Livestock breeding system and method based on multi-sensor fusion technology and intelligent equipment

Abstract

The invention relates to the technical field of animal husbandry management, and discloses a livestock breeding system and method based on a multi-sensor fusion technology and intelligent equipment, and the livestock breeding method comprises the steps: monitoring pasture environment parameters, livestock behaviors and resource consumption data in real time, and obtaining a livestock breeding data set through the multi-sensor fusion technology; constructing a pasture Internet of Things real-time monitoring network to collect and transmit a livestock breeding data set in real time; the pasture environment is regulated and controlled in real time; when the monitored livestock breeding data set exceeds a preset safety threshold value, it is indicated that equipment breaks down, early warning information is remotely output, and an emergency plan is executed; by remotely monitoring pasture environment parameters, checking livestock behaviors, receiving early warning information and regulating and controlling environment regulating equipment, livestock breeding environment data and resource data are intelligently and accurately regulated and controlled, so that pasture environment monitoring and regulation and control can be quickly and accurately realized under the condition that manual observation and a basic sensor are not completely reliable.

Description

Technical Field

[0001] The present invention relates to the technical field of animal husbandry management, and more specifically, to an animal husbandry breeding system and method based on multi-sensor fusion technology and intelligent equipment. Background Art

[0002] Traditional pasture farming methods often rely on manual observation and basic data recording, viewing pasture management as a simple management structure consisting of manual operations and basic sensors. Most of these methods are based on existing basic data recording technology and are difficult to handle in modern animal husbandry, where there is a higher demand for efficient and precise management.

[0003] Therefore, how to quickly and accurately monitor and control pasture environment when manual observation and basic sensors are not completely reliable has become an urgent problem to be solved. Summary of the Invention

[0004] The present invention provides a livestock breeding system and method based on multi-sensor fusion technology and intelligent equipment, which solves the technical problem in the existing technology of how to quickly and accurately monitor and control the pasture environment when manual observation and basic sensors are not completely reliable.

[0005] The present invention provides a livestock breeding system and method based on multi-sensor fusion technology and intelligent equipment, including:

[0006] First, livestock breeding methods based on multi-sensor fusion technology and intelligent equipment include:

[0007] Real-time monitoring of pasture environmental parameters, livestock behavior, and resource consumption data, and acquisition of livestock breeding data sets through multi-sensor fusion technology;

[0008] Building a ranch IoT real-time monitoring network based on the livestock breeding data set to collect and transmit the livestock breeding data set in real time;

[0009] Based on the real-time monitoring network of the ranch Internet of Things, the ranch environment is regulated in real time according to the real-time monitoring livestock breeding data set;

[0010] When the monitored livestock breeding data set exceeds the preset safety threshold, it indicates that the equipment has failed. Through the local buzzer and LED indicator, the warning information is remotely output and the emergency plan is executed;

[0011] By remotely monitoring pasture environmental parameters, checking livestock behavior, receiving early warning information and adjusting environmental conditioning equipment, livestock breeding environmental data and resource data can be intelligently and accurately controlled.

[0012] Further, including:

[0013] Collect pasture environmental data, including light intensity, CO2 concentration, NH3 concentration, air temperature, air humidity parameters, and livestock behavior data, through multi-sensor fusion technology;

[0014] The collected data is transmitted to the main controller in real time through the wireless network in the real-time monitoring network of the Internet of Things, realizing real-time transmission and centralized processing of data;

[0015] The main controller determines whether the environmental parameters exceed the appropriate range according to the preset environmental parameter control rules. When the environmental parameters exceed the threshold of the environmental suitable range, the main controller automatically controls the corresponding environmental adjustment equipment to adjust, realizing intelligent control of the environment;

[0016] Upload environmental data and control records to the adaptive knowledge service cloud platform module for storage and analysis.

[0017] Further, including:

[0018] The main controller automatically controls the operating status of the environmental conditioning equipment according to the preset environmental parameter control rules;

[0019] Environmental parameter control rules set environmental suitable range thresholds for light intensity, CO2 concentration, NH3 concentration, air temperature and air humidity parameters. When one or more parameters exceed or fall below the environmental suitable range threshold, the main controller automatically starts or shuts down the corresponding environmental adjustment device to enable precise control of environmental parameters.

[0020] The main controller uses the parameter deviation determination algorithm to determine whether the environmental parameters exceed the appropriate range threshold. :

[0021] ,in, Indicates the current environment parameter value. Indicates the standard parameter value, Indicates the parameter's allowable fluctuation range. or When the corresponding control action is triggered, Indicates the threshold of the environmental suitability range.

