Temperature monitoring and control system in goat breeding house

Abstract

The invention discloses a temperature monitoring and control system in a goat breeding house. The system comprises an environmental data sensing unit, an edge data preprocessing unit, an optimized long-short term memory neural network processing unit, an air internal and external circulation regulation and control strategy generation unit, an equipment linkage execution unit and a system operation state monitoring unit. Environmental data in a breeding house is collected in a self-adaptive mode through multiple types of high-precision sensors, after edge processing, temperature spatial-temporal characteristics are extracted deeply through an optimized long-short-term memory neural network, an air circulation regulation and control strategy is generated in combination with a multi-target optimization model, equipment is driven to execute precisely, and system dynamic optimization is achieved through closed-loop feedback. The system overcomes the defects of insufficient monitoring coverage and extensive regulation and control in the prior art, considers the coupling relationship between temperature and air components, realizes multi-factor collaborative accurate regulation and control, effectively reduces energy consumption, creates a stable and comfortable growth environment for goats, and improves the breeding benefit and intelligent management level.

Description

Technical Field

[0001] The present invention relates to the field of temperature monitoring in goat breeding, and in particular to a temperature monitoring and control system in a goat breeding house. Background Art

[0002] As modern animal husbandry shifts toward large-scale, intensive operations, environmental control within goat farms has become crucial for ensuring both profitability and animal welfare. Goats are sensitive to temperature fluctuations, and a suitable temperature environment not only improves their growth and health but also reduces breeding costs. However, traditional temperature monitoring and control technologies are no longer sufficient for the demands of precision farming, necessitating the need for more intelligent and efficient systems to optimize the breeding environment.

[0003] Existing temperature monitoring technology for goat farms has significant limitations. First, data collection methods are limited and lack coverage. Most farms rely on a small number of scattered temperature sensors, failing to fully capture temperature differences between vertical and horizontal areas within the farm. It's even more difficult to track temperature fluctuations in areas where goats gather in real time. This one-sided data collection method results in deviations from the actual temperature environment, making it difficult to support accurate regulatory decisions.

[0004] On the other hand, existing temperature control systems lack the ability to coordinate environmental factors and dynamically adjust. Traditional equipment simply raises or lowers the temperature based on preset temperature thresholds, ignoring the coupled relationship between environmental parameters such as humidity, carbon dioxide, and ammonia concentrations, and temperature. Furthermore, the system is unable to flexibly adjust control strategies based on the differentiated temperature requirements of goats throughout their growth cycles, as well as the dynamic changes in the indoor environment. This results in excessive energy consumption and makes it difficult to maintain a stable and comfortable breeding environment, hindering both breeding efficiency and economic benefits. Summary of the Invention

[0005] In order to overcome the shortcomings and deficiencies of the prior art, the present invention provides a temperature monitoring and control system for goat breeding houses.

[0006] The technical solution adopted by the present invention is a temperature monitoring and control system in a goat breeding house, comprising:

[0007] The breeding house environmental data sensing unit is equipped with multiple high-precision temperature and humidity sensors, carbon dioxide concentration sensors, and ammonia concentration sensors, which are deployed at different heights and locations in the breeding house. Each sensor is connected to a data aggregation node via a wired data transmission link. The data aggregation node packages the collected multi-dimensional environmental data and transmits it to the next unit via a dedicated communication protocol.

[0008] The edge data preprocessing unit receives environmental data from the data perception unit, identifies and corrects outliers based on a preset historical data benchmark library, downsamples the data by establishing a sliding time window model, and transmits the processed data to subsequent units via a high-speed data bus.

[0009] The long short-term memory neural network processing unit is optimized. This unit is constructed with a bidirectional long short-term memory network structure and introduces an attention mechanism module in the hidden layer. It receives data transmitted by the edge data preprocessing unit, deeply mines the time series characteristics and spatial distribution characteristics of the temperature in the breeding house, and transmits the processed feature data to the decision-making unit through the feature vector output interface.

[0010] The air internal and external circulation control strategy generation unit receives the characteristic data output by the optimized long short-term memory neural network processing unit, combines the preset goat breeding temperature comfort range and air composition standard range, and generates the air internal and external circulation control strategy by constructing a multi-objective optimization function. The control strategy data is transmitted to the execution unit through a dedicated control instruction protocol;

[0011] The air circulation equipment linkage execution unit receives control instructions from the control strategy generation unit and controls the operating status of the fan, ventilation duct valves, and air purification equipment according to the instructions. The operating status feedback data of each device is transmitted back to the edge data preprocessing unit through the status monitoring line;

[0012] The system operation status monitoring unit monitors the operation status of each system unit and the working parameters of the air circulation equipment in real time. By establishing a system status evaluation model, the monitoring data and evaluation results are fed back to the edge data preprocessing unit and the control strategy generation unit.

[0013] Furthermore, in the optimized long short-term memory neural network processing unit, the temperature spatiotemporal feature fusion model formula is constructed as follows:

[0014] S tf =α×LSTM temporal (T seq )+(1-α)×CNN spatial (T map )

[0015] Among them, S tf is the fused temperature spatiotemporal feature vector; α is the fusion weight coefficient of time feature and spatial feature, and its value range is [0, 1]; LSTM temporal (T seq ) represents the temperature time series T based on the bidirectional long short-term memory network seq Extracted temporal feature vector; CNN spatial (T map) represents the temperature spatial distribution matrix T based on convolutional neural network map Extracted spatial feature vectors;

[0016] In the air internal and external circulation control strategy generation unit, the formula for constructing the air circulation control multi-objective optimization model is as follows:

[0017]

[0018] Where u is the air circulation equipment control parameter vector; w i is the weight coefficient of the i-th objective function, f i is the i-th objective function, corresponding to the temperature deviation objective function, carbon dioxide concentration deviation objective function, and ammonia concentration deviation objective function respectively; T is the real-time temperature in the breeding house; Real-time carbon dioxide concentration in the breeding house; It is the real-time ammonia concentration in the breeding house.

