Agricultural machine operation data low-latency updating method and device based on remote communication

Abstract

The application discloses a kind of based on long-distance communication's agricultural machine operation data low latency update method and device, belong to wisdom agriculture and agricultural informatization field.The method is in view of traditional cycle polling and cloud centralization scheme exists's high update delay, data is not consecutive etc., constructs " hierarchical transmission-prediction compensation-GPU collaborative rendering " three-level acceleration framework: first, according to delay sensitivity, data is divided into safety alarm, real-time operation, agricultural machinery state three levels and is endowed with dynamic priority;Then, according to network bandwidth, priority weight is adjusted in real time and bandwidth is adaptively distributed, to ensure that key data is sent first;Client triggers agricultural machine motion model prediction compensation through 200ms heartbeat monitoring, and after transmission recovery, the latest observation is used to correct historical trajectory in reverse smoothing;Finally, on the GPU side, quadtree blocking and heading preloading strategy are used to realize the rapid rendering of map engine.

Description

Technical Field

[0001] This invention belongs to the field of smart agriculture and agricultural informatization, and particularly relates to a method and device for low-latency updating of agricultural machinery operation data based on remote communication. Background Technology

[0002] In remote control and monitoring systems for unmanned agricultural machinery, data such as operating paths, operating speeds, task progress, operating width, and fault alarms are typically transmitted to remote monitoring terminals via wireless communication networks (e.g., 4G / 5G, satellite communication, LoRa) to enable remote monitoring and decision support for agricultural machinery operation. Common agricultural machinery data update schemes include periodic polling updates, event-triggered updates, and centralized cloud processing. Periodic polling updates involve the agricultural machinery terminal uploading data at fixed time intervals (e.g., 1-5 seconds), with the remote control and monitoring system passively receiving the data. This method has a low data update frequency and cannot meet the need for real-time updates and display of operation status data. Event-triggered updates only trigger data uploads when the agricultural machinery status changes significantly (e.g., fault alarms), resulting in high delays in updating non-critical data and inconsistent data updates. Furthermore, this update method relies on status threshold settings, leading to the problem of missed alarms near the threshold. Centralized cloud processing involves uploading all agricultural machinery operation data to a cloud server before distributing it to the control and monitoring terminals. The long communication link results in high latency and delayed operation data updates.

[0003] In current remote control and monitoring systems for unmanned agricultural machinery, there is a significant delay in updating agricultural machinery operation parameters. This causes the monitoring interface to fail to accurately reflect the real-time status, affecting emergency fault response and precise operational decision-making. Traditional solutions mainly rely on periodic polling or centralized cloud processing, which have outdated data update mechanisms. These mechanisms suffer from problems such as inconsistent data updates, asynchronous data from multiple sources, low bandwidth utilization, and long response times for progressive faults, making it difficult to meet the requirements for real-time updates of agricultural machinery operation data. For example, in scenarios such as automatic obstacle avoidance or variable seeding, high latency can lead to delayed decision-making, reducing operational safety and efficiency. Summary of the Invention

[0004] To overcome the problems existing in the prior art, this invention addresses the issue that traditional periodic polling and cloud-centralized solutions have poor data update timeliness, failing to meet the real-time display requirements of agricultural machinery operation data. It proposes a low-latency update method and device for agricultural machinery operation data based on remote communication. The specific technical solution is as follows:

[0005] A low-latency update method for agricultural machinery operation data based on remote communication includes the following steps:

[0006] Step 1: On the agricultural machinery terminal, the operation data is divided into three levels according to the latency sensitivity from high to low: safety alarm, real-time operation, and agricultural machinery status, and corresponding transmission priorities are assigned.

[0007] Step 2: Based on the aforementioned priorities, assign importance coefficients to each category of subdivided data, and calculate dynamic priorities by combining the rate of change of control instructions with the current network bandwidth availability.

[0008] Step 3: Real-time detection of network bandwidth availability, and dynamic adjustment of the weight coefficients involved in priority calculation in Step 2 accordingly, so that the priority adapts to the network conditions in real time.

[0009] Step 4: Based on the dynamic priority obtained in Step 2 and the minimum bandwidth requirements of various types of data, adaptively allocate the total bandwidth, and ensure that the total allocation does not exceed the network capacity limit.

