Agricultural production irrigation resource scheduling method and system based on edge computing

Abstract

The invention relates to the technical field of agricultural production irrigation resource scheduling based on edge computing, in particular to an agricultural production irrigation resource scheduling method and system based on edge computing, and the method comprises the following steps: sending an irrigation instruction, and obtaining the communication delivery state information of the irrigation instruction; physical action state information of irrigation execution equipment is obtained; acquiring measurement data of irrigation measurement equipment; performing logic cross validation based on the communication delivery state information, the physical action state information and the measurement data to identify the fault type of the irrigation task; and taking corresponding irrigation resource scheduling measures according to the fault type. The accuracy of fault identification and the reliability of irrigation resource scheduling are improved.

Description

Technical Field

[0001] This invention relates to the technical field of edge computing-based agricultural irrigation resource scheduling, specifically to a method and system for edge computing-based agricultural irrigation resource scheduling. Background Technology

[0002] In modern agricultural production, smart irrigation systems widely utilize edge computing technology to achieve precise water resource management and optimized crop yields. These systems typically deploy edge computing nodes in the field to collect sensor data and control irrigation equipment. However, the complexity and dynamism of the agricultural environment, particularly the challenges of wireless communication and the aging of physical infrastructure, often affect the reliability and stability of these systems. Summary of the Invention

[0003] The purpose of this invention is to address the aforementioned shortcomings by proposing a method and system for scheduling agricultural irrigation resources based on edge computing.

[0004] The present invention adopts the following technical solution: A method for scheduling agricultural irrigation resources based on edge computing, comprising the following steps: Send irrigation commands and obtain communication delivery status information for the irrigation commands; Obtain the physical motion status information of the irrigation equipment; Acquire measurement data from irrigation measurement equipment; Logical cross-validation is performed based on communication delivery status information, physical action status information, and measurement data to identify the fault type of the irrigation task. Based on the type of fault, corresponding irrigation resource scheduling measures will be taken.

[0005] Through this technical solution, this application effectively solves the problem of inaccurate state judgment caused by a single information source in traditional irrigation systems by cross-validating multi-dimensional information. It significantly improves the accuracy of fault identification and the reliability of irrigation resource scheduling, and overcomes the shortcomings of existing technologies such as frequent state machine switching and inability to stably judge the actual operating state.

[0006] Furthermore, the steps for obtaining the physical motion status information of the irrigation execution equipment include: After receiving the open or close command from the edge node, the solenoid valve controller activates the power drive circuit and applies voltage to the solenoid valve coil. The solenoid valve controller performs high-frequency sampling of the transient electrical response of the solenoid valve coil to obtain sampling data; The solenoid valve controller extracts transient electromagnetic features from the sampled data; The solenoid valve controller compares the transient electromagnetic characteristics with the preset valve core position characteristics to obtain the comparison result; The solenoid valve controller determines whether the valve core is fully open based on the comparison results. The solenoid valve controller generates a physical action confirmation signal based on whether the valve core is fully open, and uses the physical action confirmation signal as the physical action status information of the irrigation execution equipment.

[0007] Furthermore, the steps for identifying the fault type of the irrigation task by performing logical cross-validation based on communication delivery status information, physical action status information, and measurement data include: A consistency assessment is performed on communication delivery status information, physical action status information, and measurement data to obtain the consistency assessment results. When there is a mismatch or ambiguity in the consistency assessment results, the initial confidence level is assigned to each possible failure type by combining the priority of the current irrigation task, the success rate of similar historical tasks, and the current environmental parameters. The communication delivery status information, physical action status information, and measurement data are weighted according to the preset fuzzy evidence parsing rules to correct the initial confidence level and obtain the corresponding comprehensive confidence level. Activate the behavior observation window to continuously monitor key feedback information, and dynamically update the overall confidence level based on the changing trends of key feedback information; Based on the dynamically updated overall confidence distribution, the fault scenario with the highest overall confidence is selected to make the final judgment.

[0008] Furthermore, the steps for identifying the fault type of the irrigation task by performing logical cross-validation based on communication delivery status information, physical action status information, and measurement data include: A consistency assessment is performed on communication delivery status information, physical action status information, and measurement data to obtain the consistency assessment results. When there is incomplete matching or ambiguity in the consistency assessment results, the current environmental parameters are continuously monitored. Based on the changes in the current environmental parameters, the weighting coefficients of each communication delivery status information, physical action status information, and measurement data in the fuzzy evidence parsing rules are dynamically adjusted. Based on the changes in the current environmental parameters, the initial confidence of each possible fault type is dynamically adjusted. Based on the current environmental parameters and weighting coefficients, the initial confidence is corrected to obtain the corresponding comprehensive confidence. Activate the behavior observation window to continuously monitor key feedback information, and dynamically update the overall confidence level based on the changing trends of key feedback information; Based on the dynamically updated overall confidence distribution, the fault scenario with the highest overall confidence is selected to make the final judgment.

[0009] Furthermore, the steps for identifying the fault type of the irrigation task by performing logical cross-validation based on communication delivery status information, physical action status information, and measurement data include: A consistency assessment is performed on communication delivery status information, physical action status information, and measurement data to obtain the consistency assessment results. When there is incomplete matching or ambiguity in the consistency assessment results, the irrigation area is divided into multiple sub-regions, and the local environmental parameters of each sub-region are obtained. Based on the local environmental parameters of each sub-region, the weighting coefficients of each communication delivery status information, physical action status information, and measurement data in the fuzzy evidence parsing rules of each sub-region are dynamically adjusted. Based on the local environmental parameters of each sub-region, the initial confidence of each possible fault type in each sub-region is dynamically adjusted. Based on the local environmental parameters and weighting coefficients of each sub-region, the initial confidence of each sub-region is corrected to obtain the corresponding comprehensive confidence. Activate the behavior observation window to continuously monitor key feedback information, and dynamically update the overall confidence level based on the changing trends of key feedback information; Based on the dynamically updated overall confidence distribution, the fault scenario with the highest overall confidence is selected to make the final judgment.

[0010] Furthermore, the steps for selecting the fault scenario with the highest overall confidence level to make a final judgment include: When the overall confidence difference of multiple fault scenarios is less than a preset threshold, the previously determined fault type judgment result is maintained. At the same time, a short-term data smoothing mechanism is activated to perform a moving average on the key feedback information received subsequently, which helps to improve the stability of irrigation. During the operation of the short-term data smoothing mechanism, the trend of the overall confidence level change of the fault scenario with the highest overall confidence level is continuously monitored; When the overall confidence level of the fault scenario with the highest overall confidence level is consistently higher than that of other fault scenarios, and the duration exceeds the preset time, the system switches to a new fault type for judgment.

[0011] Furthermore, maintaining the previously determined fault type assessment result can also be achieved through the following steps: Start the stabilization timer; Activate the disturbance intensity assessment module to monitor environmental disturbance parameters; When the stabilization timer detects that the overall intensity of the environmental disturbance parameters exceeds the preset disturbance threshold within the preset stabilization duration, and the difference in overall confidence between the fault scenario with the highest overall confidence and the fault scenario with the second highest overall confidence fluctuates frequently, the stabilization timer duration is extended, and the previously determined fault type judgment result is maintained. When the stabilization timer ends and the overall intensity of the environmental disturbance parameters does not exceed the preset disturbance threshold within the entire preset stabilization period, and the overall confidence of the fault scenario with the highest overall confidence remains consistently higher than the overall confidence of other fault scenarios, the system switches to a new fault type for judgment.

[0012] Furthermore, the steps of extending the duration of the stabilization timer and maintaining the previously determined fault type judgment result include: Monitor the type, intensity, and duration of environmental disturbance parameters; Monitor the fluctuation frequency and amplitude of the difference in overall confidence between the fault scenario with the highest overall confidence and the fault scenario with the second highest overall confidence; Based on the type, intensity, and duration of the environmental disturbance parameters, as well as the fluctuation frequency and amplitude of the difference in the overall confidence level between the fault scenario with the highest overall confidence level and the fault scenario with the second highest overall confidence level, calculate the specific value of extending the timing duration of the judgment stabilization timer. Based on the specific value of the extended stabilization timer, the duration of the stabilization timer is extended, and the previously determined fault type judgment result is maintained.