[0019] Furthermore, in the breeding house environmental data perception unit, each sensor is set with differentiated sampling frequency adjustment parameters according to the deployment position and height. The parameter is dynamically adjusted according to the temperature gradient change rate and air flow speed in the breeding house to perform adaptive optimization of environmental data collection.

[0020] Furthermore, in the edge data preprocessing unit, an outlier recognition model based on temperature-humidity correlation is established. When the correlation between the monitored temperature data and humidity data deviates from the statistical law of historical data and exceeds a preset threshold range, the data group is marked and corrected using the adjacent time data interpolation method;

[0021] In the optimized long short-term memory neural network processing unit, a temperature fluctuation intensity factor is introduced in the attention mechanism module. This factor is dynamically calculated based on the standard deviation and mean of the temperature time series and is used to adjust the weight distribution of temperature data at different times in the feature extraction process.

[0022] Furthermore, in the air internal and external circulation control strategy generation unit, a temperature-air composition coupling influence model is constructed. This model uses the temperature change in the breeding house as the independent variable and the carbon dioxide concentration change rate and the ammonia concentration change rate as the dependent variables. A regression model is established through historical data training to predict the air composition change trend under different temperature conditions, thereby assisting in generating a more accurate air internal and external circulation control strategy.

[0023] In the air circulation equipment linkage execution unit, a coordinated adjustment parameter table of the fan speed and the ventilation duct valve opening is set. The parameter table is predefined according to different temperature intervals and air component concentration ranges to perform linkage optimization control between devices.

[0024] Furthermore, a system fault warning model is constructed in the system operation status monitoring unit. The model takes the operation status parameters of each unit and the working parameters of the air circulation equipment as input, calculates the degree and duration of the deviation of the parameters from the normal operation range, and outputs the system fault warning level;

[0025] The breeding house environment data perception unit is configured with a sensor data reliability assessment module, which calculates the reliability score of the data collected by each sensor based on the consistency and stability of the sensor data and the correlation with the surrounding sensor data, and is used to dynamically adjust the weight distribution in the data fusion process.

[0026] Furthermore, in the breeding house environment data perception unit, the data aggregation node has data caching and priority management functions. When the data transmission link is congested, data is cached and sent in the priority order of temperature and humidity data, carbon dioxide concentration data, and ammonia concentration data.

[0027] Furthermore, in the edge data preprocessing unit, the historical data benchmark library adopts a distributed storage architecture, regularly performs aging processing on the stored data, and deletes historical data that exceeds a preset time period.

[0028] Furthermore, in the air circulation equipment linkage execution unit, the air purification equipment is equipped with a consumables usage status monitoring module, which predicts the consumables replacement time by monitoring the resistance changes and pollutant adsorption amount of the consumables, and feeds back the prediction results to the system operation status monitoring unit.

[0029] The temperature monitoring and control system in the goat breeding house includes the following steps:

[0030] Step S1: Multiple high-precision sensors deployed in the breeding house environmental data sensing unit collect temperature, humidity, carbon dioxide concentration, and ammonia concentration data at different heights and locations in the breeding house according to a dynamically adjusted sampling frequency, and package the data and transmit it to the edge data preprocessing unit;

[0031] Step S2: The edge data preprocessing unit processes the received data based on a preset historical data benchmark library using an outlier identification and correction algorithm and a sliding time window downsampling algorithm, and transmits the processed data to the optimized long short-term memory neural network processing unit;

[0032] Step S3: Optimizing the long short-term memory neural network processing unit to use the constructed bidirectional long short-term memory network structure and attention mechanism module to deeply extract and fuse the time series characteristics and spatial distribution characteristics of temperature;

[0033] Step S4: The air internal and external circulation control strategy generation unit receives the processed feature data, combines the preset goat breeding environment standards, and generates an air internal and external circulation control strategy through a constructed multi-objective optimization function;

[0034] Step S5: The air circulation equipment linkage execution unit receives the control strategy, controls the fan, ventilation duct valve, and air purification equipment to operate according to the strategy, and transmits the equipment operation status feedback data back to the edge data preprocessing unit;

[0035] Step S6: The system operation status monitoring unit monitors the operation status of each unit and equipment in the system in real time, and feeds back the monitoring data and evaluation results to the edge data preprocessing unit and the control strategy generation unit to form a closed-loop process of temperature monitoring and control.

[0036] Beneficial Effects: The present invention proposes a temperature monitoring and control system for goat breeding houses. The system uses a breeding house environmental data sensing unit to deploy multiple high-precision sensors at different heights and areas in the breeding house. The sampling frequency is dynamically adjusted according to the temperature gradient change rate and air flow speed, achieving comprehensive and accurate collection of data such as temperature, humidity, carbon dioxide, and ammonia concentrations. Compared with the traditional collection method of a small number of scattered sensors, it can more accurately reflect the actual environmental conditions in the breeding house. In terms of data processing and control, the long short-term memory neural network processing unit is optimized, and a bidirectional long short-term memory network is combined with an attention mechanism to deeply explore the time series and spatial distribution characteristics of temperature. The air internal and external circulation control strategy generation unit comprehensively considers the temperature comfort range and air composition standards for goat breeding, constructs a multi-objective optimization model, and generates a control strategy. This changes the shortcomings of the traditional system that only relies on temperature threshold control and ignores the coupling relationship between environmental factors. At the same time, the air circulation equipment linkage execution unit accurately controls the fan, ventilation duct valve, and air purification equipment according to the control strategy, and can also make dynamic adjustments based on the feedback data of the equipment operation status, realizing refined and intelligent environmental control. The system operation status monitoring unit monitors the operation status of each system unit and equipment in real time, and provides timely feedback and optimization to ensure the efficient and stable operation of the entire system. It not only reduces energy consumption, but also creates a stable and comfortable growth environment for goats, significantly improving breeding efficiency and economic benefits. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 It is a diagram of the system unit composition of the present invention;

