Soil monitoring method, device and equipment based on internet of things and storage medium

Abstract

The present application relates to the technical field of information processing, and especially relates to a soil monitoring method, device and equipment based on the Internet of Things and a storage medium, a node deployment density suggestion is generated according to a soil difference atlas, a plurality of intelligent sensing nodes are deployed in a target soil monitoring area based on the node deployment density suggestion, the rationality of node deployment can be improved, abnormal data detection is performed on multi-parameter data based on multi-parameter space-time correlation, dynamic adaptive filtering processing is performed on the multi-parameter data according to the detection result, the reliability of data processing is improved, LoRa ad hoc network or NB-IoT network is selected according to the output result of the decision matrix to transmit effective monitoring data to a cloud platform, the flexibility of data transmission is improved, a soil parameter early warning threshold is dynamically updated according to growth stage information and short-term weather forecast information, the accuracy of early warning is improved, and farming operation suggestions are generated according to effective monitoring data and the soil parameter early warning threshold, and operable intelligent decision support is provided.

Description

Technical Field

[0001] This invention relates to the field of soil monitoring technology, and in particular to a soil monitoring method, device, equipment and storage medium based on the Internet of Things. Background Technology

[0002] With the widespread adoption of IoT technology in agriculture, soil monitoring systems have become a key means of achieving precision irrigation and scientific fertilization. Existing solutions primarily focus on the hardware integration of sensor nodes and remote data transmission. However, in practical applications, especially for small and medium-sized farmers, these solutions still face significant challenges. First, sensor nodes are often deployed uniformly or empirically, failing to fully consider the inherent spatial variability of soil moisture and nutrient parameters in the field. This results in inefficient monitoring networks, excessively high costs, or insufficient monitoring accuracy. Second, node data processing capabilities are limited, typically employing simple fixed threshold filtering, which struggles to effectively distinguish between false anomalies caused by instantaneous sensor interference and minor environmental changes, and genuine agricultural anomalies, thus requiring improved data reliability. Third, node communication strategies are simplistic, often relying on fixed LoRa or NB-IoT networks, failing to dynamically balance communication coverage, energy consumption, and cost based on real-time operating conditions. Furthermore, cloud-based early warning models often use static thresholds, failing to integrate with the dynamic water and fertilizer requirements of crop growth and short-term weather changes, leading to untimely or false alarms. Finally, most existing systems are limited to data display and lack the intelligent decision-making ability to integrate multi-source monitoring data, agronomic knowledge and real-time environmental information to generate specific and actionable agricultural recommendations. Summary of the Invention

[0003] In order to overcome the shortcomings of the prior art, the present invention aims to provide an Internet of Things-based soil monitoring method, device, equipment and storage medium that can improve the rationality of node deployment, improve the reliability of data processing, improve the flexibility of data transmission, improve the accuracy of early warning and provide operable intelligent decision support.

[0004] The first aspect of this invention provides an Internet of Things (IoT)-based soil monitoring method, comprising: generating a node deployment density suggestion based on a soil difference map of a target soil monitoring area; deploying multiple intelligent sensing nodes in the target soil monitoring area based on the node deployment density suggestion; and collecting multi-parameter data of the soil using the intelligent sensing nodes; performing anomaly detection on the multi-parameter data based on the spatiotemporal correlation of the multi-parameters to obtain detection results; performing dynamic adaptive filtering on the multi-parameter data based on the detection results to obtain effective monitoring data; constructing a decision matrix based on data priority, remaining node power, and network signal strength; selecting a LoRa ad hoc network or an NB-IoT network to transmit the effective monitoring data to a cloud platform based on the output of the decision matrix; acquiring real-time information on the current growth stage of the target crop and short-term weather forecast information for the target soil monitoring area; dynamically updating soil parameter warning thresholds based on the growth stage information and the short-term weather forecast information; generating agricultural operation suggestions in the cloud platform using a preset agricultural decision model based on the effective monitoring data and the soil parameter warning thresholds; and pushing the agricultural operation suggestions to a user terminal.

[0005] Optionally, in a first implementation of the first aspect of the present invention, the step of generating a node deployment density suggestion based on a soil difference map of the target soil monitoring area, deploying multiple intelligent sensing nodes in the target soil monitoring area based on the node deployment density suggestion, and using the intelligent sensing nodes to collect multi-parameter soil data includes: analyzing the spatial variability of soil parameters in the target soil monitoring area based on historical remote sensing data to generate a soil difference map; generating a node deployment density suggestion based on the soil difference map; deploying multiple intelligent sensing nodes in the target soil monitoring area based on the node deployment density suggestion, and using the intelligent sensing nodes to collect multi-parameter soil data, the multi-parameter data including moisture content, temperature, pH value, and nitrogen, phosphorus, and potassium content; and performing data preprocessing on the multi-parameter data.

[0006] Optionally, in a second implementation of the first aspect of the present invention, the step of detecting abnormal data in the multi-parameter data based on multi-parameter spatiotemporal correlation to obtain detection results, and performing dynamic adaptive filtering on the multi-parameter data according to the detection results to obtain effective monitoring data, includes: constructing a multi-parameter empirical correlation model based on the short-term change correlation between different soil parameters at the same node; when the data value of any soil parameter in the multi-parameter data exceeds a preset fixed threshold, the multi-parameter empirical correlation model is used to retrieve the change trends of other correlated parameters in the multi-parameter data to verify the logical rationality of the data value, and a verification result is obtained; when the verification result is logically unreasonable, the data value is marked as high-confidence abnormal data; when the verification result is logically reasonable, it is determined to be an environmental disturbance, and the data value is subjected to enhanced filtering.

[0007] Optionally, in a third implementation of the first aspect of the present invention, the step of constructing a decision matrix based on data priority, remaining node power, and network signal strength, and selecting a LoRa ad hoc network or an NB-IoT network to transmit the effective monitoring data to the cloud platform based on the output of the decision matrix, includes: determining the data type of the multi-parameter data, wherein the data type includes emergency abnormal data, important periodic data, and regular data; matching the corresponding data priority according to the data type, and identifying the remaining node power and network signal strength of the node corresponding to the multi-parameter data; constructing a decision matrix based on the data priority, the remaining node power, and the network signal strength, and selecting a LoRa ad hoc network or an NB-IoT network based on the output of the decision matrix to obtain a selection result; and using the network selected by the selection result to transmit the effective monitoring data to the cloud platform.

[0008] Optionally, in the fourth implementation of the first aspect of the present invention, the step of acquiring the current growth stage information of the target crop and the short-term weather forecast information of the target soil monitoring area in real time, and dynamically updating the soil parameter warning threshold according to the growth stage information and the short-term weather forecast information, includes: determining the current growth stage information of the target crop based on the sowing date and variety of the target crop through a preset growth period model, and matching the soil parameter warning threshold corresponding to the growth stage information from the crop demand database; calling a meteorological data interface to acquire gridded meteorological forecast data of the target soil monitoring area for a future period of time in real time, wherein the gridded meteorological forecast data includes precipitation information, temperature information and sunshine information; calculating a water replenishment factor and an evapotranspiration stress factor based on the precipitation information, the temperature information and the sunshine information; and updating the soil parameter warning threshold according to the water replenishment factor and the evapotranspiration stress factor.

[0009] Optionally, in a fifth implementation of the first aspect of the present invention, the step of generating agricultural operation suggestions based on the effective monitoring data and the soil parameter early warning threshold using a preset agricultural decision-making model on the cloud platform, and pushing the agricultural operation suggestions to the user terminal, includes: comparing the effective monitoring data and the soil parameter early warning threshold using the preset agricultural decision-making model on the cloud platform to obtain a comparison result, and matching agricultural rules according to the comparison result to obtain agricultural operation suggestions; merging the effective monitoring data, the comparison result, and the agricultural operation suggestions to obtain merged information, and generating a structured monitoring report based on the merged information; and pushing the monitoring report to the user terminal so that the user terminal generates and displays a visual report page based on the monitoring report, and provides the user terminal with a feedback entry for the agricultural operation suggestions.

