An adaptive pasture active cooling regulation method and system

Abstract

The application discloses a kind of self-adapting pasture active cooling regulation methods and systems, including obtaining pasture real-time environment, livestock state and equipment operating parameter, transmission to pasture digital twin temperature control platform constructs digital twin model;Call pasture multi-physical field coupling thermal simulation model simulates each regional thermal field distribution;Based on thermal field distribution, determine each regional cooling demand level by fuzzy logic dynamic temperature control algorithm;Call Markov load distribution model to calculate the optimal load distribution scheme;According to scheme, generate active cooling regulation instruction containing equipment power adjustment, operation time length and start sequence;Issue instruction and feedback real-time parameter update model.The system corresponds to the above method setting acquisition, processing, calculation, regulation and control unit.The application is combined by multiple models and digital twin, realizes accurate regulation and control of cooling, overcomes the simulation deviation and regulation rigidity problem of prior art, improves cooling efficiency, reduces energy consumption, and protects the suitable environment of pasture.

Description

Technical Field

[0001] This invention relates to the field of pasture cooling and control technology, and in particular to an adaptive pasture active cooling and control method and system. Background Technology

[0002] With the popularization of large-scale ranching, livestock density on ranches is constantly increasing, and the impact of ambient temperature on livestock growth, health, and production performance is becoming increasingly significant. Traditional ranch cooling relies heavily on manual experience or single-device start-stop control, which is insufficient to cope with the complex thermal field distribution caused by differences in spatial structure, uneven livestock distribution, and changes in the external environment within different areas of the ranch. Simultaneously, the dual demands of ranch operation for energy cost control and cooling efficiency have prompted the industry to seek control solutions that can dynamically adapt to changes in the ranch environment and precisely match cooling needs. The mature application of digital twins, multiphysics simulation, and intelligent algorithms provides technical support for building an active ranch cooling control system that balances accuracy and economy. Therefore, there is an urgent need to develop an adaptive active ranch cooling control method and system that integrates multiple technology modules.

[0003] Existing pasture cooling control technologies have two significant drawbacks: First, existing technologies mostly rely on a single heat conduction model or simple environmental parameter thresholds for cooling decisions, failing to fully integrate the coupling effects of multiple physical fields such as heat conduction, convection, and radiation. This makes it impossible to accurately simulate the complex thermal field distribution within the pasture, resulting in a large discrepancy between the cooling demand assessment and the actual regional thermal environment, easily leading to over-cooling or under-cooling. Second, existing technologies lack dynamic iterative logic in the load allocation of cooling equipment, relying mostly on fixed equipment power or preset sequences for control. They do not dynamically optimize based on the real-time operating status of the equipment and the regional cooling demand level. This not only makes it difficult to achieve a reasonable allocation of loads to each piece of equipment but may also lead to energy waste or poor cooling effects due to mismatches between equipment operating parameters and actual needs. Furthermore, the inability to update control strategies through real-time data feedback makes it difficult to form a closed-loop control system. Summary of the Invention

[0004] In order to overcome the shortcomings and deficiencies of existing technologies, this invention provides an adaptive active cooling control method and system for pastures.

[0005] The technical solution adopted in this invention is an adaptive active cooling control method for pastures, comprising the following steps: S1, acquiring real-time environmental parameters and livestock activity status parameters in different areas of the pasture, transmitting the parameters to a digital twin temperature control platform for the pasture, and establishing a digital twin model consistent with the actual spatial structure of the pasture; S2, based on historical environmental data and cooling control data stored in the digital twin temperature control platform for the pasture, calling a multi-physics coupled thermal simulation model for the pasture to simulate and calculate the heat conduction, heat convection, and heat radiation processes in different areas of the pasture, and outputting the thermal field distribution results for each area; S3, based on the thermal field distribution results output in S2, calling a fuzzy logic dynamic temperature control algorithm to perform fuzzy inference between the thermal field distribution results and a preset temperature control threshold to determine the cooling demand level for each area. Simultaneously, extract the current operating parameters of cooling equipment in each area; S4, call the Markov load allocation model, using the cooling demand level determined in S3 and the current operating parameters of the cooling equipment as input, calculate the load allocation coefficient under different combinations of cooling equipment, and select the optimal load allocation scheme; S5, based on the optimal load allocation scheme selected in S4, generate active cooling control instructions for the ranch, the control instructions including the operating power adjustment parameters, running time parameters, and start-up sequence parameters of each cooling equipment; S6, send the active cooling control instructions generated in S5 to the corresponding cooling equipment in the ranch, and at the same time, feed back the environmental parameters and equipment operating parameters collected in real time during the execution of the control instructions to the ranch digital twin temperature control platform, update the digital twin model, and perform dynamic iteration of the cooling control process.

[0006] Furthermore, the expression for the multiphysics coupled thermal simulation model of the pasture is as follows: ,in, Coordinates within the ranch The temperature value at time t. The air thermal diffusivity of the pasture. For the Laplace operator, The airflow velocity influence coefficient. coordinates The airflow velocity at time t For temperature gradient, The influence coefficient of livestock heat dissipation. coordinates Heat dissipation per unit area of ​​livestock at time t. The heat exchange coefficient of the pasture enclosure structure. coordinates The ambient temperature at the location. Further, the fuzzy inference output expression of the fuzzy logic dynamic temperature control algorithm is: ,in, The cooling requirement level for region i is... Let i be the real-time temperature of the i-th region. The real-time humidity of region i is given. The temperature membership function The membership function is the humidity. The number of temperature-fuzzy subsets. The number of humidity-fuzzy subsets. For the fuzzy rules of temperature and humidity, This is the threshold for the maximum cooling demand level.

[0007] Furthermore, the expression for calculating the load allocation coefficients of the Markov load allocation model is as follows: ,in, for The load distribution coefficient of the j-th cooling device in the i-th region at time i. This is the load adjustment factor. Let be the energy consumption coefficient of the j-th cooling device in the i-th region at time t. Let be the difference between the actual temperature and the target temperature in region i. Let be the load distribution coefficient of the j-th cooling device in the i-th region at time t. Let represent the number of cooling devices in region i.

[0008] Furthermore, the model update expression for the ranch digital twin temperature control platform is: ,in, For the parameters of the time-space digital twin model, The parameters of the digital twin model at time t are... For model update rate coefficients, This refers to the number of environmental parameter collection points. Let be the actual temperature of the i-th sampling point at time t. Let be the simulated temperature at the i-th sampling point at time t. For the gradient of the model parameters, These are the weighting coefficients for equipment parameters. The number of cooling devices, Let be the actual power of the j-th device at time t. Let be the simulated power of the j-th device at time t. Further, the calculation expression for the operating power adjustment parameters of the active cooling control system in the pasture is: ,in, This is the power adjustment value for the j-th cooling device. The reference power for the j-th cooling device is... This is the power adjustment coefficient. The cooling requirement level for region i is... This represents the average cooling requirement level for pastures. Let i be the real-time temperature of the i-th region. For the target temperature, This is the highest temperature ever recorded on the ranch. This is the lowest temperature ever recorded on the ranch.

[0009] Further, step S3 includes the following sub-steps: S31, extracting the thermal field distribution data of each region output by S2 from the digital twin temperature control platform database of the ranch. This data includes temperature values, temperature change rates, and heat flux density values ​​under different spatial coordinates. The extracted data is classified and stored according to the regional division rules; S32, calling the membership function library of the fuzzy logic dynamic temperature control algorithm, selecting the corresponding temperature membership function and humidity membership function according to the livestock species and density in different regions of the ranch, and determining the boundary threshold of each fuzzy subset; S33, inputting the classified thermal field distribution data and the real-time collected humidity data into the selected membership function, calculating the membership values ​​corresponding to different parameters, and then performing fuzzy synthesis operation on the membership values ​​according to the preset fuzzy rule library; S34, based on the fuzzy synthesis operation results, converting the fuzzy quantity into an accurate cooling demand level value through a defuzzification algorithm, and simultaneously recording the inference process data corresponding to the cooling demand level of each region, storing it in the algorithm operation log.

