Farmland Soil Multi-Sensor Collaborative Calibration and Zonal Irrigation Control Method and System
By dividing irrigation grids in farmland and using edge computing for neighborhood consistency verification and calibration, the problems of abnormal sensor readings and coarse-grained irrigation control are solved, achieving high-precision zoned irrigation control and data accuracy, which is suitable for weak grid environments in large fields.
Patent Information
- Application Number
- CN202611216538.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-08-12
- Publication Date
- 2026-11-10
AI Technical Summary
In large-area field and multi-node monitoring scenarios, soil water, heat and salt sensors lack neighborhood collaborative verification, making it difficult to identify abnormal sensor readings. This leads to the system misjudging or missing water shortages, and the irrigation control is too coarse, making it impossible to independently control local grid differences, resulting in ineffective irrigation and water and fertilizer leaching.
By dividing farmland into irrigation grids, configuring water, heat, and salt sensor nodes and irrigation mechanisms, using edge computing gateways to establish neighborhood sets for data consistency verification, marking real data and candidate abnormal data, performing calibration, and judging irrigation conditions based on irrigation grid data, zoned irrigation control is achieved.
It improves the accuracy of farmland monitoring data and irrigation precision, effectively avoids interference from abnormal data, identifies irrigation anomalies, reduces the risk of mis-irrigation under high temperature and salt stress, and is suitable for operation in weak grid environments in large fields.
Smart Images

Figure CN123058243A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of soil moisture monitoring technology, specifically a method and system for multi-sensor collaborative calibration and zoned irrigation control of farmland soil. Background Technology
[0002] In field crops, orchards, greenhouse agriculture, and large-scale farms, soil moisture, temperature, and salinity are important parameters affecting crop water requirements, root zone salt stress, and irrigation regimes. With the development of the Internet of Things (IoT) in agriculture, more and more fields are being equipped with soil water, heat, and salt sensors, and the data is uploaded to a platform via wireless communication for soil moisture monitoring and irrigation management.
[0003] While existing soil water, heat, and salt monitoring and irrigation control technologies can achieve a certain degree of automation, they still have the following shortcomings in large-scale field and multi-node monitoring scenarios: Sensor nodes operate independently, lacking neighborhood collaborative verification. Multiple soil water, heat, and salt sensors typically collect and upload data independently, with the system only performing simple display or threshold judgment, failing to fully utilize the spatial continuity and consistency between adjacent sensors. When a node generates abnormal readings due to local soil disturbance, poor contact, electrode contamination, wireless packet loss, or sensor drift, the system struggles to identify it in a timely manner, easily leading to misjudgments or missed judgments of water shortage.
[0004] Irrigation decisions are coarse-grained. Traditional irrigation control often relies on a single representative sensor, field average, or human experience to uniformly control the irrigation of the entire field. It cannot independently control local grid differences, resulting in even water-sufficient areas being irrigated, leading to ineffective irrigation and water and fertilizer leaching. Summary of the Invention
[0005] To address the technical problems existing in the prior art, this invention provides a method and system for collaborative calibration and zoned irrigation control of farmland soil using multiple sensors. This method can verify the data of sensor nodes by constructing a sensor neighborhood set, and irrigate the corresponding farmland grids according to the irrigation grid data, thereby improving the accuracy of farmland monitoring data and the precision of farmland irrigation.
[0006] To achieve the above objectives, the present invention provides the following technical solution: This invention discloses a method for multi-sensor collaborative calibration and zoned irrigation control of farmland soil, comprising: S1. Divide the farmland into irrigation grids and configure water, heat and salt sensor nodes and irrigation mechanisms for each irrigation grid; S2. The edge computing gateway establishes a neighborhood set for each sensor node based on the location of the sensor node, receives soil moisture, temperature and salinity data uploaded by the sensor node, and performs neighborhood consistency verification on the soil moisture, temperature and salinity data within the neighborhood set. Data that passes the neighborhood consistency verification is marked as real data, and data that fails the neighborhood consistency verification is marked as candidate abnormal data. Candidate abnormal data calibration is performed, and irrigation grid data is calculated based on real data and calibrated data. S3. Determine whether the irrigation grid data meets the irrigation conditions or the conditions for stopping irrigation. If the irrigation conditions are met, start the irrigation mechanism corresponding to the irrigation grid. If the conditions for stopping irrigation are met, stop the irrigation mechanism corresponding to the irrigation grid.
[0007] Preferably, S2 specifically includes: S201, Edge computing gateway receives sensor nodes Uploaded soil moisture, temperature, and salinity data, and based on data from each sensor node. Location information for each sensor node Establish neighborhood set ; S202, Obtain the neighborhood set Data from all sensor nodes in the system, and calculation parameters. neighborhood median and standard deviation Parameters of sensor nodes reading A neighborhood consistency check is performed. Data that passes the neighborhood consistency check is marked as real data; sensor node data that fails the neighborhood consistency check is marked as candidate anomalous data, and candidate anomalous data calibration is performed. The neighborhood consistency check is as follows:
[0008] in, For sensor nodes parameters Readings; parameters It is one of the soil's moisture, temperature, and salinity; For neighborhood set Intrinsic parameters The median value; For neighborhood set Intrinsic parameters Standard deviation; This is the threshold coefficient; Number the sensor nodes; S203. Calculate irrigation grid data based on real data and calibrated sensor node data.
