Smart agriculture management strategy optimization system

Abstract

This invention relates to the field of agricultural information management, specifically to a smart agricultural management strategy optimization system. The system includes: a device access module for establishing standardized protocols to access all types of agricultural sensing devices; a data management module for constructing a unified database of agricultural equipment; a device management module for full-process control of equipment and data; a GIS visualization module for building an agricultural management visualization platform; an agricultural decision-making module for constructing a multi-factor fusion quantitative agricultural decision-making model to determine irrigation triggering rules and execution strategies; a water and fertilizer control module for using PID and feedforward control algorithms for closed-loop regulation of water and fertilizer ratios; and a device operation and maintenance module for establishing an operation and maintenance management system for full lifecycle management of equipment.

Description

Technical Field

[0001] The present invention belongs to the field of agricultural information management, and specifically relates to a system for optimizing intelligent agricultural management strategies. Background Art

[0002] With the large-scale development of digital agriculture and intelligent agriculture, the demand for the whole-link control of the Internet of Things and data-driven farming decisions in intensive agricultural parks has been continuously increasing. Technologies such as Internet of Things perception, GIS geographic information, big data, and AI computer vision have been gradually applied to agricultural production links, but there are still many technical bottlenecks in the existing technical solutions. In the current agricultural scenario, the communication protocols of perception devices of multiple categories and brands have strong heterogeneity, and there are significant protocol barriers, making it difficult to achieve standardized access and unified management of devices; spatial geographic data and farming business data are mutually separated, and it is impossible to achieve the fusion visualization and spatialized linkage control of data; multi-source time-series monitoring data are mostly analyzed in isolation, and it is difficult to form a quantitative farming decision model with multi-factor linkage, and it cannot support precise farming management. At the same time, the existing water and fertilizer control technologies have problems such as insufficient real-time dynamic adjustment accuracy and unbalanced irrigation pipe network pressure scheduling, making it difficult to achieve closed-loop precise control of water and fertilizer ratio; the difficulty of extracting characteristics from agricultural image data is large, the warning of pests and diseases lags, the value of a large amount of agricultural time-series data is not fully exploited, and equipment operation and maintenance are mostly passive fault handling, lacking an active control system for the whole life cycle, which seriously restricts the transformation of agricultural production from experience-based planting to data-driven intelligent transformation. Summary of the Invention

[0003] The purpose of the present invention is to provide a system for optimizing intelligent agricultural management strategies to solve the problems mentioned in the above background art.

[0004] To solve the above technical problems, the present invention provides the following technical solutions: A system for optimizing intelligent agricultural management strategies, comprising: An equipment access module for establishing a standardized protocol to access all categories of agricultural perception devices; A data management module for normalizing the collected data and constructing a unified database for agricultural equipment; An equipment management module for performing the whole-process control of equipment and data; A GIS visualization module for constructing a three-dimensional geographical model of farmland based on the GIS system, integrating farming business data, building a visualization platform for agricultural management, and performing spatial display and full-dimensional viewing of data; A farming decision module for integrating multi-dimensional time-series monitoring data, constructing a multi-factor fusion quantitative farming decision model, determining irrigation trigger rules and execution strategies, and making precise farming decisions driven by data; The water and fertilizer control module is used to perform closed-loop regulation of water and fertilizer ratio using PID and feedforward control algorithms, and to perform dynamic balance scheduling of irrigation network pressure through group polling and constant pressure control algorithms. The equipment operation and maintenance module is used to complete equipment fault alarms based on the SCADA system, establish an operation and maintenance management system, and conduct full life cycle management of equipment.

[0005] Furthermore, it also includes: an intelligent early warning module, used to preprocess and extract features from agricultural image data, and to perform crop growth analysis and pest and disease identification through deep learning algorithms; and to build a prediction model based on meteorological data to provide early warning of pests and diseases. The data analysis module is used to establish a multi-dimensional data statistical analysis system and to conduct in-depth mining of agricultural data.

[0006] Furthermore, the establishment of standardized protocols for accessing all types of agricultural sensing devices includes: formulating standardized access specifications for all types of agricultural sensing devices based on the mainstream communication protocols of agricultural devices; determining the interaction and management rules for device access; formulating protocol conversion rules for devices of the same type but different brands and models; completing the unified conversion of non-standard protocols to standardized protocols; eliminating protocol barriers between devices; assigning a unique identification code containing multi-dimensional information to each access device; completing device permission verification, identity recognition, and access link binding management; formulating full-process verification rules for device access; completing formal device access after verification; and implementing a matching device online status heartbeat detection mechanism to periodically detect device online status and update the device online ledger in real time.

[0007] Furthermore, the normalization processing of the collected data and the construction of a unified database for agricultural equipment include: formulating full-dimensional classification rules for the collected data, dividing the data into multiple categories according to their sources and determining their corresponding dimensions, formulating invalid data cleaning rules, completing the processing and anomaly marking synchronization of abnormal, redundant, and missing data, formulating heterogeneous data normalization processing rules, completing the alignment of data time dimensions, unifying the units of measurement for numerical data, standardizing the coding of status data, and unifying the coordinates and formats of spatial geographic data, constructing a hierarchical unified database for agricultural equipment with the unique identification code of the equipment as the core, establishing multiple related sub-databases, and formulating database access control, archiving and backup, and full-link data association mapping rules.

[0008] Furthermore, the full-process management and control of equipment and data includes: establishing a full lifecycle electronic file for the accessed equipment based on the unique identification code of the equipment; standardizing the management rules and access control and change recording mechanism for basic equipment information; synchronizing the corresponding information to the unified agricultural equipment database; collecting and updating the core operating status of the equipment in real time; completing classification and collection and multi-dimensional filtering and searching; synchronizing the status data to the corresponding module of the platform; managing the full-link data and full-process traceability of the equipment with the unique identification code of the equipment; standardizing the permission verification and operation recording of data calls; formulating full-process management rules for remote control commands of the equipment; standardizing the closed-loop management of the entire process of fault handling; collecting equipment and data by monitoring station; and formulating differentiated management and display rules.

[0009] Furthermore, the construction of a 3D geographic model of farmland based on a GIS system, the integration of agricultural business data, and the establishment of an agricultural management visualization platform for spatial data display and full-dimensional viewing include: determining a unified spatial coordinate system compatible with the GIS system; completing the coordinate collection and vectorization processing of all spatial elements of farmland; assigning unique identifier codes to spatial elements and establishing association mappings with equipment, field, and station numbers; building a 3D geographic model of the entire farmland area based on multi-source spatial data; completing the refined modeling and mounting of surface entities and core equipment; configuring model interactive operations and multi-condition filtering and positioning rules; establishing a classification mounting and real-time synchronization mechanism between agricultural business data and the 3D model; and building a single-map visualization platform for agricultural management with the 3D model as the core.

[0010] Furthermore, the method employs PID and feedforward control algorithms for closed-loop regulation of water and fertilizer ratios, and uses group polling and constant pressure control algorithms for dynamic balancing and scheduling of irrigation network pressure. This includes: setting water and fertilizer control targets based on crop characteristics, growth stage, and soil properties; pre-adjusting initial operating parameters of pumps and valves through feedforward control; dynamically correcting water and fertilizer ratio deviations based on real-time monitoring data and PID algorithms; implementing a safety interlock mechanism; retaining data throughout the entire process; dividing independent irrigation zones based on GIS; formulating group polling execution rules for electric valves; dynamically adjusting pump operating parameters based on real-time network pressure data to maintain stable network pressure; linking with agricultural decision-making to optimize scheduling strategies; synchronizing operating status to corresponding modules; verifying irrigation effects; and continuously optimizing control parameters.

[0011] Furthermore, the aforementioned equipment fault alarm based on the SCADA system, the establishment of an operation and maintenance management system, and the full lifecycle management of equipment include: establishing a real-time data synchronization link with a unified database of agricultural equipment based on the SCADA system interface platform modules, formulating equipment and data anomaly judgment rules and graded alarm standards, generating full-dimensional alarm records and pushing them through multiple channels according to level, establishing a closed-loop handling mechanism for alarm work orders, building a differentiated operation and maintenance system for all types of equipment, formulating operation and maintenance work order management rules, constructing an equipment health assessment model, and conducting multi-dimensional statistical analysis of equipment operation and maintenance data.

[0012] Furthermore, the preprocessing and feature extraction of agricultural image data, and the completion of crop growth analysis and pest and disease identification through deep learning algorithms, include: full-link collection, metadata binding and classification archiving of multi-category agricultural image data; formulation of differentiated preprocessing and feature extraction rules for different types of agricultural images; completion of core feature extraction and quantitative calculation of crop growth, pests and diseases, and fungal spores through corresponding algorithms; completion of crop growth status analysis, trend prediction and abnormal cause judgment based on feature parameters; generation of water and fertilizer optimization suggestions and push to the agricultural decision-making module; completion of pest and disease and fungal spore classification and identification through deep learning algorithms; completion of pest and disease outbreak risk quantitative analysis and level determination based on corresponding data; generation of early warning information and plant protection suggestions; and simultaneous completion of data archiving and algorithm iteration optimization.

[0013] Furthermore, the method of constructing a prediction model based on meteorological data for early warning of pests and diseases includes: extracting time-series meteorological monitoring data corresponding to the occurrence of pests and diseases, completing data time dimension alignment and invalid data processing, connecting with meteorological forecast data and completing format normalization verification, determining core meteorological influencing factors and weight allocation rules for different pests and diseases, constructing a pest and disease outbreak probability prediction model and completing training, verification and iterative optimization, classifying warning levels and setting trigger rules based on model output, field monitoring data and control thresholds, generating full-dimensional warning information and pushing it according to level, generating plant protection operation work orders and completing closed-loop processing.

[0014] Furthermore, the establishment of a multi-dimensional data statistical analysis system for in-depth agricultural data mining includes: collecting core agricultural data across all categories based on a unified agricultural equipment database; using multi-dimensional unique codes as anchors to complete data association and time dimension alignment, ensuring complete and consistent data source association; building a hierarchical statistical architecture across three dimensions: time, space, and business; formulating detailed statistical rules for all scenarios; constructing a multi-scenario association analysis model based on statistical results; mining implicit data associations and outputting optimal management solutions; and standardizing the display, archiving, backup, and access control rules for statistical results. This application also discloses a smart agriculture management strategy optimization system, including: This application also discloses an electronic device, including: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, which, when executed by the at least one processor, enables the at least one processor to execute the smart agriculture management strategy optimization system of the present invention.