[0010] Further, S4 includes the following sub-steps: S41, collecting real-time operating parameters of all cooling equipment in the pasture, including current power, running time, fault warning status, and energy consumption data, and matching the collected parameters with the cooling demand level output in S3; S42, calling the state transition matrix of the Markov load allocation model, and calculating the state transition probability of each equipment under different load allocation ratios based on the historical operating data of the equipment and the current matching results; S43, constructing a load allocation optimization model with the objective functions of minimizing equipment energy consumption and optimizing cooling effect, and inputting the state transition probability as a constraint condition into the optimization model for multi-objective optimization calculation; S44, evaluating the multiple load allocation schemes obtained from the optimization calculation, and selecting the scheme with the highest comprehensive benefit value as the optimal load allocation scheme by calculating the comprehensive benefit value of the schemes, and outputting it to the next step.

[0011] Further, S5 includes the following sub-steps: S51, parse the optimal load allocation scheme output by S4, extract the load allocation coefficient, target operating parameters, and control priority of each cooling device in the scheme, and group them according to the area where the devices are located; S52, based on the grouped device information and combined with the real-time power supply data of the ranch, calculate the total power demand when the cooling devices in each area are started simultaneously, and determine whether the total power demand exceeds the power supply threshold; S53, if the total power demand does not exceed the threshold, directly generate the operating power adjustment parameters, running time parameters, and start-up sequence parameters of each device according to the load allocation coefficient; if the total power demand exceeds the threshold, adjust the operating parameters of high-priority devices, reduce the number of low-priority devices started, and recalculate the total power demand until the power supply constraint is met; S54, encapsulate the generated control parameters according to the preset data format to form a standardized active cooling control command, with an instruction generation timestamp and device identification information attached.

[0012] An adaptive active cooling control system for pastures, applied to an adaptive active cooling control method for pastures, includes: a pasture multi-parameter acquisition unit, which consists of temperature sensors, humidity sensors, livestock activity monitors, and equipment operation monitoring modules distributed in different areas of the pasture. Each sensor and monitoring module is connected via a wireless communication link to acquire pasture environmental parameters, livestock status parameters, and equipment operation parameters, and transmits the acquired data to a data preprocessing unit; a data preprocessing unit, connected to the pasture multi-parameter acquisition unit via Ethernet, which incorporates a data filtering module and an outlier removal module to filter noise and identify abnormal data in the acquired data, and transmits the processed data to a pasture digital twin temperature control platform unit; and a pasture digital twin temperature control platform unit, connected to the data preprocessing unit and the model calculation unit via industrial Ethernet, which incorporates a digital twin modeling module, a data storage module, and a model update module to construct a pasture digital twin model, store historical and real-time data, and receive model calculations. The system includes: a unit that receives data feedback and updates the digital twin model; a model calculation unit, connected to the ranch digital twin temperature control platform unit via a high-speed data bus, integrating a ranch multi-physics coupled thermal simulation model calculation module, a fuzzy logic dynamic temperature control algorithm running module, and a Markov load distribution model solving module; a control command generation unit, connected to the model calculation unit and equipment control unit via a wireless communication module, with built-in command parsing and parameter configuration modules; a unit that receives the optimal load distribution scheme output by the model calculation unit, generates standardized active cooling control commands, and sends them to the equipment control unit; and an equipment control unit, connected to the control command generation unit and ranch cooling equipment via power line carrier communication, with built-in command execution and status feedback modules; a unit that receives control commands and controls the operation of the cooling equipment, collects equipment operating status data, and feeds it back to the ranch digital twin temperature control platform unit for closed-loop system control.

[0013] Beneficial Effects: This invention proposes an adaptive active cooling control method and system for pastures. By integrating a multi-physics coupled thermal simulation model of the pasture, a fuzzy logic dynamic temperature control algorithm, a Markov load distribution model, and a digital twin temperature control platform for the pasture, the multi-physics coupled thermal simulation model comprehensively simulates the heat conduction, convection, and radiation processes within the pasture, accurately outputting the thermal field distribution in each area. This effectively solves the problem that existing technologies, with their single models, cannot accurately simulate complex thermal fields, leading to biased judgments of cooling demand. Furthermore, the fuzzy logic dynamic temperature control algorithm performs fuzzy inference on the thermal field distribution and temperature control thresholds, combined with Markov load... The allocation model dynamically calculates the optimal load allocation scheme for equipment, achieving a precise match between cooling demand and equipment operation. This overcomes the shortcomings of existing technologies that rely on fixed parameter control and lack dynamic iterative logic, thus avoiding energy waste and poor cooling effects. Simultaneously, a closed-loop control system is constructed through a digital twin temperature control platform for the ranch, updating the model in real time and feeding back equipment operation data. This ensures that the control continuously adapts to changes in the ranch environment, improving the accuracy and efficiency of cooling control, ensuring the stability of the environment required for livestock growth, reducing ranch energy costs, and balancing breeding efficiency and operational economy, thereby comprehensively optimizing the ranch cooling control system. Attached Figure Description

[0014] Figure 1 This is a flowchart of the method steps of the present invention; Figure 2 This is a diagram showing the system unit composition of the present invention. Detailed Implementation

[0015] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. The application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0016] like Figure 1 As shown, an adaptive active cooling control method for pastures includes the following steps: S1. Acquire real-time environmental parameters and livestock activity status parameters in different areas of the pasture, and transmit these parameters to the pasture's digital twin temperature control platform to establish a digital twin model consistent with the actual pasture spatial structure. Specifically, step S1 completes the basic data collection and digital twin model initialization of the pasture. In practice, temperature and humidity sensors are first arranged in the pasture at 5m x 5m intervals. One livestock activity monitor is set up for every 100 square meters of livestock pen area. At the same time, each cooling device (such as a fan or sprinkler) is equipped with an operation monitoring module. The acquisition frequency of the sensors and monitors is set to 1 time / minute to ensure real-time data capture. The collected parameters include: temperature (measurement range -10℃ to 50℃, accuracy ±0.5℃), humidity (measurement range 20%RH to 95%RH, accuracy ±3%RH), livestock activity frequency (number of times livestock move per unit time, statistical accuracy ±1 time / 10 minutes), current power of cooling equipment (measurement range 0kW to 15kW, accuracy ±0.1kW), and operating time (statistical accuracy ±1 minute). After data acquisition, the data is transmitted to the data preprocessing module via a LoRa wireless communication link (transmission rate 125kbps to 500kbps, transmission distance ≤3km). After filtering, the data is sent to the ranch digital twin temperature control platform. Based on the actual CAD drawings of the ranch, the platform constructs a digital twin model at a 1:1 scale, including the spatial structure, equipment location, and livestock distribution area. The acquired data is mapped to the corresponding location in the model, completing the model initialization and providing basic data support for subsequent simulation calculations. This step ensures the consistency between the digital twin model and the actual ranch environment, avoiding the impact of missing data or model deviation on the accuracy of subsequent control.

[0017] S2. Based on the historical environmental data and cooling regulation data stored in the ranch digital twin temperature control platform, the ranch multi-physics field coupled thermal simulation model is called to simulate and calculate the heat conduction, heat convection and heat radiation processes in different areas of the ranch, and output the thermal field distribution results of each area. Specifically, in step S2, the ranch multi-physics field coupled thermal simulation model is called and the thermal field distribution is calculated. During implementation, the ranch digital twin temperature control platform first extracts the historical environmental data of the past 30 days (including hourly temperature, humidity, and outside wind speed) and the equipment operation data of the past 10 cooling regulation (equipment power, running time, and cooling area) from the database, and uses these data as model input parameters. The multiphysics coupled thermal simulation model includes three core physical processes: heat conduction, heat convection, and heat radiation. The heat conduction calculation is based on the thermophysical properties of pasture building materials (such as the thermal conductivity of wall insulation materials of 0.03 W / (m·K) and the thermal conductivity of ground concrete of 1.74 W / (m·K)). The heat convection calculation is combined with the airflow pattern in the pasture (the airflow velocity is set to 0.5 m / s to 3 m / s when the fan is on and 0.1 m / s to 0.3 m / s when natural ventilation is used). The heat radiation calculation considers the influence of solar radiation intensity (set to 1000 W / m² at noon in summer and 600 W / m² at noon in winter) and the heat dissipation of livestock themselves (100 W to 150 W of heat dissipation power per adult animal and 50 W to 80 W of heat dissipation power per young animal). During model execution, simulations are performed region-by-region across 100 sub-regions (each region is 10m x 10m) within the pasture, with a time step of 10 minutes. The simulation outputs the temperature distribution, heat flux density (W / m²), and temperature change rate (°C / 10 minutes) for each region at the current moment and every 10 minutes for the next hour, and displays these values ​​as a heat map on the digital twin platform. This step, through multiphysics coupling analysis, accurately recreates the complex thermal environment within the pasture, avoiding thermal field distribution errors caused by simulations of single physical processes, and providing a scientific basis for subsequent assessment of cooling needs.