[0009] Preferably, the candidate anomaly data calibration includes: a. Determine the accuracy of candidate outlier data; The sensor nodes corresponding to the candidate abnormal data are resampled, and the data change trend of their neighborhood set is acquired simultaneously; it is determined whether the change trend of the resampled data is consistent with the data change trend of the neighborhood set; if they are consistent, the candidate abnormal data is converted into real data; otherwise, abnormal data calibration is performed. b. Abnormal data calibration: Determine whether the resampled data passes the neighborhood consistency check; If the resampled data passes the neighborhood consistency check, the resampled data will be used as the calibrated data. If the resampled data fails the neighborhood consistency check, the neighborhood weighted average of the neighborhood set is used as the calibrated data.
[0010] Preferably, the neighborhood weighted average is calculated as follows:
[0011] in, It is a neighborhood-weighted average; Sensor nodes within the neighborhood set Data readings; For sensor nodes The weight in the neighborhood set; m is the neighborhood set number.
[0012] Preferably, the edge computing gateway stores a grid mapping table; The grid mapping table includes irrigation grid number, sensor node number, irrigation agency number, crop type, growth period parameters, irrigation start threshold, and irrigation stop threshold; The edge computing gateway determines whether the irrigation grid data meets the irrigation conditions or the conditions for stopping irrigation based on the grid mapping table, and controls the irrigation mechanism to start or stop irrigation.
[0013] Preferably, determining whether the irrigation grid data meets the irrigation conditions or the conditions for stopping irrigation specifically includes: The irrigation grid data includes soil moisture data, soil temperature data, and soil salinity data; Meeting the irrigation conditions includes: soil moisture data in the irrigation grid is below the irrigation start threshold; and soil temperature data is below the high temperature constraint threshold and soil salinity data is below the salinity constraint threshold. Irrigation can be stopped when: the soil moisture data in the irrigation grid reaches the stop irrigation threshold, the irrigation duration reaches the maximum limit, or the valve feedback is abnormal.
[0014] Preferably, the sensor node communicates with the edge computing gateway using one or more of the following communication methods: LoRa, NB-IoT, ZigBee, Bluetooth Mesh, RS485, and Ethernet. The edge computing gateway communicates with the irrigation mechanism using one or more of the following communication methods: LoRa, RS485, and relay control.
[0015] Preferably, the sensor node collects data at a first sampling frequency and collects data at a second sampling frequency after the irrigation mechanism is turned on; wherein the second sampling frequency is higher than the first sampling frequency. The data from sensor nodes is used to monitor the rate of change in soil moisture after the irrigation system is activated, and to determine if there are any irrigation anomalies.
[0016] This invention also proposes a multi-sensor collaborative calibration and zoned irrigation control system for farmland soil, comprising: The initialization module is used to divide farmland into irrigation grids and configure sensor nodes and irrigation mechanisms for each irrigation grid. Hydrothermal-salt sensor nodes are used to collect and upload soil moisture, temperature, and salinity data to the edge computing gateway; The edge computing gateway includes a neighborhood consistency verification and data fusion module and a gridded irrigation decision module. The neighborhood consistency verification and data fusion module is used to establish a neighborhood set for each sensor node based on its location; receive soil moisture, temperature, and salinity data uploaded by the sensor nodes and perform neighborhood consistency verification within the neighborhood set; mark data that passes the neighborhood consistency verification as real data; mark sensor node data that fails the neighborhood consistency verification as candidate anomalous data and perform candidate anomalous data calibration; and calculate irrigation grid data based on real data and calibrated sensor node data. The grid-based irrigation decision module is used to determine whether the irrigation grid data meets the irrigation conditions or the conditions for stopping irrigation. When the irrigation conditions are met, it sends a valve opening command to the irrigation mechanism corresponding to the irrigation grid. When the conditions for stopping irrigation are met, it sends a valve closing command to the irrigation mechanism corresponding to the irrigation grid. Irrigation mechanism, used to execute valve opening or closing commands sent by the edge computing gateway.
[0017] The present invention also proposes a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of any one of the methods for multi-sensor collaborative calibration and zoned irrigation control of farmland soil.
[0018] Compared with the prior art, the beneficial effects of the present invention are: (1) The farmland soil multi-sensor collaborative calibration and zoned irrigation control method and system proposed in this invention can realize zoned irrigation of farmland by dividing farmland into irrigation grids, thereby improving the accuracy of farmland irrigation; and can effectively improve the accuracy of sensor node data by verifying the neighborhood consistency of sensor nodes.