[0010] Step 5: The client performs heartbeat monitoring on the real-time trajectory data of the agricultural machinery at fixed intervals. If no data is received within a time limit, it is determined that the transmission is abnormal and the CPU-side agricultural machinery motion model is activated to predict the position of the agricultural machinery in order to compensate for the missing trajectory data.

[0011] Step 6: When the heartbeat monitoring receives data again within a specified period of time, it is determined that the transmission has been restored. Then, the parameters of the agricultural machinery motion model used in Step 5 are recalibrated using the latest data, and reverse smoothing is performed between the predicted segment and the measured segment to correct the historical data.

[0012] Step 7: On the GPU side, divide the basic map data into quadtree blocks according to the current zoom level and create an index;

[0013] Step 8: Based on the real-time location and heading of the agricultural machinery and the index obtained in Step 7, dynamically load the required data blocks and complete the client rendering.

[0014] A low-latency update device for agricultural machinery operation data based on remote communication includes the following modules:

[0015] The module divides the operational data at the agricultural machinery end into three levels based on latency sensitivity, from high to low: safety alarm, real-time operation, and agricultural machinery status, and assigns corresponding transmission priorities.

[0016] The calculation module assigns an importance coefficient to each type of subdivided data based on the aforementioned priority, and calculates dynamic priority by combining the rate of change of control instructions with the current network bandwidth availability.

[0017] The adaptation module detects network bandwidth availability in real time and dynamically adjusts the weight coefficients participating in priority calculation accordingly, so that the priority adapts to network conditions in real time.

[0018] The allocation module adaptively allocates the total bandwidth based on dynamic priorities and the minimum bandwidth requirements of various types of data, and the total allocation amount does not exceed the network capacity limit.

[0019] The monitoring module allows the client to perform heartbeat monitoring on the real-time trajectory data of agricultural machinery at fixed intervals. If no data is received within a time limit, it is determined that the transmission is abnormal and the central processing unit (CPU) side agricultural machinery motion model is activated to predict the position of the agricultural machinery in order to compensate for the missing trajectory data.

[0020] The correction module determines that the transmission has been restored when the heartbeat monitoring receives data again within a specified period of time. Then, it uses the latest data to recalibrate the parameters of the agricultural machinery motion model and performs reverse smoothing between the predicted segment and the measured segment to correct the historical data.

[0021] The indexing module, on the GPU side, divides the basic map data into quadtree blocks according to the current zoom level and creates an index;

[0022] The rendering module dynamically loads the required data blocks and completes client-side rendering based on the real-time location, heading, and index of the agricultural machinery.

[0023] An electronic device includes: one or more processors; and a memory for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to implement the method.

[0024] A computer-readable storage medium having executable instructions stored thereon, which, when executed by a processor, cause the processor to implement the method described thereon.

[0025] The present invention has the following beneficial effects:

[0026] This invention proposes a dynamic priority calculation method that integrates fixed priority, importance coefficient and current available bandwidth. By incorporating network congestion into the priority calculation factor, the scheduling strategy can both ensure the priority transmission of critical data and adapt to network fluctuations in real time, thereby significantly reducing the transmission delay and packet loss rate of critical data and improving the robustness of the system in harsh network environments.

[0027] This invention proposes an adaptive network bandwidth allocation algorithm that guarantees both network capacity thresholds and minimum bandwidth. The algorithm strictly adheres to two core constraints: first, it ensures a minimum available bandwidth for each data stream or data type, fundamentally preventing critical data transmission failures due to bandwidth contention; second, it controls the total bandwidth allocation upper limit within the current network's actual capacity threshold, preventing network congestion caused by over-allocation. This dual-constraint adaptive bandwidth mechanism effectively optimizes network resource utilization, maximizing overall data transmission efficiency and reducing average transmission latency while ensuring transmission stability.

[0028] This invention introduces a client-side CPU-GPU collaborative rendering technology based on a prediction compensation mechanism at the client-side rendering level. When data transmission anomalies (such as latency or loss) are detected, the CPU employs an active compensation mechanism, performing real-time prediction and calculation based on historical data streams, and performing reverse smoothing on the prediction results to fill data gaps and ensure the continuity and naturalness of the rendering curves. Meanwhile, the GPU performs parallel processing of data chunking and loading of the map engine's basic data, freeing up CPU resources to focus on highly logical prediction compensation tasks. This CPU-GPU collaborative architecture effectively compensates for potential latency caused by network transmission at the rendering level, ensuring the continuity, trajectory accuracy, and real-time updates of agricultural machinery data movement on the user interface.

