Ruminant methane emission estimation method, system and equipment and storage medium

Abstract

The invention relates to a ruminant methane emission estimation method, system and device and a storage medium, and the method comprises the steps: obtaining real-time activity data and environmental factor data of a ruminant, collecting animal individual data, and constructing an input data set according to the environmental factor data and the animal individual data; inputting the input data set into a preset methane emission estimation model, and estimating the methane emission factor of the ruminant to obtain a spatio-temporal change result of the emission factor; based on the estimation result of the emission factors, combining real-time activity data to quantify methane emissions of different regions, analyzing key factors for determining regional distribution characteristics, and generating a regional difference analysis result; and performing visualization processing on the spatio-temporal change result of the emission factor and the regional difference analysis result, displaying spatial distribution and time change of the spatio-temporal change result of the emission factor, and generating a corresponding regional emission list. The method has the effect of improving the methane emission accuracy of the ruminant.

Description

Technical Field

[0001] The present application relates to the technical field of methane emissions from ruminants, and in particular to a method, system, device and storage medium for estimating methane emissions from ruminants. Background Art

[0002] At present, livestock methane emissions are the main source of agricultural greenhouse gas emissions. As one of the countries with the largest number of ruminants in the world, ruminants, mainly cattle and sheep, are the primary source of agricultural methane emissions in my country. Ruminant methane emissions account for about 58.8% of my country's agricultural methane emissions.

[0003] The existing methods for calculating methane emissions from ruminants are mostly based on the internationally accepted IPCC Greenhouse Gas Inventory Guidelines (Tier 2 method) or the simplified algorithms provided by the OECD. These methods usually rely on measuring individual annual emission factors and extrapolating them to regional or national scales. However, these methods do not take into account the regional representativeness of population characteristics and the dynamic changes of environmental factors. They mainly rely on international empirical values or local observation data, which limits the accuracy of the calculations. In addition, although some studies have revised important coefficients based on actual conditions, the classification of existing methods is not detailed in terms of the spatial and temporal changes in ruminant methane emissions. They are only roughly divided according to ruminant species (such as buffaloes and dairy cows), without going deep into sub-classes or regional specific breeds. There are significant changes in the quality and quantity of ruminant feed, feeding methods, and environmental factors (such as temperature and precipitation) in various regions of my country, resulting in regional and seasonal differences in methane emission factors.

[0004] The above-mentioned existing technical solutions have the following defects: the existing ruminant methane emission accounting method does not adequately consider the subdivision of population characteristics and the dynamic changes of environmental factors, resulting in poor regional and seasonal adaptability of the accounting results and low accuracy, so there is room for improvement. Summary of the invention

[0005] In order to improve the accuracy of methane emissions from ruminants, the present application provides a method, system, device and storage medium for estimating methane emissions from ruminants.

[0006] The above-mentioned invention objective of the present application is achieved through the following technical solutions: A dynamic methane emission estimation method, the dynamic methane emission estimation method comprising the following steps: Acquire real-time activity data and environmental factor data of ruminants, collect individual animal data, and construct an input data set based on the environmental factor data and the individual animal data; Inputting the input data set into a preset methane emission estimation model, estimating the methane emission factor of the ruminant, and obtaining the spatiotemporal variation results of the emission factor; Based on the spatio-temporal variation results of the emission factors, combined with the real-time activity data, quantify the methane emissions in different regions, analyze the key factors determining the regional distribution characteristics, and generate the regional difference analysis results; Visualize the spatio-temporal variation results of the emission factors and the regional difference analysis results, display the spatial distribution and temporal variation of the spatio-temporal variation results of the emission factors, and generate the corresponding regional emission inventory.

[0007] By adopting the above technical solutions, by obtaining the real-time activity data and environmental factor data of ruminants, collecting animal individual data, and constructing an input data set according to the environmental factor data and animal individual data, it is possible to comprehensively consider the influence of the population characteristics of ruminants and the dynamic changes of the environment, provide a basis for the accurate prediction of subsequent models, and thus significantly improve the accuracy of methane emission estimation; by inputting the input data set into a preset methane emission estimation model to estimate the methane emissions of ruminants, it is possible to dynamically calculate the emissions of livestock based on a data-driven method, reduce the uncertainty brought by empirical values in traditional methods, and thus improve the regional adaptability of the estimation results; by quantifying the methane emissions in different regions based on the spatio-temporal variation results of the emission factors, analyzing the key influencing factors within the region, and generating the regional difference analysis results, it is possible to clarify the methane emission levels and their influencing factors in different regions, and thus provide a scientific basis for regional emission reduction measures; by visualizing the spatio-temporal variation results of the emission factors and the regional difference analysis results, displaying the spatial distribution and temporal variation of the spatio-temporal variation results of the emission factors, and generating the corresponding regional emission inventory, it is possible to intuitively display the spatial distribution and temporal dynamic changes of the emissions, and thus facilitate policy making and management optimization.

[0008] In one example of the present application, it can be further configured that: constructing the input data set according to the environmental factor data and the animal individual data specifically includes: Perform alignment processing on the environmental factor data and the animal individual data in the time and space dimensions to ensure the consistency and integrity of the data; Clean the synchronized data, remove the missing data and outliers generated during the monitoring process, and complete the missing information by means of data interpolation and / or historical data supplementation; Integrate the data after synchronization and anomaly processing, perform standardization processing on time, space, unit and format, and generate the input data set that meets the model input requirements.