[0038] Figure 2 It is a flow chart of the system operation steps of the present invention. DETAILED DESCRIPTION

[0039] It should be noted that, unless there is a conflict, the embodiments in this application and the features described in the embodiments can be combined with each other. The application is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0040] like Figure 1 As shown in the figure, the temperature monitoring and control system in the goat breeding house includes:

[0041] The breeding house environmental data sensing unit is equipped with multiple high-precision temperature and humidity sensors, carbon dioxide concentration sensors, and ammonia concentration sensors, which are deployed at different heights and locations in the breeding house. Each sensor is connected to a data aggregation node via a wired data transmission link. The data aggregation node packages the collected multi-dimensional environmental data and transmits it to the next unit via a dedicated communication protocol.

[0042] Specifically, this unit serves as the data collection foundation for the entire temperature monitoring and control system. Its core function is to obtain comprehensive and accurate environmental data within the breeding house. The unit is equipped with multiple high-precision temperature and humidity sensors, as well as carbon dioxide concentration sensors and ammonia concentration sensors. These sensors are rigorously calibrated to ensure the accuracy of the collected data. The sensors are positioned at different heights and locations within the breeding house, covering the different vertical spaces where the goats move and evenly distributed throughout the breeding house to capture environmental parameters at different locations. Each sensor is connected to a data aggregation node via a wired data transmission link. This wired transmission method ensures stable and reliable data transmission, avoiding signal interference and loss that may occur with wireless transmission. The data aggregation node packages the collected multi-dimensional environmental data and transmits it to the next unit using a dedicated communication protocol. This dedicated communication protocol incorporates data encryption and verification functions to ensure the security and integrity of the data during transmission.

[0043] In terms of implementation, based on an adaptive weight adjustment mechanism, each sensor is set with differentiated sampling frequency adjustment parameters depending on its deployment location and altitude. Specifically, in areas and altitudes where goats are active and environmental changes may be more drastic, the sensor's sampling frequency is increased accordingly to capture changes in environmental parameters more promptly; in areas with relatively stable environments, the sampling frequency is appropriately reduced to conserve data transmission and processing resources. This adaptive adjustment is dynamically performed based on the temperature gradient change rate and air flow speed within the breeding house. A built-in algorithm module analyzes these environmental factors in real time to adjust the sampling frequency, achieving adaptive optimization of environmental data collection and providing more valuable data support for subsequent data processing and system decision-making.

[0044] The edge data preprocessing unit receives environmental data from the data perception unit, identifies and corrects outliers based on a preset historical data benchmark library, downsamples the data by establishing a sliding time window model, and transmits the processed data to subsequent units via a high-speed data bus.

[0045] Specifically, this unit is primarily responsible for performing preliminary processing on the raw data from the environmental data sensing unit to improve data quality. After receiving the environmental data, it identifies and corrects outliers based on a preset historical data benchmark library. This historical data benchmark library stores a large amount of environmental data from farmhouses over different time periods and environmental conditions. Through statistical analysis of this data, a model of the range and variation patterns of normal data is established. When newly collected data deviates significantly from this model, the system identifies it as an outlier and applies a corresponding correction algorithm, such as interpolating data from adjacent moments. At the same time, the unit downsamples the data by establishing a sliding time window model. This reduces the data volume without affecting the key characteristics of the data, reduces the pressure on subsequent data processing and the consumption of computing resources, and improves data processing efficiency. The processed data is transmitted to subsequent units via a high-speed data bus, which ensures fast and stable data transmission.

[0046] During the actual implementation process, the dynamic threshold calibration function of the unit will be continuously optimized as the system operates. With the continuous collection and accumulation of new data, the historical data benchmark library will be regularly updated to include more new environmental data samples, making the benchmark library more in line with the actual environmental conditions of the current breeding house. In the process of identifying outliers, in addition to being based on the historical data benchmark library, a comprehensive judgment will be made based on the correlation between the data. For example, there is usually a certain correlation between parameters such as temperature and humidity, and carbon dioxide concentration. When a parameter is abnormal and its correlation with other parameters does not conform to the normal rules, it will be further confirmed whether the data is an outlier, thereby improving the accuracy of outlier identification. Through this dynamic calibration and comprehensive judgment mechanism, the data input into subsequent units is ensured to be true and reliable, providing a guarantee for the precise operation of the system.

[0047] The long short-term memory neural network processing unit is optimized. This unit is constructed with a bidirectional long short-term memory network structure and introduces an attention mechanism module in the hidden layer. It receives data transmitted by the edge data preprocessing unit, deeply mines the time series characteristics and spatial distribution characteristics of the temperature in the breeding house, and transmits the processed feature data to the decision-making unit through the feature vector output interface.

