Agricultural data retention and history tracking system based on unmanned inspection

Abstract

The application discloses an agricultural data retention and history tracing system based on unmanned inspection, and belongs to the technical field of agriculture. The system comprises a data acquisition module, a data record storage module, a history tracing analysis module and a user interaction module. The data acquisition module is connected with a drone, an environmental sensor and an unmanned work vehicle, and synchronously collects crop images, soil air environment data, operation parameters and equipment states. The data record storage module adopts a distributed architecture to classify and index data, long-term encrypt storage and off-site backup. The history tracing analysis module supports multi-dimensional retrieval of historical data according to time, land, crop type and the like, realizes cross-period comparison and analysis, and outputs a planting optimization scheme. The user interaction module provides a Web end and a mobile APP end, which are respectively used for system configuration and data management, real-time data viewing and scheme receiving. The application solves the problems that agricultural data is easy to be lost, difficult to be traced and planting decisions depend on experience, improves yield and reduces waste.

Description

Technical Field

[0001] This invention relates to agricultural information technology methods, and in particular to an agricultural data retention and historical traceability system based on unmanned inspection. Background Technology

[0002] With the acceleration of agricultural mechanization and intelligentization, intelligent equipment such as drone inspections and unmanned fertilization vehicles have been gradually applied to core aspects of agricultural production, including inspection, fertilization, and watering, playing a significant role in improving operational efficiency. However, the application of existing intelligent agricultural equipment still has significant technical shortcomings, with the core issue being the lack of a closed-loop data management system.

[0003] In subsequent planting cycles, growers cannot access key historical data (such as soil moisture and fertilizer application rates during the same period in previous years), nor can they compare the differences in growth environment and operational parameters for the same crop in different years, making it difficult to summarize planting patterns. Due to the lack of data support, growers can only set current planting operational parameters (such as watering and fertilizer application rates) based on past experience. However, experience-based decision-making has strong randomness and adaptability limitations: for example, if the previous year had abundant rainfall but this year has less, using historical watering parameters can easily lead to crop wilting due to water shortage or excessive water waste; if soil fertility changes due to crop rotation, climate, or other factors, continuing to apply historical fertilizer rates may result in reduced crop yields due to insufficient fertility or root burn due to excessive fertility. The "trial and error costs" caused by such experience biases not only increase the material and labor costs of agricultural production but also directly affect crop yields, severely restricting the long-term optimization of planting plans and the improvement of agricultural production efficiency.

[0004] In existing technologies, some agricultural data management tools attempt to solve the above problems, but they still have functional limitations: one type of tool focuses only on single-dimensional data collection (such as only monitoring soil moisture or only recording operational parameters), and cannot cover the entire chain of data from "crop images to environmental data to operational parameters to equipment status", resulting in insufficient data integrity; another type of tool has basic data storage functions, but lacks core capabilities such as cross-period comparison of historical data and analysis of the adaptability of operational parameters to the environment, and cannot transform data into effective information to support planting decisions, making it difficult to promote the transformation of planting decisions from "experience-driven" to "data-driven". Summary of the Invention

[0005] Purpose of the invention: To overcome the technical problems of scattered and lost agricultural data, lack of historical traceability, and reliance on experience for planting decisions in the existing technology, this invention provides an agricultural data retention and historical traceability system based on unmanned inspection.

