Air environment detection system and method for livestock farms

Abstract

The present invention relates to the field of air environment detection technology, and in particular to an air environment detection system and method for animal husbandry farms. The method comprises the following steps: by deploying a multi-spectral imaging sensor in an animal husbandry farm, collecting air spectrum data in multiple time periods and identifying gas components, obtaining gas component spectrum change data, thereby performing similarity fitting of phase change conditions and simulation of microbial activity loss, analyzing the decline of microbial decomposition ability, and identifying the trend of air deterioration; finally, constructing an air environment detection model using a policy gradient algorithm, and uploading the model to a cloud platform for execution, so as to realize air environment monitoring. The present invention makes the air environment detection technology more perfect by optimizing the air environment detection technology.

Description

Technical Field

[0001] The present invention relates to the technical field of air environment detection, and in particular to an air environment detection system and method for a livestock farm. Background Art

[0002] With the development of animal husbandry, the scale of farms is constantly expanding, and the problem of air pollution is becoming increasingly serious. There are a large number of harmful gases in livestock farms, such as ammonia, hydrogen sulfide, methane, etc. These gases not only threaten the health of animals, but also have a negative impact on the living environment of surrounding residents. At the same time, factors such as odor pollution, temperature and humidity changes, and microbial activity in farms are also constantly affecting air quality and ecological balance. Traditional air quality monitoring methods mostly rely on manual regular detection, with a low monitoring frequency, and are not responsive to instantaneous changes in air pollution, which can easily miss potential air environment deterioration problems. In order to meet this challenge, it is particularly important to use advanced technical means to conduct continuous, real-time and accurate air environment monitoring. The combination of multispectral imaging technology and sensors provides a new way to collect gas composition data in the air in real time. These sensors can effectively identify changes in gas composition in different time periods, and through the analysis of data, they can warn of the deterioration trend of the air environment in advance. At the same time, microorganisms play an important role in the process of air pollution, and their activity changes and decomposition capacity decline are also one of the important factors for the deterioration of air quality. However, traditional air environment monitoring methods have inaccurate analysis of the decline in microbial activity and decomposition capacity in livestock farms, resulting in large errors in air environment quality monitoring. Summary of the invention

[0003] Based on this, it is necessary to provide an air environment detection system and method for livestock farms to solve at least one of the above technical problems.

[0004] To achieve the above object, a method for detecting the air environment of a livestock farm is provided, the method comprising the following steps:

[0005] Step S1: deploy a multi-spectral imaging sensor in a livestock farm, and collect air spectrum data in multiple time periods to obtain air spectra in multiple time periods; identify gas component types of the air spectra in multiple time periods to obtain gas component type data; perform time period feature sampling based on the gas component type data to obtain gas component spectrum change sampling data;

[0006] Step S2: performing similar fitting of phase change conditions based on the sampling data of gas component spectrum changes to obtain similar fitting data of phase change conditions between different gas components; performing equal loss of microbial activity simulation based on the similar fitting data of phase change conditions to obtain equal loss of microbial activity data;

[0007] Step S3: performing a microbial decomposition capacity decline analysis based on the microbial activity equal loss data to obtain microbial decomposition capacity decline data; identifying the air deterioration trend of the livestock farm based on the microbial decomposition capacity decline data to obtain air deterioration trend data;

[0008] Step S4: construct an air environment detection model for the air deterioration trend data based on the policy gradient algorithm to obtain an air environment detection model; send the air environment detection model to the cloud platform to perform air environment detection in the livestock farm.

[0009] Preferably, step S1 comprises the following steps:

[0010] Step S11: deploying a multi-spectral imaging sensor in the livestock farm, and collecting air spectrum data in multiple time periods to obtain air spectra in multiple time periods;

[0011] Step S12: performing time series variation trend analysis on the air spectra in multiple time periods to obtain the air spectra with variation trends in multiple time periods;

[0012] Step S13: Identify the gas component type of the air spectrum with the changing trend in multiple time periods to obtain gas component type data;

[0013] Step S14: based on the gas component type data, the air spectrum with multiple time period change trends is sampled for time period characteristics to obtain gas component spectrum change sampling data.

[0014] Preferably, step S2 comprises the following steps:

[0015] Step S21: analyzing the concentration variation between different gas components on the gas component spectrum variation sampling data to obtain the concentration variation data between different gas components;

[0016] Step S22: performing phase change intensity coupling based on the concentration change data between different gas components to obtain phase change intensity coupling data between different gas components;

[0017] Step S23: calculating the toxic coupling release index between different gas components on the phase change strip intensity coupling data to obtain the toxic coupling release index between different gas components;

[0018] Step S24: performing a simulation of the equivalent loss of microbial activity according to the toxicity coupling release index and the phase change strip intensity coupling data to obtain the equivalent loss data of microbial activity.

[0019] Preferably, step S24 includes the following steps:

[0020] Step S241: performing release hysteresis analysis on the toxicity coupling release index to obtain toxicity release hysteresis data;

[0021] Step S242: calculating the specific heat tolerance between different gas components on the phase change strip intensity coupling data to obtain the phase conversion specific heat tolerance between different gas components;

[0022] Step S243: performing a simulation of the molecular group conformational change between different gas components on the phase change strip intensity coupling data based on the phase conversion specific heat tolerance to obtain the molecular group conformational change data between different gas components;

[0023] Step S244: performing a toxicity release nonlinear gradient analysis according to the toxicity release hysteresis data and the toxicity coupling release index to obtain toxicity release nonlinear gradient data;

[0024] Step S245: performing gas toxicity molecular penetration enhancement isoquant gradient evaluation based on the molecular group conformation change data and the toxicity release nonlinear gradient data to obtain molecular toxicity penetration enhancement isoquant data;

[0025] Step S246: Perform a simulation of the equivalent loss of microbial activity based on the molecular toxicity penetration enhancement equivalent data to obtain the equivalent loss data of microbial activity.

[0026] Preferably, step S3 comprises the following steps:

[0027] Step S31: performing loss convolution processing on the microbial activity equal loss data to obtain microbial activity equal loss convolution data;

[0028] Step S32: performing microbial decomposition capacity decay analysis based on the microbial activity equal loss convolution data to obtain microbial decomposition capacity decay data;

[0029] Step S33: Identify the air deterioration trend of the livestock farm based on the microbial decomposition capacity decline data to obtain air deterioration trend data.

[0030] Preferably, step S4 comprises the following steps:

[0031] Step S41: performing logic learning on the air deterioration trend data to obtain air deterioration trend logic data;

[0032] Step S42: constructing an air environment detection model based on the microbial decomposition capacity decay data and the air deterioration trend logic data based on the strategy gradient algorithm to obtain an air environment detection model;

[0033] Step S43: Send the air environment detection model to the cloud platform to perform air environment detection in the livestock farm.

[0034] Preferably, the present invention further provides an air environment detection system for a livestock farm, which is used to perform the air environment detection method for a livestock farm as described above. The air environment detection system for a livestock farm comprises:

[0035] The gas component type identification module is used to deploy multi-spectral imaging sensors in livestock farms, collect air spectrum data in multiple time periods, and obtain air spectra in multiple time periods; identify the gas component types of air spectra in multiple time periods to obtain gas component type data; perform time period feature sampling based on the gas component type data to obtain gas component spectrum change sampling data;

[0036] The microbial activity equivalent loss analysis module is used to perform similar fitting of phase change conditions based on the sampling data of gas component spectrum changes to obtain similar fitting data of phase change conditions between different gas components; perform microbial activity equivalent loss simulation based on the similar fitting data of phase change conditions to obtain microbial activity equivalent loss data;

[0037] The air deterioration trend identification module is used to analyze the decline of microbial decomposition ability based on the microbial activity equal loss data to obtain the microbial decomposition ability decline data; the air deterioration trend of the livestock farm is identified based on the microbial decomposition ability decline data to obtain the air deterioration trend data;

[0038] The model building detection module is used to build an air environment detection model for air deterioration trend data based on the policy gradient algorithm to obtain an air environment detection model; the air environment detection model is sent to the cloud platform to perform air environment detection in livestock farms.

