traffic outside of encrypted applications.

[0029] In some embodiments, pattern matching is performed based on the service level and current channel quality data to obtain an initial transmission mode, including: When the service level is normal service level, the initial transmission mode is determined to be normal single transmission mode; When the service level is high priority, obtain the current signal strength, current signal-to-noise ratio, current packet loss rate, and current frequency domain interference from the current channel quality data; The hierarchical reinforcement learning agent model generates preset quality conditions corresponding to the current signal strength, current signal-to-noise ratio, current packet loss rate, and current frequency domain interference. If the current signal strength, current signal-to-noise ratio, current packet loss rate, and current frequency domain interference all meet their respective preset quality conditions, then the initial transmission mode will be determined as the fast retransmission redundancy mode; otherwise, the initial transmission mode will be determined as the full dual-transmission mode.

[0030] In some embodiments, the current channel quality data includes the current signal strength. (Unit: dBm) Current signal-to-noise ratio (Unit: dB) Current packet loss rate (Dimensionless, value range 0-1) and current frequency domain interference. (Dimensionless, value range 0-1). These parameters are obtained from real-time statistics of the physical layer and MAC layer, reflecting the current quality status of the wireless communication link.

[0031] Since ordinary business data is not sensitive to latency and packet loss, adopting the normal single-transmission mode can minimize air interface resource overhead and avoid unnecessary redundant transmissions that may interfere with other services.

[0032] When a service is classified as a high-priority service, the transmitting device needs to further determine the appropriate redundancy protection strategy based on current channel quality data. The operating environment and channel conditions of industrial mobile platforms change dynamically over time—the level of wireless interference and service traffic patterns may differ significantly within the same factory under different shifts and production tasks. If the pattern matching threshold remains fixed, it cannot adapt to such dynamic changes. Therefore, this invention introduces a hierarchical reinforcement learning agent to perform online self-learning optimization of the decision parameters. The hierarchical reinforcement learning agent model includes an upper-layer policy agent and a lower-layer policy agent, both trained using a deep deterministic policy gradient algorithm.

[0033] The upper-level strategy agent is responsible for macro-level decision-making, based on the first preset time period. It runs once per second. The state space of the upper-level policy agent. Includes: the average interference level of the current path segment Business type identifier and the interference pattern for the current time period. Action space of the upper-level strategy agent To select the optimization objective: {maximize throughput, minimize transmission latency, balanced mode}, they are encoded as 0, 1, and 2 respectively.

[0034] The lower-level policy agent is responsible for executing specific parameter adjustments, within the second preset time period. Run once. The state space of the lower-level policy agent. Includes: smoothed RSSI, RSSI slope Current redundancy mode, most recent Number of mode switches per second, current packet loss rate Action space of the lower-level policy agent To adjust the decision threshold parameters: including adjusting the high threshold. Adjusting the low threshold Adjusting the minimum stay time .

[0035] The collaboration mechanism between the upper-level policy agent and the lower-level policy agent is as follows: the optimization objective output by the upper-level policy agent. This additional state input is passed to the lower-level policy agent to guide its reward and penalty weight configuration. The reward function of the lower-level policy agent... for:

[0036] in This represents the change in throughput. As a penalty for packet loss, This is a penalty for pattern oscillation.

[0037] , , The weights are dynamic and determined by the optimization objective chosen by the upper-level policy agent: when the upper-level policy agent chooses "maximize throughput" ( When the weight is configured as follows: , , When the upper-layer policy agent selects "minimize transmission latency" ( When the weight is configured as follows: , , When the upper-level policy agent selects "Balanced Mode" ( When the weight is configured as follows: , , .

[0038] Preset quality conditions include signal strength thresholds. Signal-to-noise ratio threshold Packet loss rate threshold Frequency domain interference threshold In the online optimization process of the hierarchical reinforcement learning agent, the lower-level policy agent can perform self-learning optimization of preset quality conditions based on the real-time network environment and business needs.