[0048] Specifically, this unit is a key component of the system's intelligent analysis, deeply analyzing the temporal and spatial distribution characteristics of the temperature within the breeding house. The unit utilizes a bidirectional long-short-term memory (LSTM) network structure, which processes temperature time series data in both forward and reverse directions, leveraging information from past and future moments. This allows for a more comprehensive capture of temperature trends and patterns than a unidirectional network. An attention mechanism is incorporated into the hidden layer, dynamically adjusting the weighting of temperature data at different moments in the feature extraction process based on their importance. For example, moments of significant temperature change are assigned higher weights, directing the network's attention to these key data points, thereby extracting more representative temporal features. Simultaneously, a convolutional neural network is used to process the spatial distribution matrix of the temperature, extracting spatial features. Ultimately, these temporal and spatial features are fused to generate feature data that comprehensively reflects the temperature status of the breeding house. This feature data is then transmitted to the decision unit via a feature vector output interface.

[0049] During implementation, this unit's network model required extensive training and optimization. The training data was derived from environmental data collected from the farmhouse during different seasons and time periods. By learning from this data, the network model was able to continuously adjust its parameters, improving the accuracy of its extraction of spatiotemporal temperature features. During actual operation, the model continuously updates and optimizes the extracted features based on real-time environmental changes in the farmhouse. To ensure efficient model operation, model compression and acceleration technologies were also employed to reduce the model's computational workload and memory usage, enabling it to run quickly on edge devices, providing timely support for subsequent decision-making and ensuring the system's ability to rapidly respond to changes in farmhouse temperature.

[0050] The air internal and external circulation control strategy generation unit receives the characteristic data output by the optimized long short-term memory neural network processing unit, combines the preset goat breeding temperature comfort range and air composition standard range, and generates the air internal and external circulation control strategy by constructing a multi-objective optimization function. The control strategy data is transmitted to the execution unit through a dedicated control instruction protocol;

[0051] Specifically, this unit's primary task is to generate a rational air circulation control strategy based on the characteristic data processed by the previous units. It receives the characteristic data output by the optimized long-short-term memory neural network processing unit and combines it with preset parameters such as the temperature comfort range and air composition standard range for goat breeding. Goats have specific requirements for temperature and air composition at different stages of their growth. These preset ranges are determined based on extensive breeding experiments and research. The unit constructs a multi-objective optimization function, comprehensively considering multiple factors such as temperature, carbon dioxide concentration, and ammonia concentration to determine the optimal control strategy. While considering temperature, it also fully accounts for the interplay between air composition and temperature. For example, changes in temperature can affect the distribution and concentration of air components, while changes in air composition can also have a certain impact on temperature. Through this multi-objective collaborative optimization approach, the generated control strategy fully meets the environmental requirements of goat breeding and achieves precise control of the breeding house environment.

[0052] During implementation, the unit continuously optimizes and adjusts the control strategy based on the actual conditions of the breeding house. As the number of goats in the breeding house changes, the seasons change, and the external environment fluctuates, the environmental requirements of the breeding house will also change. The unit monitors these changing factors in real time, dynamically adjusts the parameters in the multi-objective optimization function, and recalculates and generates a control strategy that is more suitable for the current environment. At the same time, the unit also has strategy evaluation and screening functions. For the multiple candidate control strategies generated, a comprehensive evaluation will be conducted based on factors such as their degree of satisfaction with each goal, implementation costs, and the impact on other parts of the system. The optimal strategy will be selected and transmitted to the execution unit through a dedicated control command protocol to ensure that the air circulation equipment can operate in the most reasonable manner and maintain a stable and suitable environment in the breeding house.

[0053] The air circulation equipment linkage execution unit receives control instructions from the control strategy generation unit and controls the operating status of the fan, ventilation duct valves, and air purification equipment according to the instructions. The operating status feedback data of each device is transmitted back to the edge data preprocessing unit through the status monitoring line;

[0054] Specifically, the function of this unit is to precisely control the operation of air circulation-related equipment based on the control instructions of the control strategy generation unit. After receiving the control instructions from the control strategy generation unit, the unit parses and processes the instructions, and controls the operating status of the fan, ventilation duct valves, air purification equipment, etc. according to the instructions. The fan speed, the opening of the ventilation duct valves, and the working mode of the air purification equipment can all be precisely adjusted according to the instructions. For example, when the ventilation volume needs to be increased to reduce the temperature and improve the air quality, the instructions will control the fan to increase the speed and adjust the ventilation duct valves to increase the opening; when a high concentration of air pollutants is detected, the air purification equipment will be started and adjusted to the appropriate purification mode. The operating status feedback data of each device is transmitted back to the edge data preprocessing unit through the status monitoring line, so that the system can grasp the equipment operation status in real time and provide a reference for subsequent control.

[0055] In actual implementation, the adaptive adjustment capability of this unit is reflected in many aspects. First, there is a coordinated adjustment parameter table for the fan speed and the ventilation duct valve opening. This parameter table is predefined according to different temperature ranges and air component concentration ranges. After receiving the control instruction, it can quickly determine the optimal operating parameter combination of the equipment according to the current environmental conditions, and realize the linkage optimization control between the equipment. Secondly, the air circulation equipment will make adaptive adjustments according to its own operating status and environmental changes. For example, when the fan encounters a change in resistance during operation, the equipment will automatically adjust the motor power to maintain the set ventilation volume; the air purification equipment is equipped with a consumables usage status monitoring module. By monitoring the resistance change and pollutant adsorption of the consumables, it predicts the consumables replacement time and feeds back the relevant information to the system so that the system can make maintenance preparations in advance to ensure that the equipment is always in the best working condition, to ensure the effective implementation of the control strategy and the stability of the breeding house environment.

[0056] The system operation status monitoring unit monitors the operation status of each system unit and the working parameters of the air circulation equipment in real time. By establishing a system status evaluation model, the monitoring data and evaluation results are fed back to the edge data preprocessing unit and the control strategy generation unit.