[0029] This invention constructs a remote communication model for agricultural machinery based on a wireless network. It reduces end-to-end latency of agricultural machinery operation data through hierarchical transmission, client-side pre-rendering, and GPU-CPU collaborative rendering, enabling real-time updates and display of the data. This method reduces data update latency from three aspects: data transmission, prediction compensation, and data rendering. During data transmission, latency-sensitive scheduling is implemented based on the varying latency sensitivities of different data types, and dynamic weight adjustments are made based on the priority of each data type to ensure the transmission of critical operation data under different network conditions. Simultaneously, to cope with complex network environments and reduce excessive data update latency caused by network delays, a heartbeat monitoring mechanism is introduced on the client side. If critical operation data is not received within a certain time, it is considered a transmission anomaly. The agricultural machinery motion model predicts and displays the operation status. Upon receiving the operation data, the model is dynamically updated with adjusted parameters, and the predicted data is reverse-smoothed to achieve historical data correction, ensuring continuous display of the operation data. Finally, during client-side data rendering, GPU-CPU collaborative rendering achieves efficient rendering, reducing the update latency of agricultural machinery operation data. This method can significantly improve the timeliness of data updates, ensure data synchronization and continuity, and is suitable for precision operation and emergency fault response of unmanned agricultural machinery. Attached Figure Description

[0030] Figure 1 This is a flowchart of the processing of the present invention. Detailed Implementation

[0031] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other. To achieve the above objectives, this invention adopts the following technical solution.

[0032] This invention provides a low-latency update method for agricultural machinery operation data based on remote communication, such as... Figure 1 As shown, it includes the following steps:

[0033] Step 1: Establish a hierarchical system based on real-time requirements for various types of data generated during agricultural machinery operations, according to their differences in sensitivity to transmission delay.

[0034] Specifically, the data generated during agricultural machinery operation includes safety alarm data, real-time operation data, and agricultural machinery status data. Their transmission delay requirements are <50ms, 50-200ms, and 200-500ms, respectively. Based on the transmission delay requirements of each type of data, their transmission priorities are defined as follows: safety alarm data transmission priority 1, real-time operation data transmission priority 2, and agricultural machinery status data transmission priority 3.

[0035] Step 2: Based on the fixed priorities defined for each type of data in Step 1, further divide the data according to its characteristics, and then perform dynamic priority calculation on the subdivided data.

[0036] Specifically, safety alarm data primarily characterizes the current safety factor of agricultural machinery. Based on the actual safety alarm data generated by the machinery, it is categorized into safety alarms, path deviation warnings, and emergency stop signals, with their importance coefficients (0-1, with 1 being the highest importance) set to 1, 0.8, and 0.7 respectively. Real-time operation data represents the quality of operation generated by the machinery during operation, including real-time operation trajectory coordinates, operating speed, steering angle, and specific operation parameters under different operating scenarios. Assuming a consistent update frequency, its importance coefficients are set to 0.7, 0.6, 0.6, and 0.5 respectively. Machinery status data represents the equipment status data of the machinery itself during operation, including hydraulic pressure, fuel consumption, and water temperature data, with an importance coefficient of 0.3 for each. Dynamic priorities are calculated based on the fixed priorities and importance coefficients of each data type.

[0037] Step 3: Due to fluctuations in network conditions during agricultural machinery operations, network congestion may occur when agricultural machinery transmits a large amount of data within a certain period. Therefore, based on network bandwidth availability, dynamic weighting coefficients are adjusted according to the dynamic adjustment principle to ensure that the calculated dynamic priority is adapted to the current network conditions.

[0038] Specifically, during the agricultural machinery data transmission process, the dynamic weight coefficients calculated in step 2 are dynamically adjusted according to the availability of network bandwidth to ensure that the calculated dynamic priority can adapt to the current network conditions.

[0039] Step 4: In addition to adapting to network conditions in different environments, network bandwidth is also an important factor affecting agricultural machinery data transmission. When data with lower priority is updated more frequently and the data volume is large, it is easy to cause network resource contention, which will cause transmission delay of critical agricultural machinery data. In order to ensure the transmission of critical data, network bandwidth is adaptively allocated.

[0040] Specifically, taking into account the dynamic priority calculated in step 2 and the minimum bandwidth required to ensure data transmission, network bandwidth is adaptively allocated. Simultaneously, to ensure that the bandwidth does not exceed the network's existing carrying capacity, constraints are set for bandwidth allocation.