[0010] Optionally, in a sixth implementation of the first aspect of the present invention, after generating agricultural operation suggestions based on the effective monitoring data and the soil parameter warning threshold using a preset agricultural decision-making model in the cloud platform, and pushing the agricultural operation suggestions to the user terminal, the method further includes: obtaining feedback information on the actual effect of agricultural operations sent by the user terminal; optimizing the agricultural decision-making model based on the feedback information on the actual effect of agricultural operations; periodically measuring the electrical characteristic parameters of the intelligent sensing nodes and assessing the degree of node contamination based on the electrical characteristic parameters; generating a contamination alarm when the degree of node contamination exceeds a preset tolerance range and sending the contamination alarm to the user terminal; obtaining contamination cleaning record information sent by the user terminal; and uploading the contamination cleaning record information to the cloud platform.

[0011] A second aspect of the present invention provides an Internet of Things (IoT)-based soil monitoring device, comprising: a generation, deployment, and acquisition module, configured to generate node deployment density recommendations based on a soil difference map of a target soil monitoring area, deploy multiple intelligent sensing nodes in the target soil monitoring area based on the node deployment density recommendations, and collect multi-parameter data of the soil using the intelligent sensing nodes; a detection and filtering module, configured to perform abnormal data detection on the multi-parameter data based on the spatiotemporal correlation of the multi-parameter data, obtain detection results, and perform dynamic adaptive filtering processing on the multi-parameter data based on the detection results to obtain effective monitoring data; a construction, selection, and transmission module, configured to construct a decision matrix based on data priority, remaining node power, and network signal strength, and select a LoRa self-organizing network or an NB-IoT network to transmit the effective monitoring data to a cloud platform based on the output of the decision matrix; an acquisition and update module, configured to acquire in real time the current growth stage information of the target crop and short-term weather forecast information of the target soil monitoring area, and dynamically update the soil parameter warning threshold based on the growth stage information and the short-term weather forecast information; and a generation and push module, configured to generate agricultural operation suggestions in the cloud platform using a preset agricultural decision model based on the effective monitoring data and the soil parameter warning threshold, and push the agricultural operation suggestions to a user terminal.

[0012] Optionally, in a first implementation of the second aspect of the present invention, the generation, deployment, and acquisition module includes: an analysis and generation unit, configured to analyze the spatial variability of soil parameters within the target soil monitoring area based on historical remote sensing data, and generate a soil difference map; a generation unit, configured to generate node deployment density recommendations based on the soil difference map; a deployment and acquisition unit, configured to deploy multiple intelligent sensing nodes in the target soil monitoring area based on the node deployment density recommendations, and use the intelligent sensing nodes to collect multi-parameter data of the soil, the multi-parameter data including moisture content, temperature, pH value, and nitrogen, phosphorus, and potassium content; and a preprocessing unit, configured to perform data preprocessing on the multi-parameter data.

[0013] Optionally, in a second implementation of the second aspect of the present invention, the detection filtering module includes: a construction unit, configured to construct a multi-parameter empirical correlation model based on the short-term change correlation between different soil parameters at the same node; a verification unit, configured to, when the data value of any soil parameter in the multi-parameter data exceeds a preset fixed threshold, use the multi-parameter empirical correlation model to retrieve the change trends of other correlated parameters in the multi-parameter data to perform logical rationality verification on the data value, and obtain a verification result; a marking unit, configured to, when the verification result is logically unreasonable, mark the data value as high-confidence abnormal data; and a filtering unit, configured to, when the verification result is logically reasonable, determine it as an environmental disturbance and perform enhanced filtering processing on the data value.

[0014] Optionally, in a third implementation of the second aspect of the present invention, the construction selection transmission module includes: a determining unit, configured to determine the data type of the multi-parameter data, the data type including emergency abnormal data, important periodic data, and regular data; a matching and identification unit, configured to match the corresponding data priority according to the data type, and identify the remaining node power and network signal strength of the node corresponding to the multi-parameter data; a construction selection unit, configured to construct a decision matrix according to the data priority, the remaining node power, and the network signal strength, and select a LoRa self-organizing network or an NB-IoT network according to the output result of the decision matrix, to obtain a selection result; and a transmission unit, configured to transmit the effective monitoring data to the cloud platform using the network selected by the selection result.

[0015] Optionally, in a fourth implementation of the second aspect of the present invention, the acquisition and update module includes: a matching determination unit, used to determine the current growth stage information of the target crop based on the sowing date and variety of the target crop through a preset growth stage model, and to match the soil parameter early warning threshold corresponding to the growth stage information from a crop demand database; an acquisition call unit, used to call a meteorological data interface to acquire gridded meteorological forecast data of the target soil monitoring area in real time for a period of time in the future, the gridded meteorological forecast data including precipitation information, temperature information and sunshine information; a calculation unit, used to calculate a water replenishment factor and an evapotranspiration stress factor based on the precipitation information, the temperature information and the sunshine information; and an update unit, used to update the soil parameter early warning threshold based on the water replenishment factor and the evapotranspiration stress factor.

[0016] Optionally, in a fifth implementation of the second aspect of the present invention, the generation and push module includes: a comparison and matching unit, configured to compare the effective monitoring data and the soil parameter early warning threshold using a preset agricultural decision-making model on the cloud platform to obtain a comparison result, and to match agricultural rules according to the comparison result to obtain agricultural operation suggestions; a merging and generation unit, configured to merge the effective monitoring data, the comparison result, and the agricultural operation suggestions to obtain merged information, and to generate a structured monitoring report according to the merged information; and a push and providing unit, configured to push the monitoring report to a user terminal so that the user terminal generates and displays a visual report page based on the monitoring report, and provides the user terminal with a feedback entry for the agricultural operation suggestions.

[0017] Optionally, in the sixth implementation of the second aspect of the present invention, the method further includes: a first acquisition module, used to acquire feedback information on the actual effect of agricultural operations sent by the user terminal; an optimization module, used to optimize the agricultural decision-making model based on the feedback information on the actual effect of agricultural operations; a measurement and evaluation module, used to periodically measure the electrical characteristic parameters of the intelligent sensing node and evaluate the degree of node contamination based on the electrical characteristic parameters; a generation and sending module, used to generate a contamination alarm message and send the contamination alarm message to the user terminal when the degree of node contamination exceeds a preset tolerance range; a second acquisition module, used to acquire contamination cleaning record information sent by the user terminal; and an upload module, used to upload the contamination cleaning record information to the cloud platform.

[0018] A third aspect of the present invention provides an Internet of Things (IoT)-based soil monitoring device, the IoT-based soil monitoring device comprising: a memory and at least one processor, the memory storing instructions; the at least one processor invokes the instructions in the memory to cause the IoT-based soil monitoring device to perform various steps of the IoT-based soil monitoring method described in any of the preceding claims.

[0019] A fourth aspect of the present invention provides a computer-readable storage medium storing instructions that, when executed by a processor, implement the steps of the Internet of Things-based soil monitoring method described in any of the preceding claims.

