Grain-saving livestock and poultry breeding method and system based on carbon sink cooperation

Abstract

The invention discloses a grain-saving livestock and poultry breeding method and system based on carbon sink collaboration in the technical field of intelligent livestock and poultry breeding, and solves the problems that existing livestock and poultry nutrition management lacks an environmental carbon sink collaboration mechanism, and nutrition supply and environmental benefit collaborative optimization cannot be realized. The method comprises the following steps: collecting CO2 and O2 concentration and carbon sink capacity data through a distributed gas monitoring network, measuring an animal respiration quotient RQ value by adopting an indirect calorimetric method, establishing a nutrition demand adjustment model based on carbon metabolism intensity, carrying out collaborative optimization on nutrition sufficiency, economical efficiency and environmental friendliness by adopting an improved NSGA-II multi-target genetic algorithm, and carrying out quantitative analysis on the nutrition sufficiency, the economical efficiency and the environmental friendliness. And generating a Pareto optimal nutrition formula scheme set, selecting an optimal formula based on a multi-criterion decision rule, and executing adaptive progressive regulation and control. The system comprises a distributed gas monitoring network, an edge computing processing unit, a cloud collaborative optimization platform and a modular execution system. The feed conversion rate is improved by 7.1%, the carbon emission intensity is reduced by 11.9%, and the nutrition formula precision is improved by 46.7%.

Description

Technical Field

[0001] The present invention relates to the field of intelligent livestock and poultry breeding technology, and in particular to a grain-saving livestock and poultry breeding method and system based on carbon sink synergy. Background Art

[0002] With the expansion of animal husbandry and the increasing demand for environmental protection, intelligent livestock and poultry farming technologies have experienced rapid development. Existing technologies primarily focus on environmental monitoring and automated control, enabling intelligent management of the farming environment through technologies such as the Internet of Things and big data.

[0003] Chinese patent CN112882520A discloses an IoT system for smart farming. The system comprises an environmental monitoring system, an individual monitoring system, an intelligent linkage system, and a control center. This system enables comprehensive monitoring and autonomous adjustment of the farm environment. This technical solution focuses on monitoring and controlling environmental parameters, including real-time monitoring of air temperature, humidity, light intensity, ammonia, methane, and CO2.

[0004] Chinese patent CN116627035A discloses a method and system for intelligent IoT monitoring of livestock and poultry house environments using big data. This system uses parameter detection terminals, parameter control terminals, a cloud platform, and intelligent monitoring terminals to achieve intelligent monitoring and control of environmental parameters. This technology primarily addresses the lack of environmental control solutions in existing farming technologies.

[0005] U.S. Patent No. 2022113296A1 discloses a system and method for lifecycle assessment of animal production emissions. This system uses a machine learning algorithm to accurately quantify greenhouse gas emissions from individual animals and provide individual animal emission certification. This technology primarily focuses on emission assessment but lacks integration with nutrient management.

[0006] The existing technology has the following shortcomings: First, traditional livestock and poultry nutrition management lacks a dynamic correlation mechanism between environmental factors and nutritional needs, and cannot achieve the coordinated optimization of nutritional supply, feed costs, and environmental benefits; second, the existing nutritional formula adjustment is mainly based on static models, and fails to fully utilize the real-time metabolic status information of animals for precise nutritional management; third, there is a lack of quantitative methods and technical means to convert environmental carbon sink changes into reference factors for nutritional decision-making; finally, there is a complex nonlinear coupling relationship between nutritional adequacy, economy and environmental friendliness in multi-objective optimization, which is difficult to handle with traditional single-objective optimization methods. Summary of the Invention

[0007] Technical issues

[0008] The technical problems to be solved by the present application are that the existing livestock nutrition management technology lacks an environmental carbon sink coordination mechanism, cannot realize the coordinated optimization of nutrition supply, economic benefits and environmental benefits, and leads to problems of low feed conversion efficiency, high carbon emission intensity and insufficient precision of nutrition formula.

[0009] Technical scheme

[0010] To solve the above technical problems, the present application provides a grain-saving type livestock breeding method based on carbon sink coordination, comprising the following steps:

[0011] Step one: real-time collection of CO2 concentration, O2 concentration, temperature and humidity data in the breeding environment through a distributed gas monitoring network, the distributed gas monitoring network comprising a CO2 sensor using non-dispersive infrared absorption technology, an O2 sensor using electrochemical sensing technology, a density arrangement of three monitoring points per 50-150 square meters, and simultaneous collection of environmental carbon sink capacity data through a soil organic carbon detector and an atmospheric CO2 concentration change monitor;

[0012] Further, the environmental carbon sink capacity is calculated by weighting the carbon sink capacity index according to the weights of 0.4, 0.3 and 0.3 after standardizing the vegetation coverage, soil organic carbon content and atmospheric CO2 concentration change respectively, and a dynamic balance model with breeding density is established, and when the carbon sink capacity decreases by more than 15%, the nutrition formula is automatically adjusted;

[0013] Step two: measurement of the O2 consumption VO2 and CO2 production VCO2 of the animal through a metabolic cage using an indirect calorimetry method, calculation of the respiratory quotient RQ value of the animal, wherein and calculation of the animal heat production according to the Brouwer heat production formula

[0014] HP = 16.18 x VO2 + 5.02 x VCO2 - 2.17 x VN2

[0015] wherein VN2 is the nitrogen production, which is measured by a nitrogen analyzer using gas chromatography technology, with a measurement accuracy of ±0.1 L / h, a measurement range of 0-50 L / h, and a response time of not more than 30 seconds, and the carbon metabolism intensity is evaluated;