[0039] If all four conditions are met simultaneously, it indicates that the overall channel quality is good, and the transmitting device will set the initial transmission mode to Fast Repeat Redundancy Mode. If any one of the four conditions is not met, it indicates that the current channel quality has deteriorated to the point where stronger redundancy protection is required, and the transmitting device will set the initial transmission mode to Full Dual Transmission Mode.

[0040] When the initial transmission mode is normal single-transmission mode or fast retransmission redundancy mode, S200 acquires historical motion status data of the industrial mobile platform and historical channel quality data of the transmitting end; inputs the historical motion status data and historical channel quality data into the time series prediction model to perform channel handover prediction and obtain handover prediction results.

[0041] In some embodiments, the time-series prediction model includes a feature extraction layer, an attention weighting layer, a feature fusion layer, and a mapping layer. The process of obtaining the switching prediction result includes: Historical motion state data and historical channel quality data are obtained through historical time windows. Historical motion state data includes historical motion speed and historical motion direction angle, while historical channel quality data includes historical signal strength, historical signal-to-noise ratio, historical packet loss rate, and historical frequency domain interference. Based on the combination of historical motion state data and historical channel quality data into a multi-dimensional feature sequence, the temporal dependency relationship of each feature dimension in the multi-dimensional feature sequence is extracted by the feature extraction layer to obtain the temporal feature vector; The attention-weighted layer assigns attention weights to temporal feature vectors through an attention mechanism to obtain weighted feature vectors. The last temporal feature and the weighted feature vector in the temporal feature vector are input into the feature fusion layer and concatenated to obtain the fused feature vector. The fused feature vector is input into the mapping layer for mapping calculation to obtain the handover prediction result, which includes the channel handover probability and the prediction confidence score.

[0042] In some embodiments, historical motion state data includes historical motion speed. and the direction angle of historical movement These data are provided by the industrial mobile platform's own inertial measurement unit or odometer. Historical channel quality data includes historical signal strength. Historical signal-to-noise ratio Historical packet loss rate and historical frequency domain interference subscript This indicates the sampling time within the historical time window.

[0043] The transmitting device combines historical motion state data and historical channel quality data into a multidimensional feature sequence. : Each of these moments eigenvectors A 6-dimensional vector: The multidimensional feature sequence contains both the motion state of the industrial mobile platform and information about the external channel environment, providing rich input features for subsequent time series prediction.

[0044] The feature extraction layer uses a two-layer Long Short-Term Memory (LSTM) network to process multidimensional feature sequences. Temporal dependencies are extracted for each feature dimension. The first LSTM layer has 64 hidden units, uses the tanh activation function, returns a sequence, and has a Dropout rate of 0.2. The second LSTM layer has 32 hidden units, uses the tanh activation function, returns a sequence, and has a Dropout rate of 0.2. Dropout is a regularization technique that prevents overfitting by randomly discarding some neuron outputs during training. The output of the first LSTM layer serves as the input to the second LSTM layer, and the second LSTM layer outputs a temporal feature vector. :

[0045] in For a moment The 32-dimensional time-series feature vector and the two-layer LSTM structure enable the model to simultaneously capture the dependencies of two different time scales: short-term fluctuations and long-term trends.

[0046] Attention-weighted layers generate temporal feature vectors through an attention mechanism. Attention weights are assigned; the core idea of ​​the attention mechanism is that information from different moments within a historical time window has different importance when predicting future channel quality. The attention head count is 4, and the scaling factor is [missing value]. ,in Attention weight The calculation formula is:

[0047] in For intermediate parameters, and The weight matrix is ​​a learnable matrix with all dimensions. Weighted eigenvectors for: Weighted eigenvectors It is a weighted sum of the temporal feature vectors of all moments within the entire historical time window. The weights are automatically determined by the attention mechanism, enabling the model to focus on the moment with the most information for prediction.