[0009] By adopting the above technical solutions, through the alignment processing of real-time activity data, environmental factor data, and animal individual data in the time and space dimensions, the consistency and integrity of the data are ensured, the time and space characteristics of data from different sources can be unified, and the analysis errors caused by data alignment problems can be avoided, thereby providing a reliable input basis for subsequent data processing and modeling; by cleaning the synchronized data, removing the missing data and outliers generated during the monitoring process, and improving the missing information through data interpolation and / or historical data supplementation, the quality of the input data can be improved, the influence of noise and missing information on the model performance can be avoided, and thus the accuracy of the estimation results can be guaranteed; by integrating the data that has been synchronized and processed for anomalies, standardizing the time, space, unit, and format, and generating an input data set that meets the model input requirements, the input data can meet the model calculation needs, reduce the processing difficulties caused by inconsistent formats, and thus improve the efficiency of data input and the stability of the model.

[0010] In one example, the present application can be further configured as follows: The construction of the methane emission estimation model specifically includes: Collecting the population characteristics, historical environmental factors, historical feeding management modes, etc. of ruminants in different regions across the country as historical livestock feeding data; Classifying and spatializing the historical livestock feeding data according to a preset regional scope, seasonal changes, and sub-species categories, and constructing a training data set based on the regional scale; Performing iterative training on a preset deep learning algorithm according to the training data set, dynamically modeling the spatio-temporal changes of parameters such as methane emission factors and feed digestibility, and obtaining the methane emission estimation model.

[0011] By adopting the above technical solutions, by collecting the population characteristics, historical environmental factors, historical feeding management modes, etc. of ruminants in different regions across the country as historical livestock feeding data, a comprehensive historical data foundation can be established, ensuring that the model can take into account the livestock characteristics and environmental impacts in different regions, thereby enhancing the adaptability of the model to diverse conditions; by classifying and spatializing the historical livestock feeding data according to a preset regional scope, seasonal changes, and sub-species categories, and constructing a training data set based on the regional scale, the differential characteristics of regions, seasons, and sub-species can be accurately distinguished, thereby improving the dynamic prediction ability of the model for methane emissions; by performing iterative training on a preset deep learning algorithm according to the training data set and dynamically modeling the spatio-temporal changes of parameters such as methane emission factors and feed digestibility, a high-precision dynamic methane emission estimation model can be established, thereby improving the prediction accuracy of the model for regional, seasonal, and livestock characteristic changes.

[0012] In one example, the present application can be further configured as follows: The construction of the methane emission estimation model further includes: Generate a calibration dataset by real-time monitoring of the actual methane emissions of livestock, actual activity and feeding data, and dynamic change data of environmental factors within the region; Compare the calibration dataset with the estimation results of the methane emission estimation model and calculate the estimation deviation of key parameters; Based on the estimation deviation, use the reinforcement learning algorithm to dynamically adjust the model parameter weights and optimize the adaptability of the methane emission estimation model under different regions, seasons and breeding methods.

[0013] By adopting the above technical solutions, by real-time monitoring of the actual methane emissions of livestock, actual activity and feeding data, and dynamic change data of environmental factors within the region to generate a calibration dataset, it is possible to dynamically optimize the model input using real-time data, improve the adaptability of the model to the real-time environment, and thus improve the timeliness and accuracy of methane emission estimation; by comparing the calibration dataset with the estimation results of the methane emission estimation model and calculating the estimation deviation of key parameters, it is possible to accurately identify the error sources in the model, thus providing a specific direction for model optimization; by based on the estimation deviation, using the reinforcement learning algorithm to dynamically adjust the model parameter weights and optimize the adaptability of the methane emission estimation model under different regions, seasons and breeding methods, it is possible to further improve the robustness and wide applicability of the model, thus ensuring the scientificity and reliability of the estimation results.

[0014] In one example of the present application, it can be further configured as: based on the spatio-temporal change results of the emission factor, combined with the real-time activity data, quantify the methane emissions in different regions, and analyze the key factors determining the regional distribution characteristics to generate a regional difference analysis result, specifically including: Based on the spatio-temporal change results of the emission factor and the real-time activity data, calculate the difference in methane emission intensity between different regions, and quantify the methane emissions in different regions by the ratio of the total emission amount and livestock density between regions; Use the sensitivity analysis method to quantify the contribution of each factor to methane emissions based on different regions, generate the priority ranking of the key influencing factors according to the contribution, and then generate the regional difference analysis result.

[0015] By adopting the above technical solutions, by calculating the methane emission intensity differences between different regions based on the predicted emission results, and quantifying the methane emission differences between different regions through the ratio of the total emission amount and livestock density between regions, the emission intensity levels of different regions can be clarified, providing a quantitative basis for emission comparison and analysis between regions, and thus providing data support for differentiated emission reduction policies; by using the sensitivity analysis method, quantifying the contribution degree of each factor to methane emissions based on different regions, generating the priority ranking of key influencing factors according to the contribution degree, and then generating the regional difference analysis results, the factors with the greatest impact on methane emissions can be identified, thus providing a clear direction for key point emission reduction and optimal allocation of resources.

[0016] In one example, the present application can be further configured as follows: visualizing the spatio-temporal change results of the emission factors and the regional difference analysis results, displaying the spatial distribution and temporal changes of the spatio-temporal change results of the emission factors, and generating a corresponding regional emission inventory, specifically including: Based on geographic information system technology, generating a spatial distribution map of methane emissions to visually display the methane emissions in different regions, and generating a trend curve of the change of methane emissions over time according to the estimated emission data at different time nodes; Generating the corresponding regional emission inventory according to the regional difference analysis results according to the preset regional division range.