[0016] Step three: establishment of a nutrition demand adjustment model based on the animal physiological state and environmental conditions, calculation of the nutrition demand adjustment coefficient a according to the animal heat production and environmental condition changes;

[0017] Further, the calculation formula of the nutrition demand adjustment coefficient a is:

[0018] a = min(1.20, max(0.80, β x environmental weight factor x individual physiological factor)),

[0019] Among them, β is the basic adjustment coefficient, which is determined according to the ambient temperature. When the ambient temperature is 18-22℃, β = 1.0, and β increases or decreases by 0.01 for every 1℃ deviation. The environmental weight factor and individual physiological factor reflect the impact of environmental conditions and animal physiological status on nutritional requirements respectively. Among them,

[0020] Environmental weight factor = 0.5 × temperature weight + 0.3 × humidity weight + 0.2 × CO2 concentration weight,

[0021] Each weight value is obtained through standardization;

[0022] Individual physiological factors = 0.4 × body mass index + 0.35 × growth stage coefficient + 0.25 × health status score,

[0023] Body mass index = actual weight / ideal weight,

[0024] The growth stage coefficient is determined based on the animal's growth curve, and the health status score is based on a comprehensive assessment of feed intake, activity level, and body temperature;

[0025] Step 4: Use the improved NSGA-II multi-objective genetic algorithm to collaboratively optimize nutritional adequacy, economy, and environmental friendliness, and generate a Pareto optimal nutritional formula solution set;

[0026] Furthermore, the improved NSGA-II multi-objective genetic algorithm adopts a crossover probability of 0.8 to 0.9 and a mutation probability of 0.01 to 0.05, improves the algorithm convergence performance through an elite retention strategy and an adaptive crossover mutation operator, controls data processing delay within 3 seconds through multi-threaded parallel processing and a data caching mechanism, and controls system response time within 25 minutes through an optimization algorithm and a distributed computing architecture;

[0027] Furthermore, the improved NSGA-II multi-objective genetic algorithm in step 4 adopts an elite retention strategy and an adaptive crossover mutation operator, and the objective function is:

[0028] min(F1(nutritional deviation 2 )+λ1·F2(cost increment)+λ2·F3(carbon emission intensity)),

[0029] Where λ1 and λ2 are weight coefficients. For large-scale farms, λ1 = 0.35 and λ2 = 0.25; for small and medium-sized farms, λ1 = 0.45 and λ2 = 0.35. Constraints are set differently based on the livestock and poultry species, including protein content ≥18% for broilers, ≥15% for pigs, and ≥12% for cattle, a nutritional lower limit of energy density of 2800-3200 kcal / kg, and a daily processing capacity of ≤500 kg for equipment.

[0030] Furthermore, the weight coefficients are determined according to the scale of breeding: λ1 is 0.35 and λ2 is 0.25 for large-scale farms, and λ1 is 0.45 and λ2 is 0.35 for small and medium-sized farms, in order to adapt to the economic characteristics of farms of different sizes;

[0031] Step 5: Select the optimal nutritional formula from the Pareto optimal solution set based on multi-criteria decision rules, and use an adaptive progressive control strategy to implement formula adjustment.

[0032] Furthermore, the adaptive gradual regulation strategy adopts a physiological adaptation adjustment cycle of 3 to 5 days, and the adjustment range of each nutritional formula is controlled within the range of 3 to 8%. The safety and effectiveness of nutritional adjustment are ensured by monitoring changes in animal physiological indicators.

[0033] Furthermore, the environmental carbon sink capacity data in step 1 is quantified by the carbon sink capacity index CCI, and the calculation formula is:

[0034] CCI = 0.4 × normalized value of vegetation coverage

[0035] +0.3× normalized value of soil organic carbon content,

[0036] +0.3×normalized atmospheric CO2 concentration change

[0037] A dynamic equilibrium model of carbon sink capacity and stocking density was established: when CCI>0.7, stocking density can be appropriately increased; when CCI<0.4, stocking density needs to be reduced or carbon sink measures need to be strengthened;

[0038] Furthermore, the adaptive gradual control strategy in step five includes: setting an adjustment cycle of 3 to 5 days, controlling the adjustment range of each nutritional formula within the range of 3-8%, ensuring the physiological adaptability of animals to nutritional changes by monitoring the change rate of animal feed intake, daily weight gain stability and health status indicators, and suspending formula adjustment and returning to the previous formula when the feed intake change rate is greater than 15% or the daily weight gain fluctuation is greater than 10%.

[0039] The present invention also provides a grain-saving livestock and poultry breeding system based on carbon sequestration synergy, comprising:

[0040] A distributed gas monitoring network, including CO2 sensors, O2 sensors, temperature and humidity sensors, soil organic carbon detectors, and atmospheric CO2 concentration change monitors, is used to collect real-time aquaculture environment data and carbon sink capacity data;

[0041] An edge computing processing unit, configured with a multi-core ARM processor and 4GB of memory, is connected to the distributed gas monitoring network via LoRa wireless communication, and is used to perform sensor data filtering, outlier detection, data format standardization, and preliminary calculation of carbon metabolism intensity. It is also equipped with a fault detection module for real-time monitoring of sensor working status and network connection status. When a sensor fails or the network is interrupted, the unit automatically switches to a safe mode, in which the system uses a preset standard nutritional formula to maintain basic operation.