[0048] The attention-weighted layer weights the feature vectors. Temporal characteristics of the last moment Perform concatenation and fusion to obtain the fused feature vector. : Fusion feature vectors The dimension is 64, and the feature vectors will be fused. The input mapping layer performs mapping calculations. This mapping layer consists of two fully connected layers: the first fully connected layer has 64 neurons and uses the ReLU activation function; the second fully connected layer (output layer) has 4 neurons and uses a linear activation function. The ReLU activation function is defined as follows: The output of the mapping layer is the switching prediction result:

[0049] in For channel switching probability, This is used to predict the confidence score. The channel handover probability represents the likelihood of an AP handover occurring within a future period, while the confidence score represents the reliability of that prediction. Outputting both values ​​simultaneously allows the decision engine to trigger predictive redundant handover only under high confidence conditions, effectively avoiding invalid handovers caused by model misjudgments.

[0050] When the handover prediction result does not meet the preset triggering conditions, or the initial transmission mode is full dual-transmission mode, the S300 will use the initial transmission mode as the final transmission mode; when the handover prediction result meets the preset triggering conditions, the final transmission mode will be forcibly switched to full dual-transmission mode through a hysteresis handover mechanism.

[0051] In some embodiments, when the channel switching probability is greater than a first preset threshold and the prediction confidence score is greater than a second preset threshold, it is determined that the switching prediction result meets the preset triggering condition. When the channel switching probability is less than or equal to the first preset threshold, or the prediction confidence score is less than or equal to the second preset threshold, it is determined that the switching prediction result does not meet the preset triggering condition.

[0052] In some embodiments, the handover prediction result output by the time-series prediction model includes the channel handover probability. and prediction confidence score Two key metrics are needed for the decision engine to determine whether predictive redundancy switching should be triggered.

[0053] When the channel switching probability Greater than the first preset threshold ( And the predicted confidence score Greater than the second preset threshold ( When the handover prediction result meets the preset triggering conditions, the sending device determines that the handover prediction result does not meet the preset triggering conditions; otherwise, it determines that the handover prediction result does not meet the preset triggering conditions.

[0054] The dual-condition joint triggering mechanism ensures that the full dual-mode is activated only when the model prediction results have high reliability, and the prediction accuracy can be improved by more than 30% compared with the traditional scheme that relies on only a single signal strength change rate.

[0055] When the switching prediction result does not meet the preset triggering conditions, the sending device will use the initial transmission mode as the final transmission mode, that is, keep the current transmission strategy unchanged.

[0056] When the handover prediction result meets the preset triggering conditions, the transmitting device forces the final transmission mode to full dual-transmission mode through a hysteresis handover mechanism. The specific implementation of the hysteresis handover mechanism is as follows: Set a high threshold for entering full dual-speed mode. and exiting the low threshold of the full dual-mode. There is a difference between the two. When channel quality deteriorates and the signal-to-noise ratio is lower than... When switching from other modes to full dual-transmission mode is allowed; when channel quality improves and the signal-to-noise ratio is higher than... At this time, it allows exiting from full dual-fire mode to other modes; when the signal-to-noise ratio is at and In between, maintain the current pattern to prevent frequent pattern oscillations.

[0057] In addition, it is mandatory to maintain a minimum dwell time after switching from any mode to full dual-engine mode. ,exist Switching from full dual-transmission mode to other modes is prohibited within a specified time. The minimum dwell time further enhances the system's resilience, ensuring sufficient time to complete redundant transmissions after switching to full dual-transmission mode, and preventing immediate backoff due to instantaneous improvements in channel quality.

[0058] The S400 transmitter sends the data packet to be transmitted to the industrial mobile platform according to the final transmission mode, and the industrial mobile platform executes the corresponding operation instructions according to the data packet to be transmitted.