[0006] Technical solution: An agricultural data retention and historical traceability system based on unmanned inspection, including: The data acquisition module is used to connect to the drone, environmental sensors and unmanned operation vehicle through a wireless communication network to simultaneously collect crop growth images, soil and air environment data, as well as fertilizer application, watering, operation speed parameters, and record equipment power and fault status. The data recording and storage module, connected to the data acquisition module, uses a distributed file system (HDFS) architecture to retain data. It establishes indexes for four data categories—image, environment, operation, and equipment—through a metadata classification engine. Each data entry is associated with a unique identifier field, including acquisition time, plot number, crop type code, and growth cycle stage code. The module is configured to store at least 3 years of data (preferably 3-10 years). It uses AES-256 encryption algorithm for data transmission and storage encryption, with the encryption key managed by the hardware security module (HSM). Data backup employs a local dual-copy + off-site disaster recovery mode, where the off-site backup node is at least 50km (preferably ≥50km) away from the local node. The historical traceability analysis module is connected to the data recording and storage module. It has a built-in data retrieval engine for retrieving historical data by collection time, plot number, crop type and / or growth cycle stage. It performs cross-cycle data difference analysis through data comparison algorithm and analyzes the compatibility of operation parameters with soil environment data and climate data through data analysis model, and outputs a planting optimization plan including parameter adjustment suggestions. The data analysis model is a multiple linear regression model, where the dependent variable is the yield per unit area of ​​the crop (kg / mu), and the independent variables include: operation parameters (fertilizer application, watering, operation speed, topdressing frequency, plant protection operation time nodes), soil environment (soil moisture, soil temperature, soil pH, organic matter content), climate (daily average temperature, diurnal temperature range, rainfall, sunshine duration, air humidity, wind speed), and growth cycle stage codes.

[0007] The user interaction module, connected to the historical traceability analysis module and the data acquisition module, includes a web client and a mobile app. The web client is used by administrators to manage data and configure the system, while the mobile app is used by farmers to view real-time data, retrieve historical records, and receive planting optimization plans. The operation only requires three steps: "view data - compare with history - use the plan".

[0008] The environmental sensor measures soil moisture in the range of 0-100%RH with a measurement error of ≤5%FS, and soil temperature in the range of -40℃ to 85℃ (preferably -10℃ to 60℃) with a measurement error of ≤0.5℃. The data acquisition accuracy meets agricultural industry testing standards.

[0009] The data recording and storage module supports TB-level expansion, with a single node storage capacity of ≥50TB; it uses the AES-256 encryption algorithm, with an encryption key update cycle of 7 days, and performs off-site backup operations, ensuring sufficient data security.

[0010] The historical traceability analysis module has a response time of ≤3 seconds when comparing cross-cycle data; during the analysis process, line charts and bar charts are generated through data visualization algorithms to intuitively show the correlation between operating parameters and crop growth status and yield.

[0011] The default data acquisition frequency preset on the web client is 10Hz; the preset acquisition frequency threshold alarm is specifically triggered when the acquisition frequency is lower than 5Hz or higher than 50Hz, triggering an SMS alarm.

[0012] The data acquisition module uses only LoRa network for communication and does not include a 4G / 5G wireless communication module; the data recording and storage module has a single-node storage capacity of 50TB, and the off-site backup node is 50km away from the local node; the mobile APP of the user interaction module only supports filtering data by plot number.

[0013] Furthermore, the data acquisition module connects to the drone, environmental sensors, and unmanned operation vehicle via a 4G / 5G wireless communication module and a LoRa network to simultaneously collect crop growth images, soil and air environment data, and operation parameters, while also recording equipment power and fault status. Furthermore, the drone is equipped with a high-definition camera with a resolution of ≥3840×2160, a lens pixel of ≥12 million, a shooting angle covering 80°-120°, and the accuracy of identifying crop leaf morphological details is ≤0.5mm, and the positioning error of the distribution area of ​​pests and diseases is ≤10cm. Furthermore, the data recording accuracy of the unmanned vehicle is ≤1%, and the equipment status is collected at a frequency of once per second, which can promptly report any abnormalities in equipment operation. Furthermore, the web interface of the user interaction module supports batch import and export of data in Excel format, supports adjusting the data acquisition frequency (adjustment range is 1-60Hz), and can preset the acquisition frequency threshold alarm; the mobile APP supports quick data filtering by plot number and crop type code, with a filtering response time of ≤1 second; the two ends achieve real-time data synchronization through the WebSocket protocol, with a data synchronization delay of ≤1 second to ensure information consistency, and a data verification mechanism (CRC32) is used during the synchronization process to avoid data loss or mistransmission, ensuring information consistency.