[0013] Furthermore, based on the type, intensity, and duration of the environmental disturbance parameters, as well as the fluctuation frequency and amplitude of the difference in the overall confidence level between the fault scenario with the highest overall confidence level and the fault scenario with the second highest overall confidence level, the steps for calculating the specific value of extending the timing duration of the judgment stabilization timer include: The weighted combination is obtained based on the type, intensity, and duration of environmental disturbance parameters, as well as the fluctuation frequency and amplitude of the difference in the overall confidence level between the fault scenario with the highest overall confidence level and the fault scenario with the second highest overall confidence level. Based on the weighted combination, the initial extended duration is obtained by weighted summation of the fluctuation frequency and amplitude of the fluctuation frequency and amplitude of the type, intensity, and duration of the environmental disturbance parameters, as well as the difference in comprehensive confidence between the fault scenario with the highest comprehensive confidence and the fault scenario with the second highest comprehensive confidence. Based on the instantaneous change rate of environmental disturbance parameters and the instantaneous fluctuation trend of the difference in comprehensive confidence between the fault scenario with the highest comprehensive confidence and the fault scenario with the second highest comprehensive confidence, the initial extension duration is corrected to obtain the specific value of the timing duration of the extended judgment stabilization timer.

[0014] This application also discloses an edge computing-based agricultural irrigation resource scheduling system, applied to an edge computing-based agricultural irrigation resource scheduling method. The system includes: The instruction sending module sends irrigation instructions and obtains the communication delivery status information of the irrigation instructions; The physical action acquisition module acquires the physical action status information of the irrigation execution equipment. The measurement data acquisition module acquires measurement data from the irrigation measurement equipment. The fault identification module performs logical cross-validation based on communication delivery status information, physical action status information, and measurement data to identify the fault type of the irrigation task. The scheduling module takes appropriate irrigation resource scheduling measures based on the type of fault.

[0015] This application provides a system that can effectively execute the above-mentioned methods through this technical solution. Through modular design, it realizes comprehensive monitoring and cross-verification of irrigation commands, physical actions and measurement data, thereby accurately identifying the fault type of irrigation task and taking corresponding scheduling measures. This solves the problems of inaccurate fault identification and untimely scheduling in existing systems in complex agricultural environments, and improves the intelligence and reliability of the entire irrigation system.

[0016] This application significantly improves the accuracy and reliability of fault identification through a multi-dimensional, logically cross-validation mechanism, enabling the irrigation management system to make stable and accurate judgments, effectively avoiding resource waste or insufficient irrigation caused by misjudgment, thereby enhancing the edge computing system's ability to achieve precise and efficient irrigation in agricultural production.

[0017] To further understand the features and technical content of the present invention, please refer to the following detailed description and drawings of the present invention. However, the drawings provided are for reference and illustration only and are not intended to limit the present invention. Attached Figure Description

[0018] Figure 1 This is a flowchart of a method for scheduling agricultural irrigation resources based on edge computing according to the present invention. Figure 2 This is a schematic diagram of the structure of an edge computing-based agricultural irrigation resource scheduling system according to the present invention. Detailed Implementation

[0019] The following specific embodiments illustrate the implementation of the present invention. Those skilled in the art can understand the advantages and effects of the present invention from the content disclosed in this specification. The present invention can be implemented or applied through other different specific embodiments, and various details in this specification can also be modified and changed based on different viewpoints and applications without departing from the spirit of the present invention. Furthermore, the accompanying drawings of the present invention are for simple illustrative purposes only and are not depictions of actual dimensions; this is stated in advance. The following embodiments will further describe the relevant technical content of the present invention in detail, but the disclosed content is not intended to limit the scope of protection of the present invention.

[0020] This embodiment provides a method and system for scheduling agricultural irrigation resources based on edge computing, combined with... Figure 1 and Figure 2 As shown.

[0021] refer to Figure 1 A method for scheduling agricultural irrigation resources based on edge computing, comprising the following steps: Send irrigation commands and obtain communication delivery status information for the irrigation commands; Obtain the physical motion status information of the irrigation equipment; Acquire measurement data from irrigation measurement equipment; Logical cross-validation is performed based on communication delivery status information, physical action status information, and measurement data to identify the fault type of the irrigation task. Based on the type of fault, corresponding irrigation resource scheduling measures will be taken.

[0022] This application, by introducing a multi-source information cross-validation mechanism, can effectively solve the problems of inaccurate state judgment and unstable scheduling caused by communication, execution and measurement equipment failures in traditional intelligent irrigation systems in complex agricultural environments, and significantly improves the reliability of irrigation tasks and the accuracy of resource scheduling.

[0023] In this application, "edge computing" refers to a computing paradigm that performs data processing and analysis at or near the data source, aiming to reduce data transmission latency, lower network bandwidth consumption, and improve system response speed. In agricultural irrigation scenarios, edge computing nodes are typically deployed in the field, directly interacting with irrigation execution and measurement equipment. "Irrigation command" refers to control commands issued by the edge computing node or a higher-level control system to the irrigation execution equipment, such as commands to open or close a solenoid valve. "Communication delivery status information" refers to whether the irrigation command was successfully delivered to the target device during transmission and whether confirmation feedback was received from the target device. "Physical action status information" refers to whether the physical components of the irrigation execution equipment (such as a solenoid valve) performed the actions required by the command after receiving it, such as whether the valve was fully open or closed. "Measurement data" refers to environmental or physical quantity data related to the irrigation process collected by irrigation measurement equipment (such as flow meters and humidity sensors). "Logical cross-validation" refers to comparing and analyzing information from different sources to determine whether there are contradictions or inconsistencies between these information, thereby inferring potential faults. "Fault types of irrigation tasks" refer to various abnormal situations that may occur during irrigation, such as communication failures, equipment execution failures, and measurement equipment failures. "Irrigation resource scheduling measures" refer to the automatic or semi-automatic response strategies adopted by the system based on the identified fault types to restore the normal execution of irrigation tasks or optimize resource allocation. The implementation environment of this application typically includes one or more edge computing nodes, multiple irrigation execution devices (such as solenoid valves and water pumps), multiple irrigation measurement devices (such as flow meters and soil moisture sensors), and a wireless communication network.

[0024] Specifically, various methods can be used to send irrigation commands and obtain communication delivery status information. For example, an edge computing node can send commands to the irrigation execution device to open or close a solenoid valve via a wireless communication module. After sending the command, the edge computing node can wait for an acknowledgment signal from the irrigation execution device. If an acknowledgment signal is received, the communication delivery status information can be marked as "delivered"; if no acknowledgment signal is received, or if an error feedback is received within a preset time, the communication delivery status information can be marked as "not delivered" or "delivery abnormal." As another implementation, the edge computing node can periodically send heartbeat packets to the irrigation execution device and determine the health status of the communication link based on the response to the heartbeat packets, thereby indirectly obtaining communication delivery status information. For example, if multiple consecutive heartbeat packets do not receive a response, it can be assumed that there is a problem with the communication link, thus affecting the delivery of commands.

[0025] Regarding the acquisition of physical action status information of irrigation actuators, one approach is for the actuator, after receiving a command and performing a physical action, to use a simple built-in sensor (e.g., a microswitch or a photoelectric sensor) to detect whether its physical components have reached a preset position. For example, when a solenoid valve is fully open or closed, it can trigger a microswitch, generating an electrical signal that is sent back to the edge computing node as physical action status information. Another approach is for the irrigation actuator to integrate a simple current or voltage sensor to monitor the electrical signals driving its physical components. For example, when the solenoid valve coil is energized, its current change curve is monitored, and the characteristics of the current curve are analyzed to determine whether the valve core has fully moved into position. If the current curve deviates significantly from the preset curve during normal operation, it may indicate that the physical action has not been fully executed.

[0026] Regarding the acquisition of measurement data from irrigation measurement equipment (such as flow meters and soil moisture sensors), these devices can periodically collect their data and transmit it to edge computing nodes via wireless communication modules. For example, a flow meter can measure the instantaneous flow rate every minute and send this flow rate data as measurement data. A soil moisture sensor can measure soil moisture every five minutes and send the moisture data as measurement data. After receiving this data, the edge computing node can perform preliminary storage and processing for subsequent logical cross-validation.