[0039] The beneficial effect of the present invention is that by deploying multi-spectral imaging sensors in livestock farms and collecting air spectral data in multiple time periods, the changes in gas composition in the air in the farms can be captured in real time, providing high-frequency and accurate spectral data for air environment monitoring. By identifying the gas component type of these spectral data, the types and concentration changes of harmful gases in the air can be accurately identified, providing basic data for subsequent pollution source analysis and air quality improvement. At the same time, time period feature sampling can further reveal the changing laws of gas components and provide strong support for subsequent air quality prediction and early warning. Based on the gas component spectral change data, similar fitting of phase change conditions can help establish the relationship and change law between different gas components. This process can reveal the dynamic changes of gas components and analyze the potential correlation between microbial activity and air quality changes. By simulating the equal loss of microbial activity, the role of microorganisms in the air pollution process can be better understood, and the impact of microorganisms on air quality can be further predicted, thereby providing a scientific basis for environmental management of farms. By analyzing the equal loss data of microbial activity, the decline of the decomposition ability of microorganisms can be evaluated, revealing their role in air pollution. Microorganisms play a vital role in the air quality of farms, and changes in their activity will directly affect the concentration of harmful gases in the air. Based on the data of the decline of microbial decomposition ability, the trend of air deterioration in farms can be effectively identified, and potential air pollution problems can be predicted and discovered in a timely manner, thereby providing early warning for air quality management in farms. This analysis provides an important scientific basis for environmental control in farms and helps to take targeted improvement measures. The air environment detection model is constructed using the policy gradient algorithm, which can accurately identify the changing trend of air quality in farms based on real-time monitoring data. The construction of this model is based on a large amount of historical and real-time data, combined with advanced algorithm optimization, with strong predictive and adaptive capabilities, and can continuously adjust and optimize monitoring strategies. When the air quality shows an abnormal trend, the model can automatically identify and give an early warning, thereby providing decision support for farm managers. After uploading the model to the cloud platform, data sharing and remote monitoring can be realized, ensuring real-time detection and continuous optimization of the air environment in farms, improving farm management efficiency, and reducing human intervention. Therefore, the present invention is an optimization treatment of a traditional method for detecting the air environment of a livestock farm, which solves the problem that the traditional method for detecting the air environment of a livestock farm has inaccurate analysis of the microbial activity and decomposition capacity decline of the livestock farm, thereby causing large errors in air environment quality detection, improves the accuracy of the analysis of the microbial activity and decomposition capacity decline of the livestock farm, thereby reducing the error in air environment quality detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1A schematic diagram of the steps of a method for detecting the air environment of a livestock farm;

[0041] Figure 2 for Figure 1 Detailed implementation steps of step S2 in the flowchart;

[0042] Figure 3 for Figure 1 Detailed implementation steps of step S3 in FIG. DETAILED DESCRIPTION

[0043] See also Figures 1 to 3 , a method for detecting the air environment of a livestock farm, the method comprising the following steps:

[0044] Step S1: deploy a multi-spectral imaging sensor in a livestock farm, and collect air spectrum data in multiple time periods to obtain air spectra in multiple time periods; identify gas component types of the air spectra in multiple time periods to obtain gas component type data; perform time period feature sampling based on the gas component type data to obtain gas component spectrum change sampling data;

[0045] Step S2: performing similar fitting of phase change conditions based on the sampling data of gas component spectrum changes to obtain similar fitting data of phase change conditions between different gas components; performing equal loss of microbial activity simulation based on the similar fitting data of phase change conditions to obtain equal loss of microbial activity data;

[0046] Step S3: performing a microbial decomposition capacity decline analysis based on the microbial activity equal loss data to obtain microbial decomposition capacity decline data; identifying the air deterioration trend of the livestock farm based on the microbial decomposition capacity decline data to obtain air deterioration trend data;

[0047] Step S4: construct an air environment detection model for the air deterioration trend data based on the policy gradient algorithm to obtain an air environment detection model; send the air environment detection model to the cloud platform to perform air environment detection in the livestock farm.

[0048] In the embodiment of the present invention, reference Figure 1 The above is a schematic flow chart of the steps of a method for detecting the air environment of a livestock farm according to the present invention. In this example, the method for detecting the air environment of a livestock farm comprises the following steps:

[0049] Step S1: deploy a multi-spectral imaging sensor in a livestock farm, and collect air spectrum data in multiple time periods to obtain air spectra in multiple time periods; identify gas component types of the air spectra in multiple time periods to obtain gas component type data; perform time period feature sampling based on the gas component type data to obtain gas component spectrum change sampling data;

[0050] In an embodiment of the present invention, a plurality of multispectral imaging sensors are evenly deployed in a livestock farm, and the deployment density of the sensors is adjusted according to the area and terrain of the farm to ensure the comprehensiveness and representativeness of the collected data. The sensor model is selected to be suitable for detecting common harmful gases in the farm. For example, a multispectral camera whose spectral range covers visible light and near-infrared bands can be selected, and the sampling frequency of the multispectral imaging sensor is set, for example, data is collected once every hour, and air spectrum data of multiple time periods are continuously collected, for example, data is continuously collected for 24 hours to obtain air spectra of multiple time periods, and spectral analysis software, for example, ENVI (Environment for Visualizing Images) is used to perform time series change trend analysis on the air spectra of multiple time periods, and analyze the change rules of spectral data in different time periods, for example, analyze the concentration change trend of harmful gases such as ammonia and hydrogen sulfide, and use pattern recognition algorithms, for example, support vector machines (Support Vector Machine) or artificial neural networks (Artificial Neural Networks) (Artificial Neural Network)), the gas component type recognition is performed on the air spectrum with multi-time period change trend, and the gas components contained in the air, such as ammonia, hydrogen sulfide, methane, carbon dioxide, etc., are identified. Based on the gas component type data, the time period feature sampling is performed on the air spectrum with multi-time period change trend. For example, several bands with the most significant spectral changes in each time period are selected as characteristic bands, and the gas component spectral change sampling data is extracted for subsequent analysis.

[0051] Step S2: performing similar fitting of phase change conditions based on the sampling data of gas component spectrum changes to obtain similar fitting data of phase change conditions between different gas components; performing equal loss of microbial activity simulation based on the similar fitting data of phase change conditions to obtain equal loss of microbial activity data;

[0052] In an embodiment of the present invention, based on the sampling data of the spectral changes of the gas components, a similar fitting of the phase change conditions is first performed. In terms of operation, by introducing an algorithm that supports multi-substance interaction, such as multivariate regression analysis, the relationship between the phase changes of different gas components and environmental conditions (such as temperature, humidity, air pressure, etc.) is established, and the least squares method or genetic algorithm is used to fit the change trends of different gas components to obtain the phase change conditions of each gas under different environmental conditions, and similar fitting data of the phase change conditions between different gas components are obtained by fitting. On this basis, a simulation of equal loss of microbial activity is further performed. During the simulation process, a microbial decomposition activity model is used, such as a reaction kinetic model, combined with the changes in gas component concentrations, and the activity loss of microorganisms under different environmental conditions is simulated to obtain equal loss data of microbial activity. The data reflects the influence of different gas concentrations on microbial activity, and provides a basis for the subsequent analysis of microbial decomposition capacity.

[0053] Step S3: performing a microbial decomposition capacity decline analysis based on the microbial activity equal loss data to obtain microbial decomposition capacity decline data; identifying the air deterioration trend of the livestock farm based on the microbial decomposition capacity decline data to obtain air deterioration trend data;

[0054] In an embodiment of the present invention, based on the data of equal loss of microbial activity, a decay analysis of the microbial decomposition ability is then performed. By adopting a kinetic model of microbial growth and decay, such as a Monod model or a Logistic model, combined with the obtained data of equal loss of microbial activity, the change in the microbial decomposition ability is analyzed. During the analysis, microorganisms are cultured in a laboratory environment and their ability to decompose certain organic matter is monitored. At the same time, the microbial activity is dynamically monitored under different gas concentrations and environmental conditions, thereby obtaining the data of microbial decomposition ability decay. Further based on the data, an environmental monitoring model, such as a Kalman filter model, is used to identify the air deterioration trend of livestock farms. The air deterioration trend data can accurately predict the changing trend of air quality in different time periods by comprehensively analyzing the relationship between different gas components and microbial activity, combined with real-time monitoring data, and ultimately obtain the air deterioration trend data.

[0055] Step S4: construct an air environment detection model for the air deterioration trend data based on the policy gradient algorithm to obtain an air environment detection model; send the air environment detection model to the cloud platform to perform air environment detection in the livestock farm.

[0056] In an embodiment of the present invention, an air environment detection model is constructed for air deterioration trend data based on a policy gradient algorithm. First, data preprocessing is performed on historical air deterioration trend data to remove noise and abnormal data. The processed data is learned and optimized through a reinforcement learning algorithm (such as Q-learning or Deep Q Network, DQN). During the construction process, the gradient descent method is used to optimize the model so that the model can accurately predict different gas components and environmental changes. During the training process, the loss function is set as the difference between the predicted error and the actual error, and finally an optimized air environment detection model is obtained. After the model training is completed, the constructed air environment detection model is deployed through a cloud platform, and the model is uploaded to the cloud platform using a secure encryption protocol (such as HTTPS) so as to perform air environment monitoring tasks in livestock farms in real time. The cloud platform can automatically adjust the detection parameters and optimize the detection process according to the data fed back by the real-time sensors, so as to ensure that the air environment data of the livestock farms are continuously and effectively monitored.