[0014] Compared with the prior art, the present invention has the following advantages: This invention relates to an agricultural data retention and historical traceability system based on unmanned inspection, belonging to the field of agricultural technology. It includes a data acquisition module, a data recording and storage module, a historical traceability analysis module, and a user interaction module. The data acquisition module connects to unmanned inspection equipment to collect four categories of basic data: crop, environment, operation, and equipment. The data recording and storage module retains data by category, labels key attributes, provides long-term storage and encrypted backup to ensure data security. The historical traceability analysis module retrieves historical data from multiple dimensions, enables cross-period comparison, analyzes parameter adaptability, and outputs planting optimization plans. The user interaction module allows management of data / system configuration via a web interface and viewing of data / receiving plans via an app, simplifying the operation process. This invention primarily solves the problem of scattered data loss after drone and unmanned vehicle operations in traditional farming, leading to high costs associated with relying on experience and trial-and-error in subsequent planting. This system automatically collects crop photos taken by drones, soil and air data measured by sensors, and fertilizer application information from unmanned vehicles, securely stores this data, and allows for comparison of historical data from different years, ultimately recommending the most suitable planting plan to farmers. The system is easy to operate and can be used with a mobile phone or computer, helping farmers increase yields and reduce waste.

[0015] The system of this invention constructs a complete data closed loop, solving the pain points of traditional agricultural data management; it provides comprehensive data collection and secure storage, supporting the summarization of multi-cycle planting patterns; its traceability analysis is efficient and intuitive, with a low operating threshold, facilitating large-scale promotion and application. Attached Figure Description

[0016] Figure 1 shows a flowchart of the agricultural data retention and historical traceability system based on unmanned inspection provided in an embodiment of the present invention; Figure 2 shows a schematic diagram of the communication connection relationship of the data acquisition module provided in an embodiment of the present invention. Detailed Implementation

[0017] To better understand the above-mentioned objectives, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in these embodiments can be combined with each other.

[0018] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and therefore the scope of protection of the invention is not limited to the specific embodiments disclosed below.

[0019] Figure 1 shows a flowchart of an agricultural data retention and historical tracing system based on unmanned inspection provided by an embodiment of the present invention.

[0020] The present invention will now be described in detail with reference to specific embodiments: Example 1 As attached Figure 1 As shown, the system provided in this embodiment includes four core modules: a data acquisition module, a data recording and storage module, a historical traceability and analysis module, and a user interaction module. These four modules work together to form a complete closed loop of "collection-retention-traceability-application". Specific implementation details are as follows: Data acquisition module: Adopting a dual-mode communication architecture, it integrates a 4G / 5G wireless communication module and a LoRa network communication module to establish stable data transmission connections with drones, environmental sensors, and unmanned vehicles, respectively, to achieve synchronous data acquisition and real-time uploading of data from multiple devices, while also monitoring the operating status of the equipment.

[0021] The drone used is a multi-rotor industrial-grade inspection drone, equipped with a 4K high-definition camera with a resolution of 3840×2160 and a lens with 12 million pixels. The shooting angle is adjusted to 100° (within the optimal range of 80°-120°). The preset cruise route is grid-like, and the cruise speed is set to 5m / s. A crop growth image is captured every 3 meters. In the actual collection process, the accuracy of crop leaf morphology detail recognition can reach 0.4mm (better than the requirement of ≤0.5mm), and the positioning error of the distribution area of ​​pests and diseases is controlled within 8cm (better than the requirement of ≤10cm). It can clearly capture subtle features such as crop seedling diseases and leaf nutrient deficiencies.

[0022] The environmental sensors adopt a distributed deployment mode, with one monitoring node set up at a density of 6 mu (approximately 0.4 hectares) of farmland, evenly distributed in different areas of the farmland. Each monitoring node integrates a soil moisture sensor, a soil temperature sensor, and an air temperature and humidity sensor. The soil moisture sensor has a measurement range of 0-100%RH and an actual measurement error of 3%FS. The soil temperature sensor has a measurement range of -40℃ to 85℃ and an actual measurement error of 0.3℃ (better than the requirement of ≤0.5℃). The data acquisition accuracy of all sensors meets the requirements of GB / T30470-2013 "Technical Requirements for Agricultural Meteorological Observation Equipment". The acquired data is transmitted to the data acquisition module via a low-power LoRa network with a transmission delay of ≤2 seconds.