[0057] Specifically, this unit is an important link in ensuring the stable operation of the system, and its main responsibility is to monitor the operating status of each unit in the system and the working parameters of the air circulation equipment in real time. By setting up monitoring nodes at key locations in the system, information is collected, including the working status of the sensors of the environmental data perception unit, the data processing performance of the edge data preprocessing unit, the model operating parameters of the neural network processing unit, and the motor current, speed, valve position and other information of the air circulation equipment. A system status assessment model is established, which analyzes and evaluates the collected data based on the preset normal operating parameter range and the logical relationship between the units. When the monitoring data exceeds the normal range or there is an abnormal logical relationship between the operation of the units, the model will determine that there is an operational problem in the system, and will promptly feed back the relevant information to the edge data preprocessing unit and the control strategy generation unit, so that the system can take timely measures to adjust and optimize.

[0058] During implementation, the unit's feedback closed-loop control mechanism can continuously optimize system operation. On the one hand, for problems such as equipment failure or performance degradation detected, maintenance personnel will be promptly notified for processing, and the processing results will be recorded in the system to update the parameters of the system status assessment model and improve the model's ability to identify and warn of similar problems. On the other hand, through long-term analysis and summary of system operation data, potential optimization points in the system operation process are discovered. For example, based on the relationship between equipment operation energy consumption and regulation effect, optimization suggestions for equipment operation parameters are proposed and fed back to the regulation strategy generation unit, prompting it to adjust the regulation strategy to achieve energy-saving and efficient operation of the system. At the same time, the unit will also regularly calibrate and optimize the system status assessment model to ensure that the model can always accurately assess the system operation status and ensure the stable and reliable operation of the entire temperature monitoring and control system.

[0059] Preferably, in the optimized long short-term memory neural network processing unit, the temperature spatiotemporal feature fusion model formula is constructed as follows:

[0060] S tf =α×LSTM temporal (T seq )+(1-α)×CNN spatial (T map )

[0061] Among them, S tf is the fused temperature spatiotemporal feature vector; α is the fusion weight coefficient of time feature and spatial feature, and its value range is [0, 1]; LSTM temporal (T seq ) represents the temperature time series T based on the bidirectional long short-term memory network seq Extracted temporal feature vector; CNN spatial (T map) represents the temperature spatial distribution matrix T based on convolutional neural network map The extracted spatial feature vector. In the air internal and external circulation control strategy generation unit, the formula for constructing the air circulation control multi-objective optimization model is as follows:

[0062]

[0063] Where u is the air circulation equipment control parameter vector; w i is the weight coefficient of the i-th objective function, f i is the i-th objective function, corresponding to the temperature deviation objective function, carbon dioxide concentration deviation objective function, and ammonia concentration deviation objective function respectively; T is the real-time temperature in the breeding house; Real-time carbon dioxide concentration in the breeding house; It is the real-time ammonia concentration in the breeding house.

[0064] Specifically, in the optimization of the long short-term memory neural network processing unit, by constructing a temperature spatiotemporal feature fusion model, the temperature-time features extracted based on the bidirectional long short-term memory network are organically combined with the spatial features extracted based on the convolutional neural network according to the fusion weight coefficient. This enables the system to comprehensively capture the temperature variation pattern in the breeding house from the two dimensions of time evolution and spatial distribution, providing richer feature data for subsequent accurate decision-making. In the air internal and external circulation control strategy generation unit, a multi-objective optimization model for air circulation control is established. The control parameters of the air circulation equipment are used as the optimization object. Multiple objective functions such as temperature, carbon dioxide concentration, ammonia concentration and their weight coefficients are comprehensively considered. By solving the multi-objective optimization problem, an air internal and external circulation control strategy that takes into account the various needs of the goat breeding environment is generated, realizing the multi-factor coordinated optimization control of the breeding house environment.

[0065] Preferably, in the breeding house environmental data perception unit, each sensor is provided with differentiated sampling frequency adjustment parameters according to different deployment positions and heights. The parameters are dynamically adjusted according to the temperature gradient change rate and air flow speed in the breeding house to perform adaptive optimization of environmental data collection.

[0066] Specifically, in actual farmhouse environments, the degree of environmental variability varies across different areas and heights. Each sensor has a differentiated sampling frequency adjustment parameter, which dynamically adjusts in real time based on the temperature gradient and air flow rate within the farmhouse. In areas with drastic temperature fluctuations and complex airflow, the sensor increases the sampling frequency to ensure rapid capture of subtle environmental changes; in relatively stable environments, the sampling frequency is reduced to reduce data redundancy. This adaptive adjustment ensures timely acquisition of critical environmental data while rationally allocating data transmission and processing resources, improving the overall efficiency and data quality of the data sensing unit.

[0067] Preferably, an outlier recognition model based on temperature-humidity correlation is established in the edge data preprocessing unit. When the correlation between the monitored temperature data and humidity data deviates from the statistical law of historical data and exceeds a preset threshold range, the data set is marked and corrected using the adjacent time data interpolation method. In the optimized long short-term memory neural network processing unit, a temperature fluctuation intensity factor is introduced in the attention mechanism module. This factor is dynamically calculated based on the standard deviation and mean of the temperature time series and is used to adjust the weight distribution of temperature data at different times in the feature extraction process.

[0068] Specifically, in the edge data preprocessing unit, an outlier identification model based on temperature-humidity correlation is established. This model leverages the inherent correlation between temperature and humidity data. When the correlation between the two deviates from historical statistical patterns and exceeds a preset threshold, it accurately identifies abnormal data and uses interpolation to correct it, improving data accuracy. A temperature fluctuation intensity factor is introduced into the attention mechanism module of the neural network processing unit. This factor is dynamically calculated based on the standard deviation and mean of the temperature time series. The weight of data at different times in feature extraction is automatically adjusted based on the degree of temperature fluctuation. This allows the network to focus more on key information about temperature changes, enhancing the relevance and effectiveness of feature extraction.