[0019] (2) By calibrating candidate abnormal data, the present invention can screen out real data and calibrate abnormal data according to the type of abnormality, which can effectively avoid the interference of abnormal data on irrigation grid data and improve the accuracy of data.
[0020] (3) After the irrigation command is issued, the present invention verifies the irrigation effect by using the trend of water change after irrigation, and can identify valve failure, pipeline blockage or sensor failure.
[0021] (4) When making irrigation decisions, this invention not only considers soil moisture but also soil temperature and salinity data to reduce the risk of mis-irrigation caused by abnormalities such as high temperature and salinity stress.
[0022] (5) In this invention, the edge computing gateway is not only responsible for data aggregation, but also directly completes data calibration, threshold judgment, valve control and feedback verification. It can still operate in weak network or network outage environments, does not rely on real-time cloud response, and is suitable for weak network environments in Datian. Attached Figure Description
[0023] Figure 1 This is a flowchart of the multi-sensor collaborative calibration and zoned irrigation control method for farmland soil proposed in this embodiment of the invention.
[0024] Figure 2 This is a block diagram of the farmland soil multi-sensor collaborative calibration and zoned irrigation control system proposed in this embodiment of the invention.
[0025] Figure 3 This is a comparison diagram of the calibration of each sensor node proposed in the embodiments of the present invention. Detailed Implementation
[0026] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0027] Example 1
[0028] Please see Figure 1 This embodiment provides a method for multi-sensor collaborative calibration and zoned irrigation control of farmland soil, including the following steps; S1. Divide the farmland into irrigation grids and configure water, heat and salt sensor nodes and irrigation mechanisms for each irrigation grid.
[0029] This step involves dividing the farmland to be irrigated into multiple irrigation grids, each with a unique grid number. The irrigation grid includes spatial coordinate ranges. Each irrigation grid can have one or more sensor nodes, or several adjacent grids can share a single sensor node. Each irrigation grid can establish a one-to-one or many-to-one mapping with an irrigation mechanism, forming a mapping table in the edge computing gateway between grid number, sensor number, valve number, and crop threshold. By establishing the mapping relationship between grid number, sensor number, and valve number, it is easy to achieve independent irrigation for water-scarce grids while keeping non-water-scarce grids closed.
[0030] In the one-grid-one-valve mode, the edge computing gateway can independently turn the irrigation mechanism on or off for a single irrigation grid; in the multi-grid-shared-valve mode, the edge computing gateway determines the valve status based on the water shortage level of multiple grids within the valve's coverage area.
[0031] In this embodiment, the sensor node can be a hydrothermal-salt soil sensor, used to synchronously acquire farmland moisture. ,temperature Salt Data; and sensor nodes can be deployed at different depths in farmland depending on the crop type. Soil water, heat, and salt sensor nodes can be capacitive, TDR / FDR, resistive, or MEMS interdigitated electrode sensors, and can also integrate batteries, power management, solar power supply, and low-power sleep modules.
[0032] This embodiment facilitates large-scale deployment. It supports expansion with multiple sensors, multiple grids, and multiple irrigation mechanisms, and can be flexibly configured with one valve per grid, multiple grids sharing a valve, or multiple gateways working together, depending on the size of the field.
[0033] S2. The edge computing gateway establishes a neighborhood set for each sensor node based on its location; receives soil moisture, temperature, and salinity data uploaded by the sensor nodes and performs neighborhood consistency verification within the neighborhood set; marks data that passes the neighborhood consistency verification as real data; marks sensor node data that fails the neighborhood consistency verification as candidate abnormal data and performs candidate abnormal data calibration; and calculates irrigation grid data based on the real data and the calibrated sensor node data.
[0034] In this embodiment, it can be implemented through the following steps: S201, Edge computing gateway receives sensor nodes Uploaded water ,temperature Salt Data, and based on each sensor node Location encoding for each sensor node Establish neighborhood set ; Specifically, the neighborhood set consists of the target sensor node and several spatially adjacent sensor nodes, and is used to perform neighborhood consistency verification, anomaly identification, and data calibration fusion.
[0035] To ensure the accuracy of soil moisture, temperature, and salinity data, a neighborhood set can be established for each sensor node, facilitating neighborhood consistency checks when needed or routinely.
[0036] S202, Obtain the neighborhood set All sensor nodes Calculate the neighborhood median value from the data. and standard deviation For sensor nodes parameters reading Neighborhood consistency verification is performed. Sensor node data that fails the neighborhood consistency verification is marked as candidate anomalous data, and candidate anomalous data calibration is performed. The neighborhood consistency verification is as follows:
[0037] in, For sensor nodes parameters Readings; parameters It is one of the soil's moisture, temperature, and salinity; For neighborhood set Intrinsic parameters The median value; For neighborhood set Intrinsic parameters Standard deviation; This is the threshold coefficient; Number the sensor nodes.