[0020] In the technical solution of this invention, a node deployment density suggestion is generated based on the soil difference map of the target soil monitoring area. Multiple intelligent sensing nodes are deployed in the target soil monitoring area based on this suggestion, improving the rationality of node deployment. The intelligent sensing nodes collect multi-parameter soil data, and anomaly detection is performed on the multi-parameter data based on the spatiotemporal correlation of these data. Dynamic adaptive filtering is applied to the multi-parameter data based on the detection results, improving the reliability of data processing. A decision matrix is ​​constructed based on data priority, remaining node power, and network signal strength. Based on the output of the decision matrix, either a LoRa self-organizing network or an NB-IoT network is selected to transmit the effective monitoring data to the cloud platform, improving the flexibility of data transmission. Real-time acquisition of the current growth stage information of the target crop and short-term weather forecast information for the target soil monitoring area is obtained. Soil parameter warning thresholds are dynamically updated based on the growth stage information and short-term weather forecast information, improving the accuracy of warnings. In the cloud platform, a preset agricultural decision model is used to generate agricultural operation suggestions based on the effective monitoring data and soil parameter warning thresholds, providing operable intelligent decision support. Attached Figure Description

[0021] Figure 1A first flowchart of an Internet of Things-based soil monitoring method provided in an embodiment of the present invention;

[0022] Figure 2 A second flowchart of an Internet of Things-based soil monitoring method provided in an embodiment of the present invention;

[0023] Figure 3 A third flowchart of the Internet of Things-based soil monitoring method provided in an embodiment of the present invention;

[0024] Figure 4 A fourth flowchart of the Internet of Things-based soil monitoring method provided in this embodiment of the invention;

[0025] Figure 5 A schematic diagram of a soil monitoring device based on the Internet of Things provided in an embodiment of the present invention;

[0026] Figure 6 Another structural schematic diagram of the Internet of Things-based soil monitoring device provided in an embodiment of the present invention;

[0027] Figure 7 This is a schematic diagram of the structure of an IoT-based soil monitoring device provided in an embodiment of the present invention. Detailed Implementation

[0028] This invention provides a soil monitoring method, device, equipment, and storage medium based on the Internet of Things, which can improve the rationality of node deployment, improve the reliability of data processing, improve the flexibility of data transmission, improve the accuracy of early warning, and provide operable intelligent decision support.

[0029] The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" or "having" and any variations thereof are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0030] For ease of understanding, the specific process of the embodiments of the present invention is described below. Please refer to [link / reference]. Figure 1 One embodiment of the soil monitoring method based on the Internet of Things in this invention includes:

[0031] 101. Generate node deployment density recommendations based on the soil difference map of the target soil monitoring area, deploy multiple smart sensor nodes in the target soil monitoring area based on the node deployment density recommendations, and use the smart sensor nodes to collect multi-parameter data of the soil.

[0032] In this embodiment, the system first obtains the geographical boundary of the target farmland. By calling historical satellite remote sensing data (such as the NDVI time series of Sentinel-2), it calculates the NDVI standard deviation of each pixel (such as 10m×10m) to quantify the spatial variability of vegetation growth caused by the soil in the area. Based on the degree of variability, the area is divided into high, medium and low variability zones, and a soil difference map is generated. The system automatically generates a node deployment density suggestion map containing specific location points according to preset rules (such as 2-3 nodes / acre in high variability zones and 0.5-1 node / acre in low variability zones). The user deploys intelligent sensing nodes in the field according to the map. After the nodes are powered on, they automatically collect raw data on moisture content, temperature, pH value and nitrogen, phosphorus and potassium content.

[0033] 102. Based on the spatiotemporal correlation of multiple parameters, anomaly detection is performed on multi-parameter data to obtain detection results. Based on the detection results, dynamic adaptive filtering is performed on the multi-parameter data to obtain effective monitoring data.

[0034] In this embodiment, the edge computing unit built into the node runs a dynamic adaptive filtering algorithm. When the reading of any parameter (such as moisture content) exceeds a fixed threshold, the algorithm immediately retrieves the associated parameter (such as temperature) data collected by the same node at the same time and performs logical verification based on a preset multi-parameter empirical correlation model (such as "a sudden rise in temperature is usually accompanied by a gradual decrease in humidity"). If the verification passes, it is determined to be an environmental disturbance, and an enhanced moving average filter is applied to the data. If the verification fails, it is marked as high-confidence abnormal data and the original characteristics are preserved. This process effectively filters out interference such as instantaneous sensor errors or root contact, and outputs reliable and effective monitoring data.

[0035] 103. Construct a decision matrix based on data priority, node remaining power, and network signal strength. Select either a LoRa self-organizing network or an NB-IoT network based on the output of the decision matrix to effectively transmit monitoring data to the cloud platform.

[0036] In this embodiment, nodes are prioritized based on data content: urgent and abnormal data is given high priority, important periodic data is given medium priority, and routine data is given low priority. Simultaneously, nodes monitor their remaining battery power in real time and assess the LoRa self-organizing network link quality and NB-IoT network signal strength (RSRP). The system constructs a decision matrix based on these three factors. For example, high-priority data will prioritize the use of the NB-IoT network when the NB-IoT network signal is good to ensure reliability; when the battery is low, regardless of the type of data, the lower-power LoRa self-organizing network will be prioritized. Based on the matrix output, nodes intelligently select the optimal communication network to effectively transmit monitored data to the cloud platform.

[0037] 104. Real-time acquisition of the current growth stage information of the target crop and short-term weather forecast information of the target soil monitoring area; dynamic updating of soil parameter warning thresholds based on growth stage information and short-term weather forecast information.

[0038] In this embodiment, the cloud platform calculates the current growth stage (e.g., "tomato-fruit expansion stage") based on the user's preset crop variety and sowing date using a standard growth stage model. It then retrieves the soil parameter warning threshold for this stage from the crop demand database. Simultaneously, the platform obtains the gridded weather forecast for the region for the next 24-72 hours via API, calculates the water replenishment factor based on future precipitation, and calculates the evapotranspiration stress factor based on the forecast high temperature. Finally, it uses these factors to perform a weighted correction on the soil parameter warning threshold, so that the soil parameter warning threshold is dynamically adjusted according to crop growth and weather changes.

[0039] 105. Using a pre-set agricultural decision-making model on the cloud platform, generate agricultural operation suggestions based on effective monitoring data and soil parameter early warning thresholds, and push the agricultural operation suggestions to the user terminal.

[0040] In this embodiment, the cloud platform's agricultural decision-making model receives valid monitoring data and soil parameter early warning thresholds. The core of the agricultural decision-making model is an agricultural rule base, which stores a large number of "IF-THEN" rules, such as: "IF area A's average moisture content is less than the dynamic irrigation threshold AND no rainfall in the next 24 hours THEN recommends immediate irrigation." The agricultural decision-making model matches the current data with the rules to generate structured agricultural operation suggestions that include specific operations (such as "irrigation"), quantitative indicators (such as "15 cubic meters / acre"), execution areas, and timing. These agricultural operation suggestions are integrated with data reports, trend charts, etc., into a monitoring report, which is sent to farmers' mobile terminals via APP push, SMS, etc.

[0041] In this embodiment of the invention, a complete technical closed loop from "spatial difference analysis - scientific site selection" to "data-driven decision-making" is realized. Specifically, node deployment based on soil difference maps fundamentally optimizes the spatial structure of the monitoring network, maximizing information acquisition efficiency under cost constraints; dynamic adaptive filtering based on multi-parameter spatiotemporal correlation enables intelligent identification of data credibility at the edge, effectively suppressing false alarms and improving the quality of raw data; a communication decision matrix integrating multiple factors enables the system to have adaptive networking capabilities in complex farmland environments, balancing reliability, real-time performance, and energy consumption; dynamic early warning thresholds integrating crop growth models and meteorological information upgrade static monitoring to predictive early warning with agronomic and climate perception capabilities, significantly improving timeliness and accuracy; finally, specific and actionable suggestions are generated through agricultural decision-making models, bridging the "last mile" from data to action and providing users with direct guidance on planting management.

[0042] Please see Figure 2 In the second embodiment of the soil monitoring method based on the Internet of Things in this invention, steps 101 and 102 include:

[0043] 201. Based on historical remote sensing data, analyze the spatial variability of soil parameters within the target soil monitoring area and generate a soil difference map;

[0044] In this embodiment, the system connects to the European Space Agency's Copernicus Open Access Centre and automatically downloads Sentinel-2 L2A level image data of the target area over the past 2-3 growing seasons. By calculating the Normalized Difference Vegetation Index (NDVI) of each image and analyzing the standard deviation of the NDVI value of each pixel over the entire time series, the spatial variability intensity of soil characteristics (such as moisture and fertility) is quantified. Areas with high standard deviations indicate unstable soil conditions or strong heterogeneity. Finally, all pixels are divided into different levels according to the variability intensity and visualized in map form to generate an intuitive soil difference map.