[0015] Beneficial Effects: This application effectively addresses several core bottlenecks in existing smart agriculture technologies, comprehensively improving the intelligent management and control level of agricultural parks across the entire chain. Addressing the challenge of managing multi-source heterogeneous equipment access, this solution eliminates protocol barriers between different devices, achieving standardized and stable access and unified management of all types of agricultural sensing equipment, significantly improving the compatibility and standardization of equipment management. Addressing the issue of fragmented spatial geographic data and agricultural business data, it opens up a fusion channel between the two types of data, constructing a comprehensive three-dimensional visualization management and control system, realizing spatial linkage display and multi-dimensional viewing of all elements of agricultural production, enhancing the intuitiveness and collaboration of park management. Addressing the pain point that isolated analysis of multi-source data cannot support quantitative decision-making, it breaks down data silos, constructs a multi-factor fusion agricultural decision-making model, realizes data-driven precision agricultural decision-making, and promotes the transformation of agricultural production from experience-based planting to digital management and control. Addressing the problems of insufficient precision in water and fertilizer control and imbalanced pipeline pressure scheduling, it achieves closed-loop precise adjustment of water and fertilizer ratios and dynamic balance scheduling of pipeline pressure, improving the precision of water and fertilizer management and resource utilization efficiency. To address the issues of delayed pest and disease early warning, insufficient data value mining, and passive equipment operation and maintenance, this project significantly improves the timeliness and accuracy of crop growth monitoring and pest and disease early warning, deeply explores the data value of the entire agricultural chain, constructs a proactive operation and maintenance system for the entire equipment lifecycle, and comprehensively enhances the park's production prevention and control capabilities and operational management efficiency. Attached Figure Description

[0016] Figure 1 This is a block diagram of the intelligent agricultural management strategy optimization system of the present invention; Figure 2 is a flowchart of the data normalization processing and the construction of a unified database for agricultural equipment in the data management module in an embodiment of the present invention; Figure 3 is a flowchart of the device management module for full-process control of devices and data in an embodiment of the present invention; Figure 4 is a flowchart of the construction of a 3D geographic model of farmland and the establishment of an agricultural management visualization platform in an embodiment of the present invention; Figure 5 is a flowchart of the construction of the multi-factor fusion quantitative agricultural decision-making model and the precision agricultural decision-making in the agricultural decision-making module of this invention. Figure 6 is a flowchart of the water and fertilizer ratio closed-loop regulation and irrigation network pressure dynamic balance scheduling of the water and fertilizer control module in an embodiment of the present invention. Figure 7 is a flowchart of the equipment operation and maintenance module based on the SCADA system for equipment fault alarm and full life cycle management in an embodiment of the present invention; Figure 8 is a flowchart of the intelligent early warning module for agricultural image data processing, crop growth analysis, and pest and disease identification in an embodiment of the present invention. Figure 9 is a flowchart of the intelligent early warning module's construction of a pest and disease prediction model based on meteorological data and early warning in an embodiment of the present invention; Figure 10 is a flowchart of the multi-dimensional statistical analysis and in-depth mining of agricultural data in the data analysis module of this embodiment of the invention. Detailed Implementation

[0017] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. 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 of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0018] This invention provides a smart agriculture management strategy optimization system, such as... Figure 1 As shown, it includes: The device access module is used to establish standardized protocols for accessing all types of agricultural sensing devices; The data management module is used to normalize the collected data and build a unified database for agricultural equipment. The equipment management module is used for full-process control of equipment and data; The GIS visualization module is used to build three-dimensional geographic models of farmland based on the GIS system and integrate agricultural business data. Build an agricultural management visualization platform to enable spatial display and multi-dimensional viewing of data; The agricultural decision-making module is used to integrate multi-dimensional time-series monitoring data, construct a multi-factor fusion quantitative agricultural decision-making model, determine irrigation triggering rules and execution strategies, and make data-driven precision agricultural decisions. The water and fertilizer control module is used to perform closed-loop regulation of water and fertilizer ratio using PID and feedforward control algorithms, and to perform dynamic balance scheduling of irrigation network pressure through group polling and constant pressure control algorithms. The equipment operation and maintenance module is used to complete equipment fault alarms based on the SCADA system, establish an operation and maintenance management system, and conduct full life cycle management of equipment.

[0019] The establishment of standardized protocols for accessing all types of agricultural sensing devices includes, in practice: determining mainstream device communication protocols suitable for agricultural scenarios, including Modbus-RTU, Modbus-TCP, MQTT, HTTP, ONVIF, and GB / T28181 protocol types; defining corresponding standardized access protocol specifications for each type of agricultural sensing device; and determining the message format, data transmission frequency, data verification rules, command interaction logic, and device status reporting mechanism for weather station equipment, soil moisture station equipment, insect detection equipment, integrated water and fertilizer machine equipment, electric valve control equipment, seedling monitoring equipment, spore analyzer equipment, and video surveillance equipment, respectively. For the same type of equipment of different brands and models, determining protocol adaptation and conversion rules to complete the unified conversion from non-standard protocols to standardized protocols, eliminating protocol barriers between devices.

[0020] Each connected device is assigned a unique device identification code, which includes device type, station number, installation area, and serial number. Based on this code, device access permission verification, device identification, and access link binding management are completed. During device access, a full-process verification rule is established: first, the connectivity of the device communication link is verified; then, the compliance of the device data reporting format is verified; and finally, the validity of the device command issuance and execution response is verified. Once all verifications pass, the device is officially connected. Simultaneously, an online status heartbeat detection mechanism is established for the connected devices, checking their online status at preset intervals and updating the online device access ledger in real time to ensure stable access and unified management of all types of agricultural sensing devices.

[0021] The collected data is normalized to construct a unified database for agricultural equipment, such as... Figure 2 As shown, the implementation specifically includes: determining the full-dimensional classification rules for the collected data, dividing it into six categories according to the data source: environmental monitoring time-series data, equipment operation status data, spatial geographic information data, image acquisition metadata, agricultural operation record data, and fault alarm maintenance data. Each category determines the corresponding data attribution dimension. Environmental monitoring time-series data corresponds to sensor data collected by meteorological stations, soil moisture stations, fertigation machines, insect monitoring stations, and seedling monitoring stations. Equipment operation status data corresponds to the operating parameters, start / stop status, and water and electricity consumption data of fertigation machines, electric valves, video equipment, and various monitoring station terminals. Spatial geographic information data corresponds to data on field distribution, monitoring station and equipment installation locations, and pipeline routing. Image acquisition metadata corresponds to the basic attributes, acquisition time, and equipment number data of images collected by seedling monitoring, insect monitoring, spore analyzers, and video monitoring equipment. Agricultural operation record data corresponds to the execution record data of irrigation, fertilization, plant protection, and equipment maintenance. Fault alarm maintenance data corresponds to the full-process record data of equipment anomalies, data exceeding limits, and fault warnings.

[0022] Determine the rules for cleaning invalid data, and for each type of collected data, determine the corresponding outlier judgment threshold and rejection rules. Based on the sensor range, reasonable thresholds for crop growth environment, and normal operating parameter range of equipment, remove abnormal data that exceeds the reasonable range, redundant data that is repeatedly reported, and null data caused by transmission interruption. For non-critical data missing in a single collection, linear completion is performed based on the valid data of adjacent collection cycles of the same equipment. For critical data missing for multiple consecutive cycles, mark it as data abnormality and synchronize it to the equipment fault alarm module.

[0023] To normalize heterogeneous data, a unified timestamp format was established for all collected data. Standard Beijing time was used to determine the unique time identifier for data collection, aligning the time dimension of data reported from different devices. For numerical monitoring data, a unified unit conversion rule was established. For temperature, humidity, soil moisture content, conductivity, rainfall, wind speed and direction, flow rate, pressure, and liquid level data, industry-standard units of measurement were defined, ensuring a unified conversion from non-standard units reported by different devices to standard units of measurement.

[0024] For equipment status enumeration data, a unified status coding rule is established. A unique standardized code is assigned to each equipment status (online / offline), stop operation, normal fault, and valve fully open / closed), completing the unified conversion from differentiated status descriptions reported by different equipment to standard status codes. For spatial geographic data, a unified coordinate system and format specification are established, completing coordinate conversion and format unification for satellite telemetry data and equipment positioning data from different sources, ensuring compatibility with GIS geographic information systems.

[0025] The architecture rules of the unified database for agricultural equipment are determined. Based on the unique identification code of the equipment, data classification rules, and time dimension, a hierarchical data storage architecture is established. First, a basic information database for the equipment is established to store the unique identification code, equipment type, station number, installation location, equipment model, range parameters, calibration cycle, and access protocol information of each connected device, so that one device corresponds to one complete basic information file.

[0026] Establish a time-series monitoring database to store normalized environmental monitoring and equipment operation numerical data reported by all sensors, based on equipment identification codes and acquisition times. A hierarchical index is created by year, month, day, and hour to support rapid retrieval and statistical analysis. Establish a spatial geographic information database to store spatial coordinate data and attribute information of fields, pipelines, monitoring stations, and equipment, and link it to the equipment basic information database through equipment identification codes. Establish an image and business database to store image acquisition metadata, agricultural operation records, fault alarms, and maintenance records, and link them to corresponding equipment through equipment identification codes.

[0027] Establish database read / write permission control rules, defining corresponding data query, modification, export, and archiving permissions based on user roles and management scope to ensure data access security. Define data archiving and backup rules, establishing daily aggregation and monthly archiving rules for frequently collected real-time monitoring data, and creating a tiered backup mechanism for historical archived data to ensure data storage integrity and traceability. Define data association and mapping rules, using the device's unique identification code as the core anchor point, to complete the full-link association of device basic information, time-series monitoring data, spatial location data, business operation data, and alarm and maintenance data. This enables the retrieval of all dimensions of corresponding data for a single device through its identification code, supporting the data access and analysis needs of various functional modules of the platform.

[0028] The aforementioned full-process management of equipment and data, such as Figure 3 As shown, the specific implementation includes: establishing a full lifecycle electronic file for each connected device based on its unique identification code; determining the rules for editing, updating, and verifying basic device information, including device model, range parameters, installation location, affiliated station, calibration cycle, access protocol, and maintenance records, ensuring that the device file information is completely consistent with the actual on-site deployment; supporting the editing of individual device information and the import and export of batch device information; determining the rules for information modification permission verification and change tracking; and generating unalterable change records for all information modification operations, which are synchronously stored in the unified agricultural equipment database.

[0029] Based on the equipment online status heartbeat detection mechanism, the system collects and updates the core status data of the equipment in real time, including online / offline status, operation stoppage, and normal / abnormal status. It determines the hierarchical display rules for equipment status and classifies and categorizes equipment according to equipment type, affiliated monitoring station, and operating status. The system supports precise filtering and targeted searching by multiple dimensions such as equipment type, region, online status, and installation time, enabling real-time tracking of the status of individual equipment and global control of the status of all equipment in the park. Equipment status data is synchronized to the GIS visualization module and data statistical analysis module in real time to ensure that the equipment status information obtained by each module of the platform is completely consistent.