[0057] Step S1 includes the following steps:

[0058] Step S11: deploying a multi-spectral imaging sensor in the livestock farm, and collecting air spectrum data in multiple time periods to obtain air spectra in multiple time periods;

[0059] Step S12: performing time series variation trend analysis on the air spectra in multiple time periods to obtain the air spectra with variation trends in multiple time periods;

[0060] Step S13: Identify the gas component type of the air spectrum with the changing trend in multiple time periods to obtain gas component type data;

[0061] Step S14: based on the gas component type data, the air spectrum with multiple time period change trends is sampled for time period characteristics to obtain gas component spectrum change sampling data.

[0062] In an embodiment of the present invention, in a livestock farm, multispectral imaging sensors are scientifically and reasonably deployed according to the scale of breeding, the type of livestock and poultry, and the complexity of the environment. For example, in a large pig farm, a multispectral imaging sensor can be deployed every 10 meters along key locations such as pig house corridors, vents, and manure treatment areas. The sensor model is selected to be suitable for detecting common harmful gases in the farm. For example, a multispectral camera with a spectral range covering visible light and near-infrared bands is selected, and the sampling frequency of the multispectral imaging sensor is set. For example, data is collected once every hour, and air spectrum data of multiple time periods are continuously collected. For example, data is continuously collected for 24 hours to obtain air spectra in multiple time periods. To ensure data quality, the sensor needs to be calibrated regularly during the collection process, and auxiliary information such as ambient temperature and humidity is recorded. The data collected by the multispectral imaging sensor is stored in the form of digital images or spectral data, for example, stored as .tif or .mat files. The file name contains the collection timestamp information to facilitate subsequent data processing and analysis. After the collection is completed, the multispectral image or spectral data is transmitted to the data processing center for subsequent analysis. Use spectral analysis software, such as ENVI (Environment for Visualizing Images) or ASD (Analytical Spectral Devices SpectraGryph software (spectral analysis equipment) is used to perform time series trend analysis on air spectra in multiple time periods, and analyze the changing rules of spectral data in different time periods, for example, analyzing the concentration trend of harmful gases such as ammonia and hydrogen sulfide. The specific operations include: first, preprocessing the multi-spectral data, including removing noise, correcting spectral distortion, etc. Then, selecting the spectral bands of interest, such as the characteristic absorption bands of ammonia and hydrogen sulfide, and drawing a curve of spectral intensity changes over time to observe the concentration trend of different gases. For example, the concentration of ammonia increases during the day and decreases at night, and the concentration of hydrogen sulfide increases when manure is cleaned. By analyzing these changing trends, we can preliminarily understand the air quality status of the farm and provide a basis for subsequent gas component identification and feature extraction. A variety of methods can be used for time series trend analysis, such as wavelet analysis, Fourier transform, etc. The selection of an appropriate method depends on the characteristics of the data and the purpose of the analysis. The analysis results are presented in the form of charts or data, for example, drawing a curve of spectral intensity changes over time, or outputting spectral characteristic parameters for different time periods.A pattern recognition algorithm, such as a support vector machine (SVM) or an artificial neural network (ANN), is used to identify the gas component type of the air spectrum with a multi-period change trend, and identify the gas components contained in the air, such as ammonia, hydrogen sulfide, methane, carbon dioxide, etc. The specific operations include: first, establishing a gas component spectrum database, collecting standard spectrum data of known gases as training samples, and then using the training samples to train a pattern recognition model, such as a support vector machine or an artificial neural network. The trained model can classify unknown spectrum data and identify the gas components contained therein. The accuracy of gas component type recognition depends on the quality and quantity of the training samples, as well as the selection of the pattern recognition algorithm. In order to improve the recognition accuracy, a variety of pattern recognition algorithms can be used for comparison, and the optimal algorithm can be selected. The recognition results are output in the form of labels or probabilities, for example, labeling the spectrum data as "ammonia", "hydrogen sulfide", etc., or outputting the probability that the spectrum data belongs to different gas components. The recognition results can be used for subsequent gas concentration estimation and air quality assessment. Based on the gas component type data, the air spectrum with multi-period change trends is sampled for time period characteristics to obtain the sampling data of gas component spectrum change. The specific operations include: first, according to the gas component type identification result, the gas components of interest are selected, such as ammonia and hydrogen sulfide. Then, the spectral data of each time period is extracted for characteristics, such as extracting the maximum, minimum, average, variance, etc. of the spectral intensity, or extracting the spectral intensity of a specific band, such as the characteristic absorption band of ammonia and hydrogen sulfide. The extracted characteristic parameters can form a vector as the sampling data of the gas component spectrum change in the time period. The purpose of time period characteristic sampling is to reduce the amount of data, extract key information, and facilitate subsequent analysis. The sampling method can be selected according to the characteristics of the data and the purpose of the analysis. For example, equal-interval sampling, random sampling, or sampling method based on information entropy can be used. The sampling results are stored in the form of tables or data, for example, stored as .csv or .txt files. The file content includes timestamps and extracted characteristic parameters. The sampling data can be used for subsequent microbial activity loss simulation and air quality assessment.

[0063] Step S2 includes the following steps:

[0064] Step S21: analyzing the concentration variation between different gas components on the gas component spectrum variation sampling data to obtain the concentration variation data between different gas components;

[0065] Step S22: performing phase change intensity coupling based on the concentration change data between different gas components to obtain phase change intensity coupling data between different gas components;

[0066] Step S23: calculating the toxic coupling release index between different gas components on the phase change strip intensity coupling data to obtain the toxic coupling release index between different gas components;

[0067] Step S24: performing a simulation of the equivalent loss of microbial activity according to the toxicity coupling release index and the phase change strip intensity coupling data to obtain the equivalent loss data of microbial activity.

[0068] As an example of the present invention, refer to Figure 2 As shown, in this example, step S2 includes:

[0069] Step S21: analyzing the concentration variation between different gas components on the gas component spectrum variation sampling data to obtain the concentration variation data between different gas components;

[0070] In an embodiment of the present invention, the concentration change analysis between different gas components is performed on the gas component spectral change sampling data, and the gas component spectral data is deeply analyzed by using a data mining method. First, the sampling data is preprocessed, including removing outliers and noise, and the noise part in the data is smoothed by a locally weighted regression method. Then, by calculating the concentration change of each gas component in different time periods, the concentration fluctuations of the gas components are compared using a dynamic time warping (DTW) algorithm, and the synchronization and deviation degree of the concentration change of each gas component in different time periods are analyzed. The DTW algorithm can process irregular time series data, and by comparing the concentration change curves of each gas component in different time periods, the concentration change data between different gas components are obtained. These data reflect the changes in the concentration of each gas component in multiple time periods, especially the relationship and influence between the concentrations of each gas component, and provide a data basis for subsequent phase change analysis.

[0071] Step S22: performing phase change intensity coupling based on the concentration change data between different gas components to obtain phase change intensity coupling data between different gas components;

[0072] In an embodiment of the present invention, based on the concentration change data between different gas components, phase change intensity coupling is performed to obtain phase change intensity coupling data between different gas components. The specific operations include: first, obtaining the ambient temperature and humidity data of the breeding site within the time period, for example, through a temperature and humidity sensor, the data format is a table or a matrix, including a timestamp, temperature and humidity values, and then, using a gas adsorption model, for example, a Langmuir model or a Freundlich model, to analyze the component particle adsorption coupling effect on the concentration change data between different gas components, and analyze the effect of particles on the adsorption of gas components, for example, calculate the adsorption amount of particles on gases such as ammonia and hydrogen sulfide, the adsorption model parameters need to be calibrated according to actual conditions, for example, by experimentally determining the adsorption constant, and then, based on the ambient temperature and humidity data, using a thermodynamic model, for example, the Clausius-Clapeyron equation or the Antoine equation, to perform an approximate analysis of the phase change temperature / humidity / pressure conditions on the concentration change data between different gas components. , analyze the effects of temperature and humidity on the phase changes of gas components, for example, calculate the saturated vapor pressure of gases such as ammonia and hydrogen sulfide at different temperatures and humidities, the parameters of the thermodynamic model need to be determined according to the gas properties, for example, the molar heat of vaporization in the Clausius-Clapeyron equation, and finally, based on the particle adsorption coupling data and the approximate data of the temperature / humidity / pressure conditions of the phase change, use a coupling model, for example, a multi-factor weighted model or a neural network model, to couple the intensity of the phase change and calculate the effects of the interactions between different gas components on the phase change. For example, when ammonia and hydrogen sulfide exist at the same time, there is a synergistic effect between them, which affects each other's phase changes. The coupling model parameters need to be determined according to experimental data or experience, for example, the weights of different factors, and the phase change intensity coupling results are presented in the form of numerical values ​​or matrices, for example, outputting the interaction intensity between different gas components, or constructing a phase change intensity matrix. The phase change intensity coupling data can be used for subsequent toxic coupling release index calculations and simulations of equivalent losses of microbial activity.

