Digitalized wisdom agriculture planting and breeding combined system

Abstract

This invention relates to the field of smart agriculture planting and breeding technology, and discloses a digital smart agriculture planting and breeding integrated system, including: a multi-source sensing module for collecting environmental data, biological characteristic data, and equipment operation data through sensors in the planting area, breeding area, and recycling equipment; a data verification module for dynamically verifying the collected data based on a pre-set agronomic mechanism model; a hierarchical digital twin module for collaborative decision-making and generating optimization instructions in a three-level architecture of edge computing nodes, local servers, and cloud platforms; a virtual-real decision engine module; and a closed-loop execution network module. In this invention, a lightweight digital twin model is deployed at the edge layer to achieve millisecond-level equipment control, ensuring timely response to extreme conditions; the field layer coordinates the planting and breeding material cycle through dynamic optimization algorithms; and the cloud layer integrates climate data to generate a blockchain-based carbon sink strategy. This three-level collaboration enhances real-time decision-making accuracy, resource utilization efficiency, and carbon economic value.

Description

Technical Field

[0001] This invention relates to the field of smart agriculture planting and breeding technology, and in particular to a digital smart agriculture planting and breeding integrated system. Background Technology

[0002] In the field of smart agriculture technology, the digital collaborative management of integrated crop-livestock systems is a core element for achieving efficient recycling of agricultural resources and precise control of production risks. It is particularly suitable for multi-industry integration scenarios involving livestock and poultry farming, crop cultivation, and organic fertilizer production. The accuracy of multi-source data fusion and the real-time nature of dynamic decision-making directly affect the utilization rate of manure resources, the effectiveness of biosecurity control, and the efficiency of carbon sequestration asset development, placing stringent requirements on data acquisition reliability, decision chain response speed, and execution closed-loop accuracy.

[0003] Currently, the decision-making and execution systems commonly used in the industry rely heavily on centralized server scheduling, with edge nodes only handling data forwarding. This results in delayed equipment control responses, making it impossible to meet the millisecond-level emergency scenarios such as sudden increases in ammonia concentration and sharp drops in dissolved oxygen. Furthermore, carbon sequestration is based on manual recording and processing of data using fixed conversion factors, lacking reliable traceability support across the entire supply chain.

[0004] The aforementioned existing technical solutions have at least the following technical problems: centralized decision-making architecture leads to delayed response to extreme working conditions, static resource models result in reduced nitrogen fertilizer utilization, and carbon sink assets are difficult to circulate in the market due to low measurement reliability. Summary of the Invention

[0005] To address the above shortcomings, this invention provides a digital smart agriculture integrated farming system, which aims to improve and solve the problems of delayed response to extreme operating conditions under centralized decision-making architecture, low nitrogen fertilizer utilization rate caused by static resource models, and difficulty in market circulation of carbon sink assets due to lack of measurement credibility.

[0006] In a first aspect, the present invention provides the following technical solution: a digital intelligent agricultural integrated farming system, comprising:

[0007] The multi-source sensing module is used to collect environmental data, biological characteristic data, and equipment operation data through sensors in the planting area, breeding area, and recycling equipment.

[0008] The data verification module dynamically verifies the collected data based on a pre-set agronomic mechanism model;

[0009] The hierarchical digital twin module makes collaborative decisions and generates optimization instructions in a three-tier architecture of edge computing nodes, local servers, and cloud platforms.

[0010] The virtual-real decision engine module is used to perform sandbox simulation and backtracking retry of decision commands, and output verified commands.

[0011] The closed-loop execution network module converts the verified instructions into device control signals and collects execution effect data to feed back to the multi-source sensing module.

[0012] By adopting the above technical solutions, a multi-source heterogeneous sensor network is deployed to synchronously collect data on planting environment, animal husbandry characteristics, and circulation equipment, constructing a digital foundation for the entire agricultural domain. Based on agronomic mechanism models, dynamic compensation for environmental interference and spatiotemporal consistency verification are performed, outputting a reliable dataset with confidence labels. Relying on a three-level twin architecture of edge, field, and cloud, millisecond-level equipment control, dynamic resource optimization, and blockchain carbon sink strategy generation are achieved. A virtual sandbox simulation and attenuation weight retry mechanism are introduced to intercept high-risk commands. Finally, the execution signal is adaptively converted through the protocol and fed back to the perception layer based on the deviation rate closed loop, thereby enhancing production safety, boosting resource efficiency, and adding value to the carbon economy.