[0059] In some embodiments, the sender transmits the data packet to be transmitted to the industrial mobile platform according to the final transmission mode, including: Fault detection is performed on the main communication link and the secondary communication link. When the fault detection fails, the main communication link and the secondary communication link are degraded and self-healed. When the final transmission mode is the normal single-transmission mode, the sending end sends the data packet to be transmitted to the main access point of the industrial mobile platform through the main communication link, and the main access point sends the data packet to be transmitted to the control unit of the industrial mobile platform. When the final transmission mode is fast retransmission redundancy mode, the sending end sends the data packet to be transmitted to the main access point through the main communication link and continuously monitors the transmission feedback status of the main communication link; when no acknowledgment frame is received from the main access point within the preset timeout period, it is determined that the main communication link triggers retransmission, and the sending end sends the retransmission data packet of the data packet to be transmitted to the secondary access point of the industrial mobile platform through the secondary communication link; the main access point or the secondary access point sends the data packet to be transmitted to the control unit of the industrial mobile platform; When the final transmission mode is full dual-transmission mode, the sending end copies the data packet to be transmitted to obtain the first copy data packet and the second copy data packet; the first copy data packet is sent to the primary access point through the primary communication link, and the second copy data packet is sent to the secondary access point through the secondary communication link; The central controller verifies the validity of the first and second copy data packets. If neither the first nor the second copy data packets pass the validity verification, the sending end is triggered to perform a fast retransmission operation on the data packet to be transmitted. Otherwise, the quality of the first and second copy data packets is compared to determine the final copy data packet. The central controller then sends the final copy data packet to the control unit of the industrial mobile platform.

[0060] In some embodiments, before transmitting data to the primary and secondary communication links, the transmitting device first performs fault detection on the primary and secondary communication links. The frequency band health monitoring module continuously tracks the health status of each communication link, and the health status is determined based on the following indicator: number of consecutive transmission failures. Confirmation frame response rate Hardware abnormal status flags Health of each communication link The comprehensive calculation is as follows:

[0061] in To determine the maximum number of consecutive failures allowed, , , These are the weighting coefficients for each indicator.

[0062] When the health level is below the preset downgrade threshold When the sending device determines that the fault detection fails, it performs a fault degradation on the communication link: marking the communication link as unavailable, stopping data allocation to the communication link, and automatically locking the final transmission mode to send only through available links. Simultaneously, a degradation event is generated and reported to the network management system.

[0063] After downgrading, the frequency band health monitoring module will monitor the frequency band at a preset interval. Each second, probe frames are sent through the degraded communication link. When continuously... When all probe frames are successful, the sending device determines that the communication link has been restored to normal, automatically restores the communication link to an available state, removes the degradation, restores to the normal adaptive redundancy transmission mode, and generates a recovery event to report to the network management system.

[0064] In normal single-transmission mode, the secondary communication link is in a dormant state and does not participate in data transmission, thereby maximizing the utilization efficiency of air interface resources.

[0065] When the final transmission mode is Fast Retransmit Redundancy mode, the sending device starts a timeout timer after sending each data frame, waiting for the receiving end to return an ACK confirmation frame. Within the preset timeout period... If no acknowledgment frame is received from the primary access point, the sending device determines that the primary communication link has triggered a retransmission. Without waiting for the timeout timer to completely expire, it immediately sends the retransmitted data packet of the data to be transmitted to the secondary access point of the industrial mobile platform via the secondary communication link. This mechanism of "retransmitting via another frequency band without waiting for timeout" advances the retransmission triggered by RTO timeout in the traditional TCP / IP protocol stack to within 50ms, significantly shortening the packet loss recovery time.

[0066] The primary or secondary access point sends the data packet to be transmitted to the control unit of the industrial mobile platform. That is, whichever access point successfully receives a valid data packet first will then forward it to the control unit. This first-come, first-served mechanism ensures the fastest possible packet loss recovery path.

[0067] In some embodiments, sending a first replica data packet to the primary access point via the primary communication link and sending a second replica data packet to the secondary access point via the secondary communication link includes: Obtain the first estimated completion time of the first replica data packet on the primary communication link, and the second estimated completion time of the second replica data packet on the secondary communication link; Based on the difference between the first estimated transmission completion time and the second estimated transmission completion time, the first initial transmission time of the main communication link and the second initial transmission time of the secondary communication link are offset and adjusted to obtain the first optimized transmission time and the second optimized transmission time. The primary communication link sends the first replica data packet to the primary access point according to the first optimized sending time, and the secondary communication link sends the second replica data packet to the secondary access point according to the second optimized sending time.