[0027] To identify irrigation task fault types through logical cross-validation based on communication delivery status information, physical action status information, and measurement data, edge computing nodes can design a set of logical rules. For example, when the communication delivery status information of the irrigation command shows "delivered" and the physical action status information of the irrigation execution equipment shows "activated," but the flow data of the irrigation measurement equipment remains zero or far below expectations, the system can initially determine that there may be a flow meter malfunction or pipeline blockage. As another example, if the communication delivery status information shows "not delivered," and the physical action status information and measurement data remain unchanged, it may be determined as a communication fault. Through preset logical rules, the system can compare and analyze information from different sources to identify potential fault types in the irrigation task.

[0028] Regarding the implementation of corresponding irrigation resource scheduling measures based on fault type, once the fault type is identified, the system can take appropriate scheduling measures according to preset strategies. For example, if a communication fault is identified, the system can attempt to switch to a backup communication link or increase the interval between command retryes to avoid network congestion. If a solenoid valve execution fault is identified, the system can attempt to resend the open or close command, or after multiple failed attempts, mark the solenoid valve as faulty and notify maintenance personnel for repair, while adjusting the irrigation plan to redirect water flow to other normally functioning areas. If a flowmeter fault is identified, the system can temporarily ignore the flowmeter's data and instead rely on information from other sensors (such as soil moisture sensors), or make estimates based on historical data and crop models to maintain irrigation continuity, and schedule the calibration or replacement of the flowmeter.

[0029] This application further proposes steps for obtaining physical motion status information of irrigation execution equipment, including: After receiving the open or close command from the edge node, the solenoid valve controller activates the power drive circuit and applies voltage to the solenoid valve coil. The solenoid valve controller performs high-frequency sampling of the transient electrical response of the solenoid valve coil to obtain sampling data; The solenoid valve controller extracts transient electromagnetic features from the sampled data; The solenoid valve controller compares the transient electromagnetic characteristics with the preset valve core position characteristics to obtain the comparison result; The solenoid valve controller determines whether the valve core is fully open based on the comparison results. The solenoid valve controller generates a physical action confirmation signal based on whether the valve core is fully open, and uses the physical action confirmation signal as the physical action status information of the irrigation execution equipment.

[0030] Specifically, a solenoid valve controller can be understood as an integrated circuit or microcontroller unit configured to receive control commands from edge nodes and drive the solenoid valve to perform actions. Edge nodes are devices located at agricultural production sites, possessing certain computing and storage capabilities, responsible for receiving commands from the cloud and sending control commands to field equipment. When the solenoid valve controller receives an open or close command, its internal power drive circuit is activated. This circuit applies power voltage to the solenoid valve coil, thereby generating a magnetic field that drives the valve core to move.

[0031] The solenoid valve coil generates a unique transient electrical response when energized or de-energized, containing rich information about the valve spool's motion. To capture these subtle changes, the solenoid valve controller performs high-frequency sampling of the solenoid valve coil's transient electrical response to obtain detailed sampling data. High-frequency sampling ensures accurate capture of the transient process and avoids information loss.

[0032] Furthermore, the solenoid valve controller extracts transient electromagnetic features from the sampled data. These features can be parameters such as voltage, current, impedance, or their rate of change, which can reflect the position, speed, and whether the valve core is stuck during operation. For example, these features can be extracted by analyzing the rising edge, falling edge, peak value, valley value, and duration of the current waveform.

[0033] Subsequently, the solenoid valve controller compares the extracted transient electromagnetic features with preset valve spool position features. These preset valve spool position features are electromagnetic feature templates established in advance through experiments or simulations under normal solenoid valve operating conditions, corresponding to different valve spool positions such as fully open, fully closed, or partially open / closed. The comparison result indicates the degree of match between the current actual valve spool position and the desired position.

[0034] Based on the comparison results, the solenoid valve controller determines whether the valve core is fully open. For example, if the transient electromagnetic characteristics closely match the preset characteristics of "fully open," the valve core is determined to be fully open. Otherwise, there may be abnormalities such as partial opening, jamming, or no action.

[0035] Finally, the solenoid valve controller generates a physical action confirmation signal based on whether the valve core is fully open. This signal is a clear indication that the physical action of the irrigation actuator (such as the solenoid valve) has been completed as required by the instruction. This physical action confirmation signal is then reported as physical action status information of the irrigation actuator to the edge node or a higher-level control system.

[0036] The steps for identifying the fault type of irrigation task based on communication delivery status information, physical action status information, and measurement data include: A consistency assessment is performed on communication delivery status information, physical action status information, and measurement data to obtain the consistency assessment results. When there is a mismatch or ambiguity in the consistency assessment results, the initial confidence level is assigned to each possible failure type by combining the priority of the current irrigation task, the success rate of similar historical tasks, and the current environmental parameters. The communication delivery status information, physical action status information, and measurement data are weighted according to the preset fuzzy evidence parsing rules to correct the initial confidence level and obtain the corresponding comprehensive confidence level. Activate the behavior observation window to continuously monitor key feedback information, and dynamically update the overall confidence level based on the changing trends of key feedback information; Based on the dynamically updated overall confidence distribution, the fault scenario with the highest overall confidence is selected to make the final judgment.

[0037] Specifically, consistency assessment refers to comparing and verifying the communication delivery status information of irrigation instructions obtained from different sources, the physical action status information of irrigation execution equipment, and the measurement data of irrigation measurement equipment to determine whether there are logical conflicts or inconsistencies among these pieces of information. For example, if the instruction has been delivered and the physical action shows that the valve is open, but the measurement data shows no water flow, then there is an inconsistency.

[0038] When the consistency assessment results indicate incomplete matching or ambiguity in the data, the system does not immediately make a judgment but initiates a more complex fault identification process. At this point, it combines the priority of the current irrigation task (e.g., the judgment of emergency irrigation tasks requires more caution), the success rate of similar historical tasks (e.g., common fault patterns in a certain area or crop type), and current environmental parameters (e.g., temperature, humidity, wind speed, etc., factors that may affect equipment operation or sensor readings) to assign an initial confidence level to each possible fault type (e.g., valve jamming, sensor failure, communication interruption, pipeline blockage, etc.). The initial confidence level reflects the probability of a certain fault occurring under the current known conditions. Further, according to preset fuzzy evidence resolution rules, the communication delivery status information, physical action status information, and measurement data are weighted. The fuzzy evidence resolution rules can be a set of rules based on expert experience or machine learning models to handle uncertain information. Through weighted processing, the initial confidence level can be corrected to obtain a more accurate comprehensive confidence level, where the weights can be dynamically adjusted according to factors such as data reliability and importance.

[0039] Subsequently, a behavior observation window is activated. This window is a time window used to continuously monitor key feedback information. Key feedback information can include subsequent measurement data, equipment status reports, and changes in environmental parameters. By observing the trends in these key feedback information, such as whether water flow gradually recovers, whether valve status changes repeatedly, and whether sensor readings fluctuate abnormally, the previously calculated overall confidence level can be dynamically updated. This dynamic update mechanism enables the system to make more accurate judgments based on time-series data, avoiding erroneous decisions based on transiently uncertain data.

[0040] Finally, based on the dynamically updated overall confidence distribution, the system selects the fault scenario with the highest overall confidence as the final fault type determination. This means that the system will choose the most likely fault to occur as the identification result, thereby guiding subsequent irrigation resource scheduling measures.

[0041] In some preferred embodiments, it is assumed that during an irrigation task, an edge node sends an opening command to the solenoid valve controller. The system first obtains communication delivery status information of the irrigation command, indicating that the command has been successfully sent. Subsequently, it obtains the physical action status information of the irrigation execution equipment; for example, the solenoid valve controller determines, through transient electromagnetic characteristic analysis, that the valve core is not fully open. Simultaneously, the measurement data from the irrigation measuring equipment shows zero water flow within the expected timeframe.

[0042] At this point, a consistency assessment reveals an incomplete match between successful communication, valve not fully open, and no water flow. The system does not immediately diagnose a valve malfunction but instead initiates a fuzzy evidence analysis process. Assuming the current irrigation task has a high priority (e.g., crops are in a critical growth stage), historical data shows occasional flow meter drift in the area, and current environmental parameters indicate low soil moisture, the system assigns a high initial confidence level to "valve jamming," a medium initial confidence level to "flow meter malfunction," and a low initial confidence level to "pipe blockage."