[0038] In this step, neighborhood consistency checks are performed on the moisture, temperature, and salinity of the hydrothermal-salt sensor nodes. Neighborhood consistency check involves checking the sensor nodes... parameters reading The neighborhood median is calculated, and the absolute values of both are compared to a standard deviation threshold. If the absolute value is less than the standard deviation threshold, the neighborhood consistency check is passed; otherwise, it fails. The standard deviation threshold is calculated using the neighborhood standard deviation and the domain threshold coefficient. Sure; A value of 1.5-3 is acceptable, but 2 is preferred.
[0039] Candidate outlier data calibration includes: a. Determine the accuracy of candidate outlier data; The sensor nodes corresponding to the candidate abnormal data are resampled, and the data change trend of their neighborhood set is acquired simultaneously. It is then determined whether the change trend of the resampled data is consistent with the data change trend of the neighborhood set. If they are consistent, the candidate abnormal data is converted into real data; otherwise, abnormal data calibration is performed.
[0040] In this step, resampling can be performed by sending a collaborative query frame to the target sensor via an edge computing gateway, requiring it to resample rapidly n times within a short period of time. n can be 2-5, preferably 3. Simultaneously, the average trend of sensor nodes in the neighborhood set within this time window is compared. The trend refers to the overall direction of sensor node parameter readings as they increase, decrease, or oscillate steadily over time. In this step, the trend can be represented by a slope. When judging the consistency between the trend of the sampled data and the trend of the neighborhood set, the slope of the sensor node data and the slope of the median of the neighborhood set can be calculated simultaneously. Then, the Pearson correlation coefficient between the two is calculated. A coefficient higher than a set threshold, such as 0.8, indicates that the trends are consistent.
[0041] Determine whether the trend of change in the resampled data is consistent with the trend of change in the neighborhood set; if consistent, convert the candidate outlier data into real data; otherwise, perform outlier data calibration.
[0042] Taking moisture as an example, if the trend of moisture data changes is consistent across the neighborhood set, all showing a decrease, it indicates that the target node sensor is experiencing a real water shortage. Therefore, candidate abnormal data can be converted into real data.
[0043] Of course, in a preferred scenario, other parameters from the same sensor can also be used to assist in the judgment. For example, when calibrating moisture data, the trends of salinity and temperature data can be compared with the trends of salinity and temperature data in the neighboring set. If they match, it indicates that the data from that sensor node is highly reliable, and the moisture data can be converted into real data. This can prevent real data from being treated as abnormal data for calibration, thus avoiding the inability to irrigate in a timely manner based on actual data.
[0044] b. Abnormal data calibration: Determine whether the resampled data passes the neighborhood consistency check; If the resampled data passes the neighborhood consistency check, the resampled data will be used as the calibrated data. If the resampled data fails the neighborhood consistency check, the neighborhood weighted average of the neighborhood set is used as the calibrated data.
[0045] When the resampled data passes the neighborhood consistency check, it indicates that the sampling result is reliable and the resampled data can be used for calibration. Since the resampled data includes multiple samples, the average value of the resampled data can be used as the calibrated data.
[0046] If the resampled data is neither consistent with the data trend of the neighborhood set nor passes the neighborhood consistency check, the sensor data can be considered abnormal with low confidence. In this case, the weighted average of the neighborhood set can be used as the calibrated data.
[0047] Generally, sensor data anomalies can be categorized into three types: transient disturbances, persistent drift, and faults. Sensor calibration can be performed based on the type of anomaly. All three types of anomalies are calibrated using a neighborhood-weighted average. For persistent drift, the bias correction coefficient can be updated to address the overall baseline shift caused by the drift, thus completing the data calibration of the target sensor node. In subsequent calculations of irrigation grid data, since the target sensor node has already experienced drift, its weight can be reduced to minimize its impact on the irrigation grid data. For faults, its weight is reset to zero, and maintenance is notified.
[0048] It should be noted that in this embodiment, data calibration can be differentiated based on the sensor type. For hydrothermal-salt sensors, if an integrated shared electrode probe is used, then if any hydrothermal-salt data fails the neighborhood consistency check, all data will be calibrated using the neighborhood weighted average. For separate independent sensing elements, data that failed the neighborhood consistency check, such as moisture, can be calibrated first; the remaining salinity and temperature data will then undergo neighborhood consistency check.
[0049] The neighborhood weighted average is calculated as follows:
[0050] in, It is a neighborhood-weighted average; Sensor nodes within the neighborhood set Data readings; For sensor nodes The weight in the neighborhood set; m is the neighborhood set number; The weighting is related to the node's historical stability, distance weighting, and data freshness. Nodes that are closer in distance, have higher historical stability, and more recent data receive higher weights. The specific numerical values for these weights are determined according to actual needs and will not be elaborated upon in this application.
[0051] By calculating the neighborhood weighted average of moisture, temperature, and salinity as needed, the data calibration of the sensor nodes can be completed.
[0052] S203. Calculate irrigation grid data based on real data and calibrated sensor node data.
[0053] After calibration, the irrigation grid data can be calculated based on real data and calibrated sensor node data. When a single irrigation grid is configured with only one sensor node, the neighborhood-weighted average of the corresponding sensor node can be used. This serves as irrigation grid data. When a single irrigation grid is configured with multiple sensor nodes, the average value of the data from the multiple sensor nodes can be calculated to obtain the irrigation grid data.