[0068] In some embodiments, the sending device obtains the first estimated transmission completion time of the first copy data packet on the main communication link (5GHz band). The second estimated completion time for the transmission of the second copy data packet on the secondary communication link (2.4 GHz band). The estimated completion time is calculated using a dynamic estimated completion time formula: The dynamic queue waiting time is predicted based on an exponentially weighted moving average. in This represents the current queue depth. This represents the queue depth at the previous time step. Average data packet transmission time The delay of sending the previous data packet As a dynamic regulating factor It can quickly respond to sudden changes in queue depth.

[0069] For physical layer transmission time: in For data frame length, This refers to the physical layer rate. For the 5GHz band, For the 2.4GHz band, .

[0070] Delay due to anticipated retransmission: in The packet loss rate for a single transmission. Maximum number of retransmissions For the first The time to wait for the retransmission For cross-band aggregation gain: in Polymerization coefficient and These represent the physical layer transmission times for the primary and secondary communication links, respectively. When both frequency bands have low load, the aggregation gain reduces the total completion time.

[0071] Based on the first estimated delivery completion time Second estimated delivery completion time The difference between For the first initial transmission time Second initial transmission time Perform offset adjustment. If If so, the primary communication link will send the data ahead of time, with an offset of 1 / 2. ;like If so, it is sent ahead of time via the secondary communication link, with an offset of . The first optimized transmission time is obtained after offset adjustment. Second optimized transmission time :

[0072] in It is a function for maximizing the value.

[0073] When the final transmission mode is full dual-transmission mode, the primary access point and the secondary access point forward the received first copy data packet, second copy data packet, and their respective metadata to the central controller. The metadata includes at least the access point identifier, frequency band identifier, and signal strength. Signal-to-noise ratio The CRC result and hardware timestamp, these metadata, provide sufficient decision-making basis for subsequent selection.

[0074] The centralized controller performs CRC validity checks on the first and second copy data packets. CRC is a commonly used data integrity verification method. By appending a checksum to the end of the data, the receiving end recalculates and compares the results to determine whether a bit error has occurred during data transmission.

[0075] If both the first and second copy data packets fail the validity check, the centralized controller, without waiting for the sender's retransmission timeout timer to expire, constructs and returns a duplicate acknowledgment message to the sender, triggering the sender to perform a fast retransmission operation on the data packet to be transmitted. Specifically, the centralized controller constructs three duplicate acknowledgment messages, each with an acknowledgment number set to the expected sequence number of the missing data packet, and sends them back to the sender via the wired side. After receiving the third duplicate acknowledgment message, the sender immediately retransmits the data packet to be transmitted according to the fast retransmission algorithm specified in RFC 5681, without waiting for the RTO timeout. This step-by-step triggering mechanism breaks the layered constraints of the traditional protocol stack, allowing the link layer adjudication engine to directly trigger the transport layer's fast retransmission, compressing the packet loss recovery latency in double-copy failure scenarios from an average of over 350ms to 8-15ms.

[0076] If at least one of the first or second replica data packets passes the validity check, the central controller performs a quality comparison on the valid replica data packets. The quality comparison is based on factors including signal strength. Signal-to-noise ratio Based on arrival time, the overall rating is:

[0077] in , , These are the weighting coefficients for each indicator. The time of arrival of the data packet is the moment it reaches the central controller. This comprehensive score considers both signal quality and arrival time, ensuring a complete reflection of the data packet transmission quality. The central controller selects the copy with the higher score as the final copy and sends it to the control unit of the industrial mobile platform, discarding the duplicate copy and eliminating the possibility of out-of-order TCP packets at the source.