[0041] Step 5: When the client receives agricultural machinery data, the data transmission frequency is monitored in real time through a heartbeat monitoring mechanism according to the data transmission delay requirements. If agricultural machinery operation data is not received within a certain period of time, it is determined to be a transmission abnormality. The agricultural machinery operation trajectory data is predicted and compensated by using the agricultural machinery motion model on the CPU side.

[0042] Specifically, this invention sets up a heartbeat monitoring mechanism for real-time trajectory data of agricultural machinery, setting the monitoring frequency of real-time trajectory data of agricultural machinery to 200ms. If no trajectory data of agricultural machinery is received within 200ms, it is determined that the data transmission is abnormal. On the CPU side, the position of agricultural machinery during the abnormal data transmission process is predicted and compensated by the agricultural machinery motion model.

[0043] Step 6: After data transmission is restored to normal, dynamically adjust and update the agricultural machinery motion model and perform reverse smoothing on the predicted data to achieve retrospective correction of historical data and ensure the continuity of agricultural machinery operation data.

[0044] Specifically, if the client receives agricultural machinery data normally within 1 second after a data transmission anomaly alarm, it is considered that data transmission has recovered. After data transmission is recovered, its trajectory data is first compensated for with timestamp alignment. Based on the timestamp alignment, the state sequence generated by the prediction model during the data transmission anomaly in step 5 is reconstructed for the missing time period, realizing the retrospective correction of historical data and ensuring the continuity of agricultural machinery operation data.

[0045] Step 7: Based on the hierarchical transmission strategy and prediction compensation mechanism of the above data, continuous and stable agricultural machinery status data is obtained. In order to realize the rapid rendering of agricultural machinery data on the map engine on the client side, the basic data of the map engine is divided into data blocks on the GPU side according to certain rules.

[0046] Specifically, the basic data of the map engine is divided into blocks according to the current zoom level using a four-block method. When the map level is 0, the data is divided into four blocks along the 0° latitude and longitude line. Each time the zoom level is increased, the current data block is divided along the center line of latitude and longitude, and an index number is set for the total number of data blocks obtained.

[0047] Step 8: Dynamically load data based on the data blocks divided in Step 7, the current scaling level, and the real-time agricultural machinery operation data to reduce data rendering pressure and data update latency.

[0048] Specifically, the index number of the data block is first calculated based on the current position coordinates transmitted back by the agricultural machinery. Then, the set of index numbers of the data blocks to be dynamically rendered is further calculated based on the heading of the agricultural machinery. Finally, the client rendering is completed according to the set of data block index numbers.

[0049] Assuming the data is used to render real-time operational data, such as the location of an autonomous harvester during wheat harvesting in an unmanned farm, on a client-side basis, the data is transmitted via 5G internet. The client uses the Cesium 3D map engine to display the geographical location of the farm machinery, and CPU-GPU collaborative rendering is employed during the rendering process. This invention provides a low-latency update method for agricultural machinery operation data based on remote communication, comprising the following steps:

[0050] Step 1: Establish a hierarchical system based on real-time requirements for various types of data generated during agricultural machinery operations, according to their differences in sensitivity to transmission delay.

[0051] Specifically, the data generated during agricultural machinery operation includes safety alarm data, real-time operation data, and agricultural machinery status data. Their transmission delay requirements are <50ms, 50-200ms, and 200-500ms, respectively. Based on the transmission delay requirements of each type of data, their transmission priorities are defined as follows: safety alarm data transmission priority 1, real-time operation data transmission priority 2, and agricultural machinery status data transmission priority 3.

[0052] Step 2: Based on the fixed priorities defined for each type of data in Step 1, further divide the data according to its characteristics, and then perform dynamic priority calculation on the subdivided data.

[0053] Specifically, safety alarm data primarily characterizes the current safety factor of agricultural machinery. Based on the actual safety alarm data generated by the machinery, it is categorized into safety alarms, path deviation warnings, and emergency stop signals, with their importance coefficients (0-1, with 1 being the highest importance) set to 1, 0.8, and 0.7 respectively. Real-time operation data represents the quality of operation generated by the machinery during operation, including real-time operation trajectory coordinates, operating speed, steering angle, wheat header height, main clutch status, and unloading clutch status. Assuming a consistent update frequency, their importance coefficients are set to 0.7, 0.6, 0.6, and 0.5 respectively. Machinery status data represents the equipment status data of the machinery itself during operation, including hydraulic pressure, fuel consumption, and water temperature data, with an importance coefficient of 0.3 for each. Based on the fixed priorities and importance coefficients of each data type, the dynamic priority is calculated as follows:

[0054] ;

[0055] in, Dynamic priority for various types of data to be transmitted. For the fixed priority of various types of data transmitted, The importance coefficients for the various types of data transmitted. To control the rate of change of instructions, This represents the current network bandwidth availability. , and These are dynamic weighting coefficients, with default values ​​of 0.6, 0.3, and 0.1.