[0018] S3. Based on the thermal field distribution results output by S2, call the fuzzy logic dynamic temperature control algorithm to perform fuzzy reasoning between the thermal field distribution results and the preset temperature control threshold to determine the cooling demand level of each area, and at the same time extract the current operating parameters of the cooling equipment in each area. Specifically, step S3 uses a fuzzy logic dynamic temperature control algorithm to determine the cooling demand level for each area. During implementation, the thermal field distribution data of 100 areas output from step S2 is first extracted from the ranch's digital twin temperature control platform. The real-time temperature, temperature change rate, and heat flux density of each area are carefully selected, and real-time humidity data for the corresponding area is collected. These parameters are used as input variables for the fuzzy logic algorithm. The algorithm first fuzzifies the input variables, setting the fuzzy subsets of temperature as "low temperature," "medium temperature," "high temperature," and "extremely high temperature," corresponding to temperature ranges of ≤20℃, 20℃ to 28℃, 28℃ to 35℃, and >35℃, respectively; the fuzzy subsets of humidity are "low humidity," "medium humidity," and "high humidity," corresponding to ranges of ≤40%RH, 40%RH to 70%RH, and >70%RH, respectively; and the fuzzy subsets of temperature change rate are "slow," "medium," and "fast," corresponding to ranges of ≤0.2℃ / 10 minutes, 0.2℃ / 10 minutes to 0.5℃ / 10 minutes, and >0.5℃ / 10 minutes, respectively. Next, a pre-defined fuzzy rule library (including 81 rules, such as "high temperature and high humidity with rapid temperature rise → extremely high cooling demand" and "medium temperature and medium humidity with slow temperature change → medium cooling demand") is invoked to perform logical reasoning on the fuzzified variables, obtaining the fuzzy reasoning results for each region. Then, the centroid method is used to defuzzify the results, converting them into precise cooling demand levels from 0 to 10 (0 for no demand, 10 for the highest demand). Simultaneously, the current operating parameters of the cooling equipment in each region (such as the current fan speed of 1200 r / min and the current water pressure of the sprinkler system of 0.3 MPa) are extracted and stored in the algorithm result database along with the cooling demand levels. This step, through multivariate fuzzy reasoning, comprehensively considers the correlation of pasture environmental parameters, avoiding the limitations of single-parameter threshold judgments and ensuring that the cooling demand levels match the actual environment.

[0019] S4. Call the Markov load allocation model, take the cooling demand level determined in S3 and the current operating parameters of the cooling equipment as input, calculate the load allocation coefficient under different combinations of cooling equipment, and select the optimal load allocation scheme. Specifically, step S4 uses a Markov load allocation model to select the optimal load allocation scheme. During implementation, the cooling demand levels (0 to 10) for 100 regions and the real-time operating parameters of the corresponding cooling equipment are extracted from the database of results from step S3. These parameters include the current load rate of the equipment (the ratio of current power to rated power, ranging from 0 to 100%), the average number of fault warnings in the past hour (ranging from 0 to 5), and the energy consumption per unit time (ranging from 0.5 kWh / h to 2 kWh / h). These parameters are used as input state variables for the Markov model. The model first constructs a state transition matrix. The rows of the matrix represent the current operating state of the equipment (divided by load rate into "low load," "medium load," and "high load," corresponding to ranges of 0 to 30%, 30% to 70%, and 70% to 100%), and the columns represent the possible operating states at the next moment. The transition probabilities between different states are calculated using historical data (e.g., the probability of transitioning from low load to medium load is 0.6, and the probability of transitioning from medium load to high load is 0.3). Next, with constraints such as "regional cooling demand satisfaction rate ≥ 95%", "total equipment energy consumption ≤ regional rated energy consumption (set to 50 kWh / h for each region)" and "equipment failure risk rate ≤ 2%", the load allocation coefficient (the proportion of the total regional cooling load borne by each device, ranging from 0 to 1) is calculated for different combinations of cooling equipment (e.g., single fan operation, fan + spray combined operation). During model operation, each combination is iterated 10 times with a 5-minute iteration cycle. The optimal load allocation scheme is selected based on the most balanced load allocation coefficient, which meets the constraints and has the lowest overall energy consumption. The output scheme includes the target load rate, suggested operating time, and start-up priority for each device. This step, through dynamic state transition analysis and multi-constraint optimization, avoids insufficient local cooling or excessive energy consumption caused by uneven equipment load, thus improving the operational stability of the cooling system.

[0020] S5. Based on the optimal load allocation scheme selected in S4, generate active cooling control instructions for the pasture. The control instructions include the operating power adjustment parameters, running time parameters, and start-up sequence parameters of each cooling device. Specifically, step S5 generates active cooling control commands based on the optimal load allocation scheme. During implementation, the optimal scheme output from step S4 is first read through the data parsing module. The target load rate (e.g., 80% for fans, 60% for sprinkler devices), runtime (e.g., 30 minutes for fans, 20 minutes for sprinkler devices), and start-up priority (divided into 1 to 3 levels, with level 1 being the highest) of each cooling device in each area are extracted. The devices are then grouped according to the hierarchical structure of "area-equipment" (e.g., Fan No. 1 in the East Area and Sprinkler No. 2 in the East Area are in one group). Next, the real-time monitoring data of the ranch's power supply is called to obtain the total power supply of the current area (each area is set to 80kW). The total power demand when each group of devices starts simultaneously is calculated (e.g., the rated power of the fan is 15kW, and 12kW corresponds to 80% load rate; the rated power of the sprinkler device is 10kW, and 6kW corresponds to 60% load rate, with a total demand of 18kW within the group). It is then determined whether the total demand exceeds 80% of the area's power supply (i.e., 64kW). If it does not exceed this limit, the control parameters are directly generated. If the total power demand of equipment in a certain area exceeds the threshold (e.g., calculated to be 70kW), adjustments are made according to the startup priority: the target load rate of level 3 equipment is reduced (e.g., from 60% to 40%, power from 6kW to 4kW), or the number of level 2 equipment startups is reduced (e.g., from 2 to 1), and the total power demand is recalculated (adjusted to 62kW), until the constraint of "total demand ≤ 64kW" is met. After the parameters are determined, the equipment operating power adjustment parameters (e.g., the fan is adjusted from the current 10kW to 12kW, and the sprinkler device is adjusted from the current 5kW to 6kW), the running time parameters (accurate to the minute, e.g., the fan runs from 12:00 to 12:30), and the startup sequence parameters (e.g., the level 1 fan is started first, and the level 2 sprinkler device is started 5 minutes later) are encapsulated according to a standardized protocol (e.g., Modbus-RTU protocol), and a unique identifier (including area number, equipment number, and generation timestamp) is added to each instruction to ensure that the instructions are traceable. Finally, the encapsulated parameters are logically verified by the command verification module (e.g., power adjustment range ≤ ±50% of rated power, running time ≤ 2 hours). After successful verification, the final active cooling control command is generated and stored in the command sending queue. This step, through power constraint verification and dynamic parameter adjustment, can avoid power overload caused by equipment startup, while ensuring the accuracy and executability of the control command.

[0021] S6. Send the active cooling control command generated in S5 to the corresponding cooling equipment in the ranch. At the same time, feed back the environmental parameters and equipment operating parameters collected in real time during the execution of the control command to the ranch digital twin temperature control platform, update the digital twin model, and carry out dynamic iteration of the cooling control process.