[0013] Preferably, the multi-source sensing module includes:

[0014] The planting environment sensing unit collects data on soil moisture, crop growth, and pest and disease characteristics through soil temperature and humidity sensors, light intensity sensors, leaf surface humidity sensors, and multispectral imagers.

[0015] The livestock vital signs sensing unit collects biological physiological indicators and abnormal activity data through livestock and poultry body temperature monitoring ear tags, respiratory rate sensors, aquatic dissolved oxygen sensors, and behavior analysis cameras.

[0016] The recycling equipment monitoring unit collects operating status parameters of the resource recycling equipment through a biogas flow meter, a current sensor for the manure solid-liquid separator, a pH probe for water treatment, and a temperature probe for organic fertilizer fermentation.

[0017] Preferably, the data verification module includes:

[0018] Environmental coupling compensation unit, used to construct environmental interference matrix ,in, For the first Environmental factors on the first Influence coefficient of sensor type;

[0019] Perform multi-source data collaborative correction ;

[0020] in, For the first Raw values ​​of similar sensors This refers to the change in environmental factors;

[0021] Spatiotemporal consistency unit, which establishes a spatiotemporal distribution model through historical data streams. Detect abnormal deviation points;

[0022] When multiple sensors monitor the same physical quantity, the confidence weights are used to determine the weights. Calculate fusion value ;

[0023] The confidence encapsulation unit calculates the overall confidence level. ;

[0024] when or At that time, embed multi-source conflict markers and add a unified spatiotemporal stamp to the output data packets.

[0025] Preferably, the hierarchical digital twin module includes:

[0026] The edge response unit, deployed on edge computing nodes, is configured to invoke device-level twin models, including livestock and poultry environment control models and irrigation response models, and generate device control commands based on real-time data.

[0027] The field collaboration unit, deployed on a local server, is used to run the crop-livestock coupling twin model and optimize resource allocation paths through the material flow balance equation, specifically calculating the amount of manure returned to the field.

[0028] The strategy generation unit, deployed on a cloud platform, generates global strategies by integrating market, climate, and policy data, calculates carbon sink assets based on specific carbon conversion rates, and generates blockchain-verified carbon sink reports.

[0029] Preferably, the livestock and poultry environment control model in the edge response unit is a PID controller with a proportional gain. Integral coefficient ;

[0030] Material flow balance equation in the field area collaborative unit ;

[0031] in, To dynamically utilize efficiency The average monthly temperature;

[0032] The specific carbon conversion rate in the strategy generation unit And generate a uniquely linked carbon sink report through blockchain.

[0033] Preferably, the virtual-real decision engine module includes:

[0034] The pre-simulation and deduction unit loads the optimized instructions generated by the hierarchical digital twin module into the virtual twin;

[0035] Historical disaster models, including disease transmission routes and the impact of extreme weather, are used to simulate the consequences of command execution, with the rehearsal duration dynamically set. ;

[0036] The backtracking retry unit triggers policy backtracking when the pre-trial shows economic loss >15% or resource waste >20%, employing decaying weights for multiple rounds of retries. Decision weights in the second backtracking Maximum number of retries ;

[0037] The conflict arbitration unit receives multi-source conflict flags embedded in the data verification module. When a conflict flag is triggered or the number of pre-rehearsal failures exceeds 2, it freezes the command and requests manual intervention, and outputs the final verification command to the closed-loop execution network module.

[0038] Preferably, the closed-loop execution network module includes:

[0039] The protocol conversion unit converts the verified instructions into executable signals for industrial equipment and dynamically loads driver plugins based on different agricultural and aquaculture equipment.