[0043] Next, based on the preset fuzzy evidence parsing rules, the communication delivery status information (high reliability), physical action status information (medium reliability, possibly affected by the environment), and measurement data (medium reliability, possibly affected by sensor drift) are weighted and processed to correct the initial confidence level. For example, since physical action status information directly reflects the equipment's performance, its weight may be increased, further raising the overall confidence level of "valve jamming".

[0044] Subsequently, the behavior observation window is activated to continuously monitor key feedback information. Over the next 5 minutes, the system continuously monitors the flow meter data and the physical status information of the solenoid valve controller. If, during these 5 minutes, the flow meter data remains zero, and the solenoid valve controller repeatedly attempts to open the valve but the physical status information still indicates that the valve core is not fully open, the overall confidence level for "valve jamming" will remain consistently at its highest. If, during these 5 minutes, the flow meter data suddenly shows a brief non-zero reading but then returns to zero, and the solenoid valve controller reports that the valve is open, the overall confidence level for "flow meter malfunction" may dynamically increase.

[0045] Finally, based on the dynamically updated overall confidence distribution, if the overall confidence of "valve jamming" continues to be significantly higher than other fault scenarios, the system will ultimately determine it as a "valve jamming" fault and take corresponding scheduling measures, such as sending maintenance instructions or switching to the backup irrigation path.

[0046] This application further proposes steps for identifying the fault type of irrigation task by performing logical cross-validation based on communication delivery status information, physical action status information, and measurement data, including: A consistency assessment is performed on communication delivery status information, physical action status information, and measurement data to obtain the consistency assessment results. When there is incomplete matching or ambiguity in the consistency assessment results, the current environmental parameters are continuously monitored. Based on the changes in the current environmental parameters, the weighting coefficients of each communication delivery status information, physical action status information, and measurement data in the fuzzy evidence parsing rules are dynamically adjusted. Based on the changes in the current environmental parameters, the initial confidence of each possible fault type is dynamically adjusted. Based on the current environmental parameters and weighting coefficients, the initial confidence is corrected to obtain the corresponding comprehensive confidence. Activate the behavior observation window to continuously monitor key feedback information, and dynamically update the overall confidence level based on the changing trends of key feedback information; Based on the dynamically updated overall confidence distribution, the fault scenario with the highest overall confidence is selected to make the final judgment.

[0047] Specifically, during logical cross-validation, a consistency assessment is first performed on communication delivery status information, physical action status information, and measurement data to preliminarily determine whether there are any contradictions or anomalies among these information. When the consistency assessment results show incomplete matching or ambiguity, it indicates that there may be some kind of fault, but the specific fault type is not yet clear.

[0048] At this time, the system continuously monitors the current environmental parameters. These parameters can be understood as external conditions affecting the operation of the irrigation system, such as ambient temperature, humidity, wind speed, soil moisture, soil pH, and light intensity. These parameters can be acquired in real time through environmental sensors deployed on edge nodes.

[0049] Based on changes in current environmental parameters, the system dynamically adjusts the weighting coefficients of various communication delivery status information, physical action status information, and measurement data in the fuzzy evidence analysis rules. These weighting coefficients measure the importance of different types of evidence in fault identification. For example, in a hot and dry environment with high evaporation, the weight of soil moisture measurement data may be increased; while in a windy environment, communication signals may be unstable, and the weight of communication delivery status information may be appropriately reduced. This dynamic adjustment ensures that the system can more rationally utilize various pieces of evidence for fault judgment under different environmental conditions.

[0050] Simultaneously, the system dynamically adjusts the initial confidence level for each possible fault type based on changes in current environmental parameters. The initial confidence level reflects the prior probability of a fault occurring under specific environmental conditions. For example, in a low-temperature environment, the initial confidence level for a pipe freezing fault type may be increased; while in a high-temperature environment, the initial confidence level for an equipment overheating fault type may be increased.

[0051] Subsequently, the initial confidence level is corrected based on the current environmental parameters and dynamically adjusted weighting coefficients to obtain the corresponding comprehensive confidence level. The comprehensive confidence level is a comprehensive assessment of the probability of occurrence for each possible failure type.

[0052] Based on this, the system will activate a behavior observation window to continuously monitor key feedback information. Key feedback information may include real-time operating data from irrigation execution equipment, continuous measurement data from irrigation measurement equipment, and internal diagnostic logs. By observing the changing trends of this key feedback information, the obtained overall confidence level can be dynamically updated, further improving the accuracy and real-time performance of fault diagnosis.

[0053] Finally, based on the dynamically updated comprehensive confidence distribution, the fault scenario with the highest comprehensive confidence is selected as the final fault type judgment result.

[0054] In some preferred embodiments, suppose that in an agricultural irrigation area, the edge node detects that the soil moisture measurement data is consistently below a set threshold, but the communication delivery status information of the irrigation command shows that the command has been successfully sent, and the physical action status information shows that the solenoid valve has been opened. In this case, the consistency evaluation results may not be completely matched or ambiguous.

[0055] If environmental parameters detect a sharp increase in ambient temperature and wind speed, the system will dynamically adjust the fuzzy evidence resolution rules based on these changes. For example, since high temperatures and strong winds may significantly increase water evaporation, the system may increase the initial confidence level of the fault scenario "rapid evaporation leading to soil water loss" and increase the weight of soil moisture measurement data in fault judgment. Meanwhile, considering that strong winds may interfere with communication signals, the weight of communication delivery status information may be slightly reduced.

[0056] Through this dynamic adjustment, the system is more likely to determine that the problem is "rapid evaporation leading to soil water loss" rather than "solenoid valve malfunction" or "communication failure." Subsequently, the behavior observation window continuously monitors whether soil moisture recovers after irrigation and whether the solenoid valve remains open. If soil moisture does not recover significantly after a period of irrigation, the judgment of "rapid evaporation" is further reinforced. Based on the dynamically updated comprehensive confidence level, the system ultimately determines that "rapid evaporation leads to soil water loss" and takes corresponding scheduling measures, such as increasing irrigation duration or adjusting irrigation frequency. This effectively solves the potential misjudgment problem caused by environmental changes and ensures the effective execution of irrigation tasks.

[0057] The steps for identifying the fault type of irrigation task based on communication delivery status information, physical action status information, and measurement data include: A consistency assessment is performed on communication delivery status information, physical action status information, and measurement data to obtain the consistency assessment results. When there is incomplete matching or ambiguity in the consistency assessment results, the irrigation area is divided into multiple sub-regions, and the local environmental parameters of each sub-region are obtained. Based on the local environmental parameters of each sub-region, the weighting coefficients of each communication delivery status information, physical action status information, and measurement data in the fuzzy evidence parsing rules of each sub-region are dynamically adjusted. Based on the local environmental parameters of each sub-region, the initial confidence of each possible fault type in each sub-region is dynamically adjusted. Based on the local environmental parameters and weighting coefficients of each sub-region, the initial confidence of each sub-region is corrected to obtain the corresponding comprehensive confidence. Activate the behavior observation window to continuously monitor key feedback information, and dynamically update the overall confidence level based on the changing trends of key feedback information; Based on the dynamically updated overall confidence distribution, the fault scenario with the highest overall confidence is selected to make the final judgment.

[0058] Specifically, dividing an irrigation area into multiple sub-regions refers to logically or physically dividing the entire irrigation area into several smaller, relatively homogeneous units based on factors such as geographical location, soil type, crop type, topographic elevation, or sensor distribution density. For example, a large farm can be divided based on different crop planting areas, different soil moisture sensor coverage areas, or different irrigation network branches. Obtaining local environmental parameters for each sub-region can be understood as acquiring data reflecting the current environmental conditions of that sub-region in real time or periodically through environmental sensors deployed within each sub-region (such as soil moisture sensors, temperature sensors, light sensors, wind speed sensors, etc.) or through geographic information system data, remote sensing data, etc. These parameters may include soil moisture, soil temperature, air temperature, light intensity, wind speed, rainfall, etc.