[0030] Based on the classification rules and data association mapping rules of the unified database for agricultural equipment, and using the unique identification code of the equipment as the core anchor point, the system achieves full-link control of a single device from data collection, normalization processing, storage and archiving, statistical analysis to retrieval and display. It establishes full-process traceability rules for data, allowing for reverse tracing of data records to the corresponding collection device, collection time, collection location, and equipment operating status, thus defining the rules for data retrieval and use. When each functional module of the platform calls for data collection from the device, it must complete verification according to the preset permission rules. The entire data retrieval operation is recorded to ensure the compliance and traceability of data use.

[0031] Establish comprehensive control rules for the generation, verification, issuance, execution, and feedback of remote control commands for equipment. All remote control commands for integrated water and fertilizer machines, electric valves, and video monitoring equipment must be generated based on the quantitative execution strategy generated by the agricultural decision-making module. Before issuing commands, complete permission verification, equipment online status verification, and execution condition security verification. After verification, the commands are issued to the corresponding equipment according to standardized protocol specifications. Track the command execution status and results of the equipment in real time. Generate a complete execution record for successfully executed commands and immediately trigger an exception prompt for failed commands, which is simultaneously pushed to the equipment fault alarm module to ensure that the entire process of control commands is manageable, controllable, and has a closed loop execution.

[0032] Based on equipment type and operating parameters, determine the regular inspection and maintenance cycle and maintenance standards for each piece of equipment, generate equipment operation and maintenance plans and due reminders in advance, and push them to the corresponding equipment management personnel. Record all dimensions of information on equipment operation and maintenance, including execution time, personnel, operation and maintenance content, and inspection results, and update them synchronously to the equipment's full life cycle electronic file. For equipment anomaly and fault information pushed by the equipment fault alarm module, determine the full-process control rules for fault record generation, dispatch, handling, verification, and closure, track the fault handling progress in real time, complete on-site verification and information archiving after fault handling is completed, and update the equipment operating status to ensure that the entire equipment fault process is traceable and closed-loop.

[0033] Based on the type of monitoring station and its region, all equipment and corresponding collected data under the same monitoring station are uniformly collected and managed. The rules for summarizing and displaying equipment and data at the monitoring station level are determined. The location information, basic attributes, on-site images, operating status of all equipment, real-time collected data, and statistical analysis results of the monitoring station are uniformly integrated to achieve one-stop centralized management and control of all equipment and data of a single monitoring station. Differentiated management dimensions and display rules are determined for different types of monitoring stations, such as meteorological stations, soil moisture stations, video monitoring stations, irrigation control equipment monitoring stations, and water and fertilizer equipment monitoring stations, to ensure that the management content of different types of monitoring stations is fully matched with the functional attributes of the equipment.

[0034] The aforementioned method involves constructing a 3D geographic model of farmland based on a GIS system, integrating agricultural business data, and building an agricultural management visualization platform to enable spatial display and multi-dimensional viewing of data, such as... Figure 4 As shown, the specific implementation includes: determining a unified spatial coordinate system adapted to the GIS system, adopting the National Geodetic Coordinate System 2000, determining the spatial boundary mapping rules for farmland areas, completing the on-site coordinate collection and vectorization processing of field zoning, pipeline routing, irrigation facilities, various monitoring stations, and equipment installation points, determining the unique spatial identifier code for each spatial element, and establishing a one-to-one correspondence mapping relationship between the spatial identifier code and the unique identification code of the corresponding equipment, field number, and monitoring station number to ensure the anchoring and matching of spatial elements and business data.

[0035] The construction rules for the 3D geographic model are determined. Based on the collected vectorized spatial data, satellite telemetry image data, and digital elevation model data, a 3D geographic base model of the entire farmland area is built. The accuracy standards for 3D modeling of fields, roads, ditches, pipe networks, buildings, and shelterbelts are determined. The 3D fine modeling of surface entity elements is completed. For core equipment such as weather stations, soil moisture stations, insect monitoring stations, integrated water and fertilizer machines, video monitoring equipment, and electric valves, the 3D model adaptation rules for the corresponding equipment are determined. The accurate matching and mounting of equipment models with actual installation points is completed, and seamless access and model updates of 3D modeling results are supported.

[0036] The interactive operation rules of the 3D geographic model are determined, supporting translation, scaling, and rotation operations of the entire scene. The path planning and perspective switching rules of 3D roaming are determined, enabling automatic roaming and fixed-point viewing of the entire farmland scene. The spatial filtering rules of stations and equipment are determined, supporting multi-condition combination filtering based on station type, equipment type, region, and operating status. The filtering results are highlighted and displayed in the 3D geographic model, enabling step-by-step focused viewing from the macro-level scene of the park to the micro-level viewpoint of a single device.

[0037] The rules for integrating agricultural business data with the 3D geographic model are determined. Based on the association mapping relationship between spatial identification codes and equipment unique identification codes, full-dimensional agricultural business data of corresponding spatial elements are retrieved from the unified agricultural equipment database. The classification and mounting rules for data integration are determined, and environmental monitoring time series data, equipment operation status data, agricultural operation record data, fault alarm maintenance data, and image acquisition metadata are respectively mounted to the corresponding fields, stations, and equipment spatial points in the 3D geographic model. The real-time data synchronization rules are determined, and the real-time data collected by the platform and the data updated in the database are synchronously updated to the corresponding spatial elements in the 3D geographic model to ensure the consistency between the spatial display data and the actual collected data.

[0038] The rules for building an agricultural management visualization platform were established, using a 3D geographic model as the core carrier. The platform's functional zones and data display levels were defined, including a global overview zone, a real-time monitoring zone, an equipment control zone, an alarm and early warning zone, and a statistical analysis zone. The data display rules for each zone were also determined. The global overview zone displays the overall farmland layout, statistical distribution of monitoring stations and equipment, total irrigated area, cumulative water and electricity consumption, and core summary data on crop varieties. The real-time monitoring zone displays real-time environmental data collected by sensors at each monitoring station and real-time equipment operating status data. The equipment control zone displays core ledger data for the entire lifecycle management of equipment. The alarm and early warning zone displays real-time triggered equipment failures, data anomalies, and agricultural risk warnings. The statistical analysis zone displays the core results of multi-dimensional data statistics, achieving a one-stop centralized visualization display of all elements and processes of farmland data.

[0039] The rules for viewing spatial data across all dimensions are defined. For any field, station, or equipment location in the 3D geographic model, a detailed viewing panel can be triggered by clicking. The display rules for the detailed panels are defined: the field detailed panel displays field boundaries, crop information, soil properties, historical irrigation and fertilization records, cumulative water consumption, and yield statistics; the station detailed panel displays basic station information, number, installation location, on-site photos, video monitoring footage, real-time data collected by all associated equipment, and data statistical analysis results; and the equipment detailed panel displays basic equipment files, real-time operating parameters, historical data collected, maintenance records, and fault alarm records. Differentiated detailed display dimensions are determined for different types of stations and equipment to ensure that the displayed content fully matches the functional attributes of the stations and equipment, achieving seamless switching and full-dimensional viewing from macro-spatial distribution to micro-level single-equipment details.

[0040] The visualization platform establishes rules for linking and displaying abnormal data. Based on data analysis results from the SCADA system, when abnormal equipment operation, data exceeding limits, or agricultural risk events are identified, the corresponding points are highlighted with warning indicators in the 3D geographic model. Simultaneously, core information and handling suggestions for the abnormal event are pushed to the system. Clicking the warning indicator allows direct access to the details panel of the corresponding equipment or monitoring station to view complete information and handling procedures for the abnormal event, achieving spatial location, visualization, and full-process tracking of abnormal events. The visualization platform also establishes rules for data interaction and control, supporting the generation and issuance of remote control commands for equipment directly triggered from spatial points in the 3D geographic model. Before issuing commands, permission verification, equipment online status verification, and execution condition security verification are performed. After successful verification, commands are issued to the corresponding equipment according to standardized protocols. The system tracks the command execution status and results in real time, synchronously updating the status display of the corresponding points in the 3D geographic model, achieving a closed-loop linkage between spatial visualization and remote equipment control.

[0041] The system integrates multi-dimensional time-series monitoring data to construct a multi-factor fusion quantitative agricultural decision-making model, determines irrigation triggering rules and execution strategies, and enables data-driven precision agricultural decision-making. Figure 5 As shown, the specific implementation includes: determining the extraction range and time alignment rules for time-series monitoring data; extracting full-dimensional time-series monitoring data of soil temperature and humidity, soil electrical conductivity, air temperature and humidity, wind speed and direction, rainfall, solar photosynthetically active radiation, and atmospheric pressure from the unified agricultural equipment database for different soil layers of the corresponding field; determining the time dimension alignment rules for the data; using a fixed time step as a benchmark to complete the time synchronization matching of data from different sensors and different acquisition frequencies; eliminating invalid data fragments with time misalignment; and ensuring that all data participating in the model calculation are completely corresponding in the time dimension.

[0042] The core calculation rules for crop water requirement are determined. Based on synchronously matched meteorological time-series data, the evapotranspiration of the reference crop (ET0) is calculated using the Penman-Monteith formula. The stage-specific crop coefficients (Kc) for the entire growth period of the corresponding crop are determined. Based on the growth characteristics of the crop at different growth stages, the dynamic adjustment rules for the crop coefficients are determined. The actual water requirement of the target crop (CropET) is then accurately calculated. The actual water requirement of the crop is the product of the evapotranspiration of the reference crop and the crop coefficient for the corresponding growth stage.

[0043] The construction rules for the soil water balance model are determined. Based on the basic attribute parameters of the target field, such as field capacity, crop wilting point, soil bulk density, and root distribution depth, the soil water balance model is constructed. The input items and calculation rules of the model are determined. The input items include the measured volumetric water content of the soil at the current time, irrigation water volume, natural rainfall, actual water requirement of the crop, and deep infiltration. The predicted value of soil water content at the next time moment is obtained through model calculation, and the reasonable range of soil water content variation is determined, providing a core basis for irrigation decision-making.

[0044] The quantitative threshold rules for triggering irrigation are determined based on the characteristics of the target crop variety, its current growth stage, and soil type. The lower limit threshold of soil moisture content for starting irrigation is set above the crop wilting point and within the range of 60% to 70% of field capacity. The upper limit threshold of soil moisture content for stopping irrigation is set at 90% of field capacity to avoid excessive irrigation that could cause deep seepage and fertilizer loss. When the soil moisture content of the main distribution layer of the crop roots, which is monitored in real time, is lower than the set lower limit threshold, the irrigation start command is automatically triggered.