[0045] 202. Develop node deployment density recommendations based on soil difference maps;

[0046] In this embodiment, the system presets deployment density mapping rules: high-variability areas in the soil difference map are mapped to high-density deployment suggestions (e.g., 2.5 nodes / acre), medium-variability areas are mapped to medium density (e.g., 1.2 nodes / acre), and low-variability areas are mapped to low density (e.g., 0.7 nodes / acre). Based on the actual shape of the farmland, the algorithm automatically generates specific and reasonable node layout coordinates in each variation area, forming a "node deployment density suggestion map" that can directly guide field operations, and distributes it to farmers through user terminals.

[0047] 203. Based on node deployment density, it is recommended to deploy multiple smart sensor nodes in the target soil monitoring area and use the smart sensor nodes to collect multi-parameter data of the soil, including moisture content, temperature, pH value and nitrogen, phosphorus and potassium content.

[0048] In this embodiment, farmers insert smart sensor nodes into designated locations in the soil based on the node deployment density recommendation map received by the terminal. The nodes adopt an integrated probe design with a fixed insertion depth of 20 cm (mainly the active root layer). After power-on, the capacitive sensor, composite pH electrode, and ion-selective electrode inside the node start working, respectively collecting soil volumetric water content, temperature, pH value, and relative content of nitrogen, phosphorus, and potassium ions, and temporarily storing these raw multi-parameter data in a local cache.

[0049] 204. Perform data preprocessing on multi-parameter data;

[0050] In this embodiment, before edge filtering, the nodes first perform basic preprocessing on the raw acquired data, which includes: 1) invalid value removal: discarding values ​​that are obviously beyond physical possibility (such as moisture content > 100%) according to the sensor range; 2) unit standardization: converting the raw voltage or frequency signals output by each sensor into standard units (such as %, pH value, mg / kg) according to the factory calibration curve; 3) timestamp alignment: assigning a unified timestamp to different parameter data at the same acquisition time to ensure spatiotemporal consistency of data. Preprocessing provides a clean and orderly data foundation for subsequent advanced analysis.

[0051] 205. Construct a multi-parameter empirical correlation model based on the short-term variation correlation between different soil parameters at the same node;

[0052] In this embodiment, to achieve online logic verification, the system pre-configures a lightweight multi-parameter empirical correlation model in the node MCU. The multi-parameter empirical correlation model is essentially a lookup table or a set of simple linear relationships, describing the reasonable correlation between short-term changes of various parameters on the node under normal environmental conditions. For example, it defines that "within 1 hour, for every 1 degree Celsius increase in temperature, the moisture content usually decreases within the range of X% due to enhanced evaporation." This multi-parameter empirical correlation model is based on the analysis of a large amount of historical data and the summary of agronomic knowledge, and serves as the logical judgment basis for dynamic adaptive filtering.

[0053] 206. When the value of any soil parameter in the multi-parameter data exceeds the preset fixed threshold, the change trend of other related parameters in the multi-parameter data is retrieved using the multi-parameter empirical correlation model to verify the logical rationality of the data value and obtain the verification result.

[0054] In this embodiment, assuming that the original value of soil moisture content suddenly drops below a fixed threshold (e.g., 15%), triggering a preliminary anomaly, the system immediately retrieves the temperature data of the same node within the last hour. If the data shows that the temperature rises sharply at the same time, it conforms to the correlation model of "high temperature evaporation", and the verification result is marked as "logically reasonable". If the temperature is stable or drops, it cannot be explained by the correlation model, and the verification result is marked as "logically unreasonable".

[0055] 207. When the verification result is logically unreasonable, the data value is marked as high-confidence outlier data;

[0056] In this embodiment, if the moisture content drops sharply but the temperature does not rise, the verification result is "logically unreasonable". The system determines that this anomaly is not caused by environmental disturbance, but may be due to real water shortage or sensor failure. Therefore, it is marked as "high confidence abnormal data". Such data will be given the highest transmission priority and will bypass most of the smoothing filtering processing to retain its abrupt change characteristics, which will be used to trigger emergency warnings and immediate reporting.

[0057] 208. If the verification result is logically reasonable, it is determined to be an environmental disturbance, and the data value is subjected to enhanced filtering.

[0058] In this embodiment, if the decrease in moisture content is accompanied by an increase in temperature, and the verification result is "logically reasonable", the system determines that the threshold breach is likely caused by environmental disturbances such as brief strong sunlight and strong winds, rather than a substantial change in soil moisture. Therefore, the node will apply "enhanced filtering processing" to this moisture content data, such as using median filtering with a larger window or more stringent mean filtering to strongly smooth it and suppress abnormal reporting caused by this disturbance, thereby avoiding false alarms and saving communication resources.

[0059] In this embodiment of the invention, firstly, soil difference maps are generated by analyzing historical remote sensing data (such as NDVI spatiotemporal variability) to guide node deployment. This transforms the spatial variability theory of soil science into an executable, engineered deployment strategy, realizing a paradigm shift from "uniform deployment" to "on-demand focusing." It achieves precise capture of field heterogeneity with optimal hardware density. Secondly, the proposed "multi-parameter association verification" dynamic adaptive filtering algorithm creatively embeds agricultural knowledge (physiological and physical correlations between parameters) into edge data stream processing. It distinguishes between environmental disturbances and real anomalies through real-time logical reasoning. This not only improves the reliability of single-point data at the algorithm level but also enables low-cost sensors to obtain data credibility similar to high-precision sensors through algorithmic intelligence.

[0060] Please see Figure 3 In the third embodiment of the soil monitoring method based on the Internet of Things in this invention, steps 103 and 104 include:

[0061] 301. Determine the data type of multi-parameter data, including emergency abnormal data, important periodic data, and regular data;

[0062] In this embodiment, moisture content and temperature are collected at high frequency, while pH value and nitrogen, phosphorus and potassium content are collected at low frequency. Before sending data, the node first performs content analysis on the valid monitoring data cached locally to determine its data type. The specific rules are as follows: data packets of pH value and nitrogen, phosphorus and potassium content are classified as important periodic data; data packets of moisture content and temperature are classified as regular data; and any data marked as "high confidence anomaly" is classified as urgent anomaly data.

[0063] 302. Match the corresponding data priority according to the data type, and identify the remaining power and network signal strength of the node corresponding to the multi-parameter data;

[0064] In this embodiment, the system presets a data priority mapping: emergency and abnormal data corresponds to the highest priority, important periodic data corresponds to the medium priority, and regular data corresponds to the lowest priority. At the same time, the node reads the voltage information of the power management module, converts it into the remaining power percentage (e.g., 45%), measures the recent communication success rate of the LoRa self-organizing network, and measures the reference signal received power (RSRP) of the NB-IoT network, which are quantified into the LoRa self-organizing network link quality index and the NB-IoT network signal strength index, respectively.

[0065] 303. Construct a decision matrix based on data priority, remaining node power, and network signal strength. Select either a LoRa self-organizing network or an NB-IoT network based on the output of the decision matrix to obtain the selection result.

[0066] In this embodiment, the system constructs a ternary decision matrix based on data priority, remaining node power, and network signal strength. For example, data priority can be quantified into numerical scores: the highest priority is assigned 3 points, medium priority is assigned 2 points, and the lowest priority is assigned 1 point; remaining node power is divided into three levels: power ≥ 70% is "high" and assigned 3 points, 30% ≤ power < 70% is "medium" and assigned 2 points, and power < 30% is "low" and assigned 1 point; network signal strength (including link quality of LoRa self-organizing networks and RSRP of NB-IoT networks) is quantified into three levels: signal strength ≥ good threshold is "strong" and assigned 3 points, and general threshold ≤ signal strength is "strong". A signal strength score less than the "good" threshold is assigned a score of 2, and a signal strength score less than the "normal" threshold is assigned a score of 1. Based on a pre-configured decision rule table, the scores of these three dimensions are comprehensively evaluated to output the final network selection result. For example, the rule can be set as follows: if the data priority is 3 points and the NB-IoT network signal strength is ≥ 2 points, then the output selection result is the "NB-IoT network" with fast data transmission; if the node's remaining battery power score is 1 point, then regardless of the scores of other dimensions, the output selection result is the "LoRa self-organizing network" with low energy consumption; in other cases, the comprehensive weighted score of the LoRa self-organizing network and the NB-IoT network is calculated, and the network with the higher score is selected as the output result.