[0040] The execution feedback unit collects execution status data in real time and calculates the execution deviation rate. ,in For equipment feedback values, The expected value of the instruction;

[0041] Data closed-loop unit, when In real time, calibration instructions are generated, twin model parameters are updated, and structured feedback data, including spatiotemporal stamps and device IDs, is sent back to the multi-source sensing module.

[0042] Secondly, the present invention provides the following technical solution: a digital intelligent agricultural method for integrating planting and breeding, the method comprising:

[0043] S1. Multi-source data acquisition: Through the sensor clusters in the planting area, breeding area and recycling equipment, environmental data, biological characteristic data and equipment operation data are collected simultaneously.

[0044] S2. Dynamic data verification: Based on a pre-set agronomic mechanism model, cross-domain verification is performed on the collected data, and a reliable dataset with confidence labels is output.

[0045] S3, hierarchical collaborative decision-making, runs a digital twin in a three-tier architecture of edge nodes, local servers, and cloud platforms to generate optimization instructions for planting and breeding resources;

[0046] S4. Command pre-play verification: Perform sandbox simulation and backtracking retry on the optimized command, and output the verified command with controllable risk.

[0047] S5. Closed-loop execution feedback: Convert the verification command into a device control signal, collect the execution effect data, and feed it back to step S1.

[0048] Thirdly, the invention provides the following technical solution: a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-mentioned digital smart agriculture planting and breeding integration method.

[0049] Fourthly, the present invention provides the following technical solution: a readable storage medium storing a computer program, wherein the computer program, when executed by a processor, implements the above-mentioned digital intelligent agricultural planting and breeding integration method.

[0050] The present invention has the following beneficial effects:

[0051] 1. In this invention, a lightweight twin model is deployed at the edge layer to achieve millisecond-level device control, ensuring timely response to extreme working conditions. The field layer coordinates the cycle of planting and breeding materials through dynamic optimization algorithms. The cloud layer integrates climate data to generate a carbon sink strategy with blockchain evidence. The three-level collaboration improves the accuracy of real-time decision-making, the efficiency of resource utilization, and the value of the carbon economy.

[0052] 2. In this invention, multi-source sensors from planting, breeding, and recycling equipment are used to synchronously collect all-element data. An environmental interference matrix is ​​constructed based on an agronomic mechanism model to eliminate cross-domain influences. Verification data with conflict markers is output after spatiotemporal consistency fusion and reliability encapsulation, thereby achieving high-reliability data supply in complex agricultural scenarios.

[0053] 3. In this invention, the risk of instructions is verified by loading historical disaster models through sand table simulation, the strategy is optimized by attenuation weight backtracking mechanism, high-risk operations are frozen by conflict marking, and industrial signals are dynamically converted and deviation rate is calculated during the execution phase to trigger online calibration of twin model. Structured feedback data drives dynamic revision at the sensing end to realize a closed loop of decision-making, execution and calibration. Attached Figure Description

[0054] Figure 1 This is an architecture diagram of the digital intelligent agricultural integrated farming system proposed in this invention;

[0055] Figure 2 This is a flowchart of the digital smart agriculture integrated farming method proposed in this invention. Detailed Implementation

[0056] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0057] Example 1

[0058] Reference Figure 1 In the first embodiment of the present invention, the present invention provides a digital intelligent agricultural integrated farming system, comprising:

[0059] The multi-source sensing module is used to collect environmental data, biological characteristic data, and equipment operation data through sensors in the planting area, breeding area, and recycling equipment.

[0060] The data verification module dynamically verifies the collected data based on a pre-set agronomic mechanism model;

[0061] The hierarchical digital twin module makes collaborative decisions and generates optimization instructions in a three-tier architecture of edge computing nodes, local servers, and cloud platforms.

[0062] The virtual-real decision engine module is used to perform sandbox simulation and backtracking retry of decision commands, and output verified commands.

[0063] The closed-loop execution network module converts the verified instructions into device control signals and collects execution effect data to feed back to the multi-source sensing module.

[0064] Specifically, the multi-source sensing module uses a three-type sensor network deployed in planting areas, breeding areas, and recycling equipment to simultaneously collect environmental parameters (temperature and humidity / light / gas), biological signs (body temperature / respiration / activity), and equipment operation data, constructing a data foundation covering the entire agricultural chain. This provides the system with real-time, multi-dimensional, and heterogeneous raw physical world sensing information, enabling comprehensive digital mapping of agricultural production factors.