[0056] Control command change rate This represents the incremental change of control commands per unit of time:

[0057] ;

[0058] in, This refers to the control command value at the current moment. It is the command value from the previous sampling time. To control the cycle.

[0059] Step 3: Due to fluctuations in network conditions during agricultural machinery operations, network congestion may occur when agricultural machinery transmits a large amount of data within a certain period. Therefore, based on network bandwidth availability, dynamic weighting coefficients are adjusted according to the dynamic adjustment principle to ensure that the calculated dynamic priority is adapted to the current network conditions.

[0060] Specifically, during agricultural machinery data transmission, the dynamic weight coefficients of the dynamic priority calculation model in step 2 are dynamically adjusted based on network bandwidth availability. , and The dynamic adjustment principles are as follows:

[0061] ;

[0062] Step 4: In addition to adapting to network conditions in different environments, network bandwidth is also an important factor affecting agricultural machinery data transmission. When data with lower priority is updated more frequently and the data volume is large, it is easy to cause network resource contention, which will cause transmission delay of critical agricultural machinery data. In order to ensure the transmission of critical data, network bandwidth is adaptively allocated.

[0063] Specifically, taking into account the dynamic priority calculated in step 2 and the minimum bandwidth to ensure data transmission, network bandwidth is adaptively allocated. The adaptive bandwidth allocation formula is as follows:

[0064] ;

[0065] in, To allocate bandwidth, Total bandwidth To assign sharpness coefficients, The bandwidth allocation weight is typically set to 2 in this invention. The total number of data categories transmitted to agricultural machinery.

[0066] At the same time, to ensure that the bandwidth does not exceed the network's existing carrying capacity, constraints are set for bandwidth allocation:

[0067] ;

[0068] Step 5: When the client receives agricultural machinery data, the data transmission frequency is monitored in real time through a heartbeat monitoring mechanism according to the data transmission delay requirements. If agricultural machinery operation data is not received within a certain period of time, it is determined to be a transmission abnormality. The agricultural machinery operation trajectory data is predicted and compensated by using the agricultural machinery motion model on the CPU side.

[0069] Specifically, this invention sets up a heartbeat monitoring mechanism for real-time agricultural machinery trajectory data, setting the monitoring frequency to 200ms. If no trajectory data is received within 200ms, it is determined to be a data transmission anomaly. On the CPU side, the position of the agricultural machinery is predicted using an agricultural machinery motion model, and a state vector is defined:

[0070] ;

[0071] in, For the coordinates of the centroid of agricultural machinery, Linear velocity, For heading angle, ω is the angular velocity.

[0072] State prediction equation:

[0073] ;

[0074] In time State prediction, In time The state vector, This is the state transition matrix, which transfers the state from the previous moment to the current moment according to the motion model.

[0075] ;

[0076] With a prediction time step of 200ms, and The heading trigonometric function terms are used to project velocity onto position. This is the lateral position correction term caused by the change in heading. This is the longitudinal position correction term caused by changes in heading.

[0077] Step 6: After data transmission is restored to normal, dynamically adjust and update the agricultural machinery motion model and perform reverse smoothing on the predicted data to achieve retrospective correction of historical data and ensure the continuity of agricultural machinery operation data.

[0078] Specifically, if the client receives agricultural machinery data normally within 1 second after a data transmission anomaly alarm, it is considered that data transmission has recovered. After data transmission is recovered, timestamp alignment compensation is performed on its trajectory data.

[0079] ;

[0080] in, This represents the total duration of data transmission anomalies. This is the current timestamp. This is the last valid timestamp before the data transmission anomaly. Based on timestamp alignment, the state sequence generated by the prediction model during the data transmission anomaly period in step 5 is reconstructed to reconstruct the state for the missing time period:

[0081] ;

[0082] This is the last state vector before the data transmission anomaly. The duration is The state transition matrix, This is the state transition function. For integration time variable ( arrive ).