[0054] S3. Determine whether the irrigation grid data meets the irrigation conditions or the conditions for stopping irrigation. If the irrigation conditions are met, start the irrigation mechanism corresponding to the irrigation grid. If the conditions for stopping irrigation are met, stop the irrigation mechanism corresponding to the irrigation grid.
[0055] The edge computing gateway stores a grid mapping table; the grid mapping table includes irrigation grid number, sensor node number, irrigation agency number, crop type, growth period parameters, irrigation start threshold and irrigation stop threshold.
[0056] Since different crops and crops at different growth stages may have different water requirements, it is necessary to obtain crop type and growth stage parameters to facilitate the subsequent determination of irrigation start and stop thresholds.
[0057] The edge computing gateway determines whether the irrigation grid data meets the irrigation conditions or the conditions for stopping irrigation based on the grid mapping table, and controls the irrigation mechanism to start or stop irrigation.
[0058] In this embodiment, the irrigation mechanism is equipped with an electric valve, which is used to open the electric valve for grid irrigation or close the electric valve to stop irrigation according to instructions. The irrigation method can be drip irrigation, sprinkler irrigation, micro-sprinkler irrigation, pipe irrigation, or fertigation; the valve can be a solenoid valve, electric ball valve, electric butterfly valve, or a proportional valve with opening feedback.
[0059] Edge computing gateways are designed for each irrigation grid. Read the calibrated moisture content ,temperature Salt It then retrieves the crop type, growth stage, soil texture, and irrigation method parameters corresponding to that grid to form the irrigation trigger threshold. Stop irrigation threshold High temperature constraint threshold and salt constraint threshold In this application, the confidence level of the sensor nodes can also be dynamically adjusted. When a sensor node experiences a fault such as a momentary disturbance or a long-term drift, its confidence level can be lowered.
[0060] In this embodiment, the irrigation grid data includes soil moisture data, soil temperature data, and soil salinity data; The conditions for meeting irrigation requirements include: soil moisture data in the irrigation grid is below the irrigation start threshold; soil temperature data is below the high temperature constraint threshold; and soil salinity data is below the salinity constraint threshold.
[0061] Irrigation can only begin when all of the above conditions are met. Irrigation decisions are based not only on soil moisture but also on soil temperature and salinity data to reduce the risk of incorrect irrigation caused by abnormalities such as high temperature or salinity stress.
[0062] Irrigation can be stopped when: the soil moisture data in the irrigation grid reaches the stop irrigation threshold, the irrigation duration reaches the maximum limit, or the valve feedback is abnormal.
[0063] Irrigation should be stopped if any of the above conditions are met.
[0064] In this embodiment, during normal monitoring, the sensor node collects data at a first sampling frequency, and after the irrigation mechanism starts irrigating, it collects data at a second sampling frequency; wherein the second sampling frequency is higher than the first sampling frequency.
[0065] The purpose of this step is to increase the sampling frequency of the target grid to monitor the rate of change in soil moisture per unit time. Similarly, the sampling frequency of its adjacent grids can be increased simultaneously to facilitate neighborhood consistency verification. If the moisture content of the target grid does not show a significant upward trend within a preset time after the valve is opened, or if the irrigation mechanism reports an abnormal switch status, the edge computing gateway will issue an abnormal alarm and can switch to manual confirmation, backup irrigation mechanism, etc.
[0066] In one embodiment of the present invention, the sensor node and the edge computing gateway communicate using one or more of the following communication methods: LoRa, NB-IoT, ZigBee, Bluetooth Mesh, RS485, and Ethernet. The edge computing gateway and the irrigation mechanism communicate using one or more of the following communication methods: LoRa, RS485, and relay control. In this embodiment, the sensor node and the edge computing gateway are communicatively connected for data transmission and command issuance, and the transmission methods include, but are not limited to, the above-mentioned communication transmission methods. The communication connection between the edge computing gateway and the irrigation mechanism is similar.
[0067] In this embodiment of the invention, the irrigation grid is used as the control unit, which can independently open valves for areas with local water shortages, thus preventing areas with sufficient water from being irrigated simultaneously.