[0065] The data verification module dynamically corrects the raw data collected by the multi-source sensing module based on a pre-set agronomic mechanism model, eliminates environmental interference, and transforms the sensing signals into a reliable dataset that conforms to the laws of agricultural science, thus realizing a reliable input from raw physical quantities to agronomic decisions.

[0066] The hierarchical digital twin module deploys interconnected twins in a three-tier architecture of edge layer, field layer and cloud layer, integrates agronomic rules with real-time data, and generates collaborative optimization instructions covering equipment control, planting and breeding resource allocation and carbon sink asset development, realizing an intelligent leap from local response to system-level decision-making;

[0067] The virtual-real decision engine module performs virtual sandbox simulations and multiple rounds of backtracking and retries on the optimized instructions generated by the hierarchical digital twin module. It intercepts high-risk erroneous decisions and outputs reliable instructions that have been verified by risk, realizing the safe transformation from theoretical decisions to executable instructions and providing a high-confidence action plan guarantee for agricultural operations.

[0068] The closed-loop execution network module converts the verified instructions output by the virtual and real decision engine into drive signals that can be recognized by industrial control equipment. At the same time, it collects the actuator response status and physical environment change data in real time, forming a closed-loop control of instructions, execution and feedback, realizing the dynamic calibration cycle of agricultural control strategy from digital space to physical manipulation.

[0069] The multi-source sensing module includes:

[0070] The planting environment sensing unit collects data on soil moisture, crop growth, and pest and disease characteristics through soil temperature and humidity sensors, light intensity sensors, leaf surface humidity sensors, and multispectral imagers.

[0071] The livestock vital signs sensing unit collects biological physiological indicators and abnormal activity data through livestock and poultry body temperature monitoring ear tags, respiratory rate sensors, aquatic dissolved oxygen sensors, and behavior analysis cameras.

[0072] The recycling equipment monitoring unit collects operating status parameters of the resource recycling equipment through a biogas flow meter, a current sensor for the manure solid-liquid separator, a pH probe for water treatment, and a temperature probe for organic fertilizer fermentation.

[0073] Specifically, the planting environment sensing unit monitors the dynamic moisture of the root layer in real time through soil temperature and humidity sensors, captures the photosynthetically active radiation flux through light intensity sensors, provides early warning of fungal disease breeding environments through leaf humidity sensors, and analyzes crop chlorophyll content and stress characteristics through multispectral imagers to construct a holographic profile of crop growth, realizing three-dimensional monitoring from soil moisture to canopy physiology, and providing a data foundation for precision irrigation and disease prevention.

[0074] The livestock vital signs sensing unit uses animal body temperature ear tags to continuously track core body temperature fluctuations, a respiratory rate sensor to diagnose respiratory disease signs, an aquatic dissolved oxygen sensor to ensure the respiratory safety threshold of aquatic organisms, and a behavior analysis camera to capture abnormal activities (such as pecking at the anus and fighting) through posture recognition algorithms, realizing digital health management from individual physiology to group behavior.

[0075] The circulating equipment monitoring unit measures the biogas flow rate to measure the gas production efficiency of manure, the solid-liquid separator current sensor detects bearing wear and overload risk, the water treatment pH probe dynamically regulates the neutralization reaction process, and the organic fertilizer fermentation temperature probe ensures the activity threshold of thermophilic bacteria, thus realizing closed-loop optimization of the process from energy conversion to fertilizer generation.

[0076] The data verification module includes:

[0077] Environmental coupling compensation unit, used to construct environmental interference matrix ,in, For the first Environmental factors on the first Influence coefficient of sensor type;

[0078] Perform multi-source data collaborative correction ;

[0079] in, For the first Raw values ​​of similar sensors This refers to the change in environmental factors;

[0080] Spatiotemporal consistency unit, which establishes a spatiotemporal distribution model through historical data streams. Detect abnormal deviation points;

[0081] When multiple sensors monitor the same physical quantity, the confidence weights are used to determine the weights. Calculate fusion value ;

[0082] The confidence encapsulation unit calculates the overall confidence level. ;

[0083] when or At that time, embed multi-source conflict markers and add a unified spatiotemporal stamp to the output data packets.