[0073] Step S23: calculating the toxic coupling release index between different gas components on the phase change strip intensity coupling data to obtain the toxic coupling release index between different gas components;

[0074] In an embodiment of the present invention, the intensity coupling data of the phase change strip is used to calculate the toxic coupling release index between different gas components using a toxicology model, such as a dose-effect model or a toxicity index model, and calculate the degree of toxicity of different gas components on microorganisms. The specific operations include: first, establishing a gas toxicity database, collecting toxicity data of known gases, such as the median lethal concentration (LC50 (Lethal Concentration 50 (half lethal concentration))), the median inhibitory concentration (IC50 (Half maximal inhibitory concentration (half inhibitory concentration))), etc., and then, according to the intensity coupling data of the phase change strip, determining the concentration and interaction of the gas components, and then, using the toxicology model, calculating the degree of toxicity of different gas components on microorganisms. For example, according to the dose-effect model, the toxic reaction of different concentrations of gas to microorganisms is calculated, or according to the toxicity index model, the toxicity index of different gas components is calculated. The toxic coupling release index can reflect the toxicity of the gas to the microorganism and the release risk. For example, the higher the index, the greater the toxicity and the higher the release risk. The toxic coupling release index calculation can also use other methods, such as quantitative structure-activity relationship (QSAR) (Quantitative Structure-Activity Relationship) model, computational toxicology model, etc. The selection of an appropriate method depends on the characteristics of the data and the purpose of the analysis. The toxic coupling release index is presented in the form of a numerical value or a matrix. For example, the toxicity index of different gas components is output, or a toxicity index matrix is ​​constructed. The toxic coupling release index can be used for subsequent simulation of equal loss of microbial activity.

[0075] In another embodiment, a gas toxicity database is first established through HowNet and other data documents. The database contains toxicity information of common gases in livestock farms, including ammonia, hydrogen sulfide, methane, carbon dioxide, etc. The toxicity information uses LC50 (half-lethal concentration), which represents the gas concentration that causes 50% of experimental animals to die within a specific time. For example, the LC50 of ammonia is 200ppm (4 hours), and the LC50 of hydrogen sulfide is 700ppm (1 hour). The database also contains synergistic toxicity coefficients between gases. The synergistic toxicity coefficient describes the enhancement or reduction effect of toxicity when different gases are mixed. For example, when ammonia and hydrogen sulfide are mixed, synergistic toxicity is generated, resulting in a toxicity greater than the sum of the toxicities of the two gases acting alone. Based on the phase change intensity coupling data, the toxicity coupling release index between different gas components is calculated. The calculation formula is as follows: Toxicity coupling release index = (gas concentration × gas toxicity coefficient × synergistic toxicity coefficient × phase change intensity), where the gas concentration is provided by the gas concentration comparison module, the gas toxicity coefficient is provided by the gas toxicity database, the synergistic toxicity coefficient is provided by the gas toxicity database, and the phase change intensity is provided by the phase coupling unit. During the calculation process, the concentration ratio, phase change and synergistic toxicity effects between different gas components are considered, the toxicity of each gas component is weighted, and then the weighted toxicity is accumulated to obtain the final toxic coupling release index. For example, when calculating the toxic coupling release index under the conditions of an ammonia concentration of 15ppm, a hydrogen sulfide concentration of 3ppm, a methane concentration of 60ppm, and a carbon dioxide concentration of 450ppm, first query the gas toxicity database to obtain the toxicity coefficients and synergistic toxicity coefficients of various gases, and then determine the phase change intensity between various gases based on the phase change intensity coupling data. Finally, substitute these data into the formula to calculate the toxic coupling release index of 0.75, which reflects the overall toxicity level of the gas mixture under the current environmental conditions. The results of the toxic coupling release index will be used to simulate the equivalent loss of microbial activity.

[0076] Step S24: performing a simulation of the equivalent loss of microbial activity according to the toxicity coupling release index and the phase change strip intensity coupling data to obtain the equivalent loss data of microbial activity.

[0077] In an embodiment of the present invention, according to the toxic coupling release index and the phase change bar intensity coupling data, a microbial kinetic model, such as a Monod model or a Logistic model, is used to simulate the equivalent loss of microbial activity, simulate the effects of different gas components on microbial activity, and obtain the equivalent loss data of microbial activity. The specific operations include: first, determining the initial activity value of the microorganism, for example, by experimentally measuring the growth rate or metabolic rate of the microorganism, and then, according to the toxic coupling release index and the phase change bar intensity coupling data, determining the parameters affecting the gas component on the microbial activity, such as the inhibition constant, the toxicity coefficient, etc., and then, using the microbial kinetic model, simulating the change of microbial activity over time, for example, according to the Monod model, calculating the microbial activity under different gas concentrations. The growth rate of the microorganism can be calculated, or the population density of the microorganism under different gas concentrations can be calculated according to the Logistic model. The simulation of the equivalent loss of microbial activity can reflect the inhibitory effect and influence of gas on microorganisms. For example, it can simulate the inhibitory effect of different concentrations of ammonia on microbial growth, or simulate the influence of different concentrations of hydrogen sulfide on microbial metabolism. The simulation of the equivalent loss of microbial activity can also adopt other methods, such as metabolomics-based simulation, gene regulatory network-based simulation, etc. The selection of the appropriate method depends on the purpose of the study and the characteristics of the microorganism. The data of the equivalent loss of microbial activity are presented in the form of numerical values ​​or matrices. For example, the activity values ​​of microorganisms at different time points can be output, or a curve of the change of microbial activity can be drawn. The data of the equivalent loss of microbial activity can be used for the subsequent analysis of the decline of microbial decomposition ability.

[0078] In another embodiment, the toxic coupling release index is subjected to a release hysteresis analysis, and a time series analysis method, such as an autoregressive moving average (ARMA) model, is used to predict the hysteresis time of the toxic release. During the establishment of the ARMA model, historical toxic coupling release index data is required for training. The historical data includes the toxic coupling release index every 1 hour in the past 24 hours. Using these data, an ARMA model is established to predict the toxic coupling release index in the next hour, and the error between the predicted value and the actual value is calculated. If the error exceeds a preset threshold value (e.g., 10%), it is considered that there is a release hysteresis, and the hysteresis time is set to the time length that the error exceeds the threshold value. For example, it was found that the toxic coupling release index increased rapidly in a short period of time and then decreased slowly, indicating that there was a release hysteresis with a lag time of 30 minutes. The phase change intensity coupling data was used to calculate the phase transition specific heat capacity tolerance between different gas components. The calculation formula for the specific heat capacity tolerance is as follows: specific heat capacity tolerance = |(gas 1 specific heat capacity-gas 2 specific heat capacity) / (gas 1 specific heat capacity+gas 2 specific heat capacity)|, where the gas specific heat capacity is provided by the gas physical property database, which contains the physical property information of various common gases, including specific heat capacity, density, viscosity, thermal conductivity, etc. For example, the specific heat capacity of ammonia is 2.13 J / g·K, and the specific heat capacity of hydrogen sulfide is 1.02 J / g·K, the calculated specific heat tolerance of ammonia and hydrogen sulfide is 0.35. The specific heat tolerance reflects the difference in thermodynamic properties between different gas components. The larger the specific heat tolerance, the greater the difference in energy required by the two gases during phase transition. Based on the specific heat tolerance, the molecular group conformational changes between different gas components are simulated. The molecular dynamics method is used for the simulation of molecular group conformational changes. In specific implementation, a simulation box containing molecular models of different gas components is first constructed. The size of the simulation box is determined according to the gas concentration and density. The simulation box contains a sufficient number of gas molecules to ensure the accuracy of the simulation results. The interaction between molecules is described by a force field. The force field is a parameterized mathematical model used to describe the interaction potential between molecules. Commonly used force fields include AMBER , CHARMM, GROMOS, etc. The simulation was carried out under a constant temperature and constant pressure (NPT) ensemble. The temperature and pressure were set to the average temperature and pressure of the livestock farm, respectively. The simulation time was set to 10 nanoseconds and the time step was set to 1 femtosecond. During the simulation, the conformational changes of the molecular groups were recorded, including bond lengths, bond angles, dihedral angles, etc. By analyzing these data, we can understand how the interactions between different gas components affect the conformational changes of the molecular groups. The toxic release hysteresis data and the toxic coupling release index were used to perform nonlinear gradient analysis of toxic release. Numerical methods, such as the finite difference method, were used to calculate the gradient of toxic release. The gradient reflects the rate of change of toxic release over time. For example, the toxic release gradient increases rapidly in a short period of time, indicating that the toxic release rate is accelerated and a stronger toxic effect is produced on microorganisms.The molecular group conformational change data and toxic release nonlinear gradient data are used to evaluate the isocratic gradient of gas toxic molecule penetration enhancement. The evaluation adopts the dose-effect model. The dose-effect model describes the relationship between the dose of toxic substances and the biological effect. Commonly used dose-effect models include Logit model, Probit model, Weibull model, etc. Through the dose-effect model, the toxic effects of toxic substances on microorganisms under different doses and exposure times can be evaluated. For example, the inhibition rate of toxic substances on microorganisms is evaluated under the conditions of an ammonia concentration of 15ppm, a hydrogen sulfide concentration of 3ppm, and an exposure time of 24 hours. Based on the dose-effect model, the isocratic gradient of gas toxic molecule penetration enhancement is calculated. The isocratic gradient of gas toxic molecule penetration enhancement reflects the penetration ability and enhancement effect of toxic substances on microorganisms. The penetration ability refers to the ability of toxic substances to penetrate the cell membrane of microorganisms, and the enhancement effect refers to the ability of toxic substances to enhance the toxic effect on microorganisms. By evaluating the isocratic gradient of gas toxic molecule penetration enhancement, we can understand the toxicity of toxic substances. The potential harm of toxic substances to microorganisms is simulated based on the isotropic gradient of gas toxic molecule penetration enhancement. The simulation adopts the cellular automaton model. The cellular automaton is a discrete dynamic system composed of a series of discrete cells. Each cell has a finite number of states. The state of the cell is updated according to certain rules. The cellular automaton can be used to simulate complex biological systems, such as the growth, death and interaction of microbial communities. In the simulation of isotropic loss of microbial activity, the microbial community in the livestock farm is divided into cells, each cell represents a microbial individual or a microbial population. The state of the cell includes activity, death, etc. The state of the cell is updated according to the isotropic gradient of gas toxic molecule penetration enhancement. For example, if the cell is exposed to a high concentration of toxic substances, the activity of the cell will decrease or even die. By simulating the changes in the cell state, the overall activity loss of the microbial community can be understood, and finally the isotropic loss data of microbial activity can be obtained. ,