[0023] The unmanned operation vehicle uses an electric tracked operation device, which is suitable for farmland operations in different terrains. Its control system and data acquisition module communicate in real time to accurately record core operation parameters such as fertilizer application, watering, and operation speed. The data recording accuracy is 0.8% (better than the requirement of ≤1%). At the same time, the equipment has a built-in status monitoring unit that collects status data such as the remaining power, motor operating temperature, and wear and tear of operating parts once per second. When the power is detected to be below 20%, the motor temperature is above 85℃, or the wear and tear of parts exceeds the standard, an abnormal feedback signal is immediately sent to the data acquisition module, and a local audible and visual alarm is triggered simultaneously to remind the staff to deal with the situation in time.

[0024] Data recording and storage module: Built on the distributed file system (HDFS) architecture and deployed on a cloud server cluster, with a single node storage capacity of 60TB (better than the requirement of ≥50TB), supporting horizontal expansion and can be expanded to TB level or more according to actual data volume needs, meeting the long-term storage needs of 3-10 years.

[0025] The module has a built-in metadata classification engine. After receiving all the data transmitted from the data acquisition module, it automatically classifies the data into four categories: "image data, environmental data, operational data, and equipment data," and creates an independent index for each category to facilitate rapid retrieval. Each data entry is automatically associated with a unique identifier field, in the following format: acquisition time (accurate to the second, format YYYY-MM-DDHH:MM:SS), plot number (composed of a 6-digit area code + a 4-digit plot serial number, e.g., 3301000001), crop type code (referring to the coding specified in GB / T2930.1-2017 "Forage Seeds Part 1: Gramineae," e.g., wheat code is 01001), and growth cycle stage code (01-sowing period, 02-seedling period, 03-growing period, 04-flowering period, 05-fruiting period, 06-maturity period), ensuring that each data entry can be accurately located and uniquely traced.

[0026] Data transmission and storage are encrypted using the AES-256 encryption algorithm. The encryption key is managed offline through a hardware security module (HSM) to prevent key leakage. The key update cycle is set to 7 days to further enhance the data security level. At the same time, a dual backup mode of "local dual replicas + off-site disaster recovery" is adopted. The local server cluster stores two complete copies of the data, and the off-site backup node is located in an independent data center 80km away from the local node (better than the requirement of ≥50km). Incremental backups are performed every day from 2-4 am. After the backup is completed, the data integrity is automatically verified to ensure that there is no data leakage or loss. Even if the local server fails, the data can be quickly restored through the off-site backup node.

[0027] Historical data retrieval module: Built-in Elasticsearch data retrieval engine, supporting users to retrieve historical data by single or combined conditions. Retrieval conditions include collection time interval, plot number, crop type code, growth cycle stage code, etc. The retrieval response time is ≤2 seconds (better than the requirement of ≤3 seconds), and historical data of any time period, any plot, and any crop can be retrieved quickly.

[0028] The data analysis model is a multiple linear regression model, where the dependent variable is the yield per unit area of ​​the crop (kg / mu), and the independent variables include: operation parameters (fertilizer application, watering, operation speed, topdressing frequency, plant protection operation time nodes), soil environment (soil moisture, soil temperature, soil pH, organic matter content), climate (daily average temperature, diurnal temperature range, rainfall, sunshine duration, air humidity, wind speed), and growth cycle stage codes.

[0029] To overcome the nonlinearity throughout crop growth, the multiple linear regression model employs piecewise modeling: independent sub-models are trained for each growth cycle stage (seedling stage, growth stage, flowering stage, and maturity stage), ensuring that the crop's response to environmental and operational parameters approximately follows a linear relationship within each growth cycle stage. Corresponding model coefficient correction factors can be set for different plots or climate zones.

[0030] The multiple linear regression model is trained using historical data (including crop yield calibration data) stored in the system. The training dataset is divided into training, validation, and test sets in a 7:1.5:1.5 ratio, and stratified by growth cycle. Model training employs least squares optimization combined with gradient descent, with the mean squared error (MSE) as the loss function. Model performance is evaluated using 5-fold cross-validation, and the evaluation metric includes the coefficient of determination (R²). 2 Mean absolute error (MAE) and root mean square error (RMSE). After training convergence, the regression coefficients are fixed and stored in the system.