[0022] Specifically, step S6 completes the issuance of control commands and dynamic model updates. During implementation, the active cooling control command is first read from the queue through the command sending module. Based on the equipment communication protocol (RS485 protocol for fans and LoRa protocol for spray devices), the corresponding transmission link is selected, and the command is sent to the controller of the corresponding cooling equipment in the ranch (the transmission delay is set to ≤100ms to ensure real-time command performance). After receiving the command, the equipment controller adjusts the equipment operating status according to the parameters (e.g., the fan speed is adjusted from 1200r / min to 1440r / min to achieve 12kW power, and the valve opening of the spray device is adjusted from 50% to 60% to achieve 6kW power). At the same time, the status feedback module is activated to collect the actual operating parameters of the equipment (such as real-time power, speed, and valve opening) and regional environmental parameters (such as temperature and humidity changes) at a frequency of once every 2 minutes, and feeds them back to the ranch digital twin temperature control platform through the original transmission link. After receiving feedback data, the platform invokes the model update module to compare the actual parameters with the simulation parameters of the digital twin model (e.g., the model predicts a regional temperature of 28℃ in 12 minutes, while the actual feedback is 27.8℃). It calculates parameter deviations (e.g., temperature deviation -0.2℃, power deviation 0.3kW). If the deviation is within the allowable range (temperature deviation ±0.5℃, power deviation ±0.5kW), only the real-time data layer of the model is updated. If the deviation exceeds the range (e.g., temperature deviation 0.8℃), model parameter correction is triggered (e.g., adjusting the airflow velocity influence coefficient of the thermal simulation model), the model is rerun, and the simulation layer of the digital twin model is updated. Simultaneously, a complete control log is generated during the command execution process, recording "command issuance time - equipment response time - parameter adjustment result - feedback data," and stored in the historical database. This provides data support for subsequent model optimization and algorithm iteration. This step, through real-time feedback and dynamic model updates, constructs a closed-loop system for cooling control, ensuring that the control effect continuously adapts to changes in the pasture environment and avoiding control lag caused by static models. Preferably, the expression for the multiphysics coupled thermal simulation model of the pasture is: ,in, Coordinates within the ranch The temperature value at time t. The air thermal diffusivity of the pasture. For the Laplace operator, The airflow velocity influence coefficient. coordinates The airflow velocity at time t For temperature gradient, The influence coefficient of livestock heat dissipation. coordinates Heat dissipation per unit area of ​​livestock at time t. The heat exchange coefficient of the pasture enclosure structure. coordinates The ambient temperature at the location.

[0023] Specifically, the multiphysics coupled thermal simulation model for pastures is used to accurately calculate the temperature distribution at different spatial locations within the pasture at different times. During implementation, it is necessary to first determine the numerical range and basis for the various technical parameters involved in the model. Among these, the pasture air thermal diffusivity coefficient needs to be calculated based on the air density (usually taken as 1.2 kg / m³) and specific heat capacity (taken as 1005 J / (kg・K)) within the pasture, with a value ranging from 0.02 W / (m・K) to 0.03 W / (m・K); the airflow velocity influence coefficient needs to be determined in conjunction with the type of pasture fan (axial flow fan or centrifugal fan), with the coefficient for axial flow fans ranging from 0.8 to 1.2, and for centrifugal fans from 1.0 to 1.5; the influence of livestock heat dissipation... The coefficients need to be adjusted according to the livestock breed (dairy cows, beef cattle, etc.) and growth stage. The coefficients for adult dairy cows are 0.6 to 0.9, and for calves, they are 0.3 to 0.5. The heat exchange coefficients for the pasture enclosure structure are set according to the type of insulation material used for the walls and roof (such as polyurethane insulation boards and rock wool boards). The coefficients for polyurethane insulation boards are 0.3 W / (m²・K) to 0.5 W / (m²・K), and for rock wool boards, they are 0.4 W / (m²・K) to 0.6 W / (m²・K). The ambient temperature needs to be collected in real time by a weather station set up around the pasture, with data recorded and updated to the model every hour. During implementation, the pasture is first divided into several three-dimensional grid units (each unit has a side length of 2 meters). Then, the above parameters are input into the model, and the temperature values ​​of different grid units at different times are solved by numerical calculation methods. The output results are recorded in tabular form to record the temperature data of each grid unit, providing a quantitative basis for subsequent judgment of the thermal environment status of each area. The application of this model can avoid temperature calculation deviations caused by the analysis of a single physical factor, ensuring that the simulation results of thermal field distribution are highly consistent with the actual pasture environment.

[0024] Preferably, the fuzzy inference output expression of the fuzzy logic dynamic temperature control algorithm is: ,in, The cooling requirement level for region i is... Let i be the real-time temperature of the i-th region. The real-time humidity of region i is given. The temperature membership function The membership function is the humidity. The number of temperature-fuzzy subsets. The number of humidity-fuzzy subsets. For the fuzzy rules of temperature and humidity, This represents the maximum cooling demand level threshold. Specifically, the fuzzy inference output process of the fuzzy logic dynamic temperature control algorithm determines the cooling demand level for each area through multi-parameter membership calculation and rule reasoning. During implementation, the parameter settings and inference process involved in the algorithm must be clearly defined. First, the number and corresponding ranges of temperature and humidity fuzzy subsets are determined. The number of temperature fuzzy subsets is set to four, with each range divided according to the suitable temperature range for livestock growth in the pasture (18℃ to 25℃). Below 20℃ is "low temperature," 20℃ to 28℃ is "medium temperature," 28℃ to 35℃ is "high temperature," and above 35℃ is "extreme high temperature." The number of humidity fuzzy subsets is set to three, based on the livestock's tolerance range for humidity (45%RH to 65%RH). Below 40%RH is "low humidity," 40%RH to 70%RH is "medium humidity," and above 70%RH is "high humidity." The weights of the temperature-humidity fuzzy rules need to be derived from historical cooling control data. For example, the rule weight corresponding to the combination of "high temperature and high humidity" is 0.9, "medium temperature and medium humidity" is 0.5, and "low temperature and low humidity" is 0.1, forming a complete rule base of 81 rules. The maximum cooling demand level threshold is set to 10, representing the highest cooling demand. During implementation, real-time temperature and humidity data of each region are first collected. The membership degree value (range 0 to 1) of each parameter belonging to the corresponding fuzzy subset is calculated using the membership degree function. Then, matching fuzzy rules are found according to the rule base, and the corresponding rule weight is multiplied by the membership degree value and summed to obtain the fuzzy inference results of each region. Finally, the fuzzy results are converted into precise cooling demand levels of 0 to 10 using the centroid method. The implementation of this algorithm can comprehensively consider the synergistic effect of temperature and humidity on cooling demand, avoiding misjudgment of cooling demand caused by judging based on only a single parameter, and ensuring that the cooling demand level of each region can accurately reflect the urgency of cooling in the actual environment. Preferably, the expression for calculating the load allocation coefficient of the Markov load allocation model is: ,in, for The load distribution coefficient of the j-th cooling device in the i-th region at time i. This is the load adjustment factor. Let be the energy consumption coefficient of the j-th cooling device in the i-th region at time t. Let be the difference between the actual temperature and the target temperature in region i. Let be the load distribution coefficient of the j-th cooling device in the i-th region at time t. The number of cooling devices for region i is specified. Specifically, the load allocation coefficient of the Markov load allocation model is calculated by determining the optimal load allocation scheme through equipment state transition probability analysis and multi-constraint calculation. During implementation, the model parameter settings and calculation process must be clearly defined. The load adjustment coefficient needs to be set according to the aging level of the ranch equipment and energy consumption control targets. The coefficient for new equipment is 0.6 to 0.8, and for equipment used for more than 3 years, it is 0.8 to 1.0. The equipment energy consumption coefficient is obtained by calibrating the equipment nameplate parameters and actual operating data. The energy consumption coefficient for fans ranges from 0.02 kWh / (kW·h) to 0.05 kWh / (kW·h), and for sprinkler systems, it is 0.03 kWh / (kW·h) to 0.06 kWh / (kW·h). The difference between the actual temperature and the target temperature of the region needs to be determined according to the suitable temperature for different livestock breeds. The target temperature for dairy cows is set at 22℃, and for beef cattle at 20℃. The difference is obtained by subtracting the target temperature from the real-time temperature. The number of devices is determined based on the area, with 4 to 6 cooling devices configured for every 500 square meters. During implementation, the current load distribution coefficient and real-time operating data of each device are first collected. The state transition probability of different devices in their current state is calculated (e.g., the probability of transitioning from a low-load state to a medium-load state is 0.6). Then, parameters such as energy consumption coefficient and temperature difference are input into the model. The load distribution coefficient of different device combinations is calculated using an exponential function, and the final coefficient value is obtained through summation and normalization. Simultaneously, with constraints such as a regional cooling demand satisfaction rate of no less than 95% and total equipment energy consumption not exceeding the regional rated energy consumption (50 kWh / h), the scheme with the most balanced load distribution coefficient is selected. The application of this model can dynamically adjust the equipment load distribution, avoiding overload operation of some devices or idle operation of others, improving equipment operating efficiency and lifespan, while reducing overall energy consumption. Preferably, the model update expression of the ranch digital twin temperature control platform is: ,in, for Time-based digital twin model parameters The parameters of the digital twin model at time t are... For model update rate coefficients, This refers to the number of environmental parameter collection points. Let be the actual temperature of the i-th sampling point at time t. Let be the simulated temperature at the i-th sampling point at time t. For the gradient of the model parameters, These are the weighting coefficients for equipment parameters. The number of cooling devices, Let be the actual power of the j-th device at time t. Let t represent the simulated power of the j-th device. Specifically, the model update process of the ranch digital twin temperature control platform adjusts model parameters through deviation analysis between actual and simulated data to ensure that the digital twin model is synchronized with the actual ranch environment. During implementation, the update parameter settings and update process must be clearly defined. The model update rate coefficient is set according to the model accuracy requirements, ranging from 0.01 to 0.05, with a larger value used for higher accuracy requirements. The number of environmental parameter collection points is determined based on the ranch area, with 8 to 10 collection points set per 1000 square meters to ensure coverage of all areas of the ranch. The equipment parameter weight coefficient needs to be set according to the degree of influence of the equipment on the cooling effect, with a fan weight coefficient of 0.6 and a sprinkler weight coefficient of 0.4. The number of cooling devices remains consistent with weight 4, with 4 to 6 units configured per 500 square meters. During implementation, actual temperature data and actual power data of equipment in each area are first acquired through data collection points. Simultaneously, simulated temperature and power data for corresponding locations are extracted from the digital twin model. The squared difference between the actual and simulated data is calculated, and then multiplied by the model parameter gradient, update rate coefficient, and equipment parameter weight coefficient to obtain the model parameter adjustment amount. This adjustment amount is added to the current model parameters to obtain the updated model parameters. The update frequency is set to once every 30 minutes. After each update, the model is rerun to generate new simulation data and compare it with subsequent actual data. This update process can correct the deviation between the model and the actual environment in real time, avoiding the lag in simulation results caused by fixed model parameters. This ensures that the digital twin model can accurately reflect the dynamic changes in the pasture environment and equipment operating status, providing reliable model support for subsequent control decisions. Preferably, the calculation expression for the operating power adjustment parameters of the pasture active cooling control is: ,in, This is the power adjustment value for the j-th cooling device. The reference power for the j-th cooling device is... This is the power adjustment coefficient. The cooling requirement level for region i is... This represents the average cooling requirement level for pastures. Let i be the real-time temperature of the i-th region. For the target temperature, This is the highest temperature ever recorded on the ranch. This is the lowest temperature ever recorded on the ranch.