[0067] 304. The network selected using the selection results will effectively transmit monitoring data to the cloud platform;

[0068] In this embodiment, if the selected result is "LoRa self-organizing network", the node activates the LoRa radio frequency module, encapsulates the effective monitoring data into data frames conforming to the LoRaWAN protocol, and sends them to the nearest LoRa gateway (which can be wired or 4G) through the configured frequency band and spreading factor, and the gateway forwards them to the cloud platform. If the selected result is "NB-IoT network", the node activates the NB-IoT communication module, attaches to the cellular network, encapsulates the effective monitoring data into message packets conforming to the CoAP or MQTT protocol, and transmits them to the cloud platform through the base station. After data transmission, the node starts an acknowledgment timer. If it does not receive a reception confirmation from the cloud platform within a set time, it retransmits according to the preset retransmission strategy. If the retransmission fails, a network switching mechanism can be triggered to try to transmit using another network. The node records key information of this transmission in its local non-volatile memory, including timestamp, selected network, data packet size, transmission status (success / failure), etc., for subsequent communication quality analysis and maintenance.

[0069] 305. Based on the sowing date and variety of the target crop, determine the current growth stage information of the target crop through a preset growth stage model, and match the soil parameter early warning threshold corresponding to the growth stage information from the crop demand database.

[0070] In this embodiment, the user enters the crop variety (such as "summer corn") and sowing date during system initialization. The cloud platform has built-in standard growth period models for different crops (including stages such as emergence, jointing, tasseling, and grain filling, and their approximate number of days). Based on the difference between the current date and the sowing date, the platform calculates the precise growth stage of the crop (such as "tasseling and silking stage"). Subsequently, it queries the crop demand database for the appropriate range of parameters such as soil moisture content and available nitrogen under this stage, which are used as the soil parameter warning threshold.

[0071] 306. Call the meteorological data interface to obtain gridded meteorological forecast data for the target soil monitoring area in real time for a period of time in the future. The gridded meteorological forecast data includes precipitation information, temperature information and sunshine information.

[0072] In this embodiment, the cloud platform periodically (e.g., every 6 hours) calls the public API provided by the meteorological service provider, inputs the latitude and longitude coordinates of the monitoring area, and the interface returns weather forecast data for the next 72 hours with a grid precision (e.g., 5 kilometers). The platform parses the returned JSON data and extracts key fields: cumulative precipitation in the next 24 hours, daily maximum temperature, and daily sunshine hours.

[0073] 307. Based on precipitation, temperature, and sunshine information, the water replenishment factor and evapotranspiration stress factor were calculated.

[0074] In this embodiment, the water replenishment factor and evapotranspiration stress factor are calculated based on precipitation, temperature, and sunshine information. For example, the calculation logic of the water replenishment factor (α) is as follows: if the forecast indicates that there will be more than 5 mm of effective precipitation in the next 24 hours, then α is set to 0.8 (meaning that natural precipitation can partially replenish soil water, and the irrigation threshold can be appropriately relaxed); if there is no precipitation, then α is set to 1.0. The calculation logic of the evapotranspiration stress factor (β) is as follows: if the forecast daily maximum temperature exceeds the upper limit of the suitable temperature for the current growth stage of the crop, then β is set to 1.2 (meaning that high temperature exacerbates evapotranspiration, increases the risk of drought, and the irrigation threshold should be more sensitive); otherwise, β is set to 1.0. Sunshine information can be used to further fine-tune the β value.

[0075] 308. Update the early warning thresholds for soil parameters based on water replenishment factors and evapotranspiration stress factors;

[0076] In this embodiment, taking the lower limit of the irrigation warning threshold as an example, its dynamic update formula is: updated soil parameter warning threshold = previous soil parameter warning threshold × α × β. Assuming that the soil moisture content benchmark irrigation threshold for corn in the tasseling stage is 18%, and there is no rain in the future (α=1.0) but high temperature (β=1.2), then the updated dynamic threshold is 18% × 1.0 × 1.2 = 21.6%. This means that the system will trigger an irrigation warning in advance under high temperature and no rain weather, and the updated dynamic threshold will take effect immediately for data comparison and warning judgment in the current period.

[0077] In this embodiment of the invention, the intelligent communication decision-making mechanism constructs a ternary decision matrix of data priority, node power, and network signal, enabling each sensor node to become an intelligent agent with autonomous network selection capabilities. Without remote intervention, it can dynamically select between LoRa and NB-IoT networks based on its own state and task urgency, achieving real-time optimal configuration of communication resources and service quality at the edge. Simultaneously, the dynamic early warning threshold update method integrates crop growth period models with gridded weather forecast data, transforming the early warning standard from a fixed value into a variable that dynamically evolves with crop growth needs and the external environment. This solves the problem of false early warnings (too early or too late) caused by traditional static threshold models that ignore crop physiological dynamics and weather changes.

[0078] Please see Figure 4 In the fourth embodiment of the soil monitoring method based on the Internet of Things in this invention, steps 105 and 105 thereafter include:

[0079] 401. In the cloud platform, the effective monitoring data and soil parameter early warning thresholds are compared using a preset agricultural decision-making model to obtain the comparison results. Based on the comparison results, agricultural rules are matched to obtain agricultural operation suggestions.

[0080] In this embodiment, the agricultural decision-making model on the cloud platform receives spatiotemporally aligned regional effective monitoring data (e.g., average moisture content of 20.5%) from various nodes, as well as dynamically updated soil parameter warning thresholds (e.g., lower limit of irrigation threshold of 21.6%). The agricultural decision-making model first performs a comparison: comparing the actual monitored value of each parameter with the corresponding soil parameter warning threshold to determine whether the threshold is exceeded, the direction of the exceedance, and the duration, generating a structured comparison result (e.g., "Moisture content in region A is 1.1% below the irrigation threshold, and has been exceeding it for 12 hours"). Subsequently, the agricultural decision-making model compares the comparison result, which includes information such as geographical location, exceeding parameters, degree, and duration, with a preset... The model matches data against a rule base containing numerous "IF-THEN" logic statements, such as: "IF Average moisture content of region X < Dynamic irrigation threshold AND Duration > 6 hours AND No precipitation forecast for the next 24 hours THEN Recommendation: Irrigate region X, with a recommended water replenishment amount of (soil parameter warning threshold - actual monitoring value) * soil bulk density * root depth * coefficient." Upon successful matching, the agricultural decision-making model outputs structured agricultural operation recommendations. These recommendations explicitly include the operation type (irrigation / fertilization / acidity adjustment, etc.), quantitative indicators (water / fertilizer), specific execution area (geographic coordinates or plot number), and recommended timing (e.g., "It is recommended to perform the operation tomorrow morning").

[0081] 402. Merge the effective monitoring data, comparison results and agricultural operation suggestions to obtain merged information, and generate a structured monitoring report based on the merged information;

[0082] In this embodiment, the system integrates all current valid monitoring data (in the form of data tables and spatial heat maps), a detailed list of parameter comparison results, and generated agricultural operation suggestions to form a complete merged information. Based on this, the report generation engine automatically arranges the content: the first part is "Monitoring Overview," which displays the current status of core parameters and overall scores; the second part is "Anomaly and Warning Details," which lists the warning information triggered by the comparison results one by one; the third part is "Trend Analysis," which provides historical change curves of key parameters; and the fourth part is "Agricultural Guidance," which clearly presents agricultural operation suggestions and their basis. The monitoring report is finally generated as a structured document (such as JSON or HTML format) and also includes chart data for visualization, which is convenient for front-end rendering.