[0042] Furthermore, the edge computing processing unit also includes a fault detection module for real-time monitoring of the sensor working status and network connection status, and automatically switches to a safe mode when a fault is detected. In the safe mode, the system uses a preset standard nutritional formula to maintain basic operation;

[0043] A cloud-based collaborative optimization platform, which adopts a distributed computing architecture and is connected to the edge computing processing unit via a 4G / 5G mobile network, is used to perform multi-objective optimization calculations of the improved NSGA-II algorithm, generate a Pareto optimal nutritional formula set, and conduct historical data mining and analysis;

[0044] Furthermore, a distributed computing architecture is adopted, a load balancer is configured to achieve dynamic allocation of computing tasks, and failover is achieved through the master-slave node switching mechanism to ensure continuous and stable operation of the system;

[0045] Furthermore, we adopt a containerized deployment architecture based on Docker container technology to achieve rapid deployment and elastic expansion of applications, and support modular management under the microservice architecture;

[0046] The modular execution system includes a precision batching device, an automatic feeding device and a nutrition monitoring device, which is connected to the cloud-based collaborative optimization platform via an Ethernet wired network. It is used to implement nutrition formula adjustment and adaptive progressive control strategy, and is equipped with a variety identification module and a growth stage determination algorithm. The variety identification module uses image recognition technology combined with RFID tags to automatically identify livestock and poultry varieties, and the growth stage determination algorithm makes a comprehensive judgment based on weight, age and growth curve data.

[0047] Furthermore, the precision batching device includes a multi-storage system, a screw conveying mechanism and a weighing and mixing unit. The batching accuracy reaches ±0.5%, the single batching capacity is 50-500kg, and it supports the precise proportioning of more than 8 kinds of feed raw materials.

[0048] Furthermore, the breed identification module uses image recognition technology combined with RFID tags to automatically identify livestock and poultry breeds, and the growth stage determination algorithm makes a comprehensive determination based on weight, age and growth curve data;

[0049] Furthermore, the cloud-based collaborative optimization platform has the ability to process 60 concurrent optimization tasks, with a system response time of ≤25 minutes and an annual availability rate of ≥97%, and supports nutritional optimization algorithms for a variety of livestock and poultry such as pigs, chickens, and cattle.

[0050] Beneficial effects

[0051] Compared with the prior art, the present invention has the following beneficial effects:

[0052] 1. Significantly improved feed conversion efficiency: Through carbon sink-synergistic nutrition regulation technology, the feed conversion rate has been increased from the traditional 2.8:1 to 2.6:1, an increase of 7.1%, effectively reducing feed costs.

[0053] 2. Carbon emission intensity has been significantly reduced: Through the coordinated regulation of the environmental carbon sink capacity index and nutritional needs, carbon emission intensity has been reduced from 4.2kg / kg to 3.7kg / kg, a reduction of 11.9%, which is in line with the emission reduction requirements of animal husbandry.

[0054] 3. Significant improvement in the accuracy of nutritional formulas: Based on the carbon metabolism intensity assessment model of respiratory quotient dynamic assessment technology, the accuracy of nutritional formulas has been increased from the traditional ±15% to ±8%, an increase of 46.7%.

[0055] 4. The overall economic benefits of the system have been significantly improved: Through the ternary collaborative optimization decision-making engine, the overall economic benefits of the system have increased by 8.5%, while achieving a coordinated improvement in production efficiency, economic benefits and environmental benefits. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 This is a structural diagram of a grain-saving livestock and poultry breeding system based on carbon sink synergy according to an embodiment of the present invention;

[0057] Figure 2 This is a flow chart of the carbon sink collaborative nutrient management method according to an embodiment of the present invention;

[0058] Figure 3 It is a flowchart of the multi-objective optimization algorithm processing of an embodiment of the present invention. DETAILED DESCRIPTION

[0059] The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0060] Example 1

[0061] The purpose of this example is to verify the application effect of carbon sequestration-coordinated nutrition regulation technology in large-scale pig farms and to demonstrate the complete implementation process of the technical solution.

[0062] like Figure 1 and Figure 2 As shown, this embodiment follows Figure 2The carbon sequestration coordinated nutrition management method flow shown is to deploy the technical solution of the present invention in a large-scale pig farm (with 1,000 pigs in stock and a breeding area of ​​5,000 square meters).

[0063] First, if Figure 2 As shown, step one involves deploying a distributed gas monitoring network within the farm at a density of three monitoring points per 100 square meters. The CO2 sensor uses a Sensirion non-dispersive infrared absorption sensor with an accuracy of ±30 ppm + 3% of reading, a measurement range of 0-40,000 ppm, and a response time of 20 seconds. The O2 sensor uses a City Technology 4OX-V electrochemical sensor with an accuracy of ±0.8%, a measurement range of 0-25%, and a response time of 15 seconds. The temperature and humidity sensors use Sensirion integrated sensors with a temperature measurement range of -40°C to 125°C and an accuracy of ±0.2°C, and a humidity measurement range of 0-100% RH with an accuracy of ±2% RH. A soil organic carbon detector (model TOC-2000, measurement range 0-50 g / kg, accuracy ±0.1 g / kg) and an atmospheric CO2 concentration change monitor (model GMP252, measurement range 0-10,000 ppm, accuracy ±1.5% + 2 ppm) are also installed.