[0068] Example 2
[0069] like Figure 2As shown in the figure, this invention also proposes a multi-sensor collaborative calibration and zoned irrigation control system for farmland soil, including: An initialization module is used to divide farmland into irrigation grids 1, and to configure sensor nodes 2 and irrigation mechanisms 4 for each irrigation grid; Sensor node 2 is used to upload soil moisture, temperature and salinity data to the edge computing gateway; The edge computing gateway 3 includes a neighborhood consistency verification and data fusion module 31 and a gridded irrigation decision module 32. The neighborhood consistency verification and data fusion module 31 is used to establish a neighborhood set for each sensor node based on the location of the sensor node; receive soil moisture, temperature and salinity data uploaded by sensor node 2 and perform neighborhood consistency verification within the neighborhood set; mark the data that passes the neighborhood consistency verification as real data; mark the sensor node data that fails the neighborhood consistency verification as candidate abnormal data and perform candidate abnormal data calibration; and calculate irrigation grid data based on real data and calibrated sensor node data. The grid-based irrigation decision module 32 is used to determine whether the irrigation grid data meets the irrigation conditions or the conditions for stopping irrigation. When the irrigation conditions are met, it sends a valve opening command to the irrigation mechanism corresponding to the irrigation grid. When the conditions for stopping irrigation are met, it sends a valve closing command to the irrigation mechanism corresponding to the irrigation grid. Irrigation mechanism 4 is used to execute valve opening or closing commands sent by edge computing gateway 3.
[0070] The farmland soil multi-sensor collaborative calibration and zoned irrigation control system proposed in this invention may also include a cloud platform. The cloud platform is mainly used for long-term data storage, strategy parameter management, remote monitoring, and model updates. When the network is normal, the edge computing gateway uploads sensor calibration data, anomaly diagnosis results, irrigation mechanism status, and irrigation logs to the cloud platform. The cloud platform can generate or update threshold strategies for different crops, different growth stages, and different soil types based on historical data, and distribute them to the edge computing gateway.
[0071] When the network connection between the edge computing gateway and the cloud platform is interrupted, the edge computing gateway continues to perform sensor scheduling, anomaly verification, and irrigation mechanism control based on locally cached threshold policies and grid mapping tables, ensuring that farmland irrigation is not affected by network interruption. After the network is restored, the edge computing gateway automatically retransmits the data and logs from the period of network outage.
[0072] In this embodiment, the edge computing gateway may also include a local storage and offline autonomy module, which is used to cache sensor data, calibration results, irrigation strategies, valve status and abnormal logs, and maintain local irrigation control operation when communication between the edge computing gateway and the cloud platform is interrupted, and retransmit data to the cloud platform after communication is restored.
[0073] Example 3
[0074] This embodiment provides a computer-readable storage medium storing a computer program thereon. When the program is executed by a processor, it implements the steps of the farmland soil multi-sensor collaborative calibration and zoned irrigation control method as described in Embodiment 1.
[0075] Example 4: Grid-based irrigation control in 200 mu (approximately 33 hectares) of winter wheat farmland
[0076] In a 200-mu (approximately 33 acres) winter wheat field, the plot was divided into 80 irrigation grids of 50 m × 50 m each. Each grid was equipped with a soil moisture, temperature, and salinity sensor node, which collected data on soil moisture, temperature, and salinity in the 0-20 cm or 0-40 cm soil layer. Every four adjacent irrigation grids shared one electric irrigation valve, with a total of 20 valves installed. An edge computing gateway was located in an electrical control box at the field edge, communicating with the sensor nodes via LoRa and with the valve controller via RS485 or LoRa.
[0077] A grid mapping table is established in the edge computing gateway, including grid number, sensor number, valve number, crop type, growth stage, field water capacity, start threshold, and stop threshold. For winter wheat at the jointing stage, the start irrigation threshold can be configured to 55% of field water capacity, and the stop irrigation threshold to 75% of field water capacity.
[0078] When the sensor node in grid G12 measures soil moisture at 48%, and the moisture values of neighboring grids are 61%, 63%, 62%, and 60%, respectively, the edge computing gateway calculates the neighborhood median to be 61.5% and determines that the deviation between the sensor node's median and its neighborhood median exceeds a preset threshold. The edge computing gateway sends a collaborative query frame to the sensor node, requesting it to continuously and rapidly collect data three times, obtaining values of 48.2%, 47.9%, and 48.1%. If the moisture data in S12 simultaneously shows a trend consistent with the neighborhood, it is determined that there is a real local water shortage in that grid. The edge computing gateway then sends an opening command to the corresponding valve to irrigate the grid area covered by the corresponding irrigation mechanism.
[0079] After irrigation begins, the edge computing gateway reduces the sampling period for the target sensor node and adjacent nodes from 5 minutes to 1 minute. When the moisture level at the target sensor node gradually rises and reaches approximately 75%, the edge computing gateway sends a shutdown command to the irrigation system and records the start and end times, valve number, moisture change curve, and estimated water consumption for this irrigation session. Valves in other well-watered grids remain closed.
[0080] In this application, the edge computing gateway completes key judgments and controls locally in the field, without relying on real-time cloud responses, making it suitable for weak network environments in large fields. Furthermore, the edge computing gateway first completes anomaly identification, data fusion, and event judgment, uploading only necessary data, logs, and alarms, reducing the indiscriminate uploading of high-frequency raw data and lowering the pressure on the cloud.
[0081] Example 5: Anomaly Sensor Identification and Weighted Fusion
[0082] The neighborhood set of a certain sensor S03 includes sensor nodes S01, S02, S04, and S05. At a certain moment, the reported soil moisture values are 21.3%, 22.1%, 21.8%, and 22.5%, respectively, while S03's is 37.5%. The edge computing gateway calculates the neighborhood median value to be 22.1%. S03 does not meet the neighborhood consistency check and is marked as candidate abnormal data.