[0079] Step S22 includes the following steps:

[0080] Step S221: Acquire the ambient temperature and humidity data within the time period;

[0081] Step S222: performing component-particle adsorption coupling analysis on the concentration variation data between different gas components to obtain particle adsorption coupling data;

[0082] Step S223: performing phase change temperature / humidity / pressure condition approximate analysis on the concentration change data between different gas components according to the ambient temperature and humidity data, and obtaining phase change temperature / humidity / pressure condition approximate data;

[0083] Step S224: performing phase change intensity coupling based on the particle adsorption coupling effect data and the phase change temperature / humidity / pressure condition approximate data to obtain phase change intensity coupling data between different gas components.

[0084] In an embodiment of the present invention, before performing phase change intensity coupling, it is necessary to obtain the ambient temperature and humidity data of the breeding farm within a time period. The ambient temperature and humidity data can be obtained in a variety of ways, for example, a temperature and humidity sensor is set in the breeding farm to monitor the ambient temperature and humidity in real time, or the meteorological data of the area where the breeding farm is located is obtained from the meteorological department, including temperature, humidity, wind speed, air pressure, etc. The obtained ambient temperature and humidity data need to be preprocessed, for example, outliers are removed, data is smoothed, etc., to ensure the accuracy and reliability of the data. The ambient temperature and humidity data can be stored in the form of a table or data, for example, stored as a .csv or .txt file, and the file content includes a timestamp, a temperature value, and a humidity value. The accuracy of the timestamp can be adjusted according to actual needs, for example, accurate to minutes or hours, and the units of the temperature and humidity values ​​need to be clear, for example, the temperature unit is degrees Celsius (°C), and the humidity unit is percentage (%). The obtained ambient temperature and humidity data can be used for subsequent phase change temperature / humidity / pressure condition approximate analysis. The concentration change data between different gas components are analyzed for the coupling effect of component particle adsorption to obtain the particle adsorption coupling effect data. The specific operations include: first, determining the type and concentration of particles in the farm air, for example, collecting air samples through a particle sampler, and then analyzing them using equipment such as a microscope or a laser scatterer to determine the particle size distribution and chemical composition of the particles. Common particles include dust, feed residues, feces particles, etc. Different particles have different adsorption capacities for gases, which need to be measured according to actual conditions. For example, the adsorption amount of ammonia, hydrogen sulfide and other gases by particles is determined experimentally. Then, based on the concentration change data between different gas components and the particle adsorption data, the coupling effect strength of the particle adsorption on the gas component is calculated. For example, an empirical formula or model can be used for calculation, considering factors such as particle concentration, particle size distribution, and gas component concentration. The particle adsorption coupling effect analysis results can be presented in the form of numerical values ​​or matrices, for example, the amount of different gas components adsorbed by particles is output, or a particle adsorption coupling effect matrix is ​​constructed. The particle adsorption coupling effect data can be used for subsequent phase change intensity coupling.According to the ambient temperature and humidity data, the concentration change data between different gas components are analyzed approximately under the temperature / humidity / pressure conditions of phase change to obtain the approximate data of the temperature / humidity / pressure conditions of phase change. The specific operations include: first, determining the phase change type of the gas component, such as gas dissolution, volatilization, condensation, etc. The phase changes of different gases are affected differently by temperature, humidity and pressure, and need to be analyzed according to the properties of the gas. For example, ammonia is easy to volatilize under high temperature and low humidity conditions, and hydrogen sulfide is easy to dissolve under low temperature and high humidity conditions. Then, according to the ambient temperature and humidity data, calculate the gas components under the current The phase change trend under the previous conditions, for example, can be calculated using thermodynamic formulas or empirical formulas, taking into account the effects of temperature, humidity and pressure on the gas phase change, for example, calculating the saturated vapor pressure of ammonia at different temperatures and humidities, or calculating the solubility of hydrogen sulfide at different temperatures and humidities. The approximate analysis results of the phase change temperature / humidity / pressure conditions can be presented in the form of numerical values ​​or matrices, for example, outputting the phase change trends of different gas components under the current conditions, or constructing a phase change trend matrix. The approximate data of the phase change temperature / humidity / pressure conditions can be used for subsequent phase change intensity coupling. Based on the particle adsorption coupling data and the approximate data of phase change temperature / humidity / pressure conditions, phase change intensity coupling is performed to obtain the phase change strip intensity coupling data between different gas components. The specific operations include: first, determining whether there is interaction between different gas components, for example, ammonia and hydrogen sulfide have a synergistic effect, affecting each other's phase change. Then, based on the particle adsorption coupling data and the approximate data of phase change temperature / humidity / pressure conditions, calculate the effect of the interaction between different gas components on the phase change. For example, a coupling model or empirical formula can be used for calculation, considering factors such as particle adsorption, temperature, humidity, and pressure. The phase change intensity coupling results can be presented in the form of numerical values ​​or matrices, for example, outputting the interaction intensity between different gas components, or constructing a phase change intensity matrix. The phase change strip intensity coupling data can be used for subsequent toxic coupling release index calculations.

[0085] Step S24 includes the following steps:

[0086] Step S241: performing release hysteresis analysis on the toxicity coupling release index to obtain toxicity release hysteresis data;

[0087] Step S242: calculating the specific heat tolerance between different gas components on the phase change strip intensity coupling data to obtain the phase conversion specific heat tolerance between different gas components;

[0088] Step S243: performing a simulation of the molecular group conformational change between different gas components on the phase change strip intensity coupling data based on the phase conversion specific heat tolerance to obtain the molecular group conformational change data between different gas components;

[0089] Step S244: performing a toxicity release nonlinear gradient analysis according to the toxicity release hysteresis data and the toxicity coupling release index to obtain toxicity release nonlinear gradient data;

[0090] Step S245: performing gas toxicity molecular penetration enhancement isoquant gradient evaluation based on the molecular group conformation change data and the toxicity release nonlinear gradient data to obtain molecular toxicity penetration enhancement isoquant data;

[0091] Step S246: Perform a simulation of the equivalent loss of microbial activity based on the molecular toxicity penetration enhancement equivalent data to obtain the equivalent loss data of microbial activity.