[0084] Specifically, the environmental coupling compensation unit constructs an environmental interference matrix to quantify cross-system influencing factors (such as the corrosion deviation of EC sensors by ammonia in livestock and poultry houses), eliminates cross-interference between equipment through dynamic compensation formulas, restores multi-source sensing data to physical true values, and provides a clean data base that conforms to agronomic mechanisms for decision-making.

[0085] The spatiotemporal consistency unit constructs a spatiotemporal probability distribution model based on historical data streams (such as the gradual change of soil moisture with rainfall events), detects abnormal deviation points and automatically de-weights failed sensors. When multiple sensors monitor the same target, the optimal estimate is fused according to confidence weights, solving the problem of random errors in single-point data and ensuring the continuous reliability of data in the time dimension and spatial grid.

[0086] The confidence encapsulation unit calculates the overall confidence level and identifies conflicting data combinations (such as a sudden drop in dissolved oxygen but normal fish activity). For low confidence (ρ<0.7) or significantly deviating data, conflict markers and spatiotemporal stamps are embedded, and structured data packets with quality labels are output to provide risk interception anchors for the subsequent decision engine.

[0087] The hierarchical digital twin module includes:

[0088] The edge response unit, deployed on edge computing nodes, is configured to invoke device-level twin models, including livestock and poultry environment control models and irrigation response models, and generate device control commands based on real-time data.

[0089] The field collaboration unit, deployed on a local server, is used to run the crop-livestock coupling twin model and optimize resource allocation paths through the material flow balance equation, specifically calculating the amount of manure returned to the field.

[0090] The strategy generation unit, deployed on a cloud platform, generates global strategies by integrating market, climate, and policy data, calculates carbon sink assets based on specific carbon conversion rates, and generates blockchain-verified carbon sink reports.

[0091] Specifically, the edge response unit deploys lightweight twin models (such as livestock and poultry ventilation PID controllers) on edge computing nodes, generates control commands based on real-time sensor data (ammonia concentration / soil moisture), and realizes precise regulation of temperature and humidity in livestock sheds and dynamic response of irrigation valve opening, ensuring the autonomous and safe operation of equipment under extreme conditions (such as automatic oxygenation in the absence of oxygen).

[0092] The farm area collaborative unit analyzes the material flow relationship (such as poultry manure nitrogen and phosphorus - crop fertilizer requirements) through the planting and breeding coupling twin, dynamically optimizes the resource allocation path with the minimum supply and demand balance equation, and simultaneously integrates meteorological forecasts and equipment status data to achieve accurate calculation of manure return to the field and feed feeding ratio, thereby improving the resource recycling rate.

[0093] The strategy generation unit integrates futures prices, carbon sink policies, and climate models to generate quarterly production plans and carbon asset development schemes. It calculates the carbon sink value of manure treatment based on the carbon conversion rate (β=0.28) and generates an immutable report that is bound to time and space through blockchain, enabling the resource utilization of agricultural waste to generate tradable carbon credit assets.

[0094] The livestock and poultry environment control model in the edge response unit is a PID controller, and its proportional coefficient is... Integral coefficient ;

[0095] Material flow balance equations in the field collaborative unit ;

[0096] in, To dynamically utilize efficiency The average monthly temperature;

[0097] Specific carbon conversion rate in strategy generation unit And generate a uniquely linked carbon sink report through blockchain.

[0098] Specifically, the edge response unit PID controller uses a proportional coefficient To achieve rapid response to sudden environmental changes (such as a sudden increase in ammonia concentration), the integral coefficient... Eliminate steady-state control deviations, ensure millisecond-level precise adjustment of livestock and poultry house environmental parameters, and guarantee a continuous and stable biological growth environment;

[0099] The material flow balance of the field collaborative unit drives the adaptive allocation of resources across seasons with dynamic efficiency, optimizes the amount of manure returned to the field in combination with the principle of minimum supply and demand, avoids the problem of resource mismatch in the planting and breeding system, and realizes the efficient recycling of nitrogen and phosphorus nutrients.