[0017] By adopting the above technical solutions, by generating a spatial distribution map of methane emissions based on geographic information system technology to visually display the methane emissions in different regions, and generating a trend curve of the change of methane emissions over time according to the estimated emission data at different time nodes, the dynamic characteristics of methane emissions in space and time can be clearly presented, thus providing an important basis for the spatial layout and time regulation of methane emission reduction policies; by generating a corresponding regional emission inventory according to the regional difference analysis results according to the preset regional division range, a clear emission data list can be sorted out and output, facilitating managers to quickly understand the emission status of each region, and thus providing data support for the supervision and evaluation of methane emissions.

[0018] The above second invention object of the present application is achieved by the following technical solutions: A methane emission estimation system for ruminants, the methane emission estimation system for ruminants includes the following steps: A data acquisition module, configured to obtain real-time activity data and environmental factor data of ruminants, collect animal individual data, and construct an input data set according to the environmental factor data and the animal individual data; A model input module for inputting the input data set into a preset methane emission estimation model to estimate the methane emission factor of the ruminant and obtain the spatio-temporal variation result of the emission factor; A regional analysis module for quantifying the methane emissions in different regions based on the spatio-temporal variation result of the emission factor and combining with the real-time activity data, and analyzing the key factors determining the regional distribution characteristics; A result visualization module for visualizing the spatio-temporal variation result of the emission factor and the regional difference analysis result, displaying the spatial distribution and temporal variation of the spatio-temporal variation result of the emission factor, and generating a corresponding regional emission inventory.

[0019] By adopting the above technical solutions, by obtaining the real-time activity data and environmental factor data of ruminants, collecting animal individual data, and constructing an input data set according to the environmental factor data and animal individual data, the influence of ruminant population characteristics and environmental dynamic changes can be comprehensively considered, providing a basis for the accurate prediction of subsequent models, thereby significantly improving the accuracy of methane emission estimation; by inputting the input data set into a preset methane emission estimation model to estimate the methane emission trend of ruminants, the emissions of livestock can be dynamically calculated based on a data-driven method, reducing the uncertainty brought by empirical values in traditional methods, thereby enhancing the regional adaptability of the estimation results; by quantifying the methane emissions in different regions based on the spatio-temporal variation result of the emission factor, analyzing the key influencing factors within the region, and generating a regional difference analysis result, the methane emission levels and their influencing factors in different regions can be clarified, thereby providing a scientific basis for regional emission reduction measures; by visualizing the spatio-temporal variation result of the emission factor and the regional difference analysis result, displaying the spatial distribution and temporal variation of the spatio-temporal variation result of the emission factor, and generating a corresponding regional emission inventory, the spatial distribution and temporal dynamic variation of emissions can be intuitively displayed, thereby facilitating policy formulation and management optimization.

[0020] The above object three of the present application is achieved by the following technical solutions: A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the above methane emission estimation method for ruminants are implemented.

[0021] The above object four of the present application is achieved by the following technical solutions: A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above methane emission estimation method for ruminants are implemented.

[0022] In summary, the present application includes the following beneficial technical effects: 1. By obtaining the real-time activity data and environmental factor data of ruminants, collecting individual animal data, and constructing an input data set based on the real-time activity data, environmental factor data, and individual animal data, it is possible to comprehensively consider the population characteristics of ruminants and the impact of environmental dynamic changes, providing a basis for the accurate prediction of subsequent models, thus significantly improving the accuracy of methane emission estimation; by inputting the input data set into a preset methane emission estimation model to estimate the methane emission trend of ruminants, it is possible to dynamically calculate the emissions of livestock based on a data-driven method, reducing the uncertainty brought by empirical values in traditional methods, thus enhancing the regional adaptability of the estimation results; by quantifying the methane emission differences in different regions based on the spatio-temporal variation results of emission factors, analyzing the key influencing factors within the region, and generating regional difference analysis results, it is possible to clarify the methane emission levels and their influencing factors in different regions, thus providing a scientific basis for regional emission reduction measures; by visualizing the spatio-temporal variation results of emission factors and regional difference analysis results, displaying the spatial distribution and temporal changes of the spatio-temporal variation results of emission factors, and generating a corresponding regional emission inventory, it is possible to intuitively display the spatial distribution and temporal dynamic changes of emissions, thus facilitating policy formulation and management optimization; 2. By performing alignment processing on the real-time activity data, environmental factor data, and individual animal data in the time and space dimensions to ensure data consistency and integrity, it is possible to unify the time and space characteristics of data from different sources, avoiding analysis errors caused by data alignment problems, thus providing a reliable input basis for subsequent data processing and modeling; by cleaning the synchronized data, removing the missing data and outliers generated during the monitoring process, and improving the missing information through data interpolation and / or historical data supplementation, it is possible to improve the quality of the input data, avoiding noise and missing information from affecting the model performance, thus ensuring the accuracy of the estimation results; by integrating the data after synchronization and anomaly processing, standardizing the time, space, unit, and format, and generating an input data set that meets the model input requirements, it is possible to make the input data meet the model calculation needs, reducing the processing difficulties caused by inconsistent formats, thus improving the data input efficiency and the stability of the model; 3. By collecting the population characteristics, historical environmental factors, and historical feeding management patterns of ruminants in different regions across the country as historical livestock feeding data, a comprehensive historical data foundation can be established to ensure that the model can take into account the livestock characteristics and environmental impacts of different regions, thereby enhancing the model's adaptability to diverse conditions. By classifying and spatializing the historical livestock feeding data according to the preset regional scope, seasonal changes, and sub-species categories to construct a training dataset at the regional scale, the differential characteristics of regions, seasons, and sub-species can be accurately distinguished, thus improving the model's dynamic prediction ability for methane emissions. By iteratively training the preset deep learning algorithm based on the training dataset and dynamically modeling the spatio-temporal changes of parameters such as methane emission factors and feed digestibility, a high-precision dynamic methane emission estimation model can be established, thereby improving the model's prediction accuracy for changes in regions, seasons, and livestock characteristics. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 is a flowchart of a method for estimating methane emissions from ruminants in an embodiment of the present application; Figure 2 is an implementation flowchart of step S10 in the method for estimating methane emissions from ruminants in an embodiment of the present application; Figure 3 is an implementation flowchart of the construction of a methane emission estimation model in the method for estimating methane emissions from ruminants in an embodiment of the present application; Figure 4 is another implementation flowchart of the construction of a methane emission estimation model in the method for estimating methane emissions from ruminants in an embodiment of the present application; Figure 5 is an implementation flowchart of step S30 in the method for estimating methane emissions from ruminants in an embodiment of the present application; Figure 6 is an implementation flowchart of step S40 in the method for estimating methane emissions from ruminants in an embodiment of the present application; Figure 7 is a principle block diagram of a system for estimating methane emissions from ruminants in an embodiment of the present application; Figure 8 is a schematic diagram of a device in an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0024] The present application will be further described in detail below with reference to the accompanying drawings.