[0092] In an embodiment of the present invention, a release hysteresis analysis is performed on the toxic coupling release index. First, a time series analysis is performed on the toxic coupling release index data of each gas component. A time series autoregressive integrated moving average (ARIMA) model is used to analyze the hysteresis effect of the toxic coupling release index. First, the data is detrended and differentially processed to convert the series into a stable state. Then, the ARIMA model is fitted by selecting a suitable order to calculate the release hysteresis of each type of gas component. The toxic release hysteresis data of each gas component at different time points are calculated according to the model fitting results. The data contains a specific time window for the release hysteresis of each gas, and identifies the law of the hysteresis release. This data provides a time influencing factor for subsequent toxicity evaluation and microbial activity simulation. Parameters need to be adjusted in the model to ensure matching with actual monitoring results. The specific heat capacity tolerance between different gas components is calculated for the phase change intensity coupling data. First, the specific heat capacity (Heat Capacity) data of each gas component under different environmental conditions are collected, and the specific heat capacity of the gas at different pressures and temperatures is determined by an experimental method using a thermal analysis instrument. The selected gas components include ammonia, methane, carbon dioxide, etc. In the experiment, the specific heat capacity of each gas is calculated according to the temperature fluctuations of different gas components during the phase change process. Then, the tolerance calculation is performed on the specific heat capacity data of each pair of gas components. The specific heat capacity tolerance between different gas components during phase change is calculated by the Absolute Difference Method, and the phase conversion specific heat capacity tolerance data between each pair of gas components is obtained. This data reflects the difference in heat changes of different gases during phase change, and provides a physical parameter basis for the subsequent simulation of molecular group conformational changes. Based on the phase transition specific heat capacity tolerance, the phase change intensity coupling data is used to simulate the conformational changes of molecular groups between different gas components. First, according to the molecular structure characteristics of the gas components, the molecular dynamics simulation method is used to simulate the conformational changes of different gas components. The initial temperature and pressure conditions are set, and the specific heat capacity tolerance data is input as the thermodynamic input parameter to calculate the molecular motion trajectory. During the simulation, each molecule moves freely and collides with each other according to the laws of thermodynamics and quantum mechanics. The interaction force between molecules and its influence on the molecular conformation are recorded. Through long-term simulation calculations, the conformational change data of different gas molecular groups during the phase change process are obtained. The data includes the conformational change path of gas molecules during phase transition, the mechanical interaction between molecules and its change trend, which provides more accurate molecular-level data for subsequent gas toxicity assessment.Nonlinear gradient analysis of toxic release was performed based on the toxic release hysteresis data and the toxic coupling release index. First, the toxic coupling release index data and the toxic release hysteresis data were combined to construct a nonlinear gradient analysis model. The support vector machine (SVM) method was used for data classification and regression analysis. The support vector machine can capture the nonlinear characteristics of the toxic release process by selecting a suitable kernel function to perform high-dimensional mapping of the data. The model takes the time information of the toxic release hysteresis period and the size of the toxicity index as input, and obtains nonlinear gradient data through training. The data reflects the gradient changes in the toxic release process and can characterize the nonlinear effects of gas concentration changes on air quality and microbial activity. Finally, the nonlinear gradient data of toxic release are obtained, and the nonlinear law of toxic release is analyzed based on the data. Based on the molecular group conformational change data and the nonlinear gradient data of toxicity release, an isoquantitative gradient evaluation of gas toxicity molecular penetration enhancement is performed. First, the Gas Permeation Model is used to take the molecular group conformational change data as input to simulate the penetration behavior of gas in the air. A mathematical model of the gas molecule penetration enhancement effect is established, and the nonlinear gradient data is combined with the molecular conformational change data to conduct a multi-factor comprehensive analysis to calculate the penetration effect of gas under different time and space conditions. The model considers the size and shape of gas molecules as well as the effects of temperature and humidity changes in the environment on the penetration capacity. The penetration enhancement effect of gas is obtained through calculation. The data reflects the enhancement trend of gas penetration capacity and toxic diffusion under different conditions, and provides detailed data support for evaluating the impact of gas in the environment. Based on the molecular toxicity penetration enhancement equivalent data, the simulation of the equivalent loss of microbial activity is carried out. First, the molecular toxicity penetration enhancement equivalent data is used as input by using the Bioreaction Kinetics Model to simulate the gas damage process on the activity of the microbial community. The model considers the effect of gas penetration on the microbial cell membrane, and calculates the biodegradation rate after gas molecule penetration based on the penetration data. Combined with the adaptability of the microbial community to environmental changes, simulation calculations are carried out, and finally the data on the impact of each gas component on the equivalent loss of microbial activity are obtained. The simulation results show how the permeability of different gas components affects the activity of microorganisms and provide an important reference for the monitoring of air environmental quality in farms.

[0093] Step S245 includes the following steps:

[0094] Marking the rising / falling inflection points of the nonlinear gradient data of toxic release to obtain nonlinear rising / falling inflection point marking data;

[0095] According to the nonlinear rising / falling inflection point marking data, the release concentration similarity interval convergence calculation is performed on the toxic release nonlinear gradient data to obtain the release concentration similarity convergence interval;

[0096] Based on the similar convergence interval of release concentration and the chord intercept method, the toxicity release nonlinear gradient data are subjected to approximate iteration of toxicity concentration release, and the approximate iteration data of toxicity concentration release are obtained;

[0097] Perform dynamic enhancement analysis on the interaction potential energy of molecular group conformational change data to obtain dynamic enhancement data of group potential energy;

[0098] Based on the group potential dynamic enhancement data and the approximate iteration data of toxic concentration release, an isoquant gradient evaluation of gas toxicity molecular penetration enhancement was performed to obtain the molecular toxicity penetration enhancement isoquant data.

[0099] In an embodiment of the present invention, the rising / falling inflection point of the toxic release nonlinear gradient data is marked. First, the toxic release nonlinear gradient data is sorted according to the time series, and each data point is compared with its adjacent front and rear points to determine their relative size relationship. If the current data point is larger than the front and rear points, the data point is a rising inflection point; if the current data point is smaller than the front and rear points, it is a falling inflection point. The data is smoothed using a sliding window algorithm to avoid interference from abnormal data points, and the time and position of the rising and falling inflection points are marked to obtain nonlinear rising / falling inflection point marking data, which will be used for subsequent concentration change interval analysis. Based on the nonlinear rising / falling inflection point marking data, the release concentration similarity interval convergence calculation is performed on the toxic release nonlinear gradient data. First, according to the time difference between the rising and falling inflection points, the concentration change trend in each toxic release cycle is determined, and the starting and ending points of each concentration change cycle are used as the calculation interval. The concentration data in each interval is standardized, and the concentration curve is fitted using the least squares method to calculate the similarity of the concentration change in each interval. According to the fitting results, the data intervals with similar concentration changes are aggregated to obtain the release concentration similarity convergence interval. This operation requires repeated calculations in multiple time periods to ensure that the obtained intervals have high accuracy and consistency. This data provides stable reference data for subsequent concentration release approximation iterations. Based on the release concentration similarity convergence interval and secant method, the toxic concentration release approximation iteration is performed on the toxic release nonlinear gradient data. First, the starting and ending points of each concentration similarity interval are determined, and the nonlinear gradient data is approximated by the secant method. The secant method approximates the current nonlinear gradient by using the line between two points in each iteration, and repeatedly corrects it according to the data in the similar interval until the preset accuracy is reached. The maximum number of iterations is set to 500 times, and the minimum error tolerance is 0.001. During the iteration process, each calculation result is updated to the next input data until the concentration release approximation converges. This data can provide an accurate approximation of the toxic concentration release, and the concentration release model can be further optimized by comparing the iteration results of different time periods. The dynamic enhancement analysis of the interaction potential energy of the molecular group conformational change data is carried out. First, the main components of the gas molecular groups are selected. According to the molecular dynamics (MD) simulation results, the interaction potential energy of the molecules under different temperature and pressure conditions is analyzed, and the interaction force between the gas molecular groups is calculated. By calculating the energy function, the potential energy change between the molecular groups is obtained. During the simulation process, the temperature range is set to 300K to 500K, the pressure range is 1MPa to 3MPa, and the duration of each experiment is 10,000 simulation steps. By sampling the molecular states at different time points, the dynamic enhancement data of the group potential energy are obtained. This data reflects the energy fluctuation and change law of the gas molecular groups under different environmental conditions.Based on the group potential energy dynamic enhancement data and the approximate iterative data of toxic concentration release, an isoquant gradient evaluation of gas toxicity molecular penetration enhancement is performed. First, the molecular penetration model is used to evaluate gas toxicity in combination with the potential energy dynamic enhancement data and the approximate iterative data of toxic concentration release. The permeation rate and diffusion coefficient of gas molecules are set. In each concentration release interval, the penetration ability of gas molecules in the microbial cell membrane is evaluated. The classical diffusion equation is used to calculate the adsorption rate and permeation rate of molecules on the membrane surface. The potential energy data and concentration data are input to simulate the gas permeation process. Finally, the molecular toxicity penetration enhancement isoquant data is obtained. This data reflects how gas molecules diffuse through the cell membrane and affect microbial activity under different toxic concentrations and group conformations, providing a scientific basis for subsequent air quality management and environmental governance.