[0083] 403. Push the monitoring report to the user terminal so that the user terminal can generate and display a visual report page based on the monitoring report, and provide the user terminal with a feedback entry for agricultural operation suggestions;

[0084] In this embodiment, the cloud platform sends the monitoring report to the user terminal (such as a mobile APP) via mobile application push, SMS or email, according to the warning level and user settings. After receiving the report, the user terminal parses the data and instructions, calls the chart library (such as ECharts) to draw the spatial distribution heat map of soil parameters, historical trend curve, etc., and displays the agricultural operation suggestions in the form of eye-catching cards to generate an intuitive visual report page. At the same time, there are interactive buttons such as "Adopt and Execute", "Do Not Execute for Now" or "Effect Feedback" below each agricultural operation suggestion as a user feedback entry point to collect subsequent execution decisions and actual effects.

[0085] 404. Obtain feedback information on the actual effects of agricultural operations sent by user terminals;

[0086] In this embodiment, after a user takes action based on the suggestion (such as irrigation), they can submit feedback on the effect through the feedback portal of the APP. The feedback information includes: 1) Execution confirmation: The user clicks "Executed" and may add the actual amount of operation; 2) Effect evaluation: After a period of time after the suggestion is executed (such as 7 days after fertilization), the system prompts the user to rate the effect in 1-5 stars or select status labels such as "Leaves turn green", "Growth accelerates", "No obvious change", "Fertilizer damage occurs"; 3) Text notes: The user can freely describe the details observed. This feedback information is encrypted and uploaded to the cloud platform.

[0087] 405. Optimize the agricultural decision-making model based on feedback information from the actual effects of agricultural operations;

[0088] In this embodiment, feedback information on the actual effects of agricultural operations is collected. When a large number of negative feedbacks (such as low scores or being marked "invalid") are received for a similar rule (e.g., "when nitrogen content is below X, it is recommended to apply Y kg of urea"), an analysis is triggered: historical data is searched to analyze whether the rule is ineffective under certain specific soil textures, climate conditions, or crop subspecies. The analysis results form optimization suggestions, such as "in sandy soil, the original fertilizer application rate Y is ineffective, and it is recommended to correct the coefficient Z." For simple parameter calibration, the system can automatically fine-tune the quantitative parameters in the rule; for complex logical corrections (such as adding preconditions), a report is generated for review by agronomic experts. After the review is approved, the new or revised rule is deployed to the agricultural rule base to achieve iterative optimization of the agricultural decision-making model.

[0089] 406. Regularly measure the electrical characteristic parameters of the smart sensor nodes and assess the degree of node contamination based on the electrical characteristic parameters;

[0090] In this embodiment, to ensure long-term data reliability, the intelligent sensing node has a built-in self-diagnostic program. Every certain period (e.g., every 7 days), the node control unit applies a small test signal to the probe of the soil pH or electrical conductivity (EC) sensor and measures its response parameters, such as the AC impedance between electrodes, response voltage, or signal stability. In a clean state, these electrical characteristic parameters will stabilize within a specific baseline range. By comparing the real-time measurement value with the baseline value, the offset can be calculated. The larger the offset, the more serious the contaminants such as salt crystals, biofilms, or clay adhering to the probe surface. Based on this, the system quantifies the degree of contamination into "clean", "lightly contaminated", and "heavily contaminated" levels.

[0091] 407. When the degree of node contamination exceeds the preset tolerance range, a contamination alarm message is generated and sent to the user terminal.

[0092] In this embodiment, the system presets a pollution tolerance threshold (such as an electrical parameter deviation exceeding the baseline by 30%). When the assessed pollution level reaches "severe pollution", it is considered to exceed the tolerance. The node or cloud platform will immediately generate a pollution alarm message, which includes the ID of the polluted node, its location, the estimated pollution type (such as salinization, biofouling), and the recommended cleaning method. This alarm message is pushed to the corresponding user's mobile terminal with high priority via APP message and SMS, reminding the user to perform on-site maintenance in a timely manner and avoid continuous data deviation and misjudgment due to sensor inaccuracy.

[0093] 408. Obtain pollution cleaning record information sent by the user terminal;

[0094] In this embodiment, after the user performs on-site cleaning and maintenance on the sensor node based on the alarm information (such as wiping the probe with clean water and a soft cloth), the user records the maintenance operation through the relevant interface on the user terminal APP. The contamination cleaning record information includes at least: the ID of the node being cleaned, the cleaning time, and a brief description of the cleaning method used. This contamination cleaning record information is uploaded to the cloud platform as part of the equipment maintenance file.

[0095] 409. Upload the pollution cleaning record information to the cloud platform;

[0096] In this embodiment, after receiving the pollution cleaning record information, the cloud platform associates and stores it with the historical data and alarm records of the corresponding node. On the one hand, after receiving the record, the platform will temporarily suppress similar pollution alarms for that node until the next self-diagnosis cycle. On the other hand, the long-term accumulated cleaning records can be used to analyze the pollution rate of sensors in different regions and under different soil conditions, providing data support for optimizing the self-diagnosis cycle, improving sensor packaging technology, or providing preventive maintenance suggestions, thereby reducing the long-term maintenance cost of the system and improving reliability.

[0097] In this embodiment of the invention, the agricultural decision-making model generates specific agricultural suggestions and a visual report based on real-time data and dynamic thresholds, completing a closed loop from "state perception" to "decision output." More importantly, by introducing user feedback on the actual effects of agricultural operations and iteratively optimizing the agricultural decision-making model based on this feedback, the system acquires the ability to "learn from practice." It can continuously accumulate regional planting experience, constantly revise and enrich its rule base, thus becoming more accurate and adaptable with use. In addition, the sensor node self-diagnosis and pollution alarm mechanism transforms the passive maintenance of traditional IoT devices into proactive maintenance guidance based on state prediction. By monitoring electrical characteristic parameters, it can detect sensor performance degradation in advance and remind users to clean and maintain the devices, ensuring the accuracy of long-term data collection from the source and effectively reducing the risk of hidden system failures.

[0098] The above describes the soil monitoring method based on the Internet of Things (IoT) in the embodiments of the present invention. The following describes the soil monitoring device based on the IoT in the embodiments of the present invention. Please refer to [link / reference]. Figure 5 One embodiment of the Internet of Things-based soil monitoring device of the present invention includes:

[0099] The generation and deployment acquisition module 501 is used to generate node deployment density suggestions based on the soil difference map of the target soil monitoring area, deploy multiple smart sensor nodes in the target soil monitoring area based on the node deployment density suggestions, and use the smart sensor nodes to collect multi-parameter data of the soil.

[0100] The detection filtering module 502 is used to detect abnormal data in multi-parameter data based on multi-parameter spatiotemporal correlation, obtain detection results, and perform dynamic adaptive filtering on multi-parameter data according to the detection results to obtain effective monitoring data.

[0101] The selection transmission module 503 is constructed to build a decision matrix based on data priority, node remaining power and network signal strength. Based on the output of the decision matrix, the LoRa self-organizing network or NB-IoT network is selected to effectively transmit monitoring data to the cloud platform.

[0102] The update module 504 is used to acquire the current growth stage information of the target crop and the short-term weather forecast information of the target soil monitoring area in real time, and dynamically update the soil parameter warning threshold based on the growth stage information and the short-term weather forecast information.

[0103] The generation and push module 505 is used to generate agricultural operation suggestions in the cloud platform based on the preset agricultural decision-making model, effective monitoring data and soil parameter early warning thresholds, and push the agricultural operation suggestions to the user terminal.