[0069] Preferably, in the air internal and external circulation control strategy generation unit, a temperature-air composition coupling influence model is constructed. The model uses the temperature change in the breeding house as the independent variable and the carbon dioxide concentration change rate and the ammonia concentration change rate as the dependent variable. A regression model is established through historical data training to predict the trend of air composition changes under different temperature conditions and assist in generating a more accurate air internal and external circulation control strategy. In the air circulation equipment linkage execution unit, a collaborative adjustment parameter table of the fan speed and the ventilation duct valve opening is set. The parameter table is predefined according to different temperature intervals and air component concentration ranges to perform linkage optimization control between devices.

[0070] Specifically, in the control strategy generation unit, a temperature-air composition coupling impact model is constructed, with temperature changes as the independent variable and the change rates of carbon dioxide concentration and ammonia concentration as the dependent variables. By training historical data, a regression model is established to predict the trend of air composition changes at different temperatures. This allows the impact of temperature changes on air composition to be considered in advance when generating control strategies, improving the strategy's foresight and accuracy. In the equipment linkage execution unit, a coordinated adjustment parameter table for fan speed and ventilation duct valve opening is set. Parameter combinations are predefined based on different temperature ranges and air component concentration ranges to ensure that the equipment can quickly and collaboratively adjust when executing the control strategy, achieving efficient and energy-saving air circulation control.

[0071] Preferably, the system operation status monitoring unit constructs a system fault warning model. This model uses the operating status parameters of each unit and the operating parameters of the air circulation equipment as input, calculates the degree and duration of parameter deviation from the normal operating range, and outputs a system fault warning level. The breeding house environmental data perception unit is configured with a sensor data reliability assessment module. This module calculates the reliability score of the data collected by each sensor based on the consistency and stability of the sensor data and its correlation with surrounding sensor data. This module is used to dynamically adjust the weight distribution during the data fusion process.

[0072] Specifically, within the system's operational status monitoring unit, a system failure warning model is constructed. This model integrates the operational status parameters of each unit with the operating parameters of the air circulation equipment. By calculating the degree and duration of parameter deviations from the normal range, the system failure risk is quantitatively assessed and a warning level is output, enabling managers to take timely action. Within the environmental data perception unit, a sensor data reliability assessment module is configured to evaluate the reliability of each sensor's collected data from multiple dimensions, including data consistency, stability, and correlation with surrounding sensor data. Data fusion weights are dynamically adjusted based on the scores to ensure the credibility of input system data and improve the stability of the entire system's operation.

[0073] Preferably, in the breeding house environmental data perception unit, the data aggregation node has data caching and priority management functions. When congestion occurs in the data transmission link, data is cached and sent in the order of priority of temperature and humidity data, carbon dioxide concentration data, and ammonia concentration data to ensure timely transmission of key temperature data.

[0074] Specifically, the data aggregation node has added data caching and priority management functions. During the transmission of breeding house environmental data, if the data transmission link is congested, the data aggregation node sets a priority order based on the importance of temperature and humidity data, carbon dioxide concentration data, and ammonia concentration data to goat breeding environment monitoring. Prioritizing the timely transmission of critical temperature-related data such as temperature and humidity, less important data is temporarily cached and sent in an orderly manner after the link is restored. This avoids the loss or delay of critical data due to link congestion, ensures that environmental data can be transmitted completely and promptly to subsequent processing units, and provides support for accurate decision-making by the system.

[0075] Preferably, in the edge data preprocessing unit, the historical data benchmark library adopts a distributed storage architecture, regularly performs aging processing on the stored data, and deletes historical data that exceeds a preset time period to ensure the timeliness and effectiveness of the benchmark library data.

[0076] Specifically, the historical data benchmark library utilizes a distributed storage architecture, which effectively disperses data storage pressure and improves data storage and access efficiency. Furthermore, to ensure the timeliness and validity of the benchmark library data, the system regularly ages the stored data and deletes outdated data according to preset time periods. As factors such as the farm environment and farming patterns change, data that is no longer valuable is promptly removed and newly collected valid data is incorporated. This ensures that the historical data benchmark library remains consistent with the current farm conditions, providing a more accurate and reliable reference for outlier identification and data preprocessing, ensuring the accuracy of data processing and the reliability of system operation.

[0077] Preferably, in the air circulation equipment linkage execution unit, the air purification equipment is equipped with a consumables usage status monitoring module, which predicts the consumables replacement time by monitoring the resistance changes and pollutant adsorption amount of the consumables, and feeds back the prediction results to the system operation status monitoring unit.

[0078] Specifically, the air purification equipment is equipped with a consumables usage status monitoring module. This module uses a specific algorithm to predict the replacement time of consumables by monitoring the resistance changes and pollutant adsorption of consumables in real time. When it is detected that the consumables are close to saturation or the resistance exceeds the set threshold, the prediction results are promptly fed back to the system operation status monitoring unit. The system arranges a consumables replacement plan accordingly to avoid the air purification effect being affected by consumable failure, ensure the continuous and efficient operation of the air purification equipment, maintain the air quality in the breeding house, and cooperate with other air circulation equipment to create an environment suitable for goat growth.