[0078] Reference Figure 2 This invention provides a dual-frequency adaptive transmission device 20 for an industrial mobile platform, used for a dual-frequency adaptive transmission method for an industrial mobile platform. The device includes: The initial transmission mode acquisition module 21 is used to acquire the data packet to be transmitted and the current channel quality data from the sending end, identify the service of the data packet to be transmitted, and obtain the service level; perform pattern matching based on the service level and the current channel quality data to obtain the initial transmission mode, which includes normal single transmission mode, fast retransmission redundancy mode and full dual transmission mode. The handover prediction result acquisition module 22 is used to acquire historical motion status data of the industrial mobile platform and historical channel quality data of the transmitting end when the initial transmission mode is normal single transmission mode or fast retransmission redundancy mode; input the historical motion status data and historical channel quality data into the time series prediction model to perform channel handover prediction and obtain the handover prediction result. The transmission mode switching module 23 is used to use the initial transmission mode as the final transmission mode when the switching prediction result does not meet the preset triggering condition, or when the initial transmission mode is a full dual-transmission mode; and to force the final transmission mode to be switched to a full dual-transmission mode through a hysteresis switching mechanism when the switching prediction result meets the preset triggering condition. The data transmission module 24 is used by the sending end to send the data packet to be transmitted to the industrial mobile platform according to the final transmission mode, and the industrial mobile platform executes the corresponding operation instructions according to the data packet to be transmitted.

[0079] This application provides an electronic device, including a processor and a memory; the memory stores a computer program, wherein the computer program, when executed by the processor, implements the dual-frequency adaptive transmission method of the industrial mobile platform according to any of the above schemes.

[0080] Specifically, the processor may include, for example, a general-purpose microprocessor, an instruction set processor and / or an associated chipset and / or a special-purpose microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. The processor may also include onboard memory for caching purposes. The processor may be a single processing unit or multiple processing units for performing different actions of the method flow according to embodiments of this application.

[0081] Memory can be any medium capable of containing, storing, transmitting, propagating, or transmitting instructions. For example, memory can include, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, instruments, or propagation media. Specific examples of memory include: magnetic storage devices such as magnetic tape or hard disk drives (HDDs); optical storage devices such as optical discs (CD-ROMs); and also random access memory (RAM) or flash memory; and / or wired / wireless communication links.

[0082] This application also provides a computer-readable medium storing a computer program that, when executed by a processor, implements the dual-frequency adaptive transmission method for an industrial mobile platform according to any of the above-described schemes. This computer-readable medium may be included in the device / apparatus / system described in the above embodiments; or it may exist independently and not assembled into the device / apparatus / system. The aforementioned computer-readable medium carries one or more programs, which, when executed, implement the method as described in the embodiments of this application.

[0083] According to embodiments of this application, a computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this application, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In this application, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. Computer-readable signal media can also be any computer-readable medium other than computer-readable storage media, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wireless, wired, optical fiber, radio frequency signals, etc., or any suitable combination thereof.

[0084] Those skilled in the art will understand that the features described in the various embodiments and / or claims of this application can be combined and / or combined in various ways, even if such combinations or combinations are not explicitly described in this application. In particular, the features described in the various embodiments and / or claims of this application can be combined and / or combined in various ways without departing from the spirit and teachings of this application. All such combinations and / or combinations fall within the scope of this application. Therefore, the scope of this application should not be limited to the above embodiments, but should be defined not only by the appended claims, but also by their equivalents. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A dual-frequency adaptive transmission method for an industrial mobile platform, characterized in that, include: The system acquires the data packet to be transmitted and the current channel quality data from the sending end, identifies the service in the data packet to be transmitted, and obtains the service level; it then performs pattern matching based on the service level and the current channel quality data to obtain the initial transmission mode, which includes normal single transmission mode, fast retransmission redundancy mode, and full dual transmission mode. When the initial transmission mode is normal single-transmission mode or fast retransmission redundancy mode, acquire the historical motion status data of the industrial mobile platform and the historical channel quality data of the transmitting end; input the historical motion status data and historical channel quality data into the time series prediction model to perform channel handover prediction and obtain the handover prediction result. When the switching prediction result does not meet the preset triggering conditions, or the initial transmission mode is the full dual transmission mode, the initial transmission mode will be used as the final transmission mode. When the handover prediction result meets the preset triggering conditions, the final transmission mode will be forcibly switched to full dual transmission mode through the hysteresis handover mechanism; The sending end sends the data packet to be transmitted to the industrial mobile platform according to the final transmission mode, and the industrial mobile platform executes the corresponding operation instructions according to the data packet to be transmitted.