[0104] In this embodiment, a node deployment density suggestion is generated based on the soil difference map of the target soil monitoring area. Multiple intelligent sensor nodes are deployed in the target soil monitoring area based on this suggestion, improving the rationality of node deployment. The intelligent sensor nodes collect multi-parameter soil data, and anomaly detection is performed on the multi-parameter data based on the spatiotemporal correlation of these data. Dynamic adaptive filtering is applied to the multi-parameter data based on the detection results, improving the reliability of data processing. A decision matrix is ​​constructed based on data priority, remaining node power, and network signal strength. Based on the output of the decision matrix, either a LoRa self-organizing network or an NB-IoT network is selected to transmit the effective monitoring data to the cloud platform, improving the flexibility of data transmission. Real-time information on the current growth stage of the target crop and short-term weather forecasts for the target soil monitoring area are obtained. Soil parameter warning thresholds are dynamically updated based on the growth stage information and short-term weather forecasts, improving the accuracy of warnings. In the cloud platform, a preset agricultural decision model is used to generate agricultural operation suggestions based on the effective monitoring data and soil parameter warning thresholds, providing operable intelligent decision support.

[0105] Please see Figure 6 Another embodiment of the soil monitoring device based on the Internet of Things in this invention includes:

[0106] The generation and deployment acquisition module 501 is used to generate node deployment density suggestions based on the soil difference map of the target soil monitoring area, deploy multiple smart sensor nodes in the target soil monitoring area based on the node deployment density suggestions, and use the smart sensor nodes to collect multi-parameter data of the soil.

[0107] The detection filtering module 502 is used to detect abnormal data in multi-parameter data based on multi-parameter spatiotemporal correlation, obtain detection results, and perform dynamic adaptive filtering on multi-parameter data according to the detection results to obtain effective monitoring data.

[0108] The selection transmission module 503 is constructed to build a decision matrix based on data priority, node remaining power and network signal strength. Based on the output of the decision matrix, the LoRa self-organizing network or NB-IoT network is selected to effectively transmit monitoring data to the cloud platform.

[0109] The update module 504 is used to acquire the current growth stage information of the target crop and the short-term weather forecast information of the target soil monitoring area in real time, and dynamically update the soil parameter warning threshold based on the growth stage information and the short-term weather forecast information.

[0110] The generation and push module 505 is used to generate agricultural operation suggestions in the cloud platform based on the preset agricultural decision-making model, effective monitoring data and soil parameter early warning thresholds, and push the agricultural operation suggestions to the user terminal.

[0111] In this embodiment, the generation, deployment, and acquisition module 501 includes: an analysis and generation unit 5011, used to analyze the spatial variability of soil parameters within the target soil monitoring area based on historical remote sensing data, and generate a soil difference map; a generation unit 5012, used to generate node deployment density suggestions based on the soil difference map; a deployment and acquisition unit 5013, used to deploy multiple intelligent sensing nodes in the target soil monitoring area based on the node deployment density suggestions, and use the intelligent sensing nodes to collect multi-parameter data of the soil, including moisture content, temperature, pH value, and nitrogen, phosphorus, and potassium content; and a preprocessing unit 5014, used to perform data preprocessing on the multi-parameter data.

[0112] In this embodiment, the detection filtering module 502 includes: a construction unit 5021, used to construct a multi-parameter empirical correlation model based on the short-term change correlation between different soil parameters at the same node; a verification unit 5022, used to use the multi-parameter empirical correlation model to retrieve the change trend of other related parameters in the multi-parameter data to verify the logical rationality of the data value when the data value of any soil parameter in the multi-parameter data exceeds a preset fixed threshold, and obtain a verification result; a marking unit 5023, used to mark the data value as high-confidence abnormal data when the verification result is logically unreasonable; and a filtering unit 5024, used to determine that the verification result is logically reasonable as an environmental disturbance and perform enhanced filtering processing on the data value.

[0113] In this embodiment, the construction selection and transmission module 503 includes: a determination unit 5031, used to determine the data type of multi-parameter data, including emergency abnormal data, important periodic data, and regular data; a matching and identification unit 5032, used to match the corresponding data priority according to the data type, and identify the remaining power and network signal strength of the node corresponding to the multi-parameter data; a construction selection unit 5033, used to construct a decision matrix according to the data priority, the remaining power of the node, and the network signal strength, and select a LoRa self-organizing network or an NB-IoT network according to the output result of the decision matrix to obtain the selection result; and a transmission unit 5034, used to transmit the effective monitoring data to the cloud platform using the network selected by the selection result.

[0114] In this embodiment, the acquisition and update module 504 includes: a determination and matching unit 5041, used to determine the current growth stage information of the target crop based on the sowing date and variety of the target crop through a preset growth stage model, and match the soil parameter early warning threshold corresponding to the growth stage information from the crop demand database; an acquisition and call unit 5042, used to call the meteorological data interface and use the meteorological data interface to obtain gridded meteorological forecast data of the target soil monitoring area in real time for a period of time in the future, the gridded meteorological forecast data including precipitation information, temperature information and sunshine information; a calculation unit 5043, used to calculate the water replenishment factor and evapotranspiration stress factor based on the precipitation information, temperature information and sunshine information; and an update unit 5044, used to update the soil parameter early warning threshold based on the water replenishment factor and evapotranspiration stress factor.

[0115] In this embodiment, the generation and push module 505 includes: a comparison and matching unit 5051, used to compare effective monitoring data and soil parameter early warning thresholds using a preset agricultural decision-making model on a cloud platform to obtain comparison results, and to match agricultural rules according to the comparison results to obtain agricultural operation suggestions; a merging and generation unit 5052, used to merge effective monitoring data, comparison results and agricultural operation suggestions to obtain merged information, and to generate a structured monitoring report according to the merged information; and a push and providing unit 5053, used to push the monitoring report to the user terminal so that the user terminal can generate and display a visual report page based on the monitoring report, and provide the user terminal with a feedback entry for agricultural operation suggestions.

[0116] In this embodiment, the system further includes: a first acquisition module 506, used to acquire feedback information on the actual effect of agricultural operations sent by the user terminal; an optimization module 507, used to optimize the agricultural decision-making model based on the feedback information on the actual effect of agricultural operations; a measurement and evaluation module 508, used to periodically measure the electrical characteristic parameters of the intelligent sensing nodes and evaluate the degree of node contamination based on the electrical characteristic parameters; a generation and sending module 509, used to generate a contamination alarm message and send the contamination alarm message to the user terminal when the degree of node contamination exceeds a preset tolerance range; a second acquisition module 510, used to acquire contamination cleaning record information sent by the user terminal; and an upload module 511, used to upload the contamination cleaning record information to the cloud platform.

[0117] above Figure 5 and Figure 6 The soil monitoring device based on the Internet of Things in the embodiments of the present invention will be described in detail from the perspective of modular functional entities. The soil monitoring device based on the Internet of Things in the embodiments of the present invention will be described in detail from the perspective of hardware processing.

[0118] Figure 7This is a schematic diagram of the structure of an IoT-based soil monitoring device 600 provided in an embodiment of the present invention. The IoT-based soil monitoring device 600 can vary significantly due to different configurations or performance characteristics. It may include one or more central processing units (CPUs) 610 (e.g., one or more processors) and a memory 620, and one or more storage media 630 (e.g., one or more mass storage devices) for storing application programs 633 or data 632. The memory 620 and storage media 630 can be temporary or persistent storage. The program stored in the storage media 630 may include one or more modules (not shown in the diagram), each module including a series of instruction operations on the IoT-based soil monitoring device 600. Furthermore, the processor 610 may be configured to communicate with the storage media 630 and execute the series of instruction operations in the storage media 630 on the IoT-based soil monitoring device 600 to implement the steps of the IoT-based soil monitoring method provided in the above-described method embodiments.

[0119] The IoT-based soil monitoring device 600 may also include one or more power supplies 640, one or more wired or wireless network interfaces 650, one or more input / output interfaces 660, and / or one or more operating systems 631, such as Windows Server, Mac OS X, Unix, Linux, FreeBSD, etc. Those skilled in the art will understand that... Figure 7 The illustrated structure of the IoT-based soil monitoring device does not constitute a limitation on IoT-based soil monitoring devices, which may include more or fewer components than illustrated, or combine certain components, or have different component arrangements.