[0031] The specific steps for generating an optimized planting plan are as follows: (1) Based on the regression coefficients obtained from training, interpret the impact of each operation parameter on output: if the regression coefficient of a certain parameter is positive, it indicates that increasing the parameter within a certain range can increase output; if it is negative, it needs to be reduced; the absolute value of the coefficient reflects the weight of the impact.

[0032] (2) Set reasonable agricultural thresholds for each operation parameter, such as the amount of fertilizer should not exceed ±20% of the standard fertilizer amount for crop varieties, the amount of irrigation should be set with upper and lower limits based on soil water holding capacity and rainfall, and the operation time should be limited to the agricultural window of the corresponding growth cycle.

[0033] (3) Fix the current uncontrollable variables (such as meteorological and soil background values), take the maximization of output as the goal, optimize the controllable operation parameters within the threshold constraint, substitute them into the trained regression model to calculate the predicted output, and obtain the optimal parameter combination.

[0034] (4) Compare the optimal parameters with the current actual operating parameters, generate suggestions for increasing or decreasing (such as "reduce fertilizer application by 5%" and "increase irrigation by 8%), and combine them with the cross-cycle historical comparison results to form a planting optimization plan and push it to the user.

[0035] The module is equipped with a dedicated data comparison algorithm that can automatically extract soil environmental data and operational parameters from different periods (such as the seedling stage of the current year versus the same period of the previous year, or different growth cycles of the same plot of land) for difference calculation and generate a difference analysis report. At the same time, it has a built-in multiple linear regression model trained and optimized with a large amount of agricultural production data. After inputting the current operational parameters, soil environmental data, and real-time climate data (obtained through an interface with the local meteorological department's database), the model can quickly analyze the suitability of the operational parameters with the soil and climate, and output a planting optimization plan that includes key parameters such as fertilizer application, watering, and operation time. The plan is tailored to the actual needs of agricultural production and can directly guide agricultural operations.

[0036] During the analysis, line charts and bar charts are automatically generated through data visualization algorithms. Line charts are used to show the changing trend of a single parameter (such as soil moisture or fertilizer application), while bar charts are used to compare data differences between different periods and different plots. This intuitively shows the correlation between operational parameters and crop growth status and yield, making it easier for users to quickly understand the analysis results and providing data support for planting decisions.

[0037] User interaction module: Includes a web-based management platform and a mobile APP terminal. The two complement each other and synchronize data to meet the needs of different user groups. The operation process is simplified to three steps: "view data - compare with history - use the solution", which lowers the operation threshold and makes it easy for managers and farmers to get started quickly.

[0038] The web-based management platform is developed based on a B / S architecture and supports administrators to log in via a browser (compatible with mainstream browsers such as Edge and Firefox). Core functions include: data management (supports batch import and export of Excel 2007 and above formats), system configuration (adjustable data acquisition frequency, ranging from 1-60Hz, with a default setting of 10Hz), threshold alarm (preset data acquisition frequency thresholds; triggers SMS alarms when the acquisition frequency is below 5Hz or above 50Hz, reminding administrators to troubleshoot equipment malfunctions), and user management (allows the creation of user accounts with different permissions, distinguishing between administrator and farmer permissions).

[0039] The mobile app supports Android 8.0 and above, and iOS 12.0 and above. After registering and logging in, farmers can perform three core operations: view real-time data (including crop growth images, soil and air environment data, and equipment operating status), retrieve historical records (supports quick filtering by plot number and crop type code, with a filtering response time of ≤0.8 seconds, better than the requirement of ≤1 second), and receive optimization plans (the system automatically pushes planting optimization plans adapted to the current plot and the current crop growth cycle, and supports offline viewing).

[0040] The web client and mobile app use the WebSocket protocol to achieve real-time data synchronization with a data synchronization delay of ≤0.8 seconds (better than the requirement of ≤1 second). During the synchronization process, a CRC32 data verification mechanism is used to verify the integrity of the transmitted data, avoid data loss or mistransmission, ensure that the data displayed on both ends is completely consistent, and guarantee the accuracy of user operations.

[0041] Example 2: Figure 2 shows a schematic diagram of the communication connection relationship of the data acquisition module provided in an embodiment of the present invention.