[0064] Secondly, if Figure 2 As shown, in step 2, an animal metabolic monitoring system was established using indirect calorimetry. 120 fattening pigs weighing 60±5kg were selected as monitoring subjects. The sample size was determined based on statistical power analysis, with a confidence level of 95% and a test power of 80%. Indirect calorimetry was used to measure the animals' O2 consumption VO2 and CO2 production VCO2 using metabolic cages. The metabolic cage has a size of 2.0m×1.2m×1.5m and is equipped with a gas flow meter (accuracy ±1%) and a gas analyzer (O2 accuracy ±0.01%, CO2 accuracy ±0.01%). Measurement conditions: ambient temperature 20±2°C, relative humidity 60±10%, animal adaptation period 3 days, and measurement time for 24 consecutive hours. The average VO2 was measured to be 180±12L / h, VCO2 was 153±10L / h, and VN2 was 8±1L / h. The respiratory quotient was calculated. The heat production of animals was calculated according to Brouwer's heat production formula HP = 16.18 × VO2 + 5.02 × VCO2 - 2.17 × VN2, and the average heat production was 3663 ± 180 kJ / h.

[0065] like Figure 2 As shown, step three calculates the nutrient demand adjustment coefficient α. Calculation of environmental weight factor:

[0066] Temperature weight = (20-20) / 20 = 0 (normalized value 1.0), humidity weight = (60-60) / 60 = 0 (normalized value 1.0), CO2 concentration weight = (800-400) / 400 = 1.0 (normalized value 0.8),

[0067] Environmental weight factor = 0.5 × 1.0 + 0.3 × 1.0 + 0.2 × 0.8 = 0.96. Calculation of individual physiological factors:

[0068] Body mass index = 60 / 60 = 1.0, growth stage coefficient = 1.1 (fattening period), health status score = 0.95 (based on a comprehensive assessment of feed intake of 2.8 kg / d, normal activity level, and body temperature of 38.5°C),

[0069] Individual physiological factor = 0.4 × 1.0 + 0.35 × 1.1 + 0.25 × 0.95 = 1.02. Basic adjustment coefficient β = 1.0 (ambient temperature 20°C). Nutritional requirement adjustment coefficient α = min(1.20, max(0.80, 1.02 × 0.96 × 1.02)) = 0.98.

[0070] Calculation of environmental carbon sink capacity: vegetation coverage rate 30% (standardized value 0.6), soil organic carbon content 25g / kg (standardized value 0.7), atmospheric CO2 concentration change +50ppm (standardized value 0.8), carbon sink capacity index CCI = 0.4×0.6+0.3×0.7+0.3×0.8=0.69.

[0071] like Figure 3 As shown in the figure, the improved NSGA-II multi-objective genetic algorithm in step 4 is Figure 3 The processing flow shown in the figure is used for optimization calculation, and the Pareto optimal solution set is generated through steps such as population initialization, fitness evaluation, crossover mutation, and elite retention. The detailed processing flow of each step is as follows Figure 3 As shown in the figure. Algorithm parameter settings: population size 100, crossover probability 0.85, mutation probability 0.03, and iteration number 300 generations. The elite retention strategy retains the top 20% of individuals, and the adaptive crossover mutation operator dynamically adjusts parameters based on population diversity. The objective function is:

[0072] min(F1(nutritional deviation 2 )+0.35·F2(cost increment)+0.25·F3(carbon emission intensity)).

[0073] Constraints: protein content ≥15%, energy density 2800-3200kcal / kg, daily processing capacity ≤500kg.

[0074] Nutritional deviation calculation: Including 8 indicators such as crude protein, energy, calcium, phosphorus, etc. Cost increment calculation: Carbon emission intensity calculation:

[0075] F3 = Carbon emissions per unit product, including feed production, transportation, animal metabolism and other links.

[0076] After 300 generations of optimization calculations, the algorithm converged to a stable state and generated a Pareto-optimal nutritional formula set of 20 solutions. Based on a multi-criteria decision rule (weighting nutritional adequacy at 0.4, economic efficiency at 0.35, and environmental friendliness at 0.25), the optimal formula was selected: 52% corn, 18% soybean meal, 12% wheat bran, 3% fish meal, 5% premix, and 10% other ingredients. The formula's nutritional composition: crude protein 16.2%, digestible energy 3050 kcal / kg, calcium 0.65%, and phosphorus 0.55%.

[0077] like Figure 2 As shown, in step five, an adaptive incremental adjustment strategy was used to adjust the diet over a three-day period, with each adjustment controlled within 5%. On day one, the corn content was increased from 50% to 51%, and the soybean meal content was decreased from 20% to 19%. On day two, the feed intake change was monitored to be 8% (<15%), with good daily gain stability. On day three, adjustments were continued to the target diet. Feed intake increased from 2.8 kg / d to 2.9 kg / d, daily gain increased from 850 g to 880 g, and health indicators remained normal.

[0078] Test results: After 90 days of comparative testing, the feed conversion rate of the experimental group using the scheme of the present invention was 2.6:1, and that of the control group (traditional nutritional management) was 2.8:1, an increase of 7.1%, and the difference was statistically significant (P<0.05); the carbon emission intensity of the experimental group was 3.7kg / kg, and that of the control group was 4.2kg / kg, a decrease of 11.9%; the daily weight gain of the experimental group was 880g, and that of the control group was 850g, an increase of 3.5%; the nutritional formula deviation of the experimental group was ±8%, and that of the control group was ±15%, and the accuracy was increased by 46.7%. This embodiment verifies that the carbon sink-synergistic nutritional regulation technology can significantly improve feed conversion efficiency and reduce carbon emission intensity, proving the effectiveness of the technical solution.

[0079] Example 2

[0080] The purpose of this example is to verify the application effect of the technical solution of the present invention in broiler farming and to demonstrate the adaptability and boundary parameter verification of different livestock and poultry breeds.

[0081] The technical solution of the present invention was implemented at a broiler chicken farm (5,000 chickens, 2,000 square meters). A distributed gas monitoring network was deployed at a density of three monitoring points per 50 square meters (verifying the lower limit of monitoring point density), for a total of 120 monitoring points. The CO2 sensor had an accuracy of ±150 ppm, and the O2 sensor had an accuracy of ±0.8%.