[0059] The process involves dynamically adjusting the weighting coefficients of communication delivery status information, physical action status information, and measurement data in the fuzzy evidence analysis rules for each sub-region based on local environmental parameters. This aims to better adapt the fault identification model to the specific environmental conditions of each sub-region. For example, in sub-regions with poor soil permeability, the weight of measurement data (such as flow meter readings) may be increased because it more directly reflects irrigation effectiveness; while in windy sub-regions, communication delivery status information may be given higher weight to address potential wireless communication interference. Dynamically adjusting the initial confidence level for each possible fault type in each sub-region involves predicting the probability of different fault scenarios based on local environmental parameters. For example, in sub-regions with severe soil salinization, the initial confidence level for pipe blockage or sensor failure may be appropriately increased. Finally, the initial confidence level of each sub-region is corrected based on its local environmental parameters and weighting coefficients to obtain a corresponding comprehensive confidence level. This aims to comprehensively consider the impact of the local environment on evidence weights and prior fault probabilities, thereby achieving a more accurate fault confidence assessment.

[0060] In some preferred embodiments, a large vineyard is assumed to be divided into three sub-regions: sub-region A (sandy soil, higher elevation), sub-region B (clay soil, lower elevation), and sub-region C (loam soil, near water source). During irrigation, the system independently identifies faults for each sub-region.

[0061] For example, in sub-region A, due to the high permeability of sandy soil and rapid water loss, and the potential for insufficient water pressure due to its higher elevation, the system will acquire soil moisture, soil temperature, and water pressure sensor data for this sub-region as local environmental parameters. Based on these parameters, the system may dynamically increase the initial confidence level of fault types such as "pipe leakage" or "insufficient pump pressure" and increase the weight of flow meter measurement data in the fuzzy evidence parsing rules.

[0062] In sub-region B, due to the poor permeability of clay soil, water easily accumulates. The system will acquire soil moisture, soil temperature, and rainfall data for this sub-region. At this time, the system may dynamically increase the initial confidence level of fault types such as "solenoid valve blockage" or "poor drainage" and increase the weight of soil moisture sensor data in evidence analysis to more sensitively detect the risk of water accumulation.

[0063] In sub-region C, due to the moderate loam soil conditions and proximity to water sources, the system may adjust parameters according to its specific crop growth stage. For example, during peak crop water demand periods, the system may increase the initial confidence level of "insufficient irrigation" and pay more attention to real-time data from flow meters and soil moisture sensors.

[0064] In this way, the fault identification model for each sub-region can be optimized according to its unique local environmental conditions, thereby enabling the irrigation system of the entire vineyard to identify and handle various irrigation faults more accurately and effectively, ensuring the healthy growth of grapes and the economical use of water resources.

[0065] This application further proposes the following steps for selecting the fault scenario with the highest overall confidence level to make a final judgment: When the overall confidence difference of multiple fault scenarios is less than a preset threshold, the previously determined fault type judgment result is maintained. At the same time, a short-term data smoothing mechanism is activated to perform a moving average on the key feedback information received subsequently, which helps to improve the stability of irrigation. During the operation of the short-term data smoothing mechanism, the trend of the overall confidence level change of the fault scenario with the highest overall confidence level is continuously monitored; When the overall confidence level of the fault scenario with the highest overall confidence level is consistently higher than that of other fault scenarios, and the duration exceeds the preset time, the system switches to a new fault type for judgment.

[0066] Specifically, "maintaining the previously determined fault type judgment result when the overall confidence difference of multiple fault scenarios is less than a preset threshold" means that when judging the fault type, the system calculates the confidence difference between the fault scenario with the highest overall confidence and the second highest. If this difference is less than a preset threshold, it indicates that the advantage of the current highest confidence is not significant, and there is a risk of unstable judgment. In this case, to avoid frequent switching and misjudgment, the system will continue to use the previously determined fault type judgment result instead of immediately adopting the current scenario with the highest confidence. The purpose is to enhance the robustness of the judgment and prevent erroneous decisions caused by instantaneous fluctuations.

[0067] The phrase "activating a short-term data smoothing mechanism to perform moving average processing on subsequently received key feedback information" can be understood as follows: when the system is in the aforementioned state of uncertainty, a data processing module is activated to obtain a more stable and reliable decision-making basis. This module performs moving average processing on key feedback information received over a subsequent period, such as new communication delivery status information, physical action status information, and measurement data. Moving average is a commonly used signal processing technique that eliminates short-term fluctuations and noise by calculating the average value of data within a certain time window, thereby obtaining a smoother and more representative data trend. Its purpose is to provide a more stable input for subsequent confidence updates and reduce the impact of noise on the judgment result.

[0068] In practical applications, "continuously monitoring the trend of the overall confidence level of the fault scenario with the highest overall confidence level during the operation of the short-term data smoothing mechanism" specifically means that while processing data, the system closely monitors the fault scenario currently considered to have the highest confidence level and continuously tracks and analyzes the changes in its overall confidence level value. This monitoring not only focuses on the instantaneous value but also emphasizes its trend over time, such as whether it continues to rise, fall, or remain stable. Its purpose is to assess whether the confidence level of this fault scenario truly has a sustained advantage, providing a basis for the final decision to switch.

[0069] Furthermore, the statement "Switch to a new fault type judgment when the overall confidence level of the fault scenario with the highest overall confidence level is consistently and stably higher than that of other fault scenarios for a duration exceeding a preset time" means that the system will only consider a fault scenario a truly definitive fault type if it observes that the overall confidence level of a certain fault scenario is not only currently the highest, but this advantage is also consistent and stable, and this stable state has persisted for a sufficiently long time (i.e., exceeding a preset time). Only after these conditions are met will the system safely switch from the previous judgment result to the new fault type judgment. The purpose is to ensure the accuracy and stability of fault judgment and avoid hasty decisions under conditions of high uncertainty.

[0070] In some preferred embodiments, suppose that during an irrigation task, the edge node detects that the solenoid valve of the irrigation execution device fails to fully open, while the irrigation measurement device reports a lower-than-expected water flow. After preliminary logical cross-validation and confidence calculation, the system identifies two possible fault scenarios: Scenario A is "mechanical failure of the solenoid valve," with a comprehensive confidence level of 0.48; Scenario B is "insufficient water pump pressure," with a comprehensive confidence level of 0.46. At this time, the difference in comprehensive confidence levels between the two scenarios (0.02) is less than a preset threshold (e.g., 0.05). According to the scheme of this application, the system will not immediately switch to Scenario A, but will maintain the previously determined fault type judgment result (assuming the previous judgment was "insufficient water pump pressure"). At the same time, the system initiates a short-term data smoothing processing mechanism to perform moving average processing on key feedback information such as water flow and transient electrical response of the solenoid valve received in the following 5 minutes. During this period, the system continuously monitors the trend of the comprehensive confidence level change of Scenario A. If, after 3 minutes of monitoring, the overall confidence level of Scenario A remains consistently above 0.55, while the overall confidence level of Scenario B remains consistently below 0.45, and this stable advantage persists for more than a preset duration (e.g., 2 minutes), the system will finally determine that it is a "mechanical failure of the solenoid valve" and take corresponding scheduling measures, such as attempting to remotely reset the solenoid valve or notifying maintenance personnel. This approach avoids blindly switching when the confidence level is ambiguous, ensuring the accuracy of fault diagnosis and the stability of scheduling decisions.

[0071] This application further proposes that the step of maintaining the previously determined fault type judgment result can also be achieved through the following steps: Start the stabilization timer; Activate the disturbance intensity assessment module to monitor environmental disturbance parameters; When the stabilization timer detects that the overall intensity of the environmental disturbance parameters exceeds the preset disturbance threshold within the preset stabilization duration, and the difference in overall confidence between the fault scenario with the highest overall confidence and the fault scenario with the second highest overall confidence fluctuates frequently, the stabilization timer duration is extended, and the previously determined fault type judgment result is maintained. When the stabilization timer ends and the overall intensity of the environmental disturbance parameters does not exceed the preset disturbance threshold within the entire preset stabilization period, and the overall confidence of the fault scenario with the highest overall confidence remains consistently higher than the overall confidence of other fault scenarios, the system switches to a new fault type for judgment.