[0100] Step S3 includes the following steps:

[0101] Step S31: performing loss convolution processing on the microbial activity equal loss data to obtain microbial activity equal loss convolution data;

[0102] Step S32: performing microbial decomposition capacity decay analysis based on the microbial activity equal loss convolution data to obtain microbial decomposition capacity decay data;

[0103] Step S33: Identify the air deterioration trend of the livestock farm based on the microbial decomposition capacity decline data to obtain air deterioration trend data.

[0104] As an example of the present invention, refer to Figure 3 As shown, in this example, step S3 includes:

[0105] Step S31: performing loss convolution processing on the microbial activity equal loss data to obtain microbial activity equal loss convolution data;

[0106] In an embodiment of the present invention, loss convolution processing is performed on the microbial activity equal loss data. First, the microbial activity equal loss data is smoothed by applying a convolution operation, and a suitable convolution kernel is selected for calculation. The design of the convolution kernel should be based on the time series characteristics of the microbial activity loss. Commonly used convolution kernels include Gaussian kernel function (GaussianKernel) and mean filter kernel (Mean Filter Kernel), wherein the Gaussian kernel function can be used to emphasize the changes in a larger time window, and the mean filter kernel is used to smooth rapid fluctuations. In the convolution process, the data points are sampled in the neighborhood, and the weighted sum algorithm is used for convolution calculation to obtain the microbial activity equal loss convolution data, which contains the dynamic change trend of the microbial activity loss in different time windows. For the sampling time point, the size of the convolution window is set to 1000 minutes. After multiple debugging and verification, it is ensured that the accuracy of the convolution processing meets the requirements.

[0107] Step S32: performing microbial decomposition capacity decay analysis based on the microbial activity equal loss convolution data to obtain microbial decomposition capacity decay data;

[0108] In an embodiment of the present invention, a microbial decomposition capacity decay analysis is performed based on the convolution data of equal loss of microbial activity. First, based on the convolved microbial activity data, a decomposition model is used to perform a decay analysis on the microbial decomposition capacity. Model parameters are set, including a decomposition rate constant and an environmental factor influence coefficient. These parameters are obtained through multiple experimental calibrations. The analysis method used for data processing is a weighted regression analysis. The time series data is analyzed. The rate of change of the microbial decomposition capacity is calculated to obtain the value of the microbial decomposition capacity decay. The decay value will gradually increase with the passage of time. The time range is set to 24 hours during the analysis. The sliding window technology is used to gradually analyze the changing trend of the decomposition capacity in different time periods, and finally the microbial decomposition capacity decay data is obtained. The data can reflect the process of the gradual decrease of microbial activity over time.

[0109] Step S33: Identify the air deterioration trend of the livestock farm based on the microbial decomposition capacity decline data to obtain air deterioration trend data.

[0110] In an embodiment of the present invention, the trend of air deterioration in a livestock farm is identified based on the microbial decomposition capacity decay data. First, a time series cluster analysis method is used to perform pattern recognition on the microbial decomposition capacity decay data. A K-means clustering algorithm is selected for processing, and the data is divided into multiple cluster groups, each group representing a specific decomposition capacity decay pattern. Combined with the existing gas concentration and microbial decomposition data, a prediction model for air quality deterioration is obtained by fitting a curve model. The analysis window is set to 3 days. Through real-time data collection and monitoring, the environmental factors such as the concentration of gas components, humidity, and temperature in the collected air are input into the model to calculate the trend value of air deterioration. Finally, the specific time period of air deterioration in the livestock farm is determined based on the trend data to obtain air deterioration trend data.

[0111] Step S32 includes the following steps:

[0112] Step S321: performing community loss dynamic estimation on the microbial activity equal loss convolution data to obtain microbial community loss dynamic data;

[0113] Step S322: deriving the coordinated weakening change of enzyme activity based on the dynamic data of microbial community loss and the convolution data of equal loss of microbial activity to obtain the weakening data of microbial enzyme activity;

[0114] Step S323: performing microbial decomposition ability decline analysis based on the microbial enzyme activity weakening data to obtain microbial decomposition ability decline data.

[0115] In an embodiment of the present invention, the convolution data of equal loss of microbial activity is used to dynamically estimate the community loss. First, the convolved microbial activity data is processed by a time series analysis method. The sliding window method is used to set the window size to 200 minutes, and the microbial community loss in different time windows is gradually calculated. The dynamic loss data of the microbial community is obtained by calculating the loss rate in each window. The data used are the microbial activity data at different time points and under different environmental conditions in the livestock farm. The data can reflect the loss of the microbial community under changes in air quality and environmental interference, and then deduce the trend of community activity over time. The dynamic data of community loss is optimized by a weighted average algorithm to maintain high accuracy and stability when there are many data outliers. All calculations and analyses are completed in real time by configuring an efficient sensor array and a data processing system to ensure the real-time and high precision of the data. Based on the dynamic data of microbial community loss and the convolution data of equal loss of microbial activity, the coordinated weakening changes of enzyme activity are deduced. First, the collaborative filtering algorithm is used to cross-analyze the dynamic data of microbial community loss and the microbial activity data to identify the change pattern between the two, and the relationship between the concentration change of each gas component and the loss of microbial activity is fitted to obtain the change deduction model of enzyme activity weakening. Through this model, combined with the time series characteristics of the change rate of microbial community loss and the convolution data of activity loss, the enzyme activity weakening data in different time periods are gradually derived. The data reflects the dynamic changes of enzyme activity in the microbial community and can be adjusted in real time according to the fluctuations in the concentration of environmental pollutants. The tools used include efficient sensors and reaction pool analysis systems. The sensor data is transmitted in real time through the data acquisition module and dynamically analyzed through the cloud computing platform. The analysis results provide data support for the changes in microbial enzyme activity.Based on the weakening data of microbial enzyme activity, the decline of microbial decomposition ability was analyzed. First, the adaptive filtering algorithm was used to process the weakening data of microbial enzyme activity to compensate for the data anomalies caused by fluctuations in environmental factors in real time. The microbial sample data collected in the laboratory were used for calibration, and the standard microbial decomposition ability and enzyme activity thresholds were set. According to different temperatures, humidity and gas concentrations, the decomposition ability parameters of the microbial community were gradually adjusted. It was verified through experiments that the microbial decomposition ability showed different decline patterns under different environmental conditions. Combined with the decline model, the quantitative analysis of the microbial decomposition ability was carried out, and the microbial decomposition ability decline data was obtained. The data reflected the decline of microbial decomposition ability caused by different gas components and changes in microclimate conditions in livestock farms. A multi-point detection system was used to monitor the air quality in different areas, and the decomposition ability decline value was updated in real time through high-frequency data sampling to ensure the accuracy and timeliness of the analysis results.

[0116] Step S4 includes the following steps:

[0117] Step S41: performing logic learning on the air deterioration trend data to obtain air deterioration trend logic data;

[0118] Step S42: constructing an air environment detection model based on the microbial decomposition capacity decay data and the air deterioration trend logic data based on the strategy gradient algorithm to obtain an air environment detection model;

[0119] Step S43: Send the air environment detection model to the cloud platform to perform air environment detection in the livestock farm.