[0083] Step 7: Based on the hierarchical transmission strategy and prediction compensation mechanism of the above data, continuous and stable agricultural machinery status data is obtained. In order to realize the fast rendering of agricultural machinery data on Cesium on the client side, the Cesium basic terrain data and image data are divided into data blocks according to certain rules on the GPU side.

[0084] Specifically, the Cesium base terrain data and imagery data are adjusted according to the current zoom level. The data is divided into four blocks using a four-block method. At map level 0, the data is divided into four blocks along the 0° latitude and longitude line. Each time the map is zoomed in, the current data block is further divided along its center line. The total number of data blocks is [number missing]. The data block index is ,in These represent the data block numbers in the longitude and latitude directions, respectively. .

[0085] Each data block contains longitudes The range is:

[0086] ;

[0087] The dimensions contained in each data block The range is:

[0088] ;

[0089] Step 8: Dynamically load data based on the data blocks divided in Step 7, the current scaling level, and the real-time agricultural machinery operation data to reduce data rendering pressure and data update latency.

[0090] Specifically, based on the current location coordinates transmitted back by the agricultural machinery. Calculate the index number of the data block it belongs to:

[0091] ;

[0092] For ease of the next calculation, let:

[0093] ;

[0094] According to the flight direction of agricultural machinery Further calculate the set of data block index numbers for dynamic rendering:

[0095] ;

[0096] Finally, client-side rendering is completed according to the data block index set.

[0097] Another aspect of the present invention provides a low-latency update device for agricultural machinery operation data based on remote communication, comprising the following modules:

[0098] The module divides the operational data at the agricultural machinery end into three levels based on latency sensitivity, from high to low: safety alarm, real-time operation, and agricultural machinery status, and assigns corresponding transmission priorities.

[0099] The calculation module assigns an importance coefficient to each type of subdivided data based on the aforementioned priority, and calculates dynamic priority by combining the rate of change of control instructions with the current network bandwidth availability.

[0100] The adaptation module detects network bandwidth availability in real time and dynamically adjusts the weight coefficients participating in priority calculation accordingly, so that the priority adapts to network conditions in real time.

[0101] The allocation module adaptively allocates the total bandwidth based on dynamic priorities and the minimum bandwidth requirements of various types of data, and the total allocation amount does not exceed the network capacity limit.

[0102] The monitoring module monitors the real-time trajectory data of agricultural machinery at fixed intervals. If no data is received within a certain time, the transmission is deemed abnormal, and the CPU-side agricultural machinery motion model is activated to predict the position of the agricultural machinery in order to compensate for the missing trajectory data.

[0103] The correction module determines that the transmission has been restored when the heartbeat monitoring receives data again within a specified period of time. Then, it uses the latest data to recalibrate the parameters of the agricultural machinery motion model and performs reverse smoothing between the predicted segment and the measured segment to correct the historical data.

[0104] The indexing module, on the GPU side, divides the basic map data into quadtree blocks according to the current zoom level and creates an index;

[0105] The rendering module dynamically loads the required data blocks and completes client-side rendering based on the real-time location, heading, and index of the agricultural machinery.

[0106] Another aspect of the present invention provides an electronic device, comprising: one or more processors; and a memory for storing one or more programs, wherein, when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to implement the method.

[0107] Another aspect of the present invention provides a computer-readable storage medium having executable instructions stored thereon, which, when executed by a processor, cause the processor to implement the method described thereon.