[0025] In one embodiment, as Figure 1 shown, the present application discloses a method for estimating methane emissions from ruminants, which specifically includes the following steps: S10: Obtain the real-time activity data and environmental factor data of ruminants, collect individual animal data, and construct an input data set based on the environmental factor data and individual animal data.

[0026] Specifically, the acquisition of real-time activity data can record data such as the activity level, body weight, and working intensity of livestock through monitoring devices. At the same time, environmental factor information such as environmental temperature, humidity, precipitation, and soil conditions is collected through environmental monitoring devices, and individual characteristic information such as the age, gender, species, and feeding management method of animals is obtained through manual or automated means. These data are marked with time stamps and location information to ensure precise docking with subsequent processing steps.

[0027] S20: Input the input data set into a preset methane emission estimation model to estimate the methane emission factors of ruminants and obtain the spatio-temporal variation results of the emission factors.

[0028] Specifically, various parameters included in the input data set, such as body weight, feed intake, air temperature, and precipitation, will be input into the model by region. The model dynamically calculates parameters such as the total energy intake (GE), feed digestibility (DE), and methane conversion factor (Ym) of livestock, combines region-specific environmental factors. For example, in a certain region, due to higher temperature, the feed digestion efficiency may decrease, and the model will automatically adjust the feed digestibility parameter according to this regional characteristic. At the same time, combined with the individual characteristics of livestock such as body weight and milk production, the methane emissions of each livestock are dynamically estimated. Finally, the emissions of all livestock in the region are superimposed to generate the spatio-temporal variation results of the regional-level methane emission factors. For example, it is predicted that the total methane emissions of a certain species in a certain province in spring are 200 tons, providing a basis for regional methane emission management.

[0029] S30: Based on the spatio-temporal variation results of the emission factors, combined with the real-time activity data, quantify the methane emissions in different regions, analyze the key factors determining the regional distribution characteristics, and generate a regional difference analysis result.

[0030] Specifically, when calculating the regional emission differences, the spatio-temporal variation results of the methane emission factors are grouped according to the geographical location of the region, and the methane emission intensity per unit area is calculated in combination with the livestock density. At the same time, the sensitivity analysis method is used to evaluate the impact of changes in feed quality, feeding methods, livestock population characteristics, and environmental factors on the emission intensity. By changing the values of each factor one by one and recording the emission change trend, it is determined which factors contribute the most to the emission change. For example, it is found that the methane emissions in regions with a higher proportion of crude fiber in the feed increase significantly, then the feed quality can be listed as a factor with a higher priority of influence. Finally, a regional difference analysis report is generated to clarify the emission intensity and key factors in different regions.

[0031] S40: Visualize the spatiotemporal variation results of emission factors and the regional difference analysis results, display the spatial distribution and time variation of the spatiotemporal variation results of emission factors, and generate the corresponding regional emission inventory.

[0032] Specifically, visualization processing includes generating a methane emission heat map based on geographic information system technology, superimposing the total emission data of each region with the geographical boundaries and displaying them through color depth. For example, dark red represents high-emission areas, and light green represents low-emission areas, so as to intuitively show the emission distribution of each region. At the same time, the data of time nodes are used to generate a trend curve of methane emissions over time. For example, the monthly change trend of methane emissions in a certain area throughout the year is plotted, and the peak and trough periods are marked. In addition, the emissions and emission intensity are divided according to administrative regions to generate a text-based emission list. For example, the total emissions of a province in spring are 300 tons and the emission intensity is 2.5 kg / head / day. Key influencing factors such as feed digestibility and stocking density are marked in combination with the results of regional difference analysis to support regional methane emission reduction planning.

[0033] In one embodiment, if Figure 2 As shown, in step S10, an input data set is constructed based on environmental factor data and animal individual data, specifically including: S11: Align environmental factor data and animal individual data in time and space dimensions to ensure data consistency and integrity.