[0072] Specifically, the stabilization timer can be understood as a time counter used to measure the stability of the fault type determination result. Its purpose is to ensure that the system can observe the trend of confidence changes in the fault scenario within a sufficiently long time window before making a fault type switching decision, thereby avoiding misjudgments caused by instantaneous fluctuations. This timer can be configured to start each time a potential fault type switching condition is met, and to be extended or reset when specific conditions are met.

[0073] The disturbance intensity assessment module can be understood as a component responsible for collecting and analyzing various environmental factors that may affect the stability of fault diagnosis. Environmental disturbance parameters refer to external or internal factors that may cause fluctuations in system data or judgment results, such as sensor noise, network latency, instantaneous voltage fluctuations, and local climate changes (such as gusts and short-term rainfall). This module monitors these parameters and calculates their combined intensity to quantify the "uncertainty" or "degree of interference" of the current environment. For example, the disturbance intensity assessment module can integrate data from multiple sensors (such as wind speed, humidity, light intensity, power supply voltage stability, etc.) and combine them with historical data and preset models to calculate a comprehensive disturbance intensity index.

[0074] In practical applications, the preset stabilization time refers to the minimum stabilization time the system needs to continuously observe before considering switching fault types. The preset disturbance threshold is a critical value used to determine whether environmental disturbances have reached a level sufficient to affect the stability of the judgment. When the judgment stabilization timer detects that the overall strength of the environmental disturbance parameters exceeds the preset disturbance threshold within the preset stabilization time, and the difference in overall confidence between the fault scenario with the highest overall confidence and the fault scenario with the second highest overall confidence fluctuates frequently, this indicates that there may be significant uncertainty or interference in the current environment, leading to unstable confidence calculation results. In this situation, to avoid misjudgment, the system will extend the timing of the judgment stabilization timer and continue to maintain the previously determined fault type judgment result, allowing more time for observation and confirmation.

[0075] Furthermore, when the stabilization timer expires, and the overall intensity of environmental disturbance parameters does not exceed the preset disturbance threshold throughout the entire preset stabilization period, and the overall confidence level of the fault scenario with the highest overall confidence level remains consistently higher than that of other fault scenarios, this indicates that the system has been observing new fault scenarios with significant and stable high confidence levels in a relatively stable environment. At this point, the system can safely switch to the new fault type judgment to ensure that irrigation resource scheduling can respond promptly to actual fault situations.

[0076] The steps for extending the duration of the stabilization timer and maintaining the previously determined fault type assessment result include: Monitor the type, intensity, and duration of environmental disturbance parameters; Monitor the fluctuation frequency and amplitude of the difference in overall confidence between the fault scenario with the highest overall confidence and the fault scenario with the second highest overall confidence; Based on the type, intensity, and duration of the environmental disturbance parameters, as well as the fluctuation frequency and amplitude of the difference in the overall confidence level between the fault scenario with the highest overall confidence level and the fault scenario with the second highest overall confidence level, calculate the specific value of extending the timing duration of the judgment stabilization timer. Based on the specific value of the extended stabilization timer, the duration of the stabilization timer is extended, and the previously determined fault type judgment result is maintained.

[0077] "Monitoring the type, intensity, and duration of environmental disturbance parameters" refers to the real-time or near-real-time data acquisition and analysis of external factors that may affect the operational stability of the irrigation system. Environmental disturbance parameters may include, but are not limited to, wind speed, temperature, humidity, light intensity, soil moisture content, and rainfall. The type refers to the physical property of the disturbance parameter, such as temperature change or wind increase; the intensity refers to the quantitative magnitude of the disturbance parameter, such as wind speed in meters per second or temperature change in degrees Celsius; and the duration refers to the length of time the disturbance parameter remains at a certain intensity or type. These parameters can be acquired through various sensors deployed at agricultural production sites and preliminarily processed and analyzed by edge nodes.

[0078] "Monitoring the fluctuation frequency and amplitude of the difference in overall confidence levels between the fault scenario with the highest overall confidence level and the fault scenario with the second highest overall confidence level" refers to the dynamic analysis of the confidence level data output by the system's internal fault identification module. The fault scenario with the highest overall confidence level represents the most likely fault type as determined by the current system, while the fault scenario with the second highest overall confidence level represents the second most likely fault type. The difference between these two confidence levels reflects the degree of certainty in the system's current fault assessment. Fluctuation frequency refers to the number of times this difference changes significantly per unit of time, and fluctuation amplitude refers to the range of variation in this difference. High frequency and large amplitude fluctuations usually indicate significant uncertainty in the system's current fault assessment or drastic environmental changes.

[0079] "Calculating the specific value of extending the stabilization timer based on the type, intensity, and duration of environmental disturbance parameters, as well as the fluctuation frequency and amplitude of the difference in comprehensive confidence between the fault scenario with the highest and second-highest comprehensive confidence, refers to using the aforementioned multi-dimensional data as input and calculating a quantified time extension value through a preset algorithm model or rule set. This calculation process can be implemented using various methods such as fuzzy logic, machine learning models (such as regression models), and expert system rules. For example, when the environmental disturbance intensity is high, the duration is long, and the difference in confidence between fault scenarios fluctuates frequently and with large amplitude, the time extension value will be calculated to provide a longer stabilization time; conversely, when the disturbance is small and the difference in confidence is stable, the extension value will be small or not extended at all."

[0080] "Extending the duration of the stabilization timer based on the specific value of the extended duration, while maintaining the previously determined fault type judgment result" means applying the calculated value to the stabilization timer. For example, if the timer's current remaining duration is T, and the calculated extension value is ΔT, then the timer's new duration will become T+ΔT. During this extension period, the system will continue to maintain the previously determined fault type judgment result, avoiding premature switching to a new fault judgment due to instantaneous fluctuations or uncertainties, thereby ensuring the stability of irrigation resource scheduling.

[0081] In some preferred embodiments, it is assumed that in an agricultural irrigation area, an edge node is performing an irrigation task. At a certain moment, the system initially identifies two possible fault scenarios, "solenoid valve jamming" or "pipeline blockage," through logical cross-validation. Their combined confidence levels are 0.45 and 0.40, respectively, with a difference of 0.05, which is less than a preset threshold. Therefore, the system maintains the previously determined fault type judgment result (e.g., "normal operation"). At this time, the judgment stabilization timer is started.

[0082] During the timer's operation, the disturbance intensity assessment module detected changes in environmental parameters: a sudden increase in wind speed (environmental disturbance parameter type: wind force, intensity: high) that persisted for a considerable period (duration: long). Simultaneously, the fault identification module reported that the combined confidence difference between the two fault scenarios, "solenoid valve jamming" and "pipeline blockage," began to fluctuate frequently between 0.03 and 0.07 (fluctuation frequency: high, amplitude: medium).

[0083] According to the scheme in this application, the system will integrate the following monitoring data: 1. Monitoring environmental disturbance parameters: wind type, high intensity, and long duration. 2. Monitoring confidence difference fluctuations: high fluctuation frequency and medium amplitude.

[0084] The system inputs this information into a pre-defined calculation model. For example, this model might be a rule-based system: if the wind intensity is high and the duration is long, the base extension duration is increased by X seconds. If the confidence difference fluctuates frequently, an additional Y seconds are added. If the confidence difference fluctuates moderately, an additional Z seconds are added.

[0085] Through calculation, the system obtains a specific value for extending the stabilization timer's duration, for example, by 30 seconds. Subsequently, the current duration of the stabilization timer will be extended by 30 seconds. During this extension, the system will continue to maintain the fault type judgment result of "normal operation" and continuously monitor key feedback information. This dynamic extension mechanism allows the system more time for observation and analysis when facing complex environmental disturbances and uncertainties, thus avoiding premature switching to potentially inaccurate fault judgments and ensuring the stability and reliability of irrigation scheduling.