[0082] A total of 150 broiler chickens weighing 1.5 ± 0.2 kg were selected for monitoring. A sample size meeting statistical requirements was used. Respiratory metabolism was measured using small metabolic cages (0.8 m × 0.6 m × 0.8 m). Measurement conditions included an ambient temperature of 22 ± 1°C, a relative humidity of 65 ± 5%, and an acclimatization period of three days. Mean VO2 (Volume of Oxygen per liter) was 12 ± 1 L / h, VCO2 (Vanity of Carbon Dioxide per liter) was 9.8 ± 0.8 L / h, and VN2 (Venus of Noxious Oxygen per liter) was 0.5 ± 0.1 L / h. The calculated respiratory quotient (RQ) was 0.82 ± 0.04, and the average heat production was 237 ± 15 kJ / h.

[0083] Calculation of the nutritional requirement adjustment coefficient for broiler chickens: Basic adjustment coefficient β = 1.02 (ambient temperature 22°C, deviation from the standard temperature of 20°C by 2°C, β = 1.0 + 2 × 0.01 = 1.02). Environmental weighting factor 0.88, individual physiological factor 1.05, nutritional requirement adjustment coefficient α = min(1.20, max(0.80, 1.02 × 0.88 × 1.05)) = 0.94.

[0084] Multi-objective optimization parameter settings: λ1 = 0.45, λ2 = 0.35 for small and medium-sized farms. Constraint: Protein content

[0085] ≥18%, energy density 3000-3200 kcal / kg. The optimal formula was determined through optimization: 58% corn, 25% soybean meal, 4% fish meal, 3% oil, 5% premix, and 5% other ingredients. The formula's nutritional composition: crude protein 19.5%, metabolizable energy 3150 kcal / kg.

[0086] Gradual regulation strategy: The adjustment cycle is 4 days, and the adjustment range is controlled within 3% each time (to verify the lower limit of the adjustment range). The feed intake of broilers changed from 120g / day to 125g / day, with a change rate of 4.2% (<15%), and the daily weight gain increased from 45g to 47g, an increase of 4.4%.

[0087] The results showed that after 42 days of testing, the feed conversion rate of the experimental group was 1.65:1, while that of the control group was 1.78:1, a 7.3% increase, with statistically significant differences (P < 0.05). The carbon emission intensity of the experimental group was 2.8 kg / kg, while that of the control group was 3.2 kg / kg, a 12.5% ​​decrease. The accuracy of the nutritional formula in the experimental group was ±7%, while that in the control group was ±14%, a 50% increase. This example verified the adaptability of the technical solution of the present invention to different livestock and poultry breeds, and demonstrated the versatility of the nutritional demand adjustment model and the effectiveness of the boundary parameters.

[0088] Example 3

[0089] The purpose of this example is to verify the system performance and stability of the edge cloud collaborative architecture in large-scale farms, as well as the application effect of the boundary value of the nutrient demand adjustment coefficient α.

[0090] like Figure 1 As shown in the figure, a complete system architecture was deployed at a large cattle farm (with a stock of 2,000 head and a breeding area of ​​10,000 square meters). The distributed gas monitoring network was arranged at a density of three monitoring points per 150 square meters (verifying the upper limit of monitoring point density), for a total of 200 monitoring points. The edge computing processing unit uses an ARM Cortex-A78 multi-core processor with 4GB of LPDDR5 memory, supports local data preprocessing, and keeps data processing latency under 3 seconds. The cloud-based collaborative optimization platform utilizes a distributed computing architecture, equipped with a 32-core CPU and 128GB of memory, supporting 60 concurrent optimization tasks and achieving a system availability rate of 97.5%.

[0091] Parameter settings for beef cattle: Constraints: protein content ≥ 12%, energy density 2800-3000 kcal / kg. For large-scale farms, λ1 = 0.35, λ2 = 0.25. Eighty beef cattle weighing 400 ± 30 kg were selected for monitoring, with a sample size meeting statistical requirements. Large metabolic cages (4.0 m × 3.0 m × 2.5 m) were used. Average VO2 was measured at 2800 ± 150 L / h, VCO2 at 2100 ± 120 L / h, VN2 at 120 ± 10 L / h, respiratory quotient (RQ) at 0.75 ± 0.05, and average heat production at 52,540 ± 2500 kJ / h.

[0092] Application of the boundary value of the nutritional demand adjustment coefficient a: Set extreme environmental conditions (temperature 35℃, humidity 85%, CO2 concentration 1200ppm), calculate the environmental weight factor 0.65, individual physiological factor 0.85, basic adjustment coefficient β = 1.15 (high temperature environment), nutritional demand adjustment coefficient a = min(1.20, max(0.80, 1.15 × 0.65 × 0.85)) = 0.80 (reach the lower limit). Under this condition, the system automatically adjusts the nutritional formula, increases the energy density to 2950kcal / kg, and increases the vitamin and mineral content.

[0093] System performance test: The system response time is an average of 22 minutes, and the longest is not more than 25 minutes; the data acquisition frequency is once every 5 minutes, the data transmission success rate is 99.5%, the optimization algorithm convergence time is an average of 280 generations, and the calculation accuracy meets the requirements. The CPU usage rate of the edge computing unit is an average of 65%, the memory usage rate is an average of 70%, and the network delay is an average of 45ms.