[0034] Specifically, by adding timestamps and spatial tags to each set of data, the environmental factor data and individual animal data are aligned in chronological order. For example, when the weight and feed intake of a cow are collected at a certain point in time, the ambient temperature and precipitation data at the same point in time are automatically matched. At the same time, the environmental data is matched according to the animal's location information. For example, the humidity in the south is higher than that in the north, so the individual data in the same time period needs to be refined by region to ensure that data from different sources are consistent and relevant in time and space.

[0035] S12: Clean the synchronized data, remove missing data and outliers generated during the monitoring process, and complete the missing information through data interpolation and / or historical data supplementation.

[0036] Specifically, for the data missing during the monitoring process due to equipment failures or signal losses, the missing values are estimated by interpolation based on the data values at adjacent time points. For example, after collecting the data of the previous day and the next day, the average value is taken for the missing value. At the same time, the abnormal data beyond the reasonable range is eliminated. For example, when the weight data of a certain cow suddenly becomes negative, it is marked as an abnormal value and eliminated. In addition, for the data points missing for a long time, the key parameters are filled by supplementing with historical data. For example, the data of the same region and the same type of livestock under similar conditions are used for substitution.

[0037] S13: Integrate the data that has been synchronized and processed for abnormalities, and perform standardization processing on time, space, units, and formats to generate an input data set that meets the requirements of the model input.

[0038] Specifically, by unifying the time unit to hours or days, unifying the spatial position to the longitude and latitude coordinate system, and at the same time unifying the units such as feed intake and emission factors to the standard International System of Units. For example, unifying the weight data to kilograms and converting the emission factor data from percentage format to decimal format to ensure that all data formats are consistent with the model input requirements. Finally, the processed data is stored as a structured data file for subsequent calls. For example, it is stored in JSON or CSV format for direct input into the methane emission estimation model for processing.

[0039] In one embodiment, as Figure 3 shown, in step S20, that is, the construction of the methane emission estimation model, specifically includes: S201: Collect the population characteristics, historical environmental factors, and historical feeding management patterns of ruminants in different regions across the country as historical livestock feeding data.

[0040] Specifically, obtain the population characteristics of livestock in different regions through publicly available data, such as the inventory quantity, species distribution, and breeding situation. Combine the local meteorological data to collect the historical environmental factors of each region, such as temperature, humidity, precipitation, and sunshine duration. Obtain the feeding management pattern data such as feed composition, feeding method, and feeding density from feed enterprises and livestock research institutions. Establish a historical livestock feeding database across years and regions by integrating these data. For example, due to the large precipitation in the southern region, the proportion of crude fiber in the feed is relatively high, while in the northern region in winter, the feeding density is greater due to the low temperature. These differences are all recorded in detail for subsequent modeling.

[0041] S202: Classify and spatialize the historical livestock feeding data according to the preset regional scope, seasonal changes, and sub-species categories to construct a training data set based on the regional scale.

[0042] Specifically, the historical data is grouped by administrative division, for example, divided into three - level datasets of province, city, and county. At the same time, it is subdivided into four parts of spring, summer, autumn, and winter according to seasonal changes, and the data is classified according to sub - species categories such as yellow cattle, water buffalo, and yak. Combining the climatic characteristics and feeding management methods of the region, spatial markers are assigned to the data through a geographic information system. For example, longitude and latitude information is added to the summer data of yellow cattle in a certain county. Finally, a training dataset including three - dimensional features of region, season, and subspecies is generated to provide multi - dimensional support for model training.

[0043] S203: Iteratively train a preset deep - learning algorithm according to the training dataset, dynamically model the spatio - temporal changes of parameters such as methane emission factors and feed digestibility, and obtain a methane emission estimation model.

[0044] Specifically, based on the IPCC inventory guidelines Tier2 method as the basic estimation model, that is, using emission factors and activity level data to estimate the total emissions, and inputting the constructed training dataset into a preset deep - learning algorithm, such as a neural - network - based model. By training key parameters such as methane emission factors and feed digestibility in historical data, learn the relationship between each parameter and time and regional characteristics. Through multiple iterations, optimize the weight and bias values so that the model can dynamically predict the changes in methane emissions under different conditions. For example, the model can predict the specific value of the methane conversion coefficient according to the input body weight and feed composition, and at the same time record the loss function value during the training process of the model to ensure that the training converges and outputs the final methane emission estimation model.

[0045] In one embodiment, as Figure 4 shown, in step S20, that is, the construction of the methane emission estimation model, it further includes: S204: Generate a calibration dataset by real - time monitoring the actual methane emissions of livestock, actual activity feeding data, and dynamic change data of environmental factors within the region.

[0046] Specifically, real - time collect the methane gas concentration data emitted, and combine the livestock service monitoring records to obtain data such as its daily activity intensity and feed intake. Real - time obtain environmental data such as temperature, humidity, and precipitation through weather stations within the region. At the same time, conduct preliminary cleaning and processing of the monitoring data, and integrate these real - time monitoring data into a calibration dataset according to region and time. For example, correlate the daily average methane emissions of yellow cattle in a certain area in summer with its feed intake and air temperature to ensure that the calibration data can reflect the actual emission level under real - time conditions.

[0047] S205: Compare the calibration dataset with the estimation results of the methane emission estimation model and calculate the estimation deviation of key parameters.