[0042] The difference between this embodiment and Embodiment 1 is that, to adapt to scenarios with weak 4G / 5G signals in remote farmland, the data acquisition module uses only LoRa network communication, eliminating the 4G / 5G wireless communication module, reducing equipment costs while ensuring data transmission stability; the data recording and storage module has a single-node storage capacity of 50TB, with the off-site backup node 50km away from the local node, meeting short-term storage needs for 3 years; the user interaction module's mobile APP only supports filtering data by plot number, simplifying the operation process and adapting to the usage habits of elderly farmers. All other technical features are completely consistent with Embodiment 1, achieving the core functions of agricultural data retention and historical traceability, with lower costs and simpler operation.

[0043] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.

Claims

1. An agricultural data retention and historical traceability system based on unmanned inspection, characterized in that, include: The data acquisition module is used to connect to the drone, environmental sensors and unmanned operation vehicle through a wireless communication network to simultaneously collect crop growth images, soil and air environment data, as well as fertilizer application, watering, operation speed parameters, and record equipment power and fault status. The data recording and storage module, connected to the data acquisition module, uses a distributed file system (HDFS) architecture to retain data. It establishes indexes for four data categories—image, environment, operation, and equipment—through a metadata classification engine. Each data entry is associated with a unique identifier field, including acquisition time, plot number, crop type code, and growth cycle stage code. The module is configured to store at least three years of data and uses AES-256 encryption algorithm for data transmission and storage encryption. The encryption key is managed by the hardware security module (HSM). Data backup is performed using a local dual-copy + off-site disaster recovery mode, where the off-site backup node is at least 50km away from the local node. The historical traceability analysis module is connected to the data recording and storage module. It has a built-in data retrieval engine for retrieving historical data by collection time, plot number, crop type and / or growth cycle stage. It performs cross-cycle data difference analysis through data comparison algorithm and analyzes the compatibility of operation parameters with soil environment data and climate data through data analysis model, and outputs a planting optimization plan including parameter adjustment suggestions. The user interaction module, connected to the historical traceability analysis module and the data acquisition module, includes a web client and a mobile app. The web client is used by administrators to manage data and configure the system, while the mobile app is used by farmers to view real-time data, retrieve historical records, and receive planting optimization plans.

2. The system according to claim 1, characterized in that, The drone is equipped with a high-definition camera with a resolution of no less than 4K; the environmental sensor has a soil moisture measurement range of 0-100%RH and a measurement error of ≤5%FS, and a soil temperature measurement range of -40℃ to 85℃ and a measurement error of ≤0.5℃.

3. The system according to claim 1, characterized in that, The data recording and storage module supports TB-level expansion, with a single node storage capacity of ≥50TB; it uses the AES-256 encryption algorithm, with an encryption key update cycle of 7 days, and performs off-site backup operations.

4. The system according to claim 1, characterized in that, The historical traceability analysis module has a response time of ≤3 seconds when comparing cross-cycle data; during the analysis process, line charts and bar charts are generated through data visualization algorithms to intuitively show the correlation between operating parameters and crop growth status and yield.

5. The system according to claim 1, characterized in that, The web interface of the user interaction module supports batch import and export of data in Excel format, supports adjustment of data acquisition frequency with an adjustment range of 1-60Hz, and can preset acquisition frequency threshold alarm; the mobile APP supports quick data filtering by plot number and crop type code with a filtering response time of ≤1 second; the two ends achieve real-time data synchronization through the WebSocket protocol with a data synchronization delay of ≤1 second, and the CRC32 data verification mechanism is used during the synchronization process.

6. The system according to claim 5, characterized in that, The default data acquisition frequency preset on the web client is 10Hz; the preset acquisition frequency threshold alarm is specifically: when the acquisition frequency is lower than 5Hz or higher than 50Hz, an SMS alarm is triggered.

7. The system according to claim 1, characterized in that, The data acquisition module uses only LoRa network for communication and does not include a 4G / 5G wireless communication module; the data recording and storage module has a single node storage capacity of 50TB, and the off-site backup node is 50km away from the local node; the mobile APP of the user interaction module only supports filtering data by plot number.

Images