[0094] Test results: After 120 days of testing, the feed conversion rate of the test group is 6.8:1, and that of the control group is 7.3:1, which is improved by 6.8%, and the difference is statistically significant (P<0.05); the carbon emission intensity of the test group is 12.5kg / kg, and that of the control group is 14.2kg / kg, which is reduced by 12.0%; the nutritional formula accuracy of the test group is ±9%, and that of the control group is ±16%, which is improved by 43.8%. The system runs stably, the failure rate is 3.2% (<4%), and the average fault recovery time is 15 minutes. This embodiment verifies the reliability of the edge cloud collaborative architecture and the feasibility of large-scale application, and proves the rationality of the system architecture design and the effectiveness of the boundary parameters.

[0095] Example 4

[0096] The purpose of this embodiment is to verify the guiding effect of the carbon sink capacity index CCI boundary value on the adjustment of the breeding density, and to conduct a 365-day long-term test to verify the stability of the system.

[0097] A dynamic monitoring test of carbon sink capacity was conducted in a certain medium-sized pig farm (500 pigs in stock, breeding area 2500 square meters). By artificially adjusting the vegetation coverage and soil improvement, different carbon sink capacity conditions were created.

[0098] High carbon sink capacity condition: vegetation coverage rate 60% (standardized value 0.9), soil organic carbon content 35g / kg (standardized value 0.9), atmospheric CO2 concentration change -20ppm (standardized value 0.95), calculate CCI = 0.4 × 0.9 + 0.3 × 0.9 + 0.3 × 0.95 = 0.915 (>0.7). Under this condition, the system suggests that the breeding density can be appropriately increased, and the stock level is increased from 500 to 550, and the density is increased from 0.2 to 0.23 head / m2 Increased to 0.22 heads / m 2 .

[0099] Low carbon sink capacity conditions: vegetation coverage of 10% (normalized value 0.3), soil organic carbon content of 15g / kg (normalized value 0.4), atmospheric CO2 concentration change +100ppm (normalized value 0.6), calculated CCI = 0.4×0.3+0.3×0.4+0.3×0.6=0.42 (>0.4 but <0.7). The system maintains the current stocking density.

[0100] Extremely low carbon sink capacity conditions: vegetation coverage 5% (standardized value 0.2), soil organic carbon content 10g / kg (standardized value 0.3), atmospheric CO2 concentration change +150ppm (standardized value 0.4), calculated CCI = 0.4×0.2+0.3×0.3+0.3×0.4=0.29 (<0.4). The system recommends reducing the stocking density from 500 to 450 heads, and the density from 0.2 heads / m 2 Reduced to 0.18 heads / m 2 .

[0101] Long-term stability verification: After 365 days of long-term testing, the system's operational stability reached 96%, sensor accuracy drift was controlled within ±5%, and the effects of nutritional formula adjustments remained stable. Under conditions of high carbon sink capacity, increasing stocking density only increased carbon emission intensity by 3.2%, and environmental quality remained stable. Under conditions of extremely low carbon sink capacity, reducing stocking density reduced carbon emission intensity by 15.8%, significantly improving environmental quality. This example verified the effective guidance of the Carbon Capacity Index (CCI) for stocking density adjustments and the long-term stability of the system.

[0102] Comparative Example 1

[0103] The purpose of this comparative example is to verify the impact of the lack of a carbon sink synergistic mechanism on aquaculture results, and to highlight the necessity of the carbon sink synergistic nutrient regulation technology of the present invention by comparison with traditional environmental monitoring programs.

[0104] The same farm and experimental conditions as in Example 1 were used, but the nutrient management plan did not consider changes in environmental carbon sink capacity and adjusted the nutrient formula based solely on traditional ambient temperature and humidity parameters. The monitoring equipment included only temperature and humidity sensors, and did not include a soil organic carbon detector or an atmospheric CO2 concentration change monitor.

[0105] Nutritional demand adjustment only considers temperature and humidity factors, and the adjustment coefficient

[0106] α=min(1.15,max(0.85,β×temperature and humidity weight factor)),

[0107] in

[0108] Temperature and humidity weight factor = 0.5 x temperature weight + 0.5 x humidity weight.

[0109] Under the same environmental conditions (temperature 20℃, humidity 60%), it is calculated that:

[0110] Temperature and humidity weight factor = 0.5 x 1.0 + 0.5 x 1.0 = 1.0, adjustment coefficient a = min(1.15, max(0.85, 1.0 x 1.0)) = 1.0.

[0111] Using the traditional single-objective optimization method, only the nutritional adequacy is taken as the target for formula optimization, and the objective function is simplified as:

[0112] min(F1(nutritional deviation 2 )),

[0113] Without considering the synergy of cost and environmental benefits. The formula obtained by optimization is: corn 50%, soybean meal 20%, wheat bran 15%, fish meal 4%, premix 6%, and others 5%. The nutritional ingredients of the formula are: crude protein 16.8%, digestible energy 2980kcal / kg, and the cost is 8% higher than that of Example 1.

[0114] Comparative test results: after 90 days of testing, the feed conversion rate of the comparative example group is 2.75:1, the carbon emission intensity is 4.0kg / kg, the daily weight gain is 860g, and the nutritional formula deviation is ±13%. Compared with Example 1, the feed conversion rate is 5.8% lower, the carbon emission intensity is 8.1% higher, the daily weight gain is 2.3% lower, and the nutritional formula precision is 38.5% lower, and the differences are statistically significant (P<0.05). Through comparison, it can be seen that the lack of carbon sink synergy mechanism leads to a decrease in the accuracy of nutritional management, and the environmental and economic benefits are also reduced, proving the necessity and superiority of the carbon sink synergy nutritional regulation technology of the present application.