[0100] Strategy unit carbon sink assetization based on experimentally locked carbon conversion rate Calculate the carbon sequestration value of manure treatment and generate a time- and space-bound tamper-proof report (combining hash value with processing volume, timestamp, and geographic coordinates) through blockchain to create verifiable agricultural carbon sequestration revenue.

[0101] The virtual-real decision engine module includes:

[0102] The pre-simulation and deduction unit loads the optimized instructions generated by the hierarchical digital twin module into the virtual twin;

[0103] Historical disaster models, including disease transmission routes and the impact of extreme weather, are used to simulate the consequences of command execution, with the rehearsal duration dynamically set. ;

[0104] The backtracking retry unit triggers policy backtracking when the pre-trial shows economic loss >15% or resource waste >20%, employing decaying weights for multiple rounds of retries. Decision weights in the second backtracking Maximum number of retries ;

[0105] The conflict arbitration unit receives multi-source conflict flags embedded in the data verification module. When a conflict flag is triggered or the number of pre-rehearsal failures exceeds 2, it freezes the command and requests manual intervention, and outputs the final verification command to the closed-loop execution network module.

[0106] Specifically, the pre-simulation and simulation unit will load optimized instructions into the virtual twin, call historical disaster models (such as the dynamics of avian influenza transmission and the impact of rainstorms and floods) to simulate the execution consequences, dynamically set the pre-simulation duration (intelligently adjusted based on historical loss cycles), quantify the potential economic losses and resource waste risks, and realize the visual prediction of decision consequences to intercept instructions with high misjudgment rates in advance.

[0107] The backtracking retry unit triggers strategy backtracking when the pre-simulation loss exceeds the threshold (economic loss > 15% or resource waste > 20%). It uses exponentially decaying weights to perform multiple rounds of instruction tuning, limits the maximum number of retries (N=3) to avoid invalid calculations, and gradually approaches the optimal solution while retaining the core logic of the original strategy through the weight decay mechanism, thus balancing decision efficiency and reliability.

[0108] The conflict arbitration unit parses the conflict flags embedded in the data verification module in real time. When a flag is triggered or the pre-rehearsal fails continuously, the command stream is frozen, and an alarm is automatically pushed to the human decision-making end to request intervention. Only the verification command that has passed arbitration is output to ensure the security of the execution command in extreme scenarios.

[0109] The closed-loop execution network module includes:

[0110] The protocol conversion unit converts the verified instructions into executable signals for industrial equipment and dynamically loads driver plugins based on different agricultural and aquaculture equipment.

[0111] The execution feedback unit collects execution status data in real time and calculates the execution deviation rate. ,in For equipment feedback values, The expected value of the instruction;

[0112] Data closed-loop unit, when In real time, calibration instructions are generated, twin model parameters are updated, and structured feedback data, including spatiotemporal stamps and device IDs, is sent back to the multi-source sensing module.

[0113] Specifically, the protocol conversion unit dynamically converts the verified instructions into executable signals for industrial control equipment (such as PWM waves), and loads dedicated driver plugins (Modbus / CANopen protocol library) for different agricultural equipment (feeders / aerators / irrigation valves) to eliminate communication between heterogeneous devices and ensure the accurate delivery of control signals in complex agricultural scenarios.

[0114] The execution feedback unit collects actuator response data (such as valve opening and motor speed) in real time, quantifies execution reliability using the deviation rate formula, and identifies mechanical faults. Sudden increase >20% or signal distortion (Continued >10%)

[0115] Data closed-loop unit in >10% or 3 consecutive times When the error rate exceeds 5%, a twin model calibration command is automatically generated, control parameters are dynamically updated, and structured feedback data with time and space stamps and device IDs (including...) is simultaneously transmitted. The baseline values ​​(including environmental interference factors) are sent back to the sensing module, driving the dynamic revision of the sensing end's baseline values.