[0079] like Figure 2 As shown, the temperature monitoring and control system in the goat breeding house includes the following steps:

[0080] Step S1: Multiple high-precision sensors deployed in the breeding house environmental data sensing unit collect temperature, humidity, carbon dioxide concentration, and ammonia concentration data at different heights and locations in the breeding house according to a dynamically adjusted sampling frequency, and package the data and transmit it to the edge data preprocessing unit;

[0081] Step S2: The edge data preprocessing unit processes the received data based on a preset historical data benchmark library using an outlier identification and correction algorithm and a sliding time window downsampling algorithm, and transmits the processed data to the optimized long short-term memory neural network processing unit;

[0082] Step S3: Optimizing the long short-term memory neural network processing unit to use the constructed bidirectional long short-term memory network structure and attention mechanism module to deeply extract and fuse the time series characteristics and spatial distribution characteristics of temperature;

[0083] Step S4: The air internal and external circulation control strategy generation unit receives the processed feature data, combines the preset goat breeding environment standards, and generates an air internal and external circulation control strategy through a constructed multi-objective optimization function;

[0084] Step S5: The air circulation equipment linkage execution unit receives the control strategy, controls the fan, ventilation duct valve, and air purification equipment to operate according to the strategy, and transmits the equipment operation status feedback data back to the edge data preprocessing unit;

[0085] Step S6: The system operation status monitoring unit monitors the operation status of each unit and equipment in the system in real time, and feeds back the monitoring data and evaluation results to the edge data preprocessing unit and the control strategy generation unit to form a closed-loop process of temperature monitoring and control.

[0086] Traditional methods for temperature monitoring and control systems in goat breeding sheds struggle to collect comprehensive and accurate environmental data due to limited sensor layout and fixed sampling frequencies. However, this system leverages environmental data sensing units in the breeding sheds, deploying high-precision sensors at multiple heights and locations. Dynamically adjusting the sampling frequency based on temperature gradients and air flow rates enables comprehensive, high-precision collection of data such as temperature, humidity, carbon dioxide, and ammonia concentrations. This ensures that the monitoring data truly reflects the complex environmental changes within the sheds, laying a solid foundation for precise control.

[0087] In terms of data processing and control strategy generation, existing technologies lack the ability to collaboratively analyze and dynamically control environmental factors. This system's optimized long-short-term memory (LSTM) neural network processing unit utilizes a bidirectional LSTM network combined with an attention mechanism to deeply explore the temporal series and spatial distribution characteristics of temperature. The air internal and external circulation control strategy generation unit integrates the temperature comfort range and air composition standards for goat farming to construct a multi-objective optimization model, fully considering the coupling relationship between temperature and air composition to generate precise control strategies. Compared to traditional single temperature threshold control, this system can flexibly adjust according to dynamic environmental changes, achieving collaborative optimization and control of multiple factors.

[0088] At the level of equipment execution and system feedback, traditional system equipment has poor linkage and lacks an effective feedback mechanism, resulting in extensive regulation and energy waste. The air circulation equipment linkage execution unit of this system accurately controls the fan, ventilation duct valves, and air purification equipment according to the regulation strategy, and realizes dynamic adjustment through equipment operation status feedback data; the system operation status monitoring unit monitors the operation status of each unit and equipment in the system in real time, and feeds the monitoring and evaluation results back to the data processing and regulation unit to form a complete closed-loop control. This design not only improves equipment operation efficiency and reduces energy consumption, but also continuously optimizes the regulation strategy, creating a stable and comfortable growth environment for goats, and effectively improving breeding efficiency and management level.

[0089] In the description of the present invention, it should be noted that, unless otherwise expressly specified or limited, the terms "disposed," "installed," "connected," "connected," and "fixed" should be understood in a broad sense. For example, they may refer to fixed connections, detachable connections, or integral connections; they may refer to mechanical connections or electrical connections; they may refer to direct connections or indirect connections through an intermediate medium; and they may refer to internal communication between two components. Those skilled in the art will understand the specific meanings of the above terms in the present invention based on specific circumstances.

[0090] While embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that various equivalent changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. The temperature monitoring and control system in the goat breeding house is characterized by: include: The breeding house environmental data sensing unit includes temperature and humidity sensors, carbon dioxide concentration sensors, and ammonia concentration sensors, which are installed at different heights and locations in the breeding house. Each sensor is connected to the data aggregation node through a data transmission link; The edge data preprocessing unit receives environmental data from the breeding house environmental data perception unit, identifies and corrects outliers based on a preset historical data benchmark library, and downsamples the data by establishing a sliding time window model. Optimize the long short-term memory neural network processing unit, build a bidirectional long short-term memory network structure, introduce an attention mechanism module in the hidden layer, receive data transmitted by the edge data preprocessing unit, and conduct in-depth mining of the time series characteristics and spatial distribution characteristics of the temperature in the breeding house; The air internal and external circulation control strategy generation unit receives the characteristic data output by the optimized long short-term memory neural network processing unit, combines the preset goat breeding temperature comfort range and air composition standard range, and generates the air internal and external circulation control strategy by constructing a multi-objective optimization function. The control strategy data is transmitted to the execution unit through a dedicated control instruction protocol; The air circulation equipment linkage execution unit receives control instructions from the control strategy generation unit and controls the operating status of the fan, ventilation duct valves, and air purification equipment according to the instructions. The operating status feedback data of each device is transmitted back to the edge data preprocessing unit through the status monitoring line; The system operation status monitoring unit monitors the operation status of each system unit and the working parameters of the air circulation equipment in real time. By establishing a system status evaluation model, the monitoring data and evaluation results are fed back to the edge data preprocessing unit and the air internal and external circulation control strategy generation unit.