[0025] Specifically, the calculation of operating power adjustment parameters for active cooling control in ranches is based on determining the equipment power adjustment value according to the cooling demand level and temperature deviation. The parameter settings and calculation logic must be clearly defined during implementation. The power adjustment coefficient needs to be set according to the equipment's adjustment accuracy; the fan power adjustment coefficient is 0.3 to 0.5, and the sprinkler system's is 0.4 to 0.6. The equipment's base power is 50% of its rated power. For example, a fan's rated power is typically 15kW, and its base power is 7.5kW; a sprinkler system's rated power is 10kW, and its base power is 5kW. The average cooling demand level of the ranch is obtained by averaging the cooling demand levels of all areas, with a value ranging from 0 to 10. The target temperature is consistent with weight 4, set according to livestock breed: 22℃ for dairy cows and 20℃ for beef cattle. The ranch's historical highest and lowest temperatures are obtained by querying a historical database; the highest temperature is typically 40℃, and the lowest temperature is -5℃. During implementation, the real-time temperature, cooling demand level, and average cooling demand level of each region are first obtained. The difference between the real-time temperature and the target temperature, and the difference between the historical highest and lowest temperatures are calculated. Then, the power adjustment coefficient, the difference between the cooling demand level and the average cooling demand level, and the temperature difference are substituted into the calculation formula. First, the ratio of the cooling demand level difference to the temperature difference is calculated. Then, the ratio is multiplied by the power adjustment coefficient to obtain the power adjustment ratio. Finally, the ratio is added to the base power to obtain the equipment power adjustment value. The calculation results need to be range-checked to ensure that the adjustment value is not lower than 30% of the equipment's rated power and not higher than 120% of the rated power, avoiding insufficient cooling due to too low power or energy waste due to too high power. This parameter calculation method can dynamically adjust the equipment power according to the actual cooling demand of the region, ensuring that the equipment operating status is accurately matched with the cooling demand, improving cooling efficiency while controlling energy consumption costs.

[0026] Preferably, step S3 includes the following sub-steps: S31, extracting the thermal field distribution data of each region output in S2 from the digital twin temperature control platform database of the ranch. This data includes temperature values, temperature change rates, and heat flux density values ​​under different spatial coordinates. The extracted data is classified and stored according to the regional division rules; S32, calling the membership function library of the fuzzy logic dynamic temperature control algorithm, selecting the corresponding temperature membership function and humidity membership function according to the livestock species and density in different regions of the ranch, and determining the boundary threshold of each fuzzy subset; S33, inputting the classified thermal field distribution data and the real-time collected humidity data into the selected membership function, calculating the membership values ​​corresponding to different parameters, and then performing fuzzy synthesis operation on the membership values ​​according to the preset fuzzy rule library; S34, based on the fuzzy synthesis operation results, converting the fuzzy quantity into an accurate cooling demand level value through a defuzzification algorithm, and recording the inference process data corresponding to the cooling demand level of each region, storing it in the algorithm operation log.

[0027] Specifically, the implementation process of step S3 involves multiple steps to apply the fuzzy logic dynamic temperature control algorithm. Each step requires clear parameter settings and operating procedures. During the implementation of S31, the thermal field distribution data output from S2 is extracted from the farm's digital twin temperature control platform database. The data includes temperature values ​​(accuracy ±0.5℃), temperature change rate (statistical interval 10 minutes, accuracy ±0.1℃ / 10 minutes), and heat flux density values ​​(accuracy ±5W / m²) for 100 divided areas (each area is 10m × 10m). The data is categorized and stored in a dedicated data folder according to area number (e.g., East Zone 1-20, West Zone 21-40, etc.). The storage format is CSV to ensure data readability. S32 calls the algorithm's built-in membership function library to determine the function type based on the livestock breed. A triangular membership function is selected for dairy cattle farming areas, and a trapezoidal membership function is selected for beef cattle farming areas. Simultaneously, fuzzy subset boundary thresholds are set, such as a temperature threshold of ≤20℃ for "low temperature" and >35℃ for "extremely high temperature," and a humidity threshold of ≤40%RH for "low humidity" and >70%RH for "high humidity." These threshold settings should refer to past year's farm environmental suitability statistics. S33 inputs the classified thermal field data and real-time humidity data (collected once per minute, with an accuracy of ±3%RH) into the selected membership function to calculate the membership values ​​for each parameter (ranging from 0 to 1, rounded to 3 decimal places). Then, fuzzy synthesis is performed based on a preset library of 81 fuzzy rules, using a maximum-minimum synthesis method. Intermediate results (such as temperature membership 0.8, humidity membership 0.7, and rule matching results) are recorded after each region's calculation. S34 defuzzifies the fuzzy synthesis results using the centroid method, converting the fuzzy values ​​into cooling demand levels (0 to 10, integer values). Simultaneously, it stores the inference process data for each region (including input parameters, membership values, rule matching numbers, and defuzzification calculations) in the algorithm's runtime log according to a "region-time" naming convention. The log update frequency matches the data acquisition frequency to ensure subsequent traceability. This step, through segmented operations, ensures the standardization and accuracy of the fuzzy logic algorithm application, avoiding errors in cooling demand level judgment due to deviations in data processing or rule application.