[0116] Example 2:

[0117] Reference Figure 2 In a second embodiment of the present invention, the present invention provides a digital smart agriculture method for integrating planting and breeding, the method comprising:

[0118] S1. Multi-source data acquisition: Through the sensor clusters in the planting area, breeding area and recycling equipment, environmental data, biological characteristic data and equipment operation data are collected simultaneously.

[0119] S2. Dynamic data verification: Based on a pre-set agronomic mechanism model, cross-domain verification is performed on the collected data, and a reliable dataset with confidence labels is output.

[0120] S3, hierarchical collaborative decision-making, runs a digital twin in a three-tier architecture of edge nodes, local servers, and cloud platforms to generate optimization instructions for planting and breeding resources;

[0121] S4. Command pre-play verification: Perform sandbox simulation and backtracking retry on the optimized command, and output the verified command with controllable risk.

[0122] S5. Closed-loop execution feedback: Convert the verification command into a device control signal, collect the execution effect data, and feed it back to step S1.

[0123] Specifically, environmental parameters (soil moisture / crop growth) of the planting area, biological characteristics (body temperature / respiration rate) of the breeding area, and operating conditions of circulation equipment (current / flow rate) are simultaneously acquired through multi-source data collection steps. Cross-domain compensation and reliability encapsulation are then used to output the data. Tag-based data packets; In hierarchical collaborative decision-making, edge nodes generate equipment instructions through PID controllers, local servers optimize resource allocation according to the material flow balance equation, cloud platforms integrate market climate data to generate carbon sink strategies, instruction pre-simulation verification loads historical disaster models for sand table simulation, when losses > 15%, attenuation weight backtracking is triggered, and instructions are frozen if conflict flags or failure limits are exceeded. Afterwards, closed-loop execution feedback converts instructions into industrial signals and calculates the deviation rate. ,when When the error rate exceeds 10%, the twin model parameters are updated, and structured feedback data is sent back to the acquisition end to calibrate the closed loop, thereby achieving precise regulation and continuous self-optimization of agricultural system resources.

[0124] Example 3

[0125] In the third embodiment of the present invention, based on the same inventive concept, the present invention proposes a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the digital smart agriculture integrated farming method of the above embodiments.

[0126] Example 4

[0127] In the fourth embodiment of the present invention, based on the same inventive concept, the present invention proposes a computer, which includes: a processor and a memory; the processor and the memory communicate with each other; the memory is used to store instructions; the processor is used to execute the instructions in the memory to execute the digital smart agriculture planting and breeding integration method of the above embodiment.

[0128] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0129] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A digital intelligent agricultural integrated farming system, characterized in that, include: The multi-source sensing module is used to collect environmental data, biological characteristic data, and equipment operation data through sensors in the planting area, breeding area, and recycling equipment. The data verification module dynamically verifies the collected data based on a pre-set agronomic mechanism model; The data verification module includes: Environmental coupling compensation unit, used to construct environmental interference matrix ,in, For the first Environmental factors on the first Influence coefficient of sensor type; Perform multi-source data collaborative correction ; in, For the first Raw values ​​of similar sensors This refers to the change in environmental factors; Spatiotemporal consistency unit, which establishes a spatiotemporal distribution model through historical data streams. Detect abnormal deviation points; When multiple sensors monitor the same physical quantity, the confidence weights are used to determine the weights. Calculate fusion value ; The confidence encapsulation unit calculates the overall confidence level. ; when or At the same time, embed multi-source conflict markers and add unified spatiotemporal stamps to the output data packets; The hierarchical digital twin module makes collaborative decisions and generates optimization instructions in a three-tier architecture of edge computing nodes, local servers, and cloud platforms. The virtual-real decision engine module is used to perform sandbox simulation and backtracking retry of decision commands, and output verified commands. The closed-loop execution network module converts the verified instructions into device control signals and collects execution effect data to feed back to the multi-source sensing module.