[0083] The edge computing gateway sends a collaborative query frame to node S03, requesting it to report data three times consecutively. The results are 37.2%, 37.6%, and 37.4%. Meanwhile, node S03 has consistently failed to meet neighborhood consistency checks over the past 24 hours. Based on this, the edge computing gateway determines that node S03 has a risk of sensor contamination, abnormal contact, or malfunction, resets its weight to 0, and uses the weighted average of nodes S01, S02, S04, and S05 as the current decision data for this grid, while simultaneously generating a maintenance alarm.
[0084] Example 6: Closed-loop judgment of valve actuation failure
[0085] After the edge computing gateway sends an opening command to valve V08 of the irrigation system, the valve controller reports that it is open. However, the soil moisture in the target grid does not show an upward trend within 10 minutes, and there is no irrigation water response from adjacent grids. The edge computing gateway determines that there may be valve jamming, pipeline blockage, water shortage in irrigation branches, or abnormal sensor contact, issues an alarm, and suspends irrigation. If a backup valve or a valve connected to an adjacent branch is set in the system, the edge computing gateway can switch to the backup strategy when safety conditions are met.
[0086] Simulation Experiment 1
[0087] To further verify the technical effectiveness of the proposed solution, simulation experiments were conducted in a Python environment with the following computer configuration: Microsoft Windows 11 system, Intel(R) Core(TM) Ultra 7 155H processor (1.40 GHz) and 32G RAM.
[0088] A total of 20 sensor nodes (S01, S02, ..., S19, S20) were configured and arranged within 20 irrigation grids. The neighborhood set used a four-neighbor approach (top, bottom, left, right). The number of simulation samples was 240, and the sampling interval was 30 minutes. Simulation scenarios included: continuous drift of sensor node S03; instantaneous disturbance (S08); localized water shortage (S12); fault (S15); and salinity drift (S18). The root mean square error (RMSE) of each node was calculated, and the calibration results are as follows: Figure 3 As shown.
[0089] from Figure 3 It can be seen that the solution proposed in this application can effectively calibrate abnormal data such as faults, salt drift, and transient disturbances.
[0090] Through the above description of the embodiments, those skilled in the art can clearly understand that this application can be implemented by means of software plus necessary general-purpose hardware, and of course, it can also be implemented by special hardware including application-specific integrated circuits, special CPUs, special memory, special components, etc. Generally, any function performed by a computer program can be easily implemented by corresponding hardware, and the specific hardware structure used to implement the same function can also be diverse, such as analog circuits, digital circuits, or special-purpose circuits. However, for this application, software program implementation is more often the preferred implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a readable storage medium, such as a computer floppy disk, USB flash drive, mobile hard disk, ROM, RAM, magnetic disk, or optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of the various embodiments of this application.
[0091] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product. A computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the flow or function according to the embodiments of this application is generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can store or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state disk (SSD)).
[0092] Finally, it should be noted that the above description is merely a preferred embodiment of the present invention and is only used to illustrate the technical solution of the present invention, and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention are included within the scope of protection of the present invention.
Claims
1. A method for multi-sensor collaborative calibration and zoned irrigation control of farmland soil, characterized in that, include: S1. Divide the farmland into irrigation grids and configure water, heat and salt sensor nodes and irrigation mechanisms for each irrigation grid; S2. The edge computing gateway establishes a neighborhood set for each sensor node based on the location of the sensor node, receives soil moisture, temperature and salinity data uploaded by the sensor node, and performs neighborhood consistency verification on the soil moisture, temperature and salinity data within the neighborhood set. Data that passes the neighborhood consistency verification is marked as real data, and data that fails the neighborhood consistency verification is marked as candidate abnormal data. Candidate abnormal data calibration is performed, and irrigation grid data is calculated based on real data and calibrated data. S3. Determine whether the irrigation grid data meets the irrigation conditions or the conditions for stopping irrigation. If the irrigation conditions are met, start the irrigation mechanism corresponding to the irrigation grid. If the conditions for stopping irrigation are met, stop the irrigation mechanism corresponding to the irrigation grid.
2. The method for multi-sensor collaborative calibration and zoned irrigation control of farmland soil as described in claim 1, characterized in that, S2 specifically includes: S201, Edge computing gateway receives sensor nodes Uploaded soil moisture, temperature, and salinity data, and based on data from each sensor node. Location information for each sensor node Establish neighborhood set ; S202, Obtain the neighborhood set Data from all sensor nodes in the system, and calculation parameters. neighborhood median and standard deviation Parameters of sensor nodes reading A neighborhood consistency check is performed. Data that passes the neighborhood consistency check is marked as real data; sensor node data that fails the neighborhood consistency check is marked as candidate anomalous data, and candidate anomalous data calibration is performed. The neighborhood consistency check is as follows: in, For sensor nodes parameters Readings; parameters It is one of the soil's moisture, temperature, and salinity; For neighborhood set Intrinsic parameters The median value; For neighborhood set Intrinsic parameters Standard deviation; This is the threshold coefficient; Number the sensor nodes; S203. Calculate irrigation grid data based on real data and calibrated sensor node data.