[0028] Preferably, step S4 includes the following sub-steps: S41. Collect real-time operating parameters of all cooling equipment in the pasture, including current power, running time, fault warning status, and energy consumption data. Match the collected parameters with the cooling demand level output in step S3. S42. Call the state transition matrix of the Markov load allocation model. Calculate the state transition probability of each device under different load allocation ratios based on the historical operating data of the equipment and the current matching results. S43. Construct a load allocation optimization model with the objective functions of minimizing equipment energy consumption and achieving optimal cooling effect. Input the state transition probability as a constraint condition into the optimization model and perform multi-objective optimization calculations. S44. Evaluate the multiple load allocation schemes obtained from the optimization calculations. By calculating the comprehensive benefit value of the schemes, select the scheme with the highest comprehensive benefit value as the optimal load allocation scheme and output it to the next step.

[0029] Specifically, the implementation process of step S4 involves multi-stage operations to complete the calculation of the Markov load distribution model and the selection of the optimal solution. Each sub-step requires clear parameter standards and operating procedures. S41 collects real-time operating parameters of all cooling equipment in the ranch (40 units in total, 4-6 units per 500 square meters), including the current power of the equipment (measurement range 0-15kW, accuracy ±0.1kW), running time (statistical start time is 0:00 of the day, accuracy ±1 minute), fault warning status (detected by built-in sensors of the equipment, warning signals are divided into "normal", "level 1 warning" and "level 2 warning") and energy consumption data (statistical interval 1 hour, accuracy ±0.2kWh). The collected data is then matched with the corresponding area cooling demand level (0-10) output by S3 through the equipment number and area number. After the matching is completed, an "equipment-area-demand" association table is generated. S42 calls the state transition matrix of the Markov model, with a matrix dimension of 3×3 (corresponding to the three states of "low load", "medium load" and "high load" for the equipment). State transition probabilities are calculated using equipment operating data from the past three months. For example, the probability of transitioning from low load (0-30% load rate) to medium load (30%-70% load rate) is 0.6, from medium load to high load (70%-100% load rate) is 0.3, and from high load to low load is 0.1. Frequency statistics are used for probability calculation to ensure data reliability. S43 uses "lowest equipment energy consumption" (energy consumption per unit per hour ≤ 2kWh) and "optimal cooling effect" (regional temperature compliance rate ≥ 95%) as objective functions to construct a linear programming optimization model. State transition probabilities are input as constraints (e.g., high load duration ≤ 2 hours) into the model. The simplex method is used for multi-objective optimization calculations, with 20 iterations to ensure convergence to the optimal solution. Step S44 calculates the comprehensive benefit value for the 10 optimized load allocation schemes. The formula for calculating the benefit value includes the energy consumption ratio (weight 0.4) and the cooling compliance rate (weight 0.6). For example, if a scheme has an energy consumption ratio of 80% and a compliance rate of 98%, the comprehensive benefit value is 80%×0.4+98%×0.6=90.8%. The scheme with the highest comprehensive benefit value (≥85%) is selected as the optimal scheme and output to step S5. This step-by-step implementation ensures that the load allocation scheme takes into account both energy consumption and cooling effect, avoiding unreasonable schemes caused by a single objective.

[0030] Preferably, step S5 includes the following sub-steps: S51, parsing the optimal load allocation scheme output by S4, extracting the load allocation coefficient, target operating parameters, and control priority of each cooling device in the scheme, and grouping them according to the area where the devices are located; S52, based on the grouped device information and combined with the real-time power supply data of the ranch, calculating the total power demand when the cooling devices in each area are started simultaneously, and determining whether the total power demand exceeds the power supply threshold; S53, if the total power demand does not exceed the threshold, directly generating the operating power adjustment parameters, running time parameters, and start-up sequence parameters of each device based on the load allocation coefficient; if the total power demand exceeds the threshold, adjusting the operating parameters of high-priority devices, reducing the number of low-priority devices started, and recalculating the total power demand until the power supply constraint is met; S54, encapsulating the generated control parameters according to a preset data format to form a standardized active cooling control command, with an instruction generation timestamp and device identification information attached.

[0031] Specifically, step S5 generates standardized active cooling control commands through multi-stage parameter processing and verification. Each sub-step needs to determine the parameter range and operating standards. S51 analyzes the optimal load distribution scheme output by S4, extracts the load distribution coefficient (value range 0-1, retaining 2 decimal places), target operating parameters (such as the target speed of the fan 1440 r / min, the target water pressure of the spray device 0.4 MPa), and control priority (levels 1-3, with level 1 being the highest) of each device, and groups them by region (such as devices 1-10 in the North Zone, devices 11-20 in the South Zone, etc.). The grouping results need to be visualized on the digital twin platform for easy verification in subsequent operations. S52 Based on the grouped equipment information, call the ranch power supply monitoring system data to obtain the current total power supply power of the area (80kW per area, accuracy ±1kW), calculate the total power demand when each group of equipment starts simultaneously, and calculate the sum of the target power of a single device (calculated based on the load allocation coefficient and rated power). For example, the total target power of 10 devices in the North Area is 18kW. Determine whether it exceeds 80% of the area's power supply power (i.e., 64kW). The determination result must be recorded in writing. S53 If the total power demand does not exceed the threshold, directly generate parameters according to the load allocation coefficient, such as adjusting the fan from the current 10kW to 12kW (80% load rate), running time of 30 minutes, and starting the first-level device in the start-up order; if it exceeds the threshold (e.g., calculated to be 70kW), reduce the load rate of the third-level device (from 60% to 40%, power from 6kW to 4kW) or reduce the number of second-level devices started (from 3 to 2), recalculate the total power demand (adjusted to 62kW), until the "total demand ≤ 64kW" is met. The adjustment process must record the comparison of the values ​​before and after the parameter change. The S54 encapsulates the generated control parameters according to the Modbus-RTU protocol. The encapsulated content includes the device number (8-bit code), power adjustment value (accuracy ±0.1kW), runtime (accurate to minutes), and start-up sequence (levels 1-3). It also adds a command generation timestamp (accurate to seconds) and a unique device identifier (16-bit code). After encapsulation, integrity is verified using a checksum algorithm (such as CRC16). Once verification is successful, the final control command is generated and stored in the command sending queue, awaiting issuance. This step-by-step implementation ensures the accuracy, compliance, and traceability of the control commands, preventing equipment malfunctions due to parameter errors or power overload.

[0032] The multiphysics coupled thermal simulation model for ranches is a technical model that integrates three physical processes—heat conduction, heat convection, and heat radiation—within a ranch to accurately calculate temperature distribution at different spatial and temporal points. Its implementation requires first determining key parameters, such as setting the wall thermal conductivity (0.03 W / (m·K) to 1.74 W / (m·K)) based on building material properties, determining the airflow velocity (0.1 m / s to 3 m / s) based on fan type, and setting the heat dissipation power (50 W to 150 W) according to livestock breed. Simultaneously, it collects ambient temperature and solar radiation intensity (600 W / m² to 1000 W / m²). Then, the ranch is divided into 10 m × 10 m grid cells, with a time step of 10 minutes. Numerical calculations are used to solve for the temperature, heat flux density, and temperature change rate of each cell, outputting a thermal map of the thermal field distribution. This model breaks through the limitations of analyzing a single physical process, accurately recreates the complex thermal environment of a pasture, provides a quantitative basis for judging subsequent cooling needs, avoids cooling decision errors caused by thermal field simulation deviations, ensures that subsequent control measures can accurately match the actual thermal environment of each area, lays a scientific data foundation for pasture temperature control, and improves the rationality and pertinence of the overall cooling plan.