2. The temperature monitoring and control system in a goat breeding house according to claim 1, characterized in that: In the optimized long short-term memory neural network processing unit, the temperature spatiotemporal feature fusion model formula is constructed as follows: S tf =α×LSTM temporal (T seq )+(1-α)×CNN spatial (T map ) Among them, S tf is the fused temperature spatiotemporal feature vector; α is the fusion weight coefficient of time feature and spatial feature, and its value range is [0, 1]; LSTM temporal (T seq ) represents the temperature time series T based on the bidirectional long short-term memory network seq Extracted temporal feature vector; CNN spatial (T map ) represents the temperature spatial distribution matrix T based on convolutional neural network map Extracted spatial feature vectors; In the air internal and external circulation control strategy generation unit, the formula for constructing the air circulation control multi-objective optimization model is as follows: Where u is the air circulation equipment control parameter vector; w i is the weight coefficient of the i-th objective function, f i is the i-th objective function, corresponding to the temperature deviation objective function, carbon dioxide concentration deviation objective function, and ammonia concentration deviation objective function respectively; T is the real-time temperature in the breeding house; Real-time carbon dioxide concentration in the breeding house; It is the real-time ammonia concentration in the breeding house.

3. The temperature monitoring and control system in a goat breeding house according to claim 1, characterized in that: In the breeding house environmental data perception unit, each sensor is set with a differentiated sampling frequency adjustment parameter according to the installation position and height. The parameter is dynamically adjusted according to the temperature gradient change rate and air flow speed in the breeding house to perform adaptive optimization of environmental data collection.

4. The temperature monitoring and control system in a goat breeding house according to claim 1, characterized in that: In the edge data preprocessing unit, an outlier recognition model based on temperature-humidity correlation is established. When the correlation between the monitored temperature data and humidity data deviates from the statistical law of historical data and exceeds a preset threshold range, the data group is marked and corrected using the adjacent time data interpolation method; In the optimized long short-term memory neural network processing unit, a temperature fluctuation intensity factor is introduced in the attention mechanism module. This factor is dynamically calculated based on the standard deviation and mean of the temperature time series and is used to adjust the weight distribution of temperature data at different times in the feature extraction process.

5. The temperature monitoring and control system in a goat breeding house according to claim 4, characterized in that: In the air internal and external circulation control strategy generation unit, a temperature-air composition coupling influence model is constructed. This model uses the temperature change in the breeding house as the independent variable and the carbon dioxide concentration change rate and the ammonia concentration change rate as the dependent variables. A regression model is established through historical data training to predict the air composition change trend under different temperature conditions, thereby assisting in generating a more accurate air internal and external circulation control strategy. In the air circulation equipment linkage execution unit, a coordinated adjustment parameter table of the fan speed and the ventilation duct valve opening is set. The parameter table is predefined according to different temperature intervals and air component concentration ranges to perform linkage optimization control between devices.

6. The temperature monitoring and control system in a goat breeding house according to claim 1, characterized in that: In the system operation status monitoring unit, a system fault warning model is constructed. The model takes the operation status parameters of each unit and the working parameters of the air circulation equipment as input, calculates the degree and duration of the deviation of the parameters from the normal operating range, and outputs the system fault warning level; The breeding house environment data perception unit is configured with a sensor data reliability assessment module, which calculates the reliability score of the data collected by each sensor based on the consistency and stability of the sensor data and the correlation with the surrounding sensor data, and is used to dynamically adjust the weight distribution in the data fusion process.

7. The temperature monitoring and control system in a goat breeding house according to claim 6, characterized in that: In the breeding house environment data perception unit, the data aggregation node has data caching and priority management functions. When the data transmission link is congested, data is cached and sent in the order of priority of temperature and humidity data, carbon dioxide concentration data, and ammonia concentration data.

8. The temperature monitoring and control system in a goat breeding house according to claim 1, characterized in that: In the edge data preprocessing unit, the historical data benchmark library adopts a distributed storage architecture, regularly performs aging processing on the stored data, and deletes historical data that exceeds a preset time period.

9. The temperature monitoring and control system in a goat breeding house according to claim 1, characterized in that: In the air circulation equipment linkage execution unit, the air purification equipment is equipped with a consumables usage status monitoring module, which predicts the consumables replacement time by monitoring the resistance changes and pollutant adsorption amount of the consumables, and feeds back the prediction results to the system operation status monitoring unit.

10. The temperature monitoring and control system in a goat breeding house according to any one of claims 1 to 9, characterized in that: The system operation includes the following steps: Step S1: Multiple sensors deployed in the breeding house environmental data sensing unit collect temperature, humidity, carbon dioxide concentration, and ammonia concentration data at different heights and locations in the breeding house according to a dynamically adjusted sampling frequency, and package the data and transmit it to the edge data preprocessing unit; Step S2: The edge data preprocessing unit processes the received data based on a preset historical data benchmark library using an outlier identification and correction algorithm and a sliding time window downsampling algorithm, and transmits the processed data to the optimized long short-term memory neural network processing unit; Step S3: Optimizing the long short-term memory neural network processing unit to use the constructed bidirectional long short-term memory network structure and attention mechanism module to deeply extract and fuse the time series characteristics and spatial distribution characteristics of temperature; Step S4: The air internal and external circulation control strategy generation unit receives the processed feature data, combines the preset goat breeding environment standards, and generates an air internal and external circulation control strategy through a constructed multi-objective optimization function; Step S5: The air circulation equipment linkage execution unit receives the control strategy, controls the fan, ventilation duct valve, and air purification equipment to operate according to the strategy, and transmits the equipment operation status feedback data back to the edge data preprocessing unit; Step S6: The system operation status monitoring unit monitors the operation status of each unit and equipment in the system in real time, and feeds back the monitoring data and evaluation results to the edge data preprocessing unit and the control strategy generation unit to form a closed-loop process of temperature monitoring and control.

Images