[0045] To determine the precise calculation rules for irrigation water volume, based on the calculation results of the soil water balance model, the actual water requirement of crops, the root distribution layer depth, and the difference between the measured and target values ​​of soil volumetric water content, the net irrigation water volume for a single irrigation is determined. At the same time, based on the water conveyance efficiency of the irrigation network and the water use coefficient of the field irrigation method, the correction calculation rules for gross irrigation water volume are determined, and the total water volume for a single irrigation is accurately quantified to avoid water waste and insufficient irrigation.

[0046] Establish dynamic correction rules for irrigation plans based on meteorological forecasts. Connect with weather forecast data for the next 24 to 72 hours released by official meteorological departments to determine correction rules for rainfall predictions. When more than 5 mm of effective rainfall is predicted in the next 24 hours, the irrigation water volume will be automatically reduced. When more than 15 mm of heavy rain or above is predicted in the next 24 hours, the irrigation plan will be automatically canceled. At the same time, based on the predicted extreme high temperature, strong wind, and severe convective weather data, determine the rules for adjusting the irrigation plan in advance or postponement to avoid the impact of severe weather on irrigation effectiveness.

[0047] An optimization strategy for irrigation execution time was determined based on the sunrise and sunset times of the target area, hourly evaporation calculated from hourly meteorological data, and the peak-valley time division rules of the power grid. The optimal time window for irrigation execution was determined, prioritizing the low evaporation and low wind speed periods from night to early morning to reduce water loss during irrigation. At the same time, peak power consumption periods were avoided to reduce irrigation electricity costs. Execution time staggering rules for different irrigation zones were also determined to avoid insufficient pipeline pressure caused by simultaneous irrigation in multiple areas.

[0048] Differentiated execution rules for irrigation decisions are determined by zoning. Based on the differences in independent field zones, crop varieties, soil types, and growth stages as defined by the GIS geographic information system, a unique multi-factor fusion decision model is constructed for each independent irrigation zone. Differentiated setting rules for irrigation thresholds, irrigation water volume, and execution time are determined for different zones, so as to achieve personalized and precise irrigation decisions for different fields within the same park.

[0049] Establish a closed-loop linkage rule for irrigation decision-making and execution. The generated irrigation decision instructions, including irrigation start time, irrigation duration, target irrigation water volume, and electric valve control rules for the corresponding irrigation zone, are synchronously sent to the water and fertilizer control module and the equipment management module. Establish full-process tracking rules for instruction execution. Collect real-time data on pipeline flow, pressure, and soil moisture content changes during irrigation execution. Verify the irrigation execution effect based on real-time monitoring data. When the actual irrigation water volume reaches the preset target value, automatically trigger the irrigation stop instruction to complete the closed loop of irrigation decision execution.

[0050] Determine the self-optimization iteration rules of the irrigation decision model, record the execution data of each irrigation decision, the changes in soil moisture content before and after execution, crop growth status data, and corresponding yield data, determine the iterative optimization rules of the model parameters, and regularly correct the crop coefficient, irrigation threshold, and corresponding parameters of the water balance model based on historical execution data and crop growth feedback, so as to continuously improve the accuracy and adaptability of the agricultural decision model.

[0051] Establish data traceability and accountability rules for the entire irrigation decision-making process. Store the model input data, calculation process parameters, decision results, execution records, and effect verification data for each irrigation decision in a unified agricultural equipment database. This will enable the entire irrigation decision-making process to be traceable and reviewable, providing comprehensive data support for agricultural management.

[0052] The method employs PID and feedforward control algorithms for closed-loop regulation of water and fertilizer ratio, such as... Figure 6As shown, a dynamic balance scheduling of irrigation network pressure is achieved through a grouped polling and constant pressure control algorithm. Specific implementation includes: determining the core control objective of closed-loop water-fertilizer ratio regulation; setting thresholds for the target EC value and target pH value at the irrigation slurry outlet based on the characteristics of the planted crop, its current growth stage, and basic soil properties; determining the input parameters and pre-calculation rules for feedforward control; and determining the target total irrigation flow, target EC value, target pH value, concentration of the fertilizer concentrate in the mother liquor tank, and the basic EC and pH values ​​of the irrigation raw water. Based on the input parameters, the initial operating parameters of the fertilizer injection pump and acid-base regulating pump are pre-calculated, and the pre-adjustment rules for the initial frequency of the frequency converter and the initial opening of the fertilizer injection valve are determined. This pre-adjustment of the basic water-fertilizer ratio is completed in advance, reducing the deviation of subsequent feedback adjustments and lowering system fluctuations.

[0053] The real-time data acquisition rules for PID feedback regulation are determined. Using the EC sensor, pH sensor, and flow sensor mounted on the integrated water and fertilizer machine, real-time EC values, pH values, and instantaneous flow data at the irrigation pipeline outlet are acquired at fixed acquisition cycles. After data normalization, the data is synchronously stored in a unified agricultural equipment database. The deviation calculation rules between the set threshold and the real-time monitoring values ​​are determined. The deviations between the set thresholds for EC and pH values ​​and the real-time monitoring values ​​are calculated separately. The tuning rules for the proportional, integral, and derivative terms of the PID algorithm are determined. Based on the water and fertilizer requirements of different crop growth stages and pipeline flow changes, the appropriate values ​​of the three parameters are dynamically adjusted. Based on the magnitude and trend of the deviation, the frequency converter frequency of the fertilizer injection pump, the opening degree of the fertilizer injection valve, and the operating parameters of the acid-base regulating pump are dynamically adjusted to correct the deviations in EC and pH values ​​in real time, ensuring that the fertilizer concentration and pH of the irrigation solution remain stable within the set threshold range.

[0054] Establish multi-dimensional safety interlock rules for the closed-loop regulation process, collect real-time monitoring data from the liquid level and pressure sensors of the integrated water and fertilizer machine, and immediately suspend the operation of the fertilizer injection pump and irrigation water pump when the liquid level in the mother liquor tank or clear water tank is lower than the set lower limit, or the pressure in the irrigation pipeline exceeds the set safety range. Simultaneously close the electric valves on the corresponding pipelines, generate a fault alarm record and push it to the equipment fault alarm module. The closed-loop regulation process can only be resumed after the abnormal situation has been handled and verified. Establish full-process data traceability rules for the water and fertilizer ratio regulation process, and store the setting parameters, real-time monitoring data, pump and valve adjustment commands, and execution results of each adjustment in a unified agricultural equipment database to achieve traceability and replayability of the entire ratio regulation process.

[0055] Based on the field boundaries, crop types, pipeline routes, and electric valve locations defined by the GIS geographic information system, the entire farmland area is divided into several independent irrigation zones. A unique zone code is assigned to each irrigation zone, and the associated mapping relationship between the electric valves, pipeline branches, and field irrigation equipment corresponding to each zone is determined. The zone code is then bound to the unique identification code of the corresponding equipment and stored synchronously in the unified agricultural equipment database.

[0056] The execution rules for grouping and polling of electric valves are determined. Based on the designed water supply capacity of the irrigation network and the rated output power of the water pumps, the maximum number of irrigation zones that can be opened at the same time is determined. The order rules for polling execution are formulated. Based on the priority of crop water demand and the ranking results of measured soil moisture content, the opening order and single irrigation duration of each irrigation zone are determined. The electric valves of the corresponding irrigation zones are opened in the preset order. After the irrigation task of the current zone is completed, the electric valve of that zone is automatically closed, and then the electric valve of the next zone in the priority zone is opened. This avoids the problems of insufficient water supply capacity of the network and reduced irrigation uniformity caused by multiple zones being opened at the same time.

[0057] The monitoring and regulation rules for constant pipeline pressure control are established. Real-time pipeline pressure data is acquired at fixed collection cycles using pressure sensors installed on the main irrigation pipeline. Based on the pipeline's design pressure resistance and the rated operating pressure range of the field irrigation equipment, the upper and lower limit thresholds and safe operating range of the pipeline pressure are determined. When the real-time monitored main pipeline pressure is lower than the set lower limit threshold, the operation of newly added irrigation zones is immediately suspended. Simultaneously, the frequency of the irrigation pump's inverter is dynamically adjusted based on the pressure deviation value to maintain the pipeline pressure stable within the set range. After the pipeline pressure returns to the normal range, the irrigation tasks of subsequent zones are resumed according to the polling rules. When the real-time monitored main pipeline pressure is higher than the set upper limit threshold, the pump operating frequency is immediately reduced, and an overpressure alarm is triggered simultaneously. If necessary, the pump operation is suspended to avoid the risk of pipe bursting.

[0058] In a further embodiment, using fixed threshold triggering and conventional PID frequency conversion regulation is prone to low-pressure regulation lag and high-pressure overshoot; furthermore, relying solely on single-point pressure data from the main pipe without integrating distributed monitoring data from branch lines fails to address the pressure imbalance problem at the end of the pipeline network; conventional PID has poor adaptability to the large lag, strong coupling, and time-varying nonlinear systems of irrigation pipeline networks. Therefore, this application proposes a dynamic pressure balancing scheduling algorithm for agricultural irrigation scenarios characterized by time-sharing irrigation, large lag, strong coupling, and time-varying nonlinearity. The specific scheme is as follows: First, set the initial impedance for the start-up and shutdown of the zones. When the i-th zone is started for the first time, that is, when Ω(t) is added at time t, set the initial value of the branch hydraulic impedance according to the following formula: ; In the formula: The initial estimate of the branch hydraulic impedance at time t in the i-th partition; The rated design gauge pressure of the pipeline network; This is the lower limit of the rated operating pressure of the irrigation equipment in the i-th zone; This is the rated design flow rate when all irrigation emitters in the i-th zone are fully open.

[0059] Within 1-3 control cycles after the valve is fully open, the system synchronously collects real-time flow and inlet / outlet pressure data of the branch in that zone at a fixed frequency. It then uses a recursive least squares method to identify the hydraulic impedance of the branch online, eliminating jump data during valve opening and closing to ensure the identified value matches the actual water flow characteristics of the pipeline. When the flow fluctuation remains below 5% of the rated flow for two consecutive control cycles, the system switches to a steady-state adaptive filtering mode. A sliding window weighted average algorithm is used to smooth the impedance value, setting upper and lower limits of ±30% for the initial estimate. A stable impedance estimate is output and bound to the corresponding zone code for storage.