[0048] Specifically, the actual methane emissions in the calibration dataset are compared one by one with the methane emissions predicted by the model, and the estimation deviation is obtained by calculating the difference between the two. For example, for a certain group of cattle, the calibration dataset shows that their methane emissions are 1.5 kg / day, while the model prediction is 1.7 kg / day, so the deviation value is 0.2 kg / day. At the same time, the estimation results of each key parameter, such as the methane conversion factor and feed digestibility, are compared and analyzed with the actual data, so as to accurately quantify the prediction deviation of the model and output a deviation analysis report.

[0049] S206: Based on the estimation deviation, use the reinforcement learning algorithm to dynamically adjust the model parameter weights and optimize the adaptability of the methane emission estimation model under different regions, seasons and breeding methods.

[0050] Specifically, use the optimization method based on the value function in the reinforcement learning algorithm, take the deviation value as the input of the model loss function, and adjust the model parameter weights through repeated iteration. For example, increase the weights of the key parameters such as feed digestibility in the regions and seasons with large prediction errors, and at the same time reduce the weights of the parameters with small prediction errors. In addition, optimize the search strategy of the model in the parameter space according to the characteristics of different breeding methods. For example, optimize the relevant parameters of feed utilization efficiency for the intensive breeding mode with high stocking density. Finally, enable the model to achieve high-precision methane emission estimation in different regions and seasons.

[0051] In one embodiment, as Figure 5 shown, in step S30, that is, based on the spatio-temporal variation results of the emission factors, combined with the real-time activity data, quantify the methane emissions in different regions, and analyze the key factors determining the regional distribution characteristics to generate the regional difference analysis results, specifically including: S31: Based on the spatio-temporal variation results of the emission factors and the real-time activity data, calculate the difference in methane emission intensity between different regions, and quantify the methane emissions in different regions by the ratio of the total emission amount and livestock density between regions.

[0052] Specifically, statistically analyze the total methane emissions predicted by the model by region, and at the same time obtain the livestock density data of each region. For example, calculate that the average daily total methane emissions in a certain area are 200 kg and the total number of livestock is 100 heads, then the emission intensity is 2 kg / head. Use this intensity index to compare the emission levels of different regions, and at the same time draw a comparison chart of the emission intensity between regions according to the statistical results, for example, showing that the emission intensity of some regions is significantly higher than the average level, providing a basis for subsequent regional management optimization.

[0053] S32: Using the sensitivity analysis method, quantify the contribution of each factor to methane emissions based on different regions, generate a priority ranking of the key influencing factors according to the contribution, and then generate the regional difference analysis result.

[0054] Specifically, by gradually changing the input value of each key factor and observing the change in the methane emissions output by the model, quantify the impact of each factor on emissions. For example, by increasing or decreasing the crude fiber ratio in the feed to observe the change trend of methane emissions. If a 10% change in the feed crude fiber ratio results in a 30% change in emissions, then the contribution of this factor is relatively high. At the same time, conduct similar analyses on other factors such as temperature and feeding density. Sort the factors with relatively high contributions and list them as key influencing factors. Finally, output the analysis result in the form of a table or chart for generating the regional emissions difference report.

[0055] In one embodiment, as Figure 6 shown, in step S40, visualize the spatio-temporal change result of the emission factor and the regional difference analysis result, display the spatial distribution and temporal change of the spatio-temporal change result of the emission factor, and generate the corresponding regional emission inventory, specifically including: S41: Based on geographic information system technology, generate a spatial distribution map of methane emissions to visually display the methane emissions in different regions, and generate a trend curve of the change in methane emissions over time according to the estimated emission data at different time nodes.

[0056] Specifically, visually process the total regional methane emission data output by the model through the GIS platform, draw an emission heat map according to the regional boundaries, and use the depth of color to represent the emission intensity. For example, dark red represents high-emission regions, and light yellow represents low-emission regions. At the same time, plot the predicted emissions at different time nodes as a time curve. For example, display the change trend of methane emissions in a certain region in the four seasons of spring, summer, autumn, and winter to help users understand the temporal fluctuation characteristics of emissions.

[0057] S42: According to the regional difference analysis result, generate the corresponding regional emission inventory according to the preset regional division range.

[0058] Specifically, according to the emission intensity and key influencing factor data in the regional difference analysis, summarize the emissions and rankings of each region into an inventory. The inventory is divided by administrative level, such as province, city, and county, and indicates the total emissions, emission intensity, and main influencing factors of each region. For example, "The methane emissions in a certain city of a certain province are 500 tons / year, the emission intensity is 2.5 kg / head, and the main influencing factor is the crude fiber ratio of the feed". The finally generated inventory can be used for regional policy planning and methane emission reduction management.

[0059] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0060] In one embodiment, a methane emission estimation system for ruminants is provided, and the methane emission estimation system for ruminants corresponds one-to-one with the methane emission estimation method for ruminants in the above embodiment. Figure 7 As shown, the methane emission estimation system for ruminants includes a data acquisition module, a model input module, a regional analysis module, and a result visualization module. The detailed description of each functional module is as follows: A data collection module is used to obtain real-time activity data and environmental factor data of ruminants, collect individual animal data, and construct an input data set based on the environmental factor data and individual animal data; A model input module is used to input the input data set into the preset methane emission estimation model, estimate the methane emission factor of ruminants, and obtain the spatiotemporal variation results of the emission factor; The regional analysis module is used to quantify methane emissions in different regions based on the spatiotemporal variation results of emission factors and combined with real-time activity data, and to analyze the key factors that determine regional distribution characteristics and generate regional difference analysis results; The result visualization module is used to visualize the spatiotemporal variation results of emission factors and the regional difference analysis results, display the spatial distribution and temporal variation of the spatiotemporal variation results of emission factors, and generate the corresponding regional emission inventory.