[0115] Comparative Example 2

[0116] The purpose of this comparative example is to verify the impact of the lack of metabolic evaluation based on respiratory quotient dynamic evaluation technology on the accuracy of nutritional management, and the traditional body weight and feed intake evaluation method is used for comparison.

[0117] The same test conditions as in Example 1 are used, but the respiratory quotient RQ method is not used for animal nutritional status evaluation, but the traditional body weight gain rate and feed intake change are used to evaluate the nutritional status of animals. Nutritional demand adjustment is calculated based on the linear relationship between body weight gain rate and feed intake.

[0118] Traditional evaluation method:

[0119] Nutritional status score = 0.6 x body weight gain rate standardized value + 0.4 x feed intake change standardized value.

[0120] The body weight growth rate was 850 g / d (standardized value 0.85), the feed intake change was 2.8 kg / d (standardized value 0.9), the nutritional status score = 0.6 × 0.85 + 0.4 × 0.9 = 0.87, and the corresponding nutritional requirement adjustment coefficient α = 0.94.

[0121] Due to the lack of accurate metabolic status assessment, formula adjustments lagged behind the actual needs of the animals, resulting in a mismatch between nutrient supply and demand. The optimized formula is: corn 53%, soybean meal 17%, wheat bran 13%, fish meal 3.5%, premix 5%, and other 8.5%.

[0122] Comparative test results: After 90 days of testing, the feed conversion rate of the control group was 2.72:1, the nutritional formula deviation was ±12%, and the daily weight gain was 865g. Compared with Example 1, the feed conversion rate was 4.4% lower, the nutritional formula accuracy was 33% lower, and the daily weight gain was 1.7% lower. The differences were statistically significant (P < 0.05). Through comparison, it can be seen that the lack of metabolic assessment based on respiratory quotient dynamic assessment technology leads to inaccurate nutritional status judgment, affecting the accuracy of the nutritional formula, proving the advancement and necessity of the metabolic assessment method of the present invention.

[0123] Comparative Example 3

[0124] The purpose of this comparative example is to verify the difference in the effects of single-objective optimization and ternary collaborative optimization decision engines, and to compare them using an optimization scheme that only aims to minimize feed costs.

[0125] The same monitoring and evaluation system as in Example 1 was used, but the nutritional formula optimization only took minimizing feed cost as the single goal, without considering the coordination between nutritional adequacy and environmental friendliness. The optimization objective function was simplified to:

[0126] min(F2(cost increment)),

[0127] The constraints only include the lower limit of basic nutritional requirements (crude protein ≥14%, energy ≥2700kcal / kg).

[0128] Single-objective optimization resulted in a formula: 60% corn, 15% soybean meal, 18% wheat bran, 4% premix, and 3% other ingredients. This formula offers a lower cost, but lacks nutritional density and environmental friendliness. The formula's nutritional composition includes 14.2% crude protein and 2750 kcal / kg digestible energy. The cost is 12% lower than Example 1.

[0129] Comparative test results: After 90 days of testing, the feed cost of the control group was reduced by 12%, but the feed conversion rate was only 2.85:1, the carbon emission intensity was 4.3kg / kg, the daily weight gain was 820g, and the nutritional formula deviation was ±16%. Compared with Example 1, although the feed cost was reduced, the feed conversion rate was 9.6% lower, the carbon emission intensity was 16.2% higher, the daily weight gain was 6.8% lower, the nutritional formula accuracy was 50% lower, and the comprehensive economic benefit actually decreased by 5.2%. The difference was statistically significant (P<0.05). By comparison, it can be seen that although single-objective optimization can reduce single-item costs, it will lead to a decline in overall benefits, which proves the superiority and practical value of the ternary collaborative optimization decision-making engine of the present invention.

[0130] Based on the above embodiments and comparative examples, the present invention establishes a quantitative correlation model between environmental carbon sink capacity and animal nutritional needs through the pioneering carbon sink collaborative nutritional regulation technology, thereby realizing the coordinated optimization of nutritional supply, economic benefits and environmental benefits. The carbon metabolism intensity assessment model based on the dynamic assessment technology of respiratory quotient can accurately assess the nutritional status of animals in real time and significantly improve the accuracy of nutritional formulas. The improved NSGA-II multi-objective genetic algorithm effectively handles complex nonlinear coupling relationships, generates a Pareto optimal solution set, and provides scientific decision-making support for nutritional management. The edge-cloud collaborative layered architecture ensures the real-time and reliability of the system, and the adaptive progressive control strategy ensures animal welfare and system stability. The essential difference between the present invention and the prior art is that the environmental carbon sink is incorporated into the livestock and poultry nutrition decision-making system, realizing a technological leap from a single production target to multiple benefit synergy, and providing an innovative technical path for the sustainable development of animal husbandry.

Claims

1. A grain-saving livestock and poultry breeding method based on carbon sink synergy, characterized in that: The following steps are involved: Step 1: Real-time data on CO2 concentration, O2 concentration, temperature, and humidity in the aquaculture environment are collected through a distributed gas monitoring network. The distributed gas monitoring network includes a CO2 sensor using non-dispersive infrared absorption technology and an O2 sensor using electrochemical sensing technology. The network is arranged at a density of three monitoring points per 50 to 150 square meters. At the same time, environmental carbon sink capacity data is collected through a soil organic carbon detector and an atmospheric CO2 concentration change monitor. Step 2: Use indirect calorimetry to measure the animal's O2 consumption (VO2) and CO2 production (VCO2) in a metabolic cage, and calculate the animal's respiratory quotient (RQ), where According to Brouwer's heat production formula HP=16.18×VO2+5.02×VCO2-2.17×VN2 Calculate animal heat production and assess carbon metabolism intensity; Step 3: Establish a nutrient requirement adjustment model based on the animal's physiological state and environmental conditions, and calculate the nutrient requirement adjustment coefficient α according to the animal's heat production and changes in environmental conditions; Step 4: Use the improved NSGA-II multi-objective genetic algorithm to collaboratively optimize nutritional adequacy, economy and environmental friendliness, and generate a Pareto optimal nutritional formula set; Step 5: Select the optimal nutritional formula from the Pareto optimal solution set based on multi-criteria decision rules, and use an adaptive progressive control strategy to perform formula adjustment.