[0060] Based on the impedance and rated flow of the pre-changed zones, the target pressure for the next cycle is pre-calculated to achieve seamless switching during zone start-up and shutdown. The calculation formula is as follows: ; In the formula: for Pre-adjustment pressure values ​​for zoned start-up and shutdown at specific times; This is the set of partition indexes that will be launched in the next phase; This is the set of partition indexes that will be closed in the next phase; For the first Each partition Smoothed estimate of branch hydraulic impedance at time t; For the first The rated design flow rate of each zone; for Time of the first Real-time sampling of instantaneous flow rate in each irrigation zone branch.

[0061] The system receives valve action commands one control cycle in advance, substitutes them into the above formula to calculate the pre-adjustment pressure value, and simultaneously corrects the target pressure of the main pipe. In the same cycle as the valve action, the command is sent to the pump frequency converter to offset pressure surges caused by zone start-up and shutdown. Each control cycle integrates distributed pressure data from the main pipe and each branch, using the constraint that the zone end pressure meets the rated operating lower limit, to dynamically fine-tune the target pressure of the main pipe. Simultaneously, an adaptive PID algorithm is used to dynamically tune control parameters, combined with integral amplitude limiting constraints, to ensure stable network pressure. After each irrigation cycle, the system collects and archives the entire process operation data, periodically iterates and optimizes algorithm parameters, forming a complete closed loop of scheduling, execution, feedback, and optimization.

[0062] Establish linkage rules between pipeline pressure scheduling and irrigation agricultural decision-making. Based on the irrigation plan generated by the agricultural decision-making module, dynamically adjust the order and duration of polling zones to ensure consistency between pressure scheduling and irrigation decision-making objectives. Establish full-process status synchronization rules for polling irrigation and pressure control. Synchronize the opening and closing status of electric valves in each zone, real-time pipeline pressure data, and pump operating parameters to the GIS visualization module and equipment management module in real time. Display the execution status and pipeline pressure distribution of each irrigation zone in a 3D geographic model. Establish effect verification rules after irrigation execution. Record the actual irrigation water volume of each zone, pipeline pressure fluctuation data during irrigation, and soil moisture content changes before and after irrigation. Synchronously store these records in a unified agricultural equipment database. Continuously optimize polling grouping rules and pressure control parameters based on historical execution data to improve the stability of pipeline scheduling and irrigation uniformity.

[0063] The aforementioned system, based on SCADA, completes equipment fault alarms, establishes an operation and maintenance management system, and conducts full lifecycle control of equipment, such as... Figure 7 As shown, the specific implementation includes: determining the data docking rules between the SCADA system and each module of the platform; establishing a real-time data synchronization link with the equipment access module and data management module based on the hierarchical storage architecture of the unified agricultural equipment database; determining the scope and frequency of data collection, including the operating status parameters of all types of accessed equipment, real-time sensor data, command execution feedback data, and equipment heartbeat message data; determining the corresponding data collection cycle according to equipment type; ensuring that the equipment data obtained by the SCADA system is completely synchronized with the actual operating status on site; and using the unique identification code of the equipment as the core anchor point to complete the precise binding of each piece of collected data with the corresponding equipment, thereby achieving full-link traceability of the operating data of a single piece of equipment.

[0064] The identification and judgment rules for equipment anomalies and data anomalies were determined. Based on the rated operating parameter range corresponding to the equipment model, the sensor range, the reasonable data threshold of the crop growth environment, and the heartbeat cycle of normal equipment communication, corresponding anomaly judgment standards were determined for each piece of equipment and each type of collected data. Five major categories of anomaly scenarios were divided: data exceeding limits, communication interruption, command execution failure, abnormal equipment operating parameters, and equipment stall overload. Corresponding judgment logic and grading standards were determined for each category of anomaly scenario. According to the scope of impact and urgency of the anomaly, it was divided into three levels: Level 1 emergency alarm, Level 2 important alarm, and Level 3 warning alarm. Level 1 alarms correspond to anomaly scenarios that affect the production safety of the park and irreversible damage to equipment. Level 2 alarms correspond to anomaly scenarios that affect the normal function of equipment and the progress of agricultural operations. Level 3 alarms correspond to minor anomaly scenarios that do not affect core functions but require attention and handling.

[0065] The system establishes comprehensive rules for handling abnormal alarms throughout the SCADA system. When the system identifies an abnormal event that meets the judgment criteria through real-time data in-depth mining, it immediately generates a corresponding level of fault alarm record. The alarm record includes comprehensive information such as the equipment's unique identification code, equipment name, affiliated monitoring station and installation location, anomaly type, anomaly trigger time, real-time anomaly data, judgment basis, and suggested handling plan. The alarm record is simultaneously stored in the unified agricultural equipment database. Based on the alarm level, corresponding push rules are determined. Level 1 alarms are simultaneously pushed to equipment managers and park managers via platform pop-ups, SMS, and voice calls. Level 2 alarms are pushed to the corresponding equipment managers via platform pop-ups and SMS. Level 3 alarms are uniformly displayed in the platform alarm center and pushed to the platform's internal messages. At the same time, based on the association mapping relationship between spatial identification codes and equipment unique identification codes, the location of abnormal equipment is highlighted and marked in the 3D geographic model of the GIS visualization module, realizing the spatial positioning and visualization of abnormal events. Clicking on the warning mark allows direct viewing of complete alarm information and handling procedures.

[0066] Establish closed-loop management rules for alarm events, assign a unique work order number to each generated alarm record, and define full-process management rules for work order dispatch, handling, verification, and closure. Based on equipment type, region, and alarm level, work orders are automatically dispatched to the corresponding authorized personnel. The work order handling status is tracked in real time. After receiving the work order, the handling personnel must upload on-site handling photos, handling measures, and handling results to the platform. After handling, a verification application is submitted, and a designated person conducts on-site verification and data review of the handling results. After confirming that the equipment has returned to normal operation and the abnormality has been completely eliminated, the work order is archived in a closed loop, and the equipment operating status and the electronic file of the equipment's entire life cycle are updated simultaneously. For alarm work orders that have not been handled within the time limit, reminders are sent according to a preset cycle, and the escalation is pushed to the superior management personnel simultaneously, ensuring that all abnormal alarm events are traceable and closed-loop throughout the entire process.

[0067] Establish rules for building an equipment operation and maintenance management system. Based on the electronic records of the entire equipment lifecycle, determine corresponding operation and maintenance standards and maintenance cycles for each connected device. The operation and maintenance standards include four categories: daily inspection, periodic calibration, preventive maintenance, and fault repair. Differentiated maintenance cycles and operation and maintenance content are determined based on equipment type, equipment model, usage environment, and operating time. For sensor equipment in weather stations and soil moisture stations, determine the periodic calibration cycle and calibration standards. For power equipment such as integrated water and fertilizer machines, water pumps, and electric valves, determine the periodic maintenance and replacement of vulnerable parts. For image acquisition equipment such as video surveillance, insect pest monitoring, and spore analysis, determine the periodic cleaning, lens calibration, and hardware testing. Based on the preset operation and maintenance cycle, generate corresponding operation and maintenance plans and due reminders in advance and push them to the corresponding operation and maintenance personnel to ensure that operation and maintenance work is executed on time.

[0068] Establish comprehensive control rules for the entire operation and maintenance (O&M) process, generating a unique O&M work order for each O&M task. These rules cover the entire process from O&M plan development and task assignment to execution recording, result verification, and archiving updates. When performing O&M tasks, O&M personnel must upload photos of the O&M site, O&M content, equipment operating status monitoring data, and consumable replacement records to the platform. After completing the O&M task, a verification application must be submitted. Once verification confirms that the O&M work meets standards and the equipment is operating normally, the O&M work order is archived in a closed loop. Simultaneously, the complete O&M record is updated to the corresponding equipment's full lifecycle electronic archive, achieving full traceability and auditability of equipment O&M work. Furthermore, based on equipment O&M records and historical fault alarm data, an equipment health assessment model is established, defining scoring rules for equipment health. Based on equipment runtime, fault frequency, O&M execution status, and parameter drift, the equipment health status is graded and assessed. For equipment with low health, preventative maintenance reminders and equipment replacement suggestions are generated in advance, shifting from reactive fault repair to proactive preventative O&M.

[0069] Establish full-lifecycle management rules for equipment, using the unique identification code as the sole identifier. Establish a comprehensive lifecycle management system for equipment, from procurement and warehousing, installation and commissioning, connection and online operation, operation monitoring, maintenance and repair, fault repair, calibration and verification to scrapping and decommissioning. Define the rules for information entry, verification, and archiving at each lifecycle node. After equipment installation and commissioning, complete the entry of basic equipment information, connection verification, and operation parameter verification. Once all are passed, complete the online archiving. During equipment operation, synchronize the operating status, alarm records, maintenance records, and calibration records to the equipment's electronic file in real time. When equipment reaches its service life or fails to meet usage requirements after testing, define the rules for equipment scrapping approval, decommissioning, and archiving. After equipment scrapping, update the equipment status and permanently archive the equipment's full lifecycle electronic file, ensuring the integrity and traceability of data throughout the entire process from equipment connection to scrapping.

[0070] Statistical analysis rules for equipment operation and maintenance data are established. Based on equipment alarm records, maintenance work order data, and full lifecycle archive information stored in the unified agricultural equipment database, a multi-dimensional statistical analysis system is established. Rules for statistics are determined according to equipment type, region, alarm level, fault type, and maintenance cycle. The statistical analysis content includes equipment online rate, fault occurrence rate, average fault repair time, maintenance task completion rate, equipment calibration pass rate, and core indicators of equipment lifespan. The statistical analysis results are simultaneously pushed to the data statistical analysis module and data visualization module, and are displayed intuitively in the form of charts. This provides data support for equipment procurement and selection, maintenance strategy optimization, and management process upgrades in the park. At the same time, based on historical fault data and maintenance data, the anomaly judgment rules, alarm classification standards, and maintenance cycle settings are continuously optimized to continuously improve the accuracy of equipment fault early warning and the efficiency of maintenance management.

[0071] The process involves preprocessing and feature extraction of agricultural image data, followed by the use of deep learning algorithms to perform crop growth analysis and pest and disease identification. Figure 8 As shown, the specific implementation includes: determining the full-link collection and anchoring rules for agricultural image data; using the unique identification code of the acquisition equipment as the core anchor point; determining the metadata binding rules for each acquired image; the metadata includes the unique identification code of the equipment, the station number to which it belongs, the acquisition time, the spatial coordinates of the acquisition point, the equipment operating parameters, and the synchronously acquired environmental monitoring time-series data; classifying and storing crop condition monitoring images, insect condition forecasting images, fungal spore acquisition images, field crop visible light monitoring images, and crop thermal imaging monitoring images according to equipment type, acquisition time, and data category; and simultaneously archiving the image metadata to the image and business database of the unified agricultural equipment database, ensuring that each image can be traced back to the corresponding acquisition equipment, acquisition environment, and associated business data through the unique identification code of the equipment.