Claims

1. A low-latency update method for agricultural machine operation data based on remote communication, characterized by, The method comprises the following steps: Step 1: dividing the work data into three levels of safety warning, real-time work and agricultural machinery state according to the delay sensitivity from high to low at the agricultural machinery end, and assigning corresponding transmission priorities; Step 2: assigning importance coefficients to each type of subdivided data based on the priorities, and calculating dynamic priorities by combining the control instruction change rate and the current network bandwidth availability rate; ； wherein, a dynamic priority of each type of data for transmission, a fixed priority of each type of data for transmission, an importance coefficient of each type of data for transmission, a control instruction change rate, a current network bandwidth availability rate, , and a dynamic weight coefficient, control instruction change rate represents the incremental change of control instruction in unit time; Step 3: detecting the network bandwidth availability rate in real time, and dynamically adjusting the weight coefficients involved in the priority calculation in step 2 according to the network condition, so that the priority is adapted to the network condition in real time; Step 4: adaptively allocating the total bandwidth according to the dynamic priorities obtained in step 2 and the minimum bandwidth requirements of each type of data, and the total allocation amount does not exceed the upper limit of the network carrying capacity; Step 5: the client performs heartbeat monitoring on the real-time trajectory data of the agricultural machinery at a fixed period, and if the data is not received within the time limit, it is determined that the transmission is abnormal, and the CPU side agricultural machinery motion model is started to predict the position of the agricultural machinery to compensate for the missing period of trajectory data; Step 6: when the heartbeat monitoring receives data again within a continuous specified time period, it is determined that the transmission is restored, then the latest data is used to recalibrate the agricultural machinery motion model used in step 5, and reverse smoothing is performed between the predicted segment and the measured segment to correct the historical data; Step 7: dividing the map basic data into quadtree blocks according to the current zoom level at the GPU side and establishing an index; Step 8: dynamically loading the required data block and completing the client rendering according to the real-time position and heading of the agricultural machinery and the index obtained in step 7.

2. The method of claim 1, wherein, The delay requirements of the safety warning data, real-time work data and agricultural machinery state data in step 1 are less than 50 ms, 50-200 ms and 200-500 ms respectively, and the priority numbers 1, 2 and 3 decrease in turn according to the strictness of the delay requirements.

3. The method of claim 1, wherein, The importance coefficients in step 2 decrease in turn according to the order of safety warning, path deviation warning, emergency stop signal, real-time trajectory coordinates, work speed, steering angle, work width, hydraulic pressure, oil consumption and water temperature, and the coefficient values are set in stages between 0.3 and 1.

0.

4. The method of claim 1, wherein, In step 3, when the network bandwidth availability rate decreases, the weight coefficient of the safety warning type data is increased to ensure that its dynamic priority is always higher than that of other data.

5. The method of claim 1, wherein, The heartbeat monitoring period in step 5 is 200 ms, and the step length of the prediction compensation after the missing is also set to 200 ms, which is consistent with the monitoring period.

6. The method of claim 1, wherein, The continuous specified time period in step 6 is 1s, and the backtracking correction adopts the method of timestamp alignment and then smoothing, so that the predicted trajectory and the measured trajectory are continuous and derivable at the switching point.

7. The method of claim 1, wherein, The quadtree blocks are divided again into four parts along the central meridian and central parallel every time the map level increases by one, and only the block where the agricultural machinery is located and the adjacent block in front of the heading direction are loaded during dynamic loading, so that the data amount of single rendering is controlled within a constant range. The method comprises the following modules:

8. A device for updating agricultural operation data with low latency based on remote communication, which implements the method of claim 1, characterized in that, A division module divides the work data into three levels of safety warning, real-time work and agricultural machinery state according to the delay sensitivity from high to low at the agricultural machinery end, and assigns corresponding transmission priorities; ​ A calculation module, which gives each type of subdivided data an importance coefficient based on the priority, and calculates a dynamic priority by combining the control instruction change rate and the current network bandwidth availability rate; An adaptation module, which detects the network bandwidth availability rate in real time, and dynamically adjusts the weight coefficient participating in the priority calculation according to the network bandwidth availability rate, so that the priority is adapted to the network condition in real time; An allocation module, which performs adaptive allocation of the total bandwidth according to the dynamic priority and the minimum bandwidth requirement of each type of data, and the total allocation amount does not exceed the network carrying upper limit; A monitoring module, which performs heartbeat monitoring on the real-time track data of the agricultural machine at a fixed period, and determines that the transmission is abnormal if the data is not received within a timeout period, and starts the CPU-side agricultural machine motion model to predict the position of the agricultural machine to compensate for the missing track data in the period; A correction module, which determines that the transmission is restored when the heartbeat monitoring receives the data again within a continuous specified period of time, then recalibrates the parameters of the agricultural machine motion model using the latest data, and performs reverse smoothing between the predicted segment and the measured segment to perform backtracking correction on the historical data; An index module, which divides and indexes the map basic data according to the current zoom level by quadtree on the GPU side; A rendering module, which dynamically loads the required data block and completes the client rendering according to the real-time position, heading of the agricultural machine and the index.

9. An electronic device, comprising: comprise: one or more processors; a memory for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the method of any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, executable instructions stored thereon, which are executed by a processor to cause the processor to implement the method of any one of claims 1 to 7.

Images