[0033] The fuzzy logic dynamic temperature control algorithm is an intelligent algorithm that determines the cooling demand level of each area of ​​the ranch through multi-parameter fuzzy reasoning. Its implementation requires first dividing the parameters into fuzzy subsets: temperature is divided into "low temperature," "medium temperature," "high temperature," and "extremely high temperature" (≤20℃ to >35℃), and humidity is divided into "low humidity," "medium humidity," and "high humidity" (≤40%RH to >70%RH), and an 81-rule fuzzy rule base is constructed (e.g., "high temperature and high humidity → extremely high demand"). Then, real-time temperature (accuracy ±0.5℃) and humidity (accuracy ±3%RH) of each area are collected, and the parameter membership values ​​(0 to 1) are calculated. Logical reasoning is performed using the maximum-minimum synthesis method, and finally, the centroid method is used to convert the fuzzy results into precise cooling demand levels from 0 to 10, while simultaneously recording the reasoning process log. This algorithm comprehensively considers the synergistic effect of temperature and humidity on cooling demand, avoids the limitations of single parameter threshold judgment, accurately outputs the urgency of cooling in each region, solves the problems of strong subjectivity and large error in traditional manual experience judgment, provides accurate demand basis for subsequent load allocation, ensures that cooling resources can be prioritized to areas with urgent needs, and improves the targeting and efficiency of cooling regulation.

[0034] The Markov load allocation model is an optimization model that calculates the optimal load allocation scheme for cooling equipment based on the state transition laws of the equipment. Its implementation requires first determining the model parameters, such as classifying equipment load rates into "low load," "medium load," and "high load" (0 to 100%), calculating the state transition probability using data from the past three months (e.g., a low load to medium load probability of 0.6), and setting constraints on energy consumption (≤2 kWh / h) and cooling compliance rate (≥95%). Then, real-time equipment power (0 to 15 kW), operating time, and fault status are collected and correlated with the cooling demand level to construct a linear programming optimization model. Multiple load allocation schemes are calculated using the simplex method with 20 iterations. Finally, the optimal scheme (≥85%) is selected based on the comprehensive benefit value (energy consumption weight 0.4, compliance rate weight 0.6). This model dynamically balances equipment load and energy consumption, avoiding the problem of some equipment being overloaded and others being idle. It outputs an allocation scheme that takes into account both efficiency and cost, breaking through the rigid limitations of fixed load allocation and realizing the efficient use of equipment resources. This reduces equipment losses caused by uneven loads and also reduces unnecessary energy expenditures, thereby improving the economy and stability of the ranch cooling system.

[0035] The digital twin temperature control platform for ranches is the core carrier for building a 1:1 mapping with the actual ranch, realizing data storage, model updates, and visual management. Its implementation requires first constructing a digital twin model based on ranch CAD drawings, including spatial structure, equipment locations, and livestock distribution; then receiving environmental, equipment, and livestock data collected by sensors via LoRa links (transmission rates from 125kbps to 500kbps), filtering the data, and storing it in a database; comparing actual and simulated data every 30 minutes, if the deviation exceeds a threshold (temperature ±0.5℃, power ±0.5kW), adjusting model parameters (such as the thermal diffusivity coefficient), updating the model, and visually displaying the thermal field distribution and equipment status. This platform synchronizes the actual ranch status in real time, providing data support and an operating environment for each model, while achieving closed-loop management and traceability of the control process. It breaks down the barriers of traditional temperature control systems where data is scattered and cannot be linked, constructing a complete closed loop of "collection-computation-control-feedback," ensuring that each model is highly adapted to the actual environment, providing a unified and efficient management center for ranch cooling control, and promoting a shift in control from "passive response" to "proactive prediction."

[0036] like Figure 2As shown, an adaptive active cooling control system for pastures is applied to an adaptive active cooling control method for pastures. The system includes: a pasture multi-parameter acquisition unit, which consists of temperature sensors, humidity sensors, livestock activity monitors, and equipment operation monitoring modules distributed in different areas of the pasture. Each sensor and monitoring module is connected via a wireless communication link to acquire pasture environmental parameters, livestock status parameters, and equipment operation parameters, and transmits the acquired data to a data preprocessing unit; a data preprocessing unit, connected to the pasture multi-parameter acquisition unit via Ethernet, which includes a built-in data filtering module and an outlier removal module to filter noise and identify abnormal data in the acquired data, and transmits the processed data to a pasture digital twin temperature control platform unit; and a pasture digital twin temperature control platform unit, connected to the data preprocessing unit and the model calculation unit via industrial Ethernet, which includes a built-in digital twin modeling module, data storage module, and model update module to construct a pasture digital twin model, store historical and real-time data, and receive model calculation data. The system consists of a calculation unit that feeds back data and updates the digital twin model; a model calculation unit, connected to the ranch digital twin temperature control platform unit via a high-speed data bus, integrating a ranch multi-physics coupled thermal simulation model calculation module, a fuzzy logic dynamic temperature control algorithm running module, and a Markov load distribution model solving module. This unit receives digital twin model data, performs model calculations and algorithm inference, and outputs the calculation results to the control command generation unit; a control command generation unit, connected to the model calculation unit and equipment control unit via a wireless communication module, with built-in command parsing and parameter configuration modules, receives the optimal load distribution scheme output by the model calculation unit, generates standardized active cooling control commands, and sends them to the equipment control unit; and an equipment control unit, connected to the control command generation unit and ranch cooling equipment via power line carrier communication, with built-in command execution and status feedback modules, receives control commands and controls the operation of the cooling equipment, collects equipment operating status data, and feeds it back to the ranch digital twin temperature control platform unit for closed-loop system control.

[0037] An adaptive active cooling control method and system for pastures introduces a multi-physics coupled thermal simulation model of the pasture. This model goes beyond simple heat conduction analysis, comprehensively covering the complex effects of heat conduction, convection, and radiation within the pasture. It accurately outputs the thermal field distribution in each area, solving the problem of inaccurate cooling demand judgment caused by a single model in existing technologies. Simultaneously, it relies on a fuzzy logic dynamic temperature control algorithm to scientifically infer the thermal field distribution and temperature control threshold. Combined with a Markov load allocation model, it dynamically calculates the optimal load allocation scheme for the equipment, achieving precise matching between cooling demand and equipment operation. This avoids the energy waste or insufficient cooling caused by existing technologies that rely on fixed parameter control and lack dynamic iterative logic.

[0038] This method and system utilize a digital twin temperature control platform for ranches to receive real-time environmental data, livestock status data, and equipment operation data from multi-parameter acquisition units. This allows for continuous updates to the digital twin model, ensuring a high degree of synchronization between the model and the actual ranch environment. Based on this, the model calculation unit continuously optimizes the control scheme, while the control command generation unit and equipment control unit efficiently execute and provide feedback on the execution status. This forms a closed-loop process of "acquisition-calculation-control-feedback-update," enabling cooling control to continuously adapt to changes in the ranch environment. This not only improves control efficiency and stability, ensuring a suitable environment for livestock growth, but also reduces energy costs through dynamic optimization of equipment load, achieving a dual improvement in breeding efficiency and operational economy.

[0039] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "set," "install," "connect," "link," and "fix" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal communication between two components. Those skilled in the art will understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0040] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various equivalent changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An adaptive active cooling control method for pastures, characterized in that, Includes the following steps: S1. Obtain real-time environmental parameters and livestock activity status parameters in different areas of the pasture, and transmit the parameters to the pasture digital twin temperature control platform to establish a digital twin model consistent with the actual pasture spatial structure. S2. Based on the historical environmental data and cooling regulation data stored in the ranch digital twin temperature control platform, the ranch multi-physics field coupled thermal simulation model is called to simulate and calculate the heat conduction, heat convection and heat radiation processes in different areas of the ranch, and output the thermal field distribution results of each area. S3. Based on the thermal field distribution results output by S2, call the fuzzy logic dynamic temperature control algorithm to perform fuzzy reasoning between the thermal field distribution results and the preset temperature control threshold to determine the cooling demand level of each area, and at the same time extract the current operating parameters of the cooling equipment in each area. S4. Call the Markov load allocation model, take the cooling demand level determined in S3 and the current operating parameters of the cooling equipment as input, calculate the load allocation coefficient under different combinations of cooling equipment, and select the optimal load allocation scheme. S5. Based on the optimal load allocation scheme selected in S4, generate active cooling control instructions for the pasture. The control instructions include the operating power adjustment parameters, running time parameters, and start-up sequence parameters of each cooling device. S6. Send the active cooling control command generated in S5 to the corresponding cooling equipment in the ranch. At the same time, feed back the environmental parameters and equipment operating parameters collected in real time during the execution of the control command to the ranch digital twin temperature control platform, update the digital twin model, and carry out dynamic iteration of the cooling control process.