2. The dual-frequency adaptive transmission method for an industrial mobile platform according to claim 1, characterized in that, Service identification is performed on the data packets to be transmitted to obtain the service level, including: Obtain the socket identifier corresponding to the data packet to be transmitted; Execute the first identification path, read the service quality mark value associated with the socket identifier, and when the service quality mark value is a preset priority value, identify the data packet to be transmitted as critical business data; The second identification path is executed, and deep packet inspection is performed on the payload part of the data packet to be transmitted. The protocol feature fields of the payload part are matched with the preset industrial control protocol feature library. When the match is successful, the data packet to be transmitted is identified as key business data. When either the first identification path or the second identification path identifies the data packet to be transmitted as critical business data, the business level corresponding to the data packet to be transmitted is determined to be a high-priority business level; when neither the first identification path nor the second identification path identifies the data packet to be transmitted as critical business data, the business level corresponding to the data packet to be transmitted is determined to be a normal business level.

3. The dual-frequency adaptive transmission method for an industrial mobile platform according to claim 1, characterized in that, Pattern matching is performed based on the service level and current channel quality data to obtain the initial transmission mode, including: When the service level is normal service level, the initial transmission mode is determined to be normal single transmission mode; When the service level is high priority, obtain the current signal strength, current signal-to-noise ratio, current packet loss rate, and current frequency domain interference from the current channel quality data; The hierarchical reinforcement learning agent model generates preset quality conditions corresponding to the current signal strength, current signal-to-noise ratio, current packet loss rate, and current frequency domain interference. If the current signal strength, current signal-to-noise ratio, current packet loss rate, and current frequency domain interference all meet their respective preset quality conditions, then the initial transmission mode will be determined as the fast retransmission redundancy mode; otherwise, the initial transmission mode will be determined as the full dual-transmission mode.

4. The dual-frequency adaptive transmission method for an industrial mobile platform according to claim 1, characterized in that, The time-series prediction model includes a feature extraction layer, an attention weighting layer, a feature fusion layer, and a mapping layer. The process of obtaining the prediction result includes: Historical motion state data and historical channel quality data are obtained through historical time windows. Historical motion state data includes historical motion speed and historical motion direction angle, while historical channel quality data includes historical signal strength, historical signal-to-noise ratio, historical packet loss rate, and historical frequency domain interference. Based on the combination of historical motion state data and historical channel quality data into a multi-dimensional feature sequence, the temporal dependency relationship of each feature dimension in the multi-dimensional feature sequence is extracted by the feature extraction layer to obtain the temporal feature vector; The attention-weighted layer assigns attention weights to temporal feature vectors through an attention mechanism to obtain weighted feature vectors. The last temporal feature and the weighted feature vector in the temporal feature vector are input into the feature fusion layer and concatenated to obtain the fused feature vector. The fused feature vector is input into the mapping layer for mapping calculation to obtain the handover prediction result, which includes the channel handover probability and the prediction confidence score.

5. The dual-frequency adaptive transmission method for an industrial mobile platform according to claim 4, characterized in that: When the channel handover probability is greater than the first preset threshold and the prediction confidence score is greater than the second preset threshold, the handover prediction result is determined to meet the preset triggering condition. When the channel switching probability is less than or equal to the first preset threshold, or the prediction confidence score is less than or equal to the second preset threshold, it is determined that the switching prediction result does not meet the preset triggering condition.