[0086] This application further proposes steps for calculating the specific value of extending the timing duration of the judgment stabilization timer based on the type, intensity, and duration of environmental disturbance parameters, as well as the fluctuation frequency and amplitude of the difference in comprehensive confidence between the fault scenario with the highest comprehensive confidence and the fault scenario with the second highest comprehensive confidence. The weighted combination is obtained based on the type, intensity, and duration of environmental disturbance parameters, as well as the fluctuation frequency and amplitude of the difference in the overall confidence level between the fault scenario with the highest overall confidence level and the fault scenario with the second highest overall confidence level. Based on the weighted combination, the initial extended duration is obtained by weighted summation of the fluctuation frequency and amplitude of the fluctuation frequency and amplitude of the type, intensity, and duration of the environmental disturbance parameters, as well as the difference in comprehensive confidence between the fault scenario with the highest comprehensive confidence and the fault scenario with the second highest comprehensive confidence. Based on the instantaneous change rate of environmental disturbance parameters and the instantaneous fluctuation trend of the difference in comprehensive confidence between the fault scenario with the highest comprehensive confidence and the fault scenario with the second highest comprehensive confidence, the initial extension duration is corrected to obtain the specific value of the timing duration of the extended judgment stabilization timer.

[0087] Specifically, obtaining the weighted combination refers to assigning corresponding weight coefficients to multiple influencing factors, such as the type, intensity, and duration of environmental disturbance parameters, as well as the fluctuation frequency and amplitude of the difference in the comprehensive confidence of the fault scenario with the highest comprehensive confidence and the fault scenario with the second highest comprehensive confidence, based on preset rules or machine learning models. These weight coefficients can be dynamically adjusted based on historical data, expert experience, or real-time system status to reflect the degree of influence of different factors on the timer extension requirement under specific circumstances. For example, when wind speed changes drastically, the weights of wind speed intensity and instantaneous rate of change may be increased. The weighted summation of the above factors to obtain the initial extension duration can be understood as multiplying each factor and its corresponding weight and then summing them to obtain a preliminary, comprehensive estimate of the extension duration. The purpose is to quantify and integrate multiple complex factors into a unified value, providing a foundation for subsequent fine-tuning. In practical applications, adjusting the initial extension duration based on the instantaneous change rate of environmental disturbance parameters and the instantaneous fluctuation trend of the difference in overall confidence between the fault scenario with the highest overall confidence and the fault scenario with the second highest overall confidence means further considering the real-time dynamic changes of these factors after obtaining the initial extension duration. For example, the instantaneous change rate of environmental disturbance parameters can reflect the suddenness and severity of the disturbance, while the instantaneous fluctuation trend of the overall confidence difference can reveal the immediate evolution of uncertainty in fault judgment. By introducing this instantaneous dynamic information, the initial extension duration can be fine-tuned to more accurately adapt to the rapidly changing actual situation and avoid judgment bias caused by lag or over-smoothing.

[0088] In some preferred embodiments, assuming an edge node is performing an irrigation task in an agricultural irrigation area and a stabilization timer is running, the system detects environmental disturbance parameters, such as a sudden increase in wind speed from 2 m / s to 8 m / s for a short duration. Simultaneously, the combined confidence difference between the two most probable fault scenarios (e.g., "solenoid valve jamming" and "pipe blockage") fluctuates from 0.15 to 0.08 with a high frequency of fluctuation. First, the system, based on preset rules or models, obtains a weighted combination for factors such as wind speed intensity (8 m / s), duration (short), wind speed change rate (instantaneous change of 6 m / s), and the fluctuation frequency and amplitude of the combined confidence difference (high) and amplitude (0.07). For example, the instantaneous change rate of wind speed and the fluctuation frequency of the confidence difference may be given higher weights. Next, the system uses this weighted combination to perform a weighted summation of the above parameters to calculate an initial extension duration, for example, a preliminary calculation result of 15 seconds. Subsequently, the system further monitors the instantaneous rate of change of wind speed (e.g., a change of 3 m / s in wind speed within the past second) and the instantaneous fluctuation trend of the comprehensive confidence difference (e.g., the difference is rapidly decreasing). Based on this instantaneous dynamic information, the system corrects the initial extension duration of 15 seconds. If the instantaneous rate of change and fluctuation trend indicate a more urgent or unstable situation, the correction may increase the extension duration to 20 seconds; conversely, if the instantaneous dynamics tend to stabilize, it may be fine-tuned to 12 seconds. Ultimately, the system obtains a precise value for the duration of the extended judgment stabilization timer and extends the timer accordingly to ensure that hasty fault type switching judgments are not made before environmental uncertainties are eliminated or the fault scenario is further clarified, thereby ensuring the stable execution of irrigation tasks.

[0089] refer to Figure 2 This application proposes an edge computing-based agricultural irrigation resource scheduling system, applied to an edge computing-based agricultural irrigation resource scheduling method. The system includes: The instruction sending module sends irrigation instructions and obtains the communication delivery status information of the irrigation instructions; The physical action acquisition module acquires the physical action status information of the irrigation execution equipment. The measurement data acquisition module acquires measurement data from the irrigation measurement equipment. The fault identification module performs logical cross-validation based on communication delivery status information, physical action status information, and measurement data to identify the fault type of the irrigation task. The scheduling module takes appropriate irrigation resource scheduling measures based on the type of fault.

[0090] Specifically, the instruction sending module can be understood as the component in the system responsible for communicating with the irrigation execution equipment. Its main function is to issue irrigation instructions such as start or stop to the irrigation execution equipment and monitor in real time whether these instructions are successfully delivered. For example, this module can be a software unit integrating a communication protocol stack, interacting with execution equipment such as solenoid valve controllers via wireless or wired networks, and receiving confirmation information or error reports from the execution equipment to generate communication delivery status information. Its purpose is to ensure reliable transmission of irrigation instructions and status feedback.

[0091] The physical motion acquisition module can be understood as a component used to monitor the actual physical motion of irrigation equipment. Specifically, this module can acquire the actual motion status of equipment such as solenoid valves and water pumps after receiving commands, such as the opening or closing status of valves and the operating status of water pumps. Its purpose is to provide physical-level verification information independent of communication status to determine whether commands have been effectively executed. In practical applications, this module can acquire physical motion status information by integrating sensors (such as position sensors and current sensors) or by analyzing the transient electrical response of the equipment.

[0092] Furthermore, the measurement data acquisition module is responsible for collecting various environmental and irrigation-related data from irrigation measurement equipment. For example, this module can acquire real-time measurement data such as soil moisture, ambient temperature, and irrigation flow rate from devices like soil moisture sensors, temperature sensors, and flow meters. Its purpose is to provide comprehensive environmental and performance feedback data for subsequent fault identification and scheduling.

[0093] Furthermore, the fault identification module is the core intelligent component of the system. It performs logical cross-validation of multi-source information based on communication delivery status information provided by the command sending module, physical action status information provided by the physical action acquisition module, and measurement data provided by the measurement data acquisition module. This module analyzes the inherent consistency or inconsistency of this data to accurately identify potential fault types in the irrigation task. For example, if a command has been delivered but the physical action has not occurred, and the measurement data does not show the expected change, it may indicate a fault in the executing equipment. Its purpose is to provide accurate fault diagnosis, providing a basis for subsequent scheduling decisions.

[0094] Finally, the scheduling module is a component that formulates and executes corresponding irrigation resource scheduling measures based on the fault type identified by the fault identification module. For example, when a solenoid valve malfunction is identified, the scheduling module can instruct the backup valve to open, adjust the irrigation plan, or even notify maintenance personnel for repairs. Its purpose is to ensure that irrigation strategies can be adjusted promptly and effectively in the event of a fault, minimizing the impact on agricultural production and guaranteeing the continuity and effectiveness of irrigation tasks.

[0095] The content disclosed above is only a preferred and feasible embodiment of the present invention, and is not intended to limit the scope of protection of the present invention. Therefore, all equivalent technical changes made based on the content of the present invention specification and drawings are included within the scope of protection of the present invention. Furthermore, the elements therein can be updated as technology develops.

Claims

1. A method for scheduling agricultural irrigation resources based on edge computing, characterized in that, The method includes the following steps: Send irrigation commands and obtain communication delivery status information for the irrigation commands; Obtain the physical motion status information of the irrigation equipment; Acquire measurement data from irrigation measurement equipment; Logical cross-validation is performed based on communication delivery status information, physical action status information, and measurement data to identify the fault type of the irrigation task. Based on the type of fault, corresponding irrigation resource scheduling measures will be taken.