[0061] Optionally, the data acquisition module specifically includes: The data synchronization submodule is used to align the environmental factor data and the animal individual data in time and space dimensions to ensure the consistency and integrity of the data; The data cleaning submodule is used to clean the synchronized data, remove missing data and outliers generated during the monitoring process, and complete the missing information through data interpolation and / or historical data supplementation; The data standardization submodule is used to integrate the synchronized and exception-processed data, standardize the time, space, unit and format, and generate an input data set that meets the model input requirements.

[0062] Optionally, the construction of a methane emission estimation model includes: The data collection module is used to collect population characteristics, historical environmental factors, and historical feeding and management patterns of ruminants in different regions across the country as historical livestock breeding data; A data classification module for classifying and spatializing historical livestock breeding data according to a preset regional scope, seasonal changes, and sub-categories, and constructing a training dataset at the regional scale; A model training module for iteratively training a preset deep learning algorithm based on the training dataset, dynamically modeling the spatio-temporal changes of parameters such as methane emission factors and feed digestibility, and obtaining a methane emission estimation model.

[0063] Optionally, the construction of the methane emission estimation model further includes: A calibration data generation module for generating a calibration dataset by real-time monitoring of the actual methane emissions of livestock, actual activity breeding data, and dynamic change data of environmental factors within the region; A deviation calculation module for comparing the calibration dataset with the estimation results of the methane emission estimation model and calculating the estimation deviation of key parameters; A model optimization module for dynamically adjusting the model parameter weights based on the estimation deviation using a reinforcement learning algorithm and optimizing the adaptability of the methane emission estimation model under different regions, seasons, and breeding methods.

[0064] Optionally, the regional analysis module specifically includes: An emission intensity analysis sub-module for calculating the methane emission intensity differences between different regions based on the spatio-temporal change results of emission factors and real-time activity data, and quantifying the methane emissions of different regions through the ratio of the total emission amount and livestock density between regions; A sensitivity analysis sub-module for using sensitivity analysis methods to quantify the contribution of each factor to methane emissions based on different regions, generating a priority ranking of key influencing factors according to the contribution degree, and then generating a regional difference analysis result.

[0065] Optionally, the result visualization module specifically includes: A spatial distribution generation sub-module for generating a spatial distribution map of methane emissions based on geographic information system technology to visually display the methane emissions in different regions, and generating a trend curve of methane emissions over time according to the estimated emission data at different time nodes; An emission inventory generation sub-module for generating a corresponding regional emission inventory according to the regional difference analysis result according to the preset regional division scope.

[0066] Specific limitations on the methane emission estimation system for ruminants can be referred to the limitations on the methane emission estimation method for ruminants in the above text, which will not be elaborated here. Each module in the above methane emission estimation system for ruminants can be implemented in whole or in part by software, hardware, and their combinations. The above modules can be embedded in the processor of the computer device in hardware form or independent of it, or stored in the memory of the computer device in software form, so that the processor can call and execute the operations corresponding to each of the above modules.

[0067] In one embodiment, a computer device is provided. The computer device can be a server, and its internal structure diagram can be as Figure 8 shown. The computer device includes a processor, a memory, a network interface, and a database connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with external terminals through a network connection. When the computer program is executed by the processor, it implements a methane emission estimation method for ruminants.

[0068] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, the following steps are implemented: Obtain the real-time activity data and environmental factor data of ruminants, collect animal individual data, and construct an input data set according to the real-time activity data, environmental factor data, and animal individual data.

[0069] Input the input data set into a preset methane emission estimation model to estimate the methane emission trend of ruminants and obtain the spatio-temporal variation results of the emission factors.

[0070] Based on the spatio-temporal variation results of the emission factors, quantify the methane emission differences in different regions, analyze the key influencing factors within the regions, and generate regional difference analysis results.

[0071] Visualize the spatio-temporal variation results of the emission factors and the regional difference analysis results, display the spatial distribution and temporal variation of the spatio-temporal variation results of the emission factors, and generate a corresponding regional emission inventory.

[0072] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by the processor, the following steps are implemented: Obtain the real-time activity data and environmental factor data of ruminants, collect the individual animal data, and construct an input data set based on the real-time activity data, environmental factor data, and individual animal data.

[0073] Input the input data set into a preset methane emission estimation model to estimate the methane emission trend of ruminants and obtain the spatio-temporal variation results of the emission factors.

[0074] Based on the spatio-temporal variation results of the emission factors, quantify the methane emission differences in different regions, analyze the key influencing factors within the regions, and generate regional difference analysis results.

[0075] Visualize the spatio-temporal variation results of the emission factors and the regional difference analysis results, display the spatial distribution and temporal variation of the spatio-temporal variation results of the emission factors, and generate the corresponding regional emission inventory.

[0076] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, storage, database, or other medium used in the various embodiments provided in the present application can include non-volatile and / or volatile memories. Non-volatile memories can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memories can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and Rambus dynamic RAM (RDRAM), etc.

[0077] Those skilled in the art can clearly understand that for the convenience and brevity of description, only the above division of each functional unit and module is used as an example. In actual applications, the above functions can be allocated to different functional units and modules according to needs, that is, the internal structure of the system can be divided into different functional units or modules to complete all or part of the functions described above.