6. The dual-frequency adaptive transmission method for an industrial mobile platform according to claim 1, characterized in that, The sending end transmits the data packets to be transmitted to the industrial mobile platform according to the final transmission mode, including: Fault detection is performed on the main communication link and the secondary communication link. When the fault detection fails, the main communication link and the secondary communication link are degraded and self-healed. When the final transmission mode is the normal single-transmission mode, the sending end sends the data packet to be transmitted to the main access point of the industrial mobile platform through the main communication link, and the main access point sends the data packet to be transmitted to the control unit of the industrial mobile platform. When the final transmission mode is fast retransmission redundancy mode, the sending end sends the data packet to be transmitted to the main access point through the main communication link and continuously monitors the transmission feedback status of the main communication link; when no acknowledgment frame is received from the main access point within the preset timeout period, it is determined that the main communication link triggers retransmission, and the sending end sends the retransmission data packet of the data packet to be transmitted to the secondary access point of the industrial mobile platform through the secondary communication link; the main access point or the secondary access point sends the data packet to be transmitted to the control unit of the industrial mobile platform; When the final transmission mode is full dual-transmission mode, the sending end copies the data packet to be transmitted to obtain the first copy data packet and the second copy data packet; the first copy data packet is sent to the primary access point through the primary communication link, and the second copy data packet is sent to the secondary access point through the secondary communication link; The central controller verifies the validity of the first and second copy data packets. If neither the first nor the second copy data packets pass the validity verification, the sending end is triggered to perform a fast retransmission operation on the data packet to be transmitted. Otherwise, the quality of the first and second copy data packets is compared to determine the final copy data packet. The central controller then sends the final copy data packet to the control unit of the industrial mobile platform.

7. The dual-frequency adaptive transmission method for an industrial mobile platform according to claim 6, characterized in that, The first replica data packet is sent to the primary access point via the primary communication link, and the second replica data packet is sent to the secondary access point via the secondary communication link, including: Obtain the first estimated completion time of the first replica data packet on the primary communication link, and the second estimated completion time of the second replica data packet on the secondary communication link; Based on the difference between the first estimated transmission completion time and the second estimated transmission completion time, the first initial transmission time of the main communication link and the second initial transmission time of the secondary communication link are offset and adjusted to obtain the first optimized transmission time and the second optimized transmission time. The primary communication link sends the first replica data packet to the primary access point according to the first optimized sending time, and the secondary communication link sends the second replica data packet to the secondary access point according to the second optimized sending time.

8. A dual-frequency adaptive transmission device for an industrial mobile platform, used to implement the dual-frequency adaptive transmission method for an industrial mobile platform as described in any one of claims 1 to 7, characterized in that, The device includes: The initial transmission mode acquisition module is used to acquire the data packet to be transmitted and the current channel quality data from the sending end, identify the service of the data packet to be transmitted, and obtain the service level; and perform pattern matching based on the service level and the current channel quality data to obtain the initial transmission mode, which includes normal single transmission mode, fast retransmission redundancy mode and full dual transmission mode. The handover prediction result acquisition module is used to acquire historical motion status data of the industrial mobile platform and historical channel quality data of the transmitting end when the initial transmission mode is normal single transmission mode or fast retransmission redundancy mode; input the historical motion status data and historical channel quality data into the time series prediction model to perform channel handover prediction and obtain the handover prediction result. The transmission mode switching module is used to use the initial transmission mode as the final transmission mode when the switching prediction result does not meet the preset triggering conditions, or when the initial transmission mode is a full dual-transmission mode; when the switching prediction result meets the preset triggering conditions, the final transmission mode is forcibly switched to the full dual-transmission mode through a hysteresis switching mechanism. The data transmission module is used by the sending end to send the data packet to be transmitted to the industrial mobile platform according to the final transmission mode, and the industrial mobile platform executes the corresponding operation instructions according to the data packet to be transmitted.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the dual-frequency adaptive transmission method for the industrial mobile platform as described in any one of claims 1 to 7.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the dual-frequency adaptive transmission method for the industrial mobile platform as described in any one of claims 1 to 7.

Images