2. The method for scheduling agricultural irrigation resources based on edge computing as described in claim 1, characterized in that, The steps for obtaining the physical motion status information of irrigation equipment include: After receiving the open or close command from the edge node, the solenoid valve controller activates the power drive circuit and applies voltage to the solenoid valve coil. The solenoid valve controller performs high-frequency sampling of the transient electrical response of the solenoid valve coil to obtain sampling data; The solenoid valve controller extracts transient electromagnetic features from the sampled data; The solenoid valve controller compares the transient electromagnetic characteristics with the preset valve core position characteristics to obtain the comparison result; The solenoid valve controller determines whether the valve core is fully open based on the comparison results. The solenoid valve controller generates a physical action confirmation signal based on whether the valve core is fully open, and uses the physical action confirmation signal as the physical action status information of the irrigation execution equipment.

3. The method for scheduling agricultural irrigation resources based on edge computing as described in claim 1, characterized in that, The steps for identifying the fault type of irrigation task based on communication delivery status information, physical action status information, and measurement data include: A consistency assessment is performed on communication delivery status information, physical action status information, and measurement data to obtain the consistency assessment results. When there is a mismatch or ambiguity in the consistency assessment results, the initial confidence level is assigned to each possible failure type by combining the priority of the current irrigation task, the success rate of similar historical tasks, and the current environmental parameters. The communication delivery status information, physical action status information, and measurement data are weighted according to the preset fuzzy evidence parsing rules to correct the initial confidence level and obtain the corresponding comprehensive confidence level. Activate the behavior observation window to continuously monitor key feedback information, and dynamically update the overall confidence level based on the changing trends of key feedback information; Based on the dynamically updated overall confidence distribution, the fault scenario with the highest overall confidence is selected to make the final judgment.

4. The method for scheduling agricultural irrigation resources based on edge computing as described in claim 1, characterized in that, The steps for identifying the fault type of irrigation task based on communication delivery status information, physical action status information, and measurement data include: A consistency assessment is performed on communication delivery status information, physical action status information, and measurement data to obtain the consistency assessment results. When there is incomplete matching or ambiguity in the consistency assessment results, the current environmental parameters are continuously monitored. Based on the changes in the current environmental parameters, the weighting coefficients of each communication delivery status information, physical action status information, and measurement data in the fuzzy evidence parsing rules are dynamically adjusted. Based on the changes in the current environmental parameters, the initial confidence of each possible fault type is dynamically adjusted. Based on the current environmental parameters and weighting coefficients, the initial confidence is corrected to obtain the corresponding comprehensive confidence. Activate the behavior observation window to continuously monitor key feedback information, and dynamically update the overall confidence level based on the changing trends of key feedback information; Based on the dynamically updated overall confidence distribution, the fault scenario with the highest overall confidence is selected to make the final judgment.

5. The method for scheduling agricultural irrigation resources based on edge computing as described in claim 1, characterized in that, The steps for identifying the fault type of irrigation task based on communication delivery status information, physical action status information, and measurement data include: A consistency assessment is performed on communication delivery status information, physical action status information, and measurement data to obtain the consistency assessment results. When there is incomplete matching or ambiguity in the consistency assessment results, the irrigation area is divided into multiple sub-regions, and the local environmental parameters of each sub-region are obtained. Based on the local environmental parameters of each sub-region, the weighting coefficients of each communication delivery status information, physical action status information, and measurement data in the fuzzy evidence parsing rules of each sub-region are dynamically adjusted. Based on the local environmental parameters of each sub-region, the initial confidence of each possible fault type in each sub-region is dynamically adjusted. Based on the local environmental parameters and weighting coefficients of each sub-region, the initial confidence of each sub-region is corrected to obtain the corresponding comprehensive confidence. Activate the behavior observation window to continuously monitor key feedback information, and dynamically update the overall confidence level based on the changing trends of key feedback information; Based on the dynamically updated overall confidence distribution, the fault scenario with the highest overall confidence is selected to make the final judgment.

6. The method for scheduling agricultural irrigation resources based on edge computing as described in claim 3, characterized in that, The steps for selecting the fault scenario with the highest overall confidence level to make a final judgment include: When the overall confidence difference of multiple fault scenarios is less than a preset threshold, the previously determined fault type judgment result is maintained. At the same time, a short-term data smoothing mechanism is activated to perform a moving average on the key feedback information received subsequently, which helps to improve the stability of irrigation. During the operation of the short-term data smoothing mechanism, the trend of the overall confidence level change of the fault scenario with the highest overall confidence level is continuously monitored; When the overall confidence level of the fault scenario with the highest overall confidence level is consistently higher than that of other fault scenarios, and the duration exceeds the preset time, the system switches to a new fault type for judgment.

7. The method for scheduling agricultural irrigation resources based on edge computing as described in claim 6, characterized in that, Maintaining the previously determined fault type assessment result can also be achieved through the following steps: Start the stabilization timer; Activate the disturbance intensity assessment module to monitor environmental disturbance parameters; When the stabilization timer detects that the overall intensity of the environmental disturbance parameters exceeds the preset disturbance threshold within the preset stabilization duration, and the difference in overall confidence between the fault scenario with the highest overall confidence and the fault scenario with the second highest overall confidence fluctuates frequently, the stabilization timer duration is extended, and the previously determined fault type judgment result is maintained. When the stabilization timer ends and the overall intensity of the environmental disturbance parameters does not exceed the preset disturbance threshold within the entire preset stabilization period, and the overall confidence of the fault scenario with the highest overall confidence remains consistently higher than the overall confidence of other fault scenarios, the system switches to a new fault type for judgment.

8. The method for scheduling agricultural irrigation resources based on edge computing as described in claim 7, characterized in that, The steps for extending the duration of the stabilization timer and maintaining the previously determined fault type assessment result include: Monitor the type, intensity, and duration of environmental disturbance parameters; Monitor the fluctuation frequency and amplitude of the difference in overall confidence between the fault scenario with the highest overall confidence and the fault scenario with the second highest overall confidence; Based on the type, intensity, and duration of the environmental disturbance parameters, as well as the fluctuation frequency and amplitude of the difference in the overall confidence level between the fault scenario with the highest overall confidence level and the fault scenario with the second highest overall confidence level, calculate the specific value of extending the timing duration of the judgment stabilization timer. Based on the specific value of the extended stabilization timer, the duration of the stabilization timer is extended, and the previously determined fault type judgment result is maintained.

9. A method for scheduling agricultural irrigation resources based on edge computing as described in claim 8, characterized in that, The steps for calculating the specific value of extending the stabilization timer based on the type, intensity, and duration of environmental disturbance parameters, as well as the fluctuation frequency and amplitude of the difference in overall confidence between the fault scenario with the highest overall confidence and the fault scenario with the second highest overall confidence, include: The weighted combination is obtained based on the type, intensity, and duration of environmental disturbance parameters, as well as the fluctuation frequency and amplitude of the difference in the overall confidence level between the fault scenario with the highest overall confidence level and the fault scenario with the second highest overall confidence level. Based on the weighted combination, the initial extended duration is obtained by weighted summation of the fluctuation frequency and amplitude of the fluctuation frequency and amplitude of the type, intensity, and duration of the environmental disturbance parameters, as well as the difference in comprehensive confidence between the fault scenario with the highest comprehensive confidence and the fault scenario with the second highest comprehensive confidence. Based on the instantaneous change rate of environmental disturbance parameters and the instantaneous fluctuation trend of the difference in comprehensive confidence between the fault scenario with the highest comprehensive confidence and the fault scenario with the second highest comprehensive confidence, the initial extension duration is corrected to obtain the specific value of the timing duration of the extended judgment stabilization timer.

10. An edge computing-based agricultural irrigation resource scheduling system, applied to the edge computing-based agricultural irrigation resource scheduling method as described in claim 1, characterized in that, The system includes: The instruction sending module sends irrigation instructions and obtains the communication delivery status information of the irrigation instructions; The physical action acquisition module acquires the physical action status information of the irrigation execution equipment. The measurement data acquisition module acquires measurement data from the irrigation measurement equipment. The fault identification module performs logical cross-validation based on communication delivery status information, physical action status information, and measurement data to identify the fault type of the irrigation task. The scheduling module takes appropriate irrigation resource scheduling measures based on the type of fault.

Images