[0072] Differentiated preprocessing rules for different categories of agricultural images were determined. For crop monitoring images, a standardized preprocessing procedure was established. First, filtering algorithms were used to remove noise data caused by changes in lighting, lens shake, and dust obstruction during image acquisition. Then, grayscale transformation and contrast enhancement algorithms were used to optimize the distinction between crops and background soil in the image. Next, a multi-temporal image registration algorithm was used to align the spatial coordinates and match the pixel-level images of crop conditions taken at the same field, the same acquisition point, and at different times, eliminating image deviations caused by differences in shooting angle and shooting distance. Finally, image masking was used to remove non-crop interference areas within the field, achieving accurate segmentation of the crop target area.

[0073] For insect pest monitoring images and fungal spore collection images, a corresponding preprocessing workflow was determined. First, a morphological filtering algorithm was used to remove background noise, light spot interference, and artifacts from the microscopic imaging process. Then, an adaptive histogram equalization algorithm was used to enhance the edge contours and texture features of the target pests and fungal spores. After that, a threshold segmentation algorithm was used to separate the target area from the background area, extracting the independent target area of ​​a single pest or spore. At the same time, the adhering target contours were separated to ensure the complete extraction of individual targets.

[0074] For field crop thermal imaging images, a corresponding preprocessing workflow was determined. First, pixel-level registration of the thermal imaging data with concurrent visible light images was performed to remove abnormal data caused by environmental temperature interference and equipment temperature measurement deviations. Then, through temperature range normalization, the crop canopy temperature data was converted into a standardized grayscale image, achieving accurate segmentation of the crop canopy region and extracting effective data on crop temperature distribution. Differential feature extraction rules were determined for different categories of preprocessed images. For preprocessed seedling images, a semantic segmentation algorithm was chosen for accurate crop feature extraction. The feature extraction dimensions of the algorithm were determined, extracting the green vegetation area of ​​the crop in the image and calculating the pixel percentage, contour boundary, plant height features, and canopy coverage of the corresponding area. Simultaneously, the number of crop leaves, leaf area, and leaf health features were extracted, and the core growth characteristic parameters, including leaf area index, vegetation coverage, crop height, and growth uniformity, were quantitatively calculated. All extracted feature parameters were bound to the corresponding image metadata and simultaneously archived to a unified agricultural equipment database.

[0075] For the preprocessed insect pest monitoring images and fungal spore collection images, a convolutional neural network algorithm was selected to extract target features. The feature extraction dimensions of the algorithm were determined, and the outline shape, texture features, color features, size specifications, and structural features of the target pests and fungal spores were extracted to differentiate the features of different types of pests and fungal spores. At the same time, the extracted target features were quantitatively statistically analyzed to accurately calculate the core parameters of the number, density, and proportion of pests and spores. The statistical results were bound to the corresponding image metadata and synchronously archived to the unified agricultural equipment database.

[0076] For the preprocessed crop thermal imaging images, the corresponding feature extraction rules are determined to extract the temperature distribution features of the crop canopy, the temperature difference features between the canopy and the ambient air, and the crop temperature difference features in different areas of the field. The quantitative calculation of the crop water stress degree is completed, and the feature parameters are synchronously bound to the corresponding image metadata and archived to the unified agricultural equipment database.

[0077] The system establishes quantitative analysis rules for the entire growth cycle of crops. Based on growth characteristic parameters extracted from seedling images and considering the crop variety, current growth stage, and planting management standards of the corresponding field, it determines the matching rules for the standard crop growth model. This involves matching the standard growth curve and characteristic parameter threshold range for the entire growth cycle of the crop. The system then compares and analyzes the real-time extracted crop growth characteristic parameters with the standard growth model to determine the current growth stage and classify the crop into three levels: normal growth, lagging growth, and excessive growth. For abnormal states such as lagging or excessive growth, the system determines corresponding quantitative calculation rules for growth deviation, calculates the deviation between growth characteristic parameters and standard thresholds, and simultaneously correlates the corresponding water and fertilizer irrigation records and environmental monitoring time-series data. This analysis identifies the core causes of abnormal growth states, generates corresponding water and fertilizer strategy adjustment suggestions, and pushes them to the agricultural decision-making module to provide data support for optimizing irrigation and fertilization plans.

[0078] The system determines the prediction rules for crop growth trends, extracts growth characteristic parameters from seedling images of the same field over multiple time periods, completes the fitting process for crop growth curves, and predicts subsequent crop growth trends and key growth stage nodes based on crop variety characteristics, current growth stage, and future weather forecast data. Simultaneously, the growth analysis results and trend prediction data are archived into a unified agricultural equipment database and linked with the corresponding field's agricultural operation records and yield prediction models to achieve continuous tracking and quantitative analysis of crop growth status throughout its entire growth period.

[0079] The system establishes intelligent classification and identification rules for pests, diseases, and fungal spores. Based on target feature parameters extracted from pest monitoring images and spore collection images, it determines the identification and matching rules for deep learning classification algorithms. The extracted target features are matched one by one with the preset pest and fungal spore feature databases to determine the specific species of target pests and fungal spores, thus completing the classification and identification. Simultaneously, based on the statistical results of the feature extraction process, the system calculates the pest population density and fungal spore concentration in the corresponding monitoring area, determines the confidence level verification rules for the identification results, and marks identification results with confidence levels below a preset threshold as targets to be reviewed. These results are then pushed to management personnel for manual verification. The results of manual verification are synchronously updated to the feature database, completing the iterative optimization of the algorithm model.

[0080] The system establishes quantitative analysis rules for determining the risk of pest and disease outbreaks. Based on the identified pest species, density data, fungal spore species, and concentration data, and according to the economic threshold for pest and disease control for the corresponding crop varieties, risk level classification rules are determined. When the pest density or spore concentration exceeds the corresponding economic threshold, it is judged as a high-risk state for pests and diseases. Simultaneously, the system links the corresponding meteorological monitoring time series data, including air temperature and humidity, rainfall, and sunshine duration. Based on the pest and disease epidemic prediction model, the outbreak probability of the corresponding pests and diseases is calculated, and a secondary calibration of the risk level is completed. The identification results, risk level, and outbreak probability data are simultaneously archived to the unified agricultural equipment database and pushed to the intelligent early warning module to generate corresponding pest and disease early warning information and plant protection operation suggestions, providing accurate data support for green pest and disease control in the park.

[0081] The aforementioned predictive model, constructed based on meteorological data, is used for early warning of pests and diseases, such as... Figure 9 As shown, the specific implementation includes: determining the dimensions and time alignment rules of the input meteorological data for the pest and disease prediction model; extracting time-series meteorological monitoring data highly corresponding to the occurrence and development of pests and diseases from the unified agricultural equipment database, including core dimensions such as air temperature, relative humidity, hourly rainfall, sunshine duration, wind speed and direction, atmospheric pressure, and saturated vapor pressure difference; determining the time dimension alignment rules of the data; using the day as a fixed time step to complete the time synchronization matching of the daily meteorological data with the concurrent insect monitoring data, spore monitoring data, and crop growth data; eliminating invalid data fragments with time misalignment or interrupted collection; ensuring that all data participating in the model calculation are completely corresponding in the time dimension; and simultaneously connecting with the refined weather forecast data for the next 3 to 7 days released by the official meteorological department, determining the format normalization and verification rules of the forecast data, and completing the format unification of the forecast data and historical monitoring data to provide preliminary data support for model prediction.

[0082] The weighting rules for meteorological driving factors corresponding to different types of pests and diseases were determined. Based on the characteristics of the crop varieties planted in the park and the occurrence and prevalence patterns of major pests and diseases in the area, the core meteorological influencing factors for the occurrence of common pests and diseases, fungal diseases, and bacterial diseases in the park were determined. For diseases that prefer high humidity environments, relative humidity, number of consecutive rainfall days, and rainfall amount were determined as core weighting factors. For pests that prefer high temperature environments, daily average temperature and effective accumulated temperature were determined as core weighting factors. At the same time, the weighting rules for various influencing factors were determined. Based on the correlation analysis results between historical pest and disease occurrence data and meteorological data of the same period, the weight coefficients of each factor were quantified and assigned to ensure that the weighting allocation is highly matched with the occurrence patterns of the corresponding pests and diseases.

[0083] The construction and training rules for the pest and disease epidemic prediction model were determined. Based on historical synchronous meteorological data, pest and disease occurrence records, crop growth data, and field management data, the random forest algorithm was adopted to construct the pest and disease outbreak probability prediction model. The input and output terms of the model were determined. The input terms include the core meteorological factor data for the current day and several consecutive days, the current growth stage of the crop, and basic data on field insect population density and spore concentration. The output term is the outbreak probability of the corresponding pest and disease in the next 3 to 7 days. The training and validation rules of the model were determined. The complete historical dataset of the park for the past 3 years was selected and divided into training and validation sets in a 7:3 ratio. The model training, parameter tuning, and accuracy validation were completed. The qualified threshold of the model prediction accuracy was determined to ensure that the accuracy of the model prediction results is not less than 85%. At the same time, the iterative optimization rules of the model were determined. The model was retrained and the parameters were updated every quarter based on the newly added actual field pest and disease occurrence data and meteorological monitoring data to continuously improve the model's prediction accuracy and scenario adaptability.

[0084] The classification and triggering rules for pest and disease early warning levels were determined. Based on the pest and disease outbreak probability output by the model, the real-time field monitoring data of insect population density and spore concentration, and the corresponding economic threshold for pest and disease control, a four-level early warning classification rule was established: Level 1 Red (high risk), Level 2 Orange (relatively high risk), Level 3 Yellow (medium risk), and Level 4 Blue (low risk). The triggering conditions for each level were determined. When the model predicts that the probability of pest and disease outbreak within the next 3 days is greater than or equal to 90%, and the field monitoring data of insect population density or spore concentration has reached 80% or more of the economic threshold, a Level 1 Red warning is triggered. When the outbreak probability is between 70% and 89%, a Level 2 Orange warning is triggered. When the outbreak probability is between 50% and 69%, a Level 3 Yellow warning is triggered. When the outbreak probability is between 30% and 49%, a Level 4 Blue warning is triggered. At the same time, a dynamic adjustment rule for the early warning threshold for different crop growth stages was determined. For crop seedling, flowering, and fruiting stages, which are sensitive growth stages for pests and diseases, the early warning trigger threshold is lowered accordingly to improve the sensitivity of early warnings during sensitive periods.