2. The adaptive active cooling control method for pastures according to claim 1, characterized in that, The expression for the multiphysics coupled thermal simulation model of the pasture is as follows: ,in, Coordinates within the ranch The temperature value at time t. The air thermal diffusivity of the pasture. For the Laplace operator, The airflow velocity influence coefficient. coordinates The airflow velocity at time t For temperature gradient, The influence coefficient of livestock heat dissipation. coordinates Heat dissipation per unit area of ​​livestock at time t. The heat exchange coefficient of the pasture enclosure structure. coordinates The ambient temperature at the location.

3. The adaptive active cooling control method for pastures according to claim 1, characterized in that, The fuzzy inference output expression of the fuzzy logic dynamic temperature control algorithm is: ,in, The cooling requirement level for region i is... Let i be the real-time temperature of the i-th region. The real-time humidity of region i is given. The temperature membership function The membership function is the humidity. The number of temperature-fuzzy subsets. The number of humidity-fuzzy subsets. For the fuzzy rules of temperature and humidity, This is the threshold for the maximum cooling demand level.

4. The adaptive active cooling control method for pastures according to claim 1, characterized in that, The expression for calculating the load allocation coefficient in the Markov load allocation model is as follows: ,in, for The load distribution coefficient of the j-th cooling device in the i-th region at time i. This is the load adjustment factor. Let be the energy consumption coefficient of the j-th cooling device in the i-th region at time t. Let be the difference between the actual temperature and the target temperature in region i. Let be the load distribution coefficient of the j-th cooling device in the i-th region at time t. Let represent the number of cooling devices in region i.

5. The adaptive active cooling control method for pastures according to claim 1, characterized in that, The model update expression for the digital twin temperature control platform of the ranch is: ,in, for Time-based digital twin model parameters The parameters of the digital twin model at time t are... For model update rate coefficients, This refers to the number of environmental parameter collection points. Let be the actual temperature of the i-th sampling point at time t. Let be the simulated temperature at the i-th sampling point at time t. For the gradient of the model parameters, These are the weighting coefficients for equipment parameters. The number of cooling devices, Let be the actual power of the j-th device at time t. Let be the simulated power of the j-th device at time t.

6. The adaptive active cooling control method for pastures according to claim 1, characterized in that, The calculation expression for the operating power adjustment parameters of the active cooling control system in the pasture is as follows: ,in, This is the power adjustment value for the j-th cooling device. The reference power for the j-th cooling device is... This is the power adjustment coefficient. The cooling requirement level for region i is... This represents the average cooling requirement level for pastures. Let i be the real-time temperature of the i-th region. For the target temperature, This is the highest temperature ever recorded on the ranch. This is the lowest temperature ever recorded on the ranch.

7. The adaptive active cooling control method for pastures according to claim 1, characterized in that, S3 includes the following steps: S31. Extract the thermal field distribution data of each region output by S2 from the digital twin temperature control platform database of the ranch. This data includes temperature values, temperature change rate and heat flux density values ​​under different spatial coordinates. Classify and store the extracted data according to the regional division rules. S32. Call the membership function library of the fuzzy logic dynamic temperature control algorithm, select the corresponding temperature membership function and humidity membership function according to the livestock species and density in different areas of the pasture, and determine the boundary threshold of each fuzzy subset. S33. Input the classified thermal field distribution data and the real-time collected humidity data into the selected membership function, calculate the membership values ​​corresponding to different parameters, and then perform fuzzy synthesis operation on the membership values ​​according to the preset fuzzy rule library. S34. Based on the fuzzy synthesis operation results, the fuzzy quantity is converted into an accurate cooling demand level value through a defuzzification algorithm. At the same time, the inference process data corresponding to the cooling demand level of each region is recorded and stored in the algorithm operation log.

8. The adaptive active cooling control method for pastures according to claim 1, characterized in that, S4 includes the following sub-steps: S41. Collect real-time operating parameters of all cooling equipment in the ranch. These parameters include the current power of the equipment, running time, fault warning status and energy consumption data. Then, associate and match the collected parameters with the cooling demand level output by S3. S42. Call the state transition matrix of the Markov load distribution model, and calculate the state transition probability of each device under different load distribution ratios based on the historical operating data of the equipment and the current association matching results. S43. With the objective functions of minimizing equipment energy consumption and achieving optimal cooling effect, construct a load allocation optimization model, input the state transition probability as a constraint condition into the optimization model, and perform multi-objective optimization calculations. S44. Evaluate the multiple load allocation schemes obtained from the optimization calculation, and select the scheme with the highest comprehensive benefit value as the optimal load allocation scheme by calculating the comprehensive benefit value of the schemes, and output it to the next step.

9. The adaptive active cooling control method for pastures according to claim 1, characterized in that, S5 includes the following steps: S51. Analyze the optimal load distribution scheme output by S4, extract the load distribution coefficient, target operating parameters and control priority of each cooling device in the scheme, and group them according to the region where the device is located. S52. Based on the grouped equipment information and combined with the real-time power supply data of the ranch, calculate the total power demand when the cooling equipment in each area is started at the same time, and determine whether the total power demand exceeds the power supply threshold. S53. If the total power demand does not exceed the threshold, the operating power adjustment parameters, operating time parameters and start-up sequence parameters of each device are directly generated based on the load allocation coefficient. If the total power demand exceeds the threshold, the operating parameters of high-priority devices are adjusted, the number of low-priority devices started is reduced, and the total power demand is recalculated until the power supply constraint is met. S54. The generated control parameters are packaged according to the preset data format to form a standardized active cooling control command, with an attached command generation timestamp and equipment identification information.

10. An adaptive active cooling control system for pastures, characterized in that, The system is applied to the adaptive active cooling control method for pastures as described in claim 1, comprising: The ranch multi-parameter acquisition unit consists of temperature sensors, humidity sensors, livestock activity monitors, and equipment operation monitoring modules distributed in different areas of the ranch. Each sensor and monitoring module is connected via a wireless communication link to collect ranch environmental parameters, livestock status parameters, and equipment operation parameters, and transmits the collected data to the data preprocessing unit. The data preprocessing unit is connected to the ranch multi-parameter acquisition unit via Ethernet. It has a built-in data filtering module and outlier removal module, which are used to filter noise and identify outliers in the acquired data and transmit the processed data to the ranch digital twin temperature control platform unit. The ranch digital twin temperature control platform unit is connected to the data preprocessing unit and the model calculation unit via industrial Ethernet. It has built-in digital twin modeling module, data storage module and model update module, which are used to build a ranch digital twin model, store historical data and real-time data, receive feedback data from the model calculation unit and update the digital twin model. The model calculation unit is connected to the ranch digital twin temperature control platform unit via a high-speed data bus. It integrates a ranch multi-physics field coupled thermal simulation model calculation module, a fuzzy logic dynamic temperature control algorithm running module, and a Markov load distribution model solving module. It is used to receive digital twin model data, perform model calculation and algorithm inference, and output the calculation results to the control command generation unit. The control command generation unit is connected to the model calculation unit and the equipment control unit through a wireless communication module. It has a built-in command parsing module and parameter configuration module, which are used to receive the optimal load distribution scheme output by the model calculation unit, generate standardized active cooling control commands, and send them to the equipment control unit. The equipment control unit is connected to the control command generation unit and the pasture cooling equipment via power line carrier communication. It has a built-in command execution module and status feedback module, which are used to receive control commands and control the operation of the cooling equipment, collect equipment operation status data and feed it back to the pasture digital twin temperature control platform unit for closed-loop system control.

Images