3. The method for multi-sensor collaborative calibration and zoned irrigation control of farmland soil as described in claim 2, characterized in that, The candidate anomaly data calibration includes: a. Determine the accuracy of candidate outlier data; The sensor nodes corresponding to the candidate abnormal data are resampled, and the data change trend of their neighborhood set is acquired simultaneously; it is determined whether the change trend of the resampled data is consistent with the data change trend of the neighborhood set; if they are consistent, the candidate abnormal data is converted into real data; otherwise, abnormal data calibration is performed. b. Abnormal data calibration: Determine whether the resampled data passes the neighborhood consistency check; If the resampled data passes the neighborhood consistency check, the resampled data will be used as the calibrated data. If the resampled data fails the neighborhood consistency check, the neighborhood weighted average of the neighborhood set is used as the calibrated data.
4. The method for multi-sensor collaborative calibration and zoned irrigation control of farmland soil as described in claim 3, characterized in that, The neighborhood weighted average is calculated as follows: in, It is a neighborhood-weighted average; Sensor nodes within the neighborhood set Data readings; For sensor nodes The weight in the neighborhood set; m is the neighborhood set number.
5. The method for multi-sensor collaborative calibration and zoned irrigation control of farmland soil as described in claim 1, characterized in that, The edge computing gateway stores a grid mapping table; The grid mapping table includes irrigation grid number, sensor node number, irrigation agency number, crop type, growth period parameters, irrigation start threshold, and irrigation stop threshold; The edge computing gateway determines whether the irrigation grid data meets the irrigation conditions or the conditions for stopping irrigation based on the grid mapping table, and controls the irrigation mechanism to start or stop irrigation.
6. The method for multi-sensor collaborative calibration and zoned irrigation control of farmland soil as described in claim 5, characterized in that, The specific steps for determining whether the irrigation grid data meets the irrigation conditions or the conditions for stopping irrigation include: The irrigation grid data includes soil moisture data, soil temperature data, and soil salinity data; Meeting the irrigation conditions includes: soil moisture data in the irrigation grid is below the irrigation start threshold; and soil temperature data is below the high temperature constraint threshold and soil salinity data is below the salinity constraint threshold. Irrigation can be stopped when: the soil moisture data in the irrigation grid reaches the stop irrigation threshold, the irrigation duration reaches the maximum limit, or the valve feedback is abnormal.
7. The method for multi-sensor collaborative calibration and zoned irrigation control of farmland soil as described in claim 1, characterized in that, The sensor node communicates with the edge computing gateway using one or more of the following communication methods: LoRa, NB-IoT, ZigBee, Bluetooth Mesh, RS485, and Ethernet. The edge computing gateway communicates with the irrigation mechanism using one or more of the following communication methods: LoRa, RS485, and relay control.
8. The method for multi-sensor collaborative calibration and zoned irrigation control of farmland soil as described in claim 1, characterized in that, The sensor node collects data at a first sampling frequency and then collects data at a second sampling frequency after the irrigation mechanism is turned on; wherein the second sampling frequency is higher than the first sampling frequency. The data from sensor nodes is used to monitor the rate of change in soil moisture after the irrigation system is activated, and to determine if there are any irrigation anomalies.
9. A multi-sensor collaborative calibration and zoned irrigation control system for farmland soil, characterized in that, include: The initialization module is used to divide farmland into irrigation grids and configure sensor nodes and irrigation mechanisms for each irrigation grid. Hydrothermal-salt sensor nodes are used to collect and upload soil moisture, temperature, and salinity data to the edge computing gateway; The edge computing gateway includes a neighborhood consistency verification and data fusion module and a gridded irrigation decision module. The neighborhood consistency verification and data fusion module is used to establish a neighborhood set for each sensor node based on the location of the sensor node; receive soil moisture, temperature and salinity data uploaded by the sensor nodes and perform neighborhood consistency verification within the neighborhood set; and mark the data that passes the neighborhood consistency verification as real data. Sensor node data that fails the neighborhood consistency check is marked as candidate anomalous data and calibrated; irrigation grid data is calculated based on the real data and the calibrated sensor node data. The grid-based irrigation decision module is used to determine whether the irrigation grid data meets the irrigation conditions or the conditions for stopping irrigation. When the irrigation conditions are met, it sends a valve opening command to the irrigation mechanism corresponding to the irrigation grid. When the conditions for stopping irrigation are met, it sends a valve closing command to the irrigation mechanism corresponding to the irrigation grid. Irrigation mechanism, used to execute valve opening or closing commands sent by the edge computing gateway.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps of the farmland soil multi-sensor collaborative calibration and zoned irrigation control method as described in any one of claims 1 to 8.