[0085] The rules for generating, pushing, and spatially displaying early warning information are established. When the early warning trigger conditions are met, corresponding pest and disease early warning information of the corresponding level is immediately generated. The early warning information includes a full range of content, including the early warning level, pest and disease type, corresponding affected field, predicted outbreak time, core inducing factors, field monitoring data, meteorological forecast data, and recommended green control measures. The early warning information is simultaneously and completely stored in the unified agricultural equipment database. Based on the early warning level, the corresponding push rules are determined. Level 1 red warnings are pushed to the plant protection manager and production management personnel of the park through platform pop-ups, SMS, and voice calls. Level 2 orange warnings are pushed to the corresponding plant protection management personnel through platform pop-ups and SMS. Level 3 yellow warnings and Level 4 blue warnings are uniformly displayed in the platform's early warning center and pushed to the platform's internal messages. At the same time, based on the association mapping relationship between spatial identification codes and field numbers, the field area corresponding to the early warning is highlighted in the 3D geographic model of the GIS visualization module to achieve spatial positioning and visualization of early warning events. Clicking on the highlighted area allows direct viewing of the complete early warning information and control suggestions.

[0086] Establish a closed-loop linkage rule between early warning and plant protection operation execution. For the generated pest and disease early warning information, automatically match the corresponding standardized plant protection operation plan for the pest and disease, and generate a plant protection operation work order including operation time, operation area, operation method, recommended pesticides and dosages, and safety interval. Assign a unique operation number to each work order. According to the operation area and operation type, the work order is automatically dispatched to the corresponding personnel with operation permissions. The execution status of the work order is tracked in real time. After completing the plant protection operation, the operator must upload full-dimensional information such as operation site photos, operation time, operation content, and pesticide usage records to the platform and submit a verification application. After the operation is reviewed by a specialist and the field pest and disease monitoring data after the operation is completed, the control measures are confirmed to be implemented and the pest and disease risk is effectively controlled. The work order is then archived in a closed loop and the complete operation record is stored in the unified agricultural equipment database. At the same time, the pest population density, spore concentration data and meteorological data monitored in the field after the operation are simultaneously input into the pest and disease prediction model to complete the verification and optimization of the model prediction results. Establish data traceability and tracking rules for the entire early warning process. Store all model input data, calculation process parameters, predicted outbreak probability, early warning level, push records, plant protection operation execution records, and control effect verification data for each early warning in a unified agricultural equipment database. This enables traceability and review of the entire pest and disease early warning process. At the same time, based on historical early warning records, actual occurrences, and control effect data, continuously optimize the model's factor weights, early warning thresholds, and level classification rules to continuously improve the accuracy of early pest and disease warnings and the effectiveness of control guidance.

[0087] The aforementioned establishment of a multi-dimensional data statistical analysis system for in-depth mining of agricultural data, such as... Figure 10 As shown, the implementation specifically includes: Based on the hierarchical storage architecture and data association mapping rules of the unified agricultural equipment database, determining the collection scope of the full-scale statistical analysis data source, including six core data categories: environmental monitoring time-series data, equipment operating status data, spatial geographic information data, image acquisition metadata, agricultural operation record data, and fault alarm maintenance data. Using the unique equipment identification code as the core anchor point, the system completes the association and collection of data throughout the entire lifecycle of a single piece of equipment; using the spatial identification code as the core anchor point, it completes the spatial dimension data association and collection of fields, monitoring stations, and irrigation zones; and using the standard timestamp as the core anchor point, it completes the time dimension alignment and collection of data reported by different collection frequencies and different equipment, ensuring the integrity, relevance, and consistency of the statistical analysis data source. A hierarchical architecture for multi-dimensional statistical analysis is determined, divided into three core statistical levels: time dimension, spatial dimension, and business dimension, with detailed statistical rules for each level.

[0088] In terms of time dimension, the system establishes hierarchical statistical rules for hours, days, weeks, months, quarters, years, and the entire crop growth period. Hourly statistics include high-frequency summaries of real-time environmental monitoring data, equipment operating parameters, and pipeline pressure and flow data. Daily statistics include daily summaries of irrigation water consumption, electricity consumption, rainfall, crop water requirements, equipment online time, and the number of alarm events. Weekly, monthly, quarterly, and annual statistics include summaries of agricultural operations, resource consumption, equipment maintenance, crop growth data, and pest and disease occurrence within the corresponding period. Statistics for the entire crop growth period include full-cycle summaries of water and fertilizer input, environmental data, agricultural operations, pest and disease control, and yield data from sowing to harvest. Custom filtering and statistical analysis for any time interval are supported.

[0089] In terms of spatial dimensions, a five-level spatial statistical rule is established, based on the entire park area, field zones, irrigation zones, monitoring station locations, and individual equipment. The overall park-level statistics include a macro-level summary of the park's total planting scale, total number of equipment, overall equipment online rate, cumulative irrigation water consumption, cumulative electricity consumption, total number of alarm events, and overall maintenance task completion rate. Field zone-level statistics include a summary of the corresponding field's crop information, soil property data, water and fertilizer input data throughout the growth period, irrigation execution records, crop growth monitoring data, pest and disease occurrence, and yield prediction data. Irrigation zone-level statistics include the corresponding zone's pipe network. The system provides a zonal summary of deployment and electric valve equipment information, single irrigation execution data, cumulative irrigation water volume, and irrigation uniformity data. The station-level statistics include environmental monitoring data collected by all equipment at the corresponding station, equipment operation status data, data change trends, and a summary of points for horizontal comparison of data from similar stations. The single-equipment-level statistics include a comprehensive summary of the corresponding equipment's basic profile information, full-cycle data collection, operational status changes, command execution records, fault alarm records, and maintenance records. It also supports targeted statistical analysis based on multiple conditions, such as station type, equipment type, crop variety, and region.

[0090] From a business perspective, specific statistical analysis rules for six major business lines have been established. Environmental monitoring statistics are based on time-series data collected by meteorological stations, soil moisture stations, and integrated water and fertilizer machines. Statistical rules for air temperature and humidity, soil temperature and humidity, rainfall, wind speed and direction, solar radiation, soil EC value, and irrigation solution EC / pH value have been established. The statistical content includes the maximum, minimum, average, cumulative values, and trends within the corresponding period, as well as the degree of matching with the suitable range for crop growth. This enables horizontal comparison of environmental data from different stations and different fields to identify environmental differences within the region.

[0091] The water and fertilizer irrigation special statistics are based on the execution data and collected data of the agricultural decision-making module and the water and fertilizer control module. The statistical rules for irrigation water volume, fertilizer application, irrigation duration, irrigation frequency, water and fertilizer utilization rate, and irrigation water and electricity costs are determined. The statistical content includes the total irrigation water volume, zonal irrigation water volume, single irrigation water volume, total fertilizer application, usage of different fertilizer types, irrigation water and electricity consumption, irrigation plan execution completion rate, water and fertilizer ratio compliance rate, and irrigation uniformity. The data on water and fertilizer input are compared horizontally for different fields, different crops, and different growth stages, and the correlation between water and fertilizer input and crop growth status is fitted.

[0092] The crop growth-specific statistics are based on image analysis data from the seedling monitoring module, environmental monitoring data, and agricultural operation record data. The statistical rules for crop plant height, leaf area index, vegetation coverage, growth uniformity, and growth period progress are determined. The statistical content includes the changing trends of crop growth characteristic parameters within the corresponding period, the deviation from the standard growth model, the horizontal comparison of crop growth status in different fields, and the correlation between growth status and water and fertilizer input and environmental conditions. The statistics are used to fit crop growth trends and predict yield.

[0093] The special statistics on pest and disease control are based on the identification data from the pest monitoring and forecasting module, the prediction data from the pest and disease early warning module, and the execution data of plant protection operations. It determines the statistical rules for pest species and quantities, spore species and concentrations, the number of pest and disease early warning events, the distribution of early warning levels, the number of plant protection operations, the amount of pesticides used, and the control effects. The statistical content includes the types of pests and diseases that occur, the areas where they occur, the frequency of occurrence, the changing trends of insect population density and spore concentration, the accuracy rate of early warnings, the completion rate of plant protection operations, and the effectiveness of control measures within the corresponding period. It also completes the fitting of the correlation between the occurrence patterns of different pests and diseases and meteorological conditions. The equipment lifecycle-specific statistics are based on full data from the equipment management module, equipment operation and maintenance module, and fault alarm module. This establishes statistical rules for equipment online rate, equipment failure rate, average fault repair time, operation and maintenance task completion rate, equipment calibration pass rate, and equipment lifespan. Statistical content includes the total number of devices within the corresponding period, online rate and offline time for different types of devices, number and severity distribution of fault alarm events, frequency of different types of faults, fault handling timeliness, planned and completed number of operation and maintenance tasks, equipment calibration and verification pass rate, vulnerable parts replacement records, and the equipment lifecycle input-output ratio. This allows for a horizontal comparison of the operational stability of different brands and types of equipment.

[0094] The special statistics for park operation and management are based on agricultural operation records, resource consumption data, equipment operation and maintenance data, and crop production data. The statistical rules for determining the park's planting scale, crop type distribution, labor input for agricultural operations, total input of water, electricity and agricultural materials, resource utilization rate, and production plan completion rate are established. The statistical content includes the overall planting structure of the park within the corresponding period, the execution progress of various agricultural operations, the total consumption of various production resources and the consumption per unit area, the execution completion rate of the production plan, and the execution efficiency of various management processes.

[0095] Based on the results of multi-dimensional statistical analysis, the execution rules for in-depth agricultural data mining were determined. Correlation analysis models were constructed for five core scenarios: environmental monitoring, water and fertilizer irrigation, crop growth, pest and disease control, and equipment operation. These models aimed to uncover implicit relationships between data. Specifically, for the water and fertilizer irrigation scenario, the correlation between water and fertilizer input, environmental conditions, soil properties, crop growth status, and final yield was explored to determine the optimal water and fertilizer input ranges for different crops and growth stages. For the pest and disease control scenario, the correlation between meteorological conditions, field environment, crop growth stage, and pest and disease occurrence patterns was explored to identify the core precursors and critical conditions for various pest and disease outbreaks. For the equipment operation scenario, the correlation between equipment operating parameters, usage time, maintenance frequency, and equipment failure rate and lifespan was explored to determine the optimal preventative maintenance cycle and maintenance standards for different types of equipment. Finally, for the crop planting scenario, the correlation between environmental conditions, planting management measures, and crop yield and quality was explored to determine the optimal planting management plan for the corresponding region and crop. Establish standardized display rules for statistical analysis results. Based on statistical dimensions and business scenarios, use standardized chart formats such as line charts, bar charts, pie charts, heat maps, and trend curves for intuitive display. Support linked drill-down of statistical charts. It can drill down from the macro summary data of the park to the detailed data of fields, stations, and individual devices, realizing full-link data viewing from macro overview to micro details.