[0120] In an embodiment of the present invention, logic learning is performed on air deterioration trend data, which include information such as different gas concentrations, temperature and humidity, and microbial activity. These multi-dimensional data are used to systematically analyze the air change trend in the livestock farm, and the time series analysis method is used to process the data, with special attention paid to the correlation between gas concentrations, humidity changes and microbial activity. The air quality change is logically deduced through an adaptive neural network. The neural network structure is set to a three-layer network, and the number of neurons in each layer is set to 256. After multiple iterative training, the change trend of air quality is obtained, and the key influencing factors are further extracted. Finally, the air deterioration trend logic data is output, which reveals the potential laws of air changes in the farm, including the change cycle of pollution sources and the time window for air quality deterioration. The logic learning process relies on a neural network model with data self-learning ability, and continuously corrects and optimizes the learning results through feedback from real-time environmental monitoring data. Based on the policy gradient algorithm, the air environment detection model is constructed for the microbial decomposition capacity decay data and the air deterioration trend logic data. First, the policy gradient algorithm in reinforcement learning is selected as the training basis of the model, and the microbial decomposition capacity decay data and the air deterioration trend logic data are used as input. By establishing the environmental state space and action space model, the dynamic change of air quality is taken as the state, and the relationship between the decay of microbial decomposition capacity and the change of ambient air quality is taken as the reward function. The maximum state switching amount of each time step is set to 0.5. Using the policy gradient algorithm, the model continuously interacts with the environment, adjusts the policy weight, and optimizes the accuracy of the air environment detection model. The gradient descent method is used to update the policy parameters. The model parameters are updated every 100 training cycles. Finally, a detection model that can monitor the air environment quality of the farm in real time is constructed. The model can output accurate air quality prediction results according to the changes in air quality and the decay of microbial decomposition capacity. The training process of the model is supported by large-scale historical data and simulation data to ensure its application effect in real scenarios.The air environment detection model is sent to the cloud platform to perform air environment detection in livestock farms. First, the air environment detection model is connected to the local data acquisition system through the API interface (Application Programming Interface). The model is uploaded to the cloud platform through the cloud transmission protocol (such as the MQTT protocol). The data is encrypted during the transmission process to ensure data security. After receiving the model, the cloud platform uses cloud computing resources for further data analysis and real-time reasoning. The monitoring equipment includes gas sensors, humidity sensors, temperature sensors, etc. The sensor data is transmitted to the cloud through the Internet of Things (IoT) devices. The cloud platform calls the air environment detection model for real-time analysis based on the uploaded data. The air quality status output by the model is presented to the farm managers through a graphical interface to ensure real-time monitoring and management. The cloud platform continuously stores the data for long-term trend analysis and prediction. After the model is deployed, the system will continue to track changes in the air environment and dynamically adjust environmental factors such as ventilation and humidity regulation in the farm to ensure that the air quality in the farm is within a safe range.

[0121] The present invention also provides an air environment detection system for a livestock farm, which is used to execute the air environment detection method for a livestock farm as described above. The air environment detection system for a livestock farm comprises:

[0122] The gas component type identification module is used to deploy multi-spectral imaging sensors in livestock farms, collect air spectrum data in multiple time periods, and obtain air spectra in multiple time periods; identify the gas component types of air spectra in multiple time periods to obtain gas component type data; perform time period feature sampling based on the gas component type data to obtain gas component spectrum change sampling data;

[0123] The microbial activity equivalent loss analysis module is used to perform similar fitting of phase change conditions based on the sampling data of gas component spectrum changes to obtain similar fitting data of phase change conditions between different gas components; perform microbial activity equivalent loss simulation based on the similar fitting data of phase change conditions to obtain microbial activity equivalent loss data;

[0124] The air deterioration trend identification module is used to analyze the decline of microbial decomposition ability based on the microbial activity equal loss data to obtain the microbial decomposition ability decline data; the air deterioration trend of the livestock farm is identified based on the microbial decomposition ability decline data to obtain the air deterioration trend data;

[0125] The model building detection module is used to build an air environment detection model for air deterioration trend data based on the policy gradient algorithm to obtain an air environment detection model; the air environment detection model is sent to the cloud platform to perform air environment detection in livestock farms.

[0126] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A method for detecting the air environment of a livestock farm, characterized in that: The following steps are involved: Step S1: deploying a multispectral imaging sensor in a livestock farm and collecting air spectrum data in multiple time periods to obtain air spectra in multiple time periods; Identify the gas component type of the air spectrum in multiple time periods to obtain gas component type data; Based on the gas component type data, time period feature sampling is performed to obtain gas component spectrum change sampling data; Step S2: performing similar fitting of phase change conditions based on the sampling data of gas component spectrum changes to obtain similar fitting data of phase change conditions between different gas components; performing equal loss of microbial activity simulation based on the similar fitting data of phase change conditions to obtain equal loss of microbial activity data; Step S3: performing a microbial decomposition capacity decline analysis based on the microbial activity equal loss data to obtain microbial decomposition capacity decline data; identifying the air deterioration trend of the livestock farm based on the microbial decomposition capacity decline data to obtain air deterioration trend data; Step S4: construct an air environment detection model for the air deterioration trend data based on the policy gradient algorithm to obtain an air environment detection model; send the air environment detection model to the cloud platform to perform air environment detection in the livestock farm.

2. The method for detecting the air environment of a livestock farm according to claim 1, characterized in that: Step S1 includes the following steps: Step S11: deploying a multi-spectral imaging sensor in the livestock farm, and collecting air spectrum data in multiple time periods to obtain air spectra in multiple time periods; Step S12: performing time series variation trend analysis on the air spectra in multiple time periods to obtain the air spectra with variation trends in multiple time periods; Step S13: Identify the gas component type of the air spectrum with the changing trend in multiple time periods to obtain gas component type data; Step S14: based on the gas component type data, the air spectrum with multiple time period change trends is sampled for time period characteristics to obtain gas component spectrum change sampling data.

3. The method for detecting the air environment of a livestock farm according to claim 1, characterized in that: Step S2 includes the following steps: Step S21: analyzing the concentration variation between different gas components on the gas component spectrum variation sampling data to obtain the concentration variation data between different gas components; Step S22: performing phase change intensity coupling based on the concentration change data between different gas components to obtain phase change intensity coupling data between different gas components; Step S23: calculating the toxic coupling release index between different gas components on the phase change strip intensity coupling data to obtain the toxic coupling release index between different gas components; Step S24: performing a simulation of the equivalent loss of microbial activity according to the toxicity coupling release index and the phase change strip intensity coupling data to obtain the equivalent loss data of microbial activity.

4. The method for detecting the air environment of a livestock farm according to claim 3, characterized in that: Step S24 includes the following steps: Step S241: performing release hysteresis analysis on the toxicity coupling release index to obtain toxicity release hysteresis data; Step S242: calculating the specific heat tolerance between different gas components on the phase change strip intensity coupling data to obtain the phase conversion specific heat tolerance between different gas components; Step S243: performing a simulation of the molecular group conformational change between different gas components on the phase change strip intensity coupling data based on the phase conversion specific heat tolerance to obtain the molecular group conformational change data between different gas components; Step S244: performing a toxicity release nonlinear gradient analysis according to the toxicity release hysteresis data and the toxicity coupling release index to obtain toxicity release nonlinear gradient data; Step S245: performing gas toxicity molecular penetration enhancement isoquant gradient evaluation based on the molecular group conformation change data and the toxicity release nonlinear gradient data to obtain molecular toxicity penetration enhancement isoquant data; Step S246: Perform a simulation of the equivalent loss of microbial activity based on the molecular toxicity penetration enhancement equivalent data to obtain the equivalent loss data of microbial activity.

5. The method for detecting the air environment of a livestock farm according to claim 1, characterized in that: Step S3 includes the following steps: Step S31: performing loss convolution processing on the microbial activity equal loss data to obtain microbial activity equal loss convolution data; Step S32: performing microbial decomposition capacity decay analysis based on the microbial activity equal loss convolution data to obtain microbial decomposition capacity decay data; Step S33: Identify the air deterioration trend of the livestock farm based on the microbial decomposition capacity decline data to obtain air deterioration trend data.

6. The method for detecting the air environment of a livestock farm according to claim 1, characterized in that: Step S4 includes the following steps: Step S41: performing logic learning on the air deterioration trend data to obtain air deterioration trend logic data; Step S42: constructing an air environment detection model based on the microbial decomposition capacity decay data and the air deterioration trend logic data based on the strategy gradient algorithm to obtain an air environment detection model; Step S43: Send the air environment detection model to the cloud platform to perform air environment detection in the livestock farm.

7. An air environment detection system for a livestock farm, characterized in that: Used to perform the air environment detection method of a livestock farm as claimed in claim 1, the air environment detection system of the livestock farm comprises: The gas component type identification module is used to deploy multi-spectral imaging sensors in livestock farms, collect air spectrum data in multiple time periods, and obtain air spectra in multiple time periods; identify the gas component types of air spectra in multiple time periods to obtain gas component type data; perform time period feature sampling based on the gas component type data to obtain gas component spectrum change sampling data; The microbial activity equivalent loss analysis module is used to perform similar fitting of phase change conditions based on the sampling data of gas component spectrum changes to obtain similar fitting data of phase change conditions between different gas components; perform microbial activity equivalent loss simulation based on the similar fitting data of phase change conditions to obtain microbial activity equivalent loss data; The air deterioration trend identification module is used to analyze the decline of microbial decomposition ability based on the microbial activity equal loss data to obtain the microbial decomposition ability decline data; the air deterioration trend of the livestock farm is identified based on the microbial decomposition ability decline data to obtain the air deterioration trend data; The model building detection module is used to build an air environment detection model for air deterioration trend data based on the policy gradient algorithm to obtain an air environment detection model; the air environment detection model is sent to the cloud platform to perform air environment detection in livestock farms.

Images