2. The digital intelligent agricultural integrated farming system according to claim 1, characterized in that, The multi-source sensing module includes: The planting environment sensing unit collects data on soil moisture, crop growth, and pest and disease characteristics through soil temperature and humidity sensors, light intensity sensors, leaf surface humidity sensors, and multispectral imagers. The livestock vital signs sensing unit collects biological physiological indicators and abnormal activity data through livestock and poultry body temperature monitoring ear tags, respiratory rate sensors, aquatic dissolved oxygen sensors, and behavior analysis cameras. The recycling equipment monitoring unit collects operating status parameters of the resource recycling equipment through a biogas flow meter, a current sensor for the manure solid-liquid separator, a pH probe for water treatment, and a temperature probe for organic fertilizer fermentation.

3. The digital intelligent agricultural integrated farming system according to claim 1, characterized in that, The hierarchical digital twin module includes: The edge response unit, deployed on edge computing nodes, is configured to invoke device-level twin models, including livestock and poultry environment control models and irrigation response models, and generate device control commands based on real-time data. The field collaboration unit, deployed on a local server, is used to run the crop-livestock coupling twin model and optimize resource allocation paths through the material flow balance equation, specifically calculating the amount of manure returned to the field. The strategy generation unit, deployed on a cloud platform, generates global strategies by integrating market, climate, and policy data, calculates carbon sink assets based on specific carbon conversion rates, and generates blockchain-verified carbon sink reports.

4. The digital intelligent agricultural integrated farming system according to claim 3, characterized in that, The livestock and poultry environment control model in the edge response unit is a PID controller with a proportional gain. Integral coefficient ; Material flow balance equation in the field area collaborative unit ; in, To dynamically utilize efficiency The average monthly temperature; The specific carbon conversion rate in the strategy generation unit And generate a uniquely linked carbon sink report through blockchain.

5. The digital intelligent agricultural integrated farming system according to claim 1, characterized in that, The virtual-real decision engine module includes: The pre-simulation and deduction unit loads the optimized instructions generated by the hierarchical digital twin module into the virtual twin; Historical disaster models, including disease transmission routes and the impact of extreme weather, are used to simulate the consequences of command execution, with the rehearsal duration dynamically set. ; The backtracking retry unit triggers policy backtracking when the pre-trial shows economic loss >15% or resource waste >20%, employing decaying weights for multiple rounds of retries. Decision weights in the second backtracking Maximum number of retries ; The conflict arbitration unit receives multi-source conflict flags embedded in the data verification module. When a conflict flag is triggered or the number of pre-rehearsal failures exceeds 2, it freezes the command and requests manual intervention, and outputs the final verification command to the closed-loop execution network module.

6. The digital intelligent agricultural integrated farming system according to claim 1, characterized in that, The closed-loop execution network module includes: The protocol conversion unit converts the verified instructions into executable signals for industrial equipment and dynamically loads driver plugins based on different agricultural and aquaculture equipment. The execution feedback unit collects execution status data in real time and calculates the execution deviation rate. ,in For equipment feedback values, The expected value of the instruction; Data closed-loop unit, when In real time, calibration instructions are generated, twin model parameters are updated, and structured feedback data, including spatiotemporal stamps and device IDs, is sent back to the multi-source sensing module.

7. A digitalized smart agriculture method integrating planting and breeding, characterized in that, The method for the digital intelligent agricultural integrated farming system according to any one of claims 1-6 comprises: S1. Multi-source data acquisition: Through the sensor clusters in the planting area, breeding area and recycling equipment, environmental data, biological characteristic data and equipment operation data are collected simultaneously. S2. Dynamic data verification: Based on a pre-set agronomic mechanism model, cross-domain verification is performed on the collected data, and a reliable dataset with confidence labels is output. S3, hierarchical collaborative decision-making, runs a digital twin in a three-tier architecture of edge nodes, local servers, and cloud platforms to generate optimization instructions for planting and breeding resources; S4. Command pre-play verification: Perform sandbox simulation and backtracking retry on the optimized command, and output the verified command with controllable risk. S5. Closed-loop execution feedback: Convert the verification command into a device control signal, collect the execution effect data, and feed it back to step S1.

8. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the digital smart agriculture integrated farming method as described in claim 7.

9. A readable storage medium, characterized in that, The readable storage medium stores a computer program, which, when executed by a processor, implements the digital smart agriculture integrated farming method as described in claim 7.

Images