[0096] Establish rules for archiving and backing up statistical analysis data, automatically archive statistical analysis results by day, month, and year, associate and bind archived data with the original data collected in the corresponding period and agricultural operation records, store them in a unified database for agricultural equipment, and establish a hierarchical backup mechanism to ensure the integrity and traceability of statistical analysis data.

[0097] Establish export and access control rules for statistical analysis data, supporting standardized export formats for statistical reports, raw data, and charts, including general table and document formats. Based on user roles and management scope, define corresponding access control rules for data statistics, viewing, exporting, and archiving. All data operations are fully traceable to ensure compliance and security. Continuously update statistical dimensions and rules based on adjustments to the park's planting structure, equipment additions and replacements, and business process upgrades. Continuously optimize data mining models and analysis dimensions based on the implementation effects of each agricultural decision, crop growth feedback, and pest and disease control results, ensuring the statistical analysis system fully matches the park's actual production and management needs, and continuously providing comprehensive, accurate, and effective data support for optimizing the park's smart agriculture management strategy.

[0098] This application also provides an embodiment of an electronic device. The electronic device is manifested in the form of a general-purpose computing device. The components of the electronic device may include, but are not limited to: one or more processors or processing units, memory, and buses connecting different components (including memory and processing units).

[0099] A bus refers to one or more of several bus architectures, including memory buses or memory controllers, peripheral buses, graphics acceleration ports, processors, or local buses using any of the various bus architectures. Examples of these architectures include, but are not limited to, Industry Standard Architecture (ISA) buses, Micro Channel Architecture (MCA) buses, Enhanced ISA buses, Video Electronics Standards Association (VESA) local buses, and Peripheral Component Interconnect (PCI) buses.

[0100] The electronic device typically includes a variety of computer-readable media. These media can be any available media that can be accessed by the electronic device, including volatile and non-volatile media, and removable and non-removable media.

[0101] Memory may include computer-readable media in the form of volatile memory, such as random access memory (RAM) and / or cache memory. Electronic devices may further include other removable / non-removable, volatile / non-volatile computer device storage media. By way of example only, storage may be used to read and write non-removable, non-volatile magnetic media.

[0102] The electronic device can also communicate with one or more external devices (e.g., keyboard, pointing device, camera, etc.), may include a display, and may communicate with one or more devices that enable a user to interact with the electronic device, and / or with any device that enables the electronic device to communicate with one or more other computing devices (e.g., network card, modem, etc.). This communication can be performed via an input / output (I / O) interface. Furthermore, the electronic device can also communicate with one or more networks (e.g., local area network (LAN), wide area network (WAN) and / or public networks, such as the Internet) via a network adapter. The network adapter communicates with other modules of the electronic device via a bus. The processor executes various functional applications and data processing by running programs stored in memory, such as implementing the smart agriculture management strategy optimization system provided in the above embodiments of the present invention.

Claims

1. A smart agricultural management strategy optimization system, characterized in that, include: The device access module is used to establish standardized protocols for accessing all types of agricultural sensing devices; The data management module is used to normalize the collected data and build a unified database for agricultural equipment. The equipment management module is used for full-process control of equipment and data; The GIS visualization module is used to build three-dimensional geographic models of farmland based on the GIS system and integrate agricultural business data. Build an agricultural management visualization platform to enable spatial display and multi-dimensional viewing of data; The agricultural decision-making module is used to integrate multi-dimensional time-series monitoring data, construct a multi-factor fusion quantitative agricultural decision-making model, determine irrigation triggering rules and execution strategies, and make data-driven precision agricultural decisions. The water and fertilizer control module is used to perform closed-loop regulation of water and fertilizer ratio using PID and feedforward control algorithms, and to perform dynamic balance scheduling of irrigation network pressure through group polling and constant pressure control algorithms. The equipment operation and maintenance module is used to complete equipment fault alarms based on the SCADA system, establish an operation and maintenance management system, and conduct full life cycle management of equipment.

2. The intelligent agricultural management strategy optimization system according to claim 1, characterized in that, Also includes: The intelligent early warning module is used to preprocess and extract features from agricultural image data, and to perform crop growth analysis and pest and disease identification through deep learning algorithms. Predictive models are built based on meteorological data to provide early warnings of pests and diseases; The data analysis module is used to establish a multi-dimensional data statistical analysis system and to conduct in-depth mining of agricultural data.

3. The intelligent agricultural management strategy optimization system according to claim 2, characterized in that, Establish standardized protocols for accessing all types of agricultural sensing devices, including: developing standardized access specifications for all types of agricultural sensing devices based on mainstream communication protocols in agricultural scenarios; determining interaction and management rules for device access; developing protocol conversion rules for different brands and models of the same type of device; completing the unified conversion of non-standard protocols to standardized protocols; eliminating protocol barriers between devices; assigning a unique identification code containing multi-dimensional information to each access device; completing device permission verification, identity recognition, and access link binding management; developing full-process verification rules for device access; completing formal device access after verification; and implementing a supporting device online status heartbeat detection mechanism to periodically detect device online status and update the device online ledger in real time.

4. The intelligent agricultural management strategy optimization system according to claim 2, characterized in that, The collected data is normalized to construct a unified database for agricultural equipment. This includes: establishing multi-dimensional classification rules for the collected data, dividing the data into multiple categories according to their sources and determining their corresponding dimensions, formulating rules for cleaning invalid data, completing the processing and synchronization of abnormal, redundant, and missing data and anomaly marking, formulating rules for normalizing heterogeneous data, completing the alignment of data time dimensions, unifying the units of measurement for numerical data, standardizing the coding of status data, and unifying the coordinates and formats of spatial geographic data, constructing a hierarchical unified database for agricultural equipment with the unique identification code of the equipment as the core, establishing multiple related sub-databases, and formulating rules for database access control, archiving and backup, and full-link data association mapping.

5. The intelligent agricultural management strategy optimization system according to claim 1, characterized in that, The aforementioned full-process management and control of equipment and data includes: establishing a full lifecycle electronic file for access equipment based on the unique identification code of the equipment; standardizing the management rules and access control and change recording mechanism for basic equipment information; synchronizing the corresponding information to the unified agricultural equipment database; collecting and updating the core operating status of the equipment in real time; completing classification and multi-dimensional filtering and searching; synchronizing the status data to the corresponding module of the platform; managing the full-link data and full-process traceability of the equipment with the unique identification code of the equipment; standardizing the permission verification and operation recording of data calls; formulating full-process management rules for remote control commands of the equipment; standardizing the closed-loop management of the entire process of fault handling; collecting equipment and data by monitoring station; and formulating differentiated management and display rules.

6. The intelligent agricultural management strategy optimization system according to claim 1, characterized in that, The process involves constructing a 3D geographic model of farmland based on a GIS system, integrating agricultural business data, and building an agricultural management visualization platform for spatial data display and multi-dimensional viewing. This includes: determining a unified spatial coordinate system compatible with the GIS system; completing the coordinate collection and vectorization processing of all spatial elements of farmland; assigning unique identifier codes to spatial elements and establishing association mappings with equipment, field, and station numbers; building a 3D geographic model of the entire farmland area based on multi-source spatial data; completing the refined modeling and mounting of surface entities and core equipment; configuring model interactive operations and multi-condition filtering and positioning rules; establishing a classification mounting and real-time synchronization mechanism between agricultural business data and the 3D model; and building a single-map visualization platform for agricultural management with the 3D model as the core.

7. The intelligent agricultural management strategy optimization system according to claim 6, characterized in that, The method employs PID and feedforward control algorithms for closed-loop regulation of water and fertilizer ratios, and uses group polling and constant pressure control algorithms for dynamic balancing and scheduling of irrigation network pressure. This includes: setting water and fertilizer control targets based on crop characteristics, growth stage, and soil properties; pre-adjusting initial operating parameters of pumps and valves through feedforward control; dynamically correcting water and fertilizer ratio deviations based on real-time monitoring data and PID algorithms; implementing a safety interlock mechanism; retaining data throughout the entire process; dividing independent irrigation zones based on GIS; formulating group polling execution rules for electric valves; dynamically adjusting pump operating parameters based on real-time network pressure data to maintain stable network pressure; linking with agricultural decision-making to optimize scheduling strategies; synchronizing operating status to corresponding modules; verifying irrigation effects; and continuously optimizing control parameters.

8. The intelligent agricultural management strategy optimization system according to claim 1, characterized in that, The aforementioned system-based equipment fault alarm system establishes an operation and maintenance management system and conducts full lifecycle control of equipment. This includes: establishing a real-time data synchronization link with a unified database of agricultural equipment by connecting various modules of the SCADA system platform; formulating rules for judging equipment and data anomalies and hierarchical alarm standards; generating full-dimensional alarm records and pushing them to multiple channels according to levels; establishing a closed-loop handling mechanism for alarm work orders; building a differentiated operation and maintenance system for all types of equipment; formulating operation and maintenance work order management rules; constructing an equipment health assessment model; and conducting multi-dimensional statistical analysis of equipment operation and maintenance data.

9. The intelligent agricultural management strategy optimization system according to claim 1, characterized in that, The process of preprocessing and feature extraction of agricultural image data, and using deep learning algorithms to perform crop growth analysis and pest and disease identification, includes: full-link collection, metadata binding, and classification archiving of multi-category agricultural image data; development of differentiated preprocessing and feature extraction rules for different types of agricultural images; extraction and quantitative calculation of core features of crop growth, pests and diseases, and fungal spores using corresponding algorithms; analysis of crop growth status throughout its entire growth period, trend prediction, and judgment of abnormal causes based on feature parameters; generation of water and fertilizer optimization suggestions and pushing them to the agricultural decision-making module; classification and identification of pests and diseases and fungal spores using deep learning algorithms; quantitative analysis and level determination of pest and disease outbreak risks based on corresponding data; generation of early warning information and plant protection suggestions; and simultaneous data archiving and algorithm iteration optimization.

10. The intelligent agricultural management strategy optimization system according to claim 1, characterized in that, Extract time-series meteorological monitoring data corresponding to the occurrence of pests and diseases, complete data time dimension alignment and invalid data processing, connect with meteorological forecast data and complete format normalization verification, determine core meteorological influencing factors and weight allocation rules for different pests and diseases, construct a pest and disease outbreak probability prediction model and complete training, verification and iterative optimization, classify early warning levels and set trigger rules based on model output, field monitoring data and control thresholds, generate full-dimensional early warning information and push it according to level, generate plant protection operation work orders and complete closed-loop processing.

Images