[0120] The present invention also provides a computer-readable storage medium, which can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, wherein the computer-readable storage medium stores instructions that, when executed on a computer, cause the computer to perform the steps of an Internet of Things-based soil monitoring method.

[0121] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the system, device, or unit described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0122] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0123] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A soil monitoring method based on the Internet of Things, characterized in that, include: Based on the soil difference map of the target soil monitoring area, a node deployment density recommendation is generated. Based on the node deployment density recommendation, multiple smart sensor nodes are deployed in the target soil monitoring area, and multi-parameter data of the soil are collected using the smart sensor nodes. Anomaly detection is performed on the multi-parameter data based on multi-parameter spatiotemporal correlation to obtain detection results. Based on the detection results, dynamic adaptive filtering is performed on the multi-parameter data to obtain effective monitoring data. A decision matrix is ​​constructed based on data priority, node remaining power, and network signal strength. Based on the output of the decision matrix, a LoRa self-organizing network or an NB-IoT network is selected to transmit the effective monitoring data to the cloud platform. The system acquires real-time information on the current growth stage of the target crop and short-term weather forecasts for the target soil monitoring area, and dynamically updates soil parameter warning thresholds based on the growth stage information and short-term weather forecasts. In the cloud platform, a preset agricultural decision-making model is used to generate agricultural operation suggestions based on the effective monitoring data and the soil parameter early warning threshold, and the agricultural operation suggestions are pushed to the user terminal. The method involves detecting abnormal data based on multi-parameter spatiotemporal correlation of the multi-parameter data, obtaining detection results, and then performing dynamic adaptive filtering on the multi-parameter data based on the detection results to obtain effective monitoring data, including: A multi-parameter empirical correlation model is constructed based on the short-term variation correlations among different soil parameters at the same node. When the data value of any soil parameter in the multi-parameter data exceeds a preset fixed threshold, the multi-parameter empirical correlation model is used to retrieve the changing trends of other correlation parameters in the multi-parameter data to verify the logical rationality of the data value and obtain the verification result. If the verification result is logically unreasonable, the data value will be marked as high-confidence abnormal data. If the verification result is logically reasonable, it is determined to be an environmental disturbance, and the data value is subjected to enhanced filtering processing. The real-time acquisition of the target crop's current growth stage information and the short-term weather forecast information for the target soil monitoring area, and the dynamic updating of soil parameter early warning thresholds based on the growth stage information and the short-term weather forecast information, includes: Based on the sowing date and variety of the target crop, the current growth stage information of the target crop is determined through a preset growth stage model, and the soil parameter early warning threshold corresponding to the growth stage information is matched from the crop demand database. The meteorological data interface is invoked to obtain gridded meteorological forecast data for the target soil monitoring area in real time over a period of time. The gridded meteorological forecast data includes precipitation information, temperature information, and sunshine information. Based on the precipitation information, temperature information, and sunshine information, the water replenishment factor and evapotranspiration stress factor are calculated. The soil parameter warning thresholds are updated based on the water replenishment factor and the evapotranspiration stress factor.

2. The soil monitoring method based on the Internet of Things according to claim 1, characterized in that, The method involves generating a node deployment density recommendation based on the soil difference map of the target soil monitoring area, deploying multiple intelligent sensing nodes in the target soil monitoring area based on the node deployment density recommendation, and using the intelligent sensing nodes to collect multi-parameter soil data, including: Based on historical remote sensing data analysis, the spatial variability of soil parameters within the target soil monitoring area is analyzed to generate a soil difference map; Based on the aforementioned soil difference map, a node deployment density recommendation is generated; Based on the node deployment density, it is recommended to deploy multiple smart sensor nodes in the target soil monitoring area and use the smart sensor nodes to collect multi-parameter data of the soil, including moisture content, temperature, pH value and nitrogen, phosphorus and potassium content. The multi-parameter data is preprocessed.

3. The soil monitoring method based on the Internet of Things according to claim 1, characterized in that, The step of constructing a decision matrix based on data priority, remaining node power, and network signal strength, and selecting either a LoRa ad hoc network or an NB-IoT network to transmit the effective monitoring data to the cloud platform based on the output of the decision matrix, includes: Determine the data type of the multi-parameter data, which includes emergency anomaly data, important periodic data, and regular data; Match the corresponding data priority according to the data type, and identify the remaining power and network signal strength of the node corresponding to the multi-parameter data; A decision matrix is ​​constructed based on the data priority, the remaining power of the node, and the network signal strength. Based on the output of the decision matrix, a LoRa self-organizing network or an NB-IoT network is selected to obtain the selection result. The selected network, based on the selection results, transmits the effective monitoring data to the cloud platform.

4. The soil monitoring method based on the Internet of Things according to claim 1, characterized in that, The process of generating agricultural operation suggestions based on the effective monitoring data and the soil parameter early warning threshold using a preset agricultural decision-making model on the cloud platform, and pushing the agricultural operation suggestions to the user terminal, includes: In the cloud platform, the effective monitoring data and the soil parameter early warning threshold are compared using a preset agricultural decision-making model to obtain the comparison results. Based on the comparison results, agricultural rules are matched to obtain agricultural operation suggestions. The effective monitoring data, the comparison results, and the agricultural operation suggestions are merged to obtain merged information, and a structured monitoring report is generated based on the merged information. The monitoring report is pushed to the user terminal, so that the user terminal can generate and display a visualization report page based on the monitoring report, and provide the user terminal with a feedback entry for the agricultural operation suggestions.

5. The soil monitoring method based on the Internet of Things according to claim 1, characterized in that, After generating agricultural operation suggestions based on the effective monitoring data and the soil parameter early warning threshold using a preset agricultural decision-making model on the cloud platform, and pushing the agricultural operation suggestions to the user terminal, the process further includes: Obtain feedback information on the actual effects of agricultural operations sent by the user terminal; The agricultural decision-making model is optimized based on the feedback information of the actual effects of the agricultural operations. The electrical characteristic parameters of the smart sensing node are measured periodically, and the degree of node contamination is assessed based on the electrical characteristic parameters. When the pollution level of a node exceeds a preset tolerance range, a pollution alarm is generated and sent to the user terminal. Obtain the pollution cleaning record information sent by the user terminal; The pollution cleaning record information is uploaded to the cloud platform.

6. An apparatus based on the Internet of Things-based soil monitoring method as described in claim 1, characterized in that, include: A deployment and acquisition module is used to generate node deployment density suggestions based on the soil difference map of the target soil monitoring area, deploy multiple smart sensor nodes in the target soil monitoring area based on the node deployment density suggestions, and use the smart sensor nodes to collect multi-parameter data of the soil. The detection filtering module is used to detect abnormal data in the multi-parameter data based on the spatiotemporal correlation of multiple parameters, obtain detection results, and perform dynamic adaptive filtering on the multi-parameter data according to the detection results to obtain effective monitoring data. A transmission selection module is constructed to build a decision matrix based on data priority, node remaining power, and network signal strength, and select either a LoRa self-organizing network or an NB-IoT network to transmit the effective monitoring data to the cloud platform based on the output of the decision matrix. The update module is used to acquire the current growth stage information of the target crop and the short-term weather forecast information of the target soil monitoring area in real time, and dynamically update the soil parameter warning threshold based on the growth stage information and the short-term weather forecast information. A push module is used to generate agricultural operation suggestions in the cloud platform based on the effective monitoring data and the soil parameter early warning threshold using a preset agricultural decision model, and push the agricultural operation suggestions to the user terminal.

7. A soil monitoring device based on the Internet of Things, characterized in that, The IoT-based soil monitoring device includes: a memory and at least one processor, wherein the memory stores instructions; At least one of the processors invokes the instructions in the memory to cause the IoT-based soil monitoring device to perform the steps of the IoT-based soil monitoring method as described in any one of claims 1-5.

8. A computer-readable storage medium storing instructions thereon, characterized in that, When the instructions are executed by the processor, they implement the various steps of the IoT-based soil monitoring method as described in any one of claims 1-5.

Images