2. The grain-saving livestock and poultry breeding method based on carbon sequestration synergy according to claim 1 is characterized in that: The calculation formula of the nutritional requirement adjustment coefficient α is: α=min(1.20,max(0.80,β×environmental weight factor×individual physiological factor)), where β is the basic adjustment coefficient, determined according to the animal species and ambient temperature. When the ambient temperature is 18-22°C, β=1.

0. For every 1°C deviation of the temperature, β is adjusted accordingly by ±0.

01. The environmental weight factor includes the weighted average of the temperature weight of 0.5, the humidity weight of 0.3, and the CO2 concentration weight of 0.

2. The individual physiological factors include the weighted average of the body mass index, the growth stage coefficient, and the health status score.

3. The grain-saving livestock and poultry breeding method based on carbon sequestration synergy according to claim 1 is characterized in that: The adaptive gradual regulation strategy includes: adopting a physiological adaptation adjustment cycle of 3 to 5 days, controlling the adjustment range of each nutritional formula within the range of 3 to 8%, ensuring physiological adaptability to nutritional changes by monitoring the animal's feeding behavior and changes in physiological indicators, and suspending formula adjustment and returning to the previous formula when the feed intake change rate is greater than 15% or the daily weight gain fluctuation is greater than 10%.

4. The grain-saving livestock and poultry breeding method based on carbon sink synergy according to claim 1, characterized in that: The improved NSGA-II multi-objective genetic algorithm includes an elite retention strategy and an adaptive crossover mutation operator, with a crossover probability of 0.8 to 0.9 and a mutation probability of 0.01 to 0.05, and optimizes the objective function min(F1(nutritional deviation 2 )+λ1·F2(cost increment)+λ2·F3(carbon emission intensity)), Among them, λ1 and λ2 are determined according to the scale of breeding.

5. The grain-saving livestock and poultry breeding method based on carbon sink synergy according to claim 4 is characterized in that: For large-scale farms, λ1 is 0.35 and λ2 is 0.25; for small and medium-sized farms, λ1 is 0.45 and λ2 is 0.

35. The constraints include protein content ≥18% for broilers, ≥15% for pigs, ≥12% for cattle, a nutritional lower limit of energy density of 2800-3200 kcal / kg, and equipment capacity limitation of daily processing capacity ≤500 kg.

6. The grain-saving livestock and poultry breeding method based on carbon sequestration synergy according to claim 1, characterized in that: The environmental carbon sink capacity is calculated by standardizing the vegetation coverage rate, soil organic carbon content, and atmospheric CO2 concentration changes, and then weighted according to weights of 0.4, 0.3, and 0.

3. A dynamic balance model of carbon sink capacity and breeding density is established, and when the carbon sink capacity drops by more than 15%, the nutritional formula adjustment is automatically triggered.

7. A grain-saving livestock and poultry breeding system based on carbon sink synergy, characterized in that: include: A distributed gas monitoring network 1 includes a CO2 sensor using non-dispersive infrared absorption technology, an O2 sensor using electrochemical sensing technology, an integrated temperature and humidity sensor, a soil organic carbon detector, and an atmospheric CO2 concentration change monitor, for real-time collection of aquaculture environment data and carbon sink capacity data; an edge computing processing unit 2, using a multi-core processor with at least 4GB of memory, is connected to the distributed gas monitoring network 1 via LoRa wireless communication and is used to perform data preprocessing and simple calculations; The cloud collaborative optimization platform 3 adopts a distributed computing architecture and is connected to the edge computing processing unit 2 via a mobile network to perform multi-objective optimization calculations of the improved NSGA-II algorithm and generate a nutritional formula plan; The modular execution system 4, including a precision batching device, an environmental control device and a data feedback device, is connected to the cloud collaborative optimization platform 3 through a wired network and is used to execute nutritional formula adjustment and adaptive progressive control strategy.

8. The grain-saving livestock and poultry breeding system based on carbon sink synergy according to claim 7 is characterized in that: The edge computing processing unit 2 integrates a dedicated data pre-processing chip, supports real-time data acquisition and preliminary processing, the data processing delay does not exceed 3 seconds, the system response time is controlled within 25 minutes, and is equipped with a fault detection module. When a sensor fails or the network is interrupted, it automatically switches to a safe mode.

9. The grain-saving livestock and poultry breeding system based on carbon sink synergy according to claim 7 is characterized in that: The cloud-based collaborative optimization platform 3 adopts a containerized deployment architecture, supports 60 concurrent optimization tasks, and is configured with load balancing and failover mechanisms, with a system availability rate of over 97%.

10. The grain-saving livestock and poultry breeding system based on carbon sink synergy according to claim 7, characterized in that: The precision batching device of the modular execution system 4 adopts spiral conveying and weighing feedback control, and the batching accuracy reaches ±0.5%. It supports the differentiated nutritional management needs of livestock and poultry such as pigs, chickens, and cattle, and is equipped with a variety recognition module and a growth stage determination algorithm.

Images