[0078] The above-described embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included within the protection scope of the present application.

Claims

1. A method for estimating methane emissions from ruminants, characterized in that, The method for estimating methane emissions from ruminants comprises the following steps: Acquire real-time activity data and environmental factor data of ruminants, collect individual animal data, and construct an input data set based on the environmental factor data and the individual animal data; Inputting the input data set into a preset methane emission estimation model, estimating the methane emission factor of the ruminant, and obtaining the spatiotemporal variation results of the emission factor; Based on the spatiotemporal variation results of the emission factors, combined with the real-time activity data, quantify the methane emissions in different regions, analyze the key factors that determine the regional distribution characteristics, and generate regional difference analysis results; The spatiotemporal variation results of the emission factors and the regional difference analysis results are visualized to display the spatial distribution and temporal variation of the spatiotemporal variation results of the emission factors, and generate a corresponding regional emission inventory.

2. The method for estimating methane emissions from ruminants according to claim 1, wherein The constructing of the input data set according to the environmental factor data and the animal individual data specifically includes: Performing alignment processing on the environmental factor data and the animal individual data in terms of time and space to ensure the consistency and integrity of the data; Clean the synchronized data, remove missing data and outliers generated during the monitoring process, and complete the missing information through data interpolation and / or historical data supplementation; The synchronized and exception-processed data are integrated, and the time, space, unit and format are standardized to generate the input data set that meets the model input requirements.

3. The method for estimating methane emissions of ruminants according to claim 1, wherein The construction of the methane emission estimation model specifically includes: Collect population characteristics, historical environmental factors, and historical feeding and management patterns of ruminants in different regions across the country as historical livestock breeding data; Classifying and spatializing the historical livestock breeding data according to preset regional ranges, seasonal changes, and subspecies categories to construct a training data set based on regional scale; The preset deep learning algorithm is iteratively trained according to the training data set, and the spatiotemporal changes of parameters such as methane emission factor and feed digestibility are dynamically modeled to obtain the methane emission estimation model.

4. The method for estimating methane emissions of ruminants according to claim 3, characterized in that, The construction of the methane emission estimation model also includes: Generate a calibration data set by real-time monitoring of actual methane emissions from livestock, actual feeding activity data, and dynamic changes in environmental factors in the region; Comparing the calibration data set with the estimation results of the methane emission estimation model to calculate the estimation deviation of key parameters; Based on the estimation deviation, the model parameter weights are dynamically adjusted using a reinforcement learning algorithm to optimize the adaptability of the methane emission estimation model in different regions, seasons and farming methods.

5. The method for estimating methane emissions from ruminants according to claim 1, characterized in that, The spatiotemporal variation results based on the emission factors are combined with the real-time activity data to quantify the methane emissions in different regions, analyze the key factors that determine the regional distribution characteristics, and generate regional difference analysis results, including: Based on the spatiotemporal variation results of the emission factors and the real-time activity data, the differences in methane emission intensity between different regions are calculated, and the methane emissions in different regions are quantified by the ratio of the total emissions and livestock density between the regions; Using the sensitivity analysis method, quantify the contribution of each factor to methane emissions based on different regions, generate the priority ranking of the key influencing factors according to the contribution, and then generate the regional difference analysis result.

6. The method for estimating methane emissions from ruminants according to claim 1, characterized in that, Visualize the spatio-temporal change results of the emission factors and the regional difference analysis result, display the spatial distribution and temporal change of the spatio-temporal change results of the emission factors, and generate the corresponding regional emission inventory, specifically including: Based on the geographic information system technology, generate a spatial distribution map of methane emissions to visually display the methane emissions in different regions, and generate a trend curve of the change of methane emissions over time according to the estimated emission data at different time nodes; According to the regional difference analysis result, generate the corresponding regional emission inventory according to the preset regional division range.

7. A methane emission estimation system for ruminants, characterized in that, The methane emission estimation system for ruminants includes: A data acquisition module for obtaining the real-time activity data and environmental factor data of ruminants, collecting animal individual data, and constructing an input data set according to the environmental factor data and the animal individual data; A model input module for inputting the input data set into a preset methane emission estimation model to estimate the methane emission factors of the ruminants and obtain the spatio-temporal change results of the emission factors; A regional analysis module for quantifying the methane emissions in different regions based on the spatio-temporal change results of the emission factors, combining the real-time activity data, and analyzing the key factors determining the regional distribution characteristics to generate a regional difference analysis result; A result visualization module for visualizing the spatio-temporal change results of the emission factors and the regional difference analysis result, displaying the spatial distribution and temporal change of the spatio-temporal change results of the emission factors, and generating the corresponding regional emission inventory.

8. The methane emission estimation system for ruminants according to claim 7, characterized in that The data acquisition module specifically includes: A data synchronization sub-module for aligning the real-time activity data, the environmental factor data, and the animal individual data in terms of time and space dimensions to ensure the consistency and integrity of the data; A data cleaning sub-module for cleaning the synchronized data, removing the missing data and outliers generated during the monitoring process, and improving the missing information by data interpolation and / or historical data supplementation; A data standardization sub-module for integrating the data that has been synchronized and processed for anomalies, standardizing the time, space, unit, and format, and generating the input data set that meets the model input requirements.

9. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the methane emission estimation method for ruminants according to any one of claims 1 to 6.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the methane emission estimation method for ruminants according to any one of claims 1 to 6.

Images