[0022] Furthermore, the environmental parameter control rules include:

[0023] When the parameter set Existence , then execute the device control set ; That is, when the environmental parameter exceeds the threshold value of the suitable range, the first control rule is indicated;

[0024] When the parameter set Existence , then execute the device control set ; That is, when the environmental parameter is lower than the threshold value of the suitable range, the second control rule is indicated;

[0025] When the parameter set Existing simultaneously and , then execute the device control set ; That is, the control rule when some parameters exceed the limit and some are insufficient represents the third control rule,

[0026] in, Represents environmental parameters, and Represent the upper and lower thresholds of the parameters respectively, A, B and C represent the device control action sets in three different situations respectively, Indicates the The first control rule executes the device control behavior, Indicates the The second control rule executes the device control behavior, Indicates the A third control rule executes the device control behavior.

[0027] Further, including:

[0028] When multiple rules are triggered simultaneously, control actions are executed in the order of security first, resource conservation second, and efficiency last, to obtain reasonable control decisions;

[0029] A multi-objective optimization model is used to handle rule conflicts, and the decision priority score S is calculated as follows: ,in, 、 and denote the normalized scores of safety, resource conservation and efficiency, respectively. 、 、 Represent the corresponding weight coefficients respectively, and satisfy ,and , to select the rule with the highest score to execute.

[0030] Further, including:

[0031] Based on the collected image data, image recognition technology is used to analyze livestock feed intake, activity patterns, and abnormal behaviors to obtain data on livestock health and behavior.

[0032] Among them, a deep learning convolutional neural network (CNN) model is used for image recognition and analysis. The model can represent:

[0033] Feature extraction: ;

[0034] Classification prediction: ;

[0035] Behavior recognition accuracy: ;

[0036] in, represents the input image, represents the weight matrix, represents the bias vector, represents the activation function, 、 、 and Represents the number of true positive, false positive, false negative and true negative samples respectively.

[0037] Furthermore, the permission level is obtained based on the user role, operation time and operation importance. Sure: ,in, Indicates basic authority level 15, Indicates additional permission value;

[0038] The correlation model between livestock growth status and environmental parameters is expressed as follows:

[0039] ;

[0040] in, represents the growth condition index, T represents temperature, H represents humidity, C represents CO2 concentration, N represents NH3 concentration, L represents light intensity, β0, β1,…, β5 represent regression coefficients, and ε represents the error term, so as to obtain a quantitative analysis of the impact of environmental parameters on livestock growth conditions.

[0041] Adjust environmental parameter thresholds according to permission level dynamics and association models to personalize and precisely control environmental parameters.

[0042] Further, including:

[0043] Monitor smart water and electricity meter data and analyze water and electricity resource consumption;

[0044] Correlate and analyze environmental regulation records with water and electricity consumption data to assess resource utilization efficiency;

[0045] Optimize environmental control strategies to achieve dual optimization of economic and environmental benefits of pasture production;

[0046] The optimized environmental control strategy:

[0047] ;

[0048] ;

[0049] Subject to constraints ;

[0050] in, represents the total cost, Indicates power consumption, Indicates water consumption, represents the maintenance cost, 、 and Represent the weight coefficients of electricity consumption, water consumption and maintenance cost respectively, Represents a set of environmental parameters, and They represent the lower and upper thresholds of the environmental control strategy parameters respectively. The optimal control parameter values ​​are obtained by optimizing the environmental control strategy to obtain livestock growth conditions while minimizing resource consumption.

[0051] Further, including:

[0052] Intelligent alarm mechanism, when environmental parameters exceed preset safety thresholds or equipment failure occurs, early warning is issued in the following ways:

[0053] Start local buzzer and LED indicator to give on-site alarm;

[0054] Send early warning information to managers' mobile devices through the adaptive knowledge service cloud platform;

[0055] Automatically record alarm events and generate suggestions for handling abnormal situations to provide support for management decisions;

[0056] In an emergency, the system automatically executes emergency plans, starts backup equipment or adjusts environmental control strategies.

[0057] The warning level calculation formula is as follows: ,when When the alarm is triggered;

[0058] in, Indicates the current parameter value. represents the parameter safety threshold, Indicates the normal range of parameters. represents the parameter weight coefficient, Indicates the alarm threshold, and the warning level indicates a general warning and emergency warnings There are two warning levels, and different response measures are taken respectively.

[0059] The second aspect is a livestock breeding system based on multi-sensor fusion technology and intelligent equipment, which is used to implement a livestock breeding method based on multi-sensor fusion technology and intelligent equipment, including:

[0060] Multi-sensor acquisition module, used to monitor pasture environmental parameters, livestock behavior and resource consumption data in real time, and obtain livestock breeding data sets through multi-sensor fusion technology;

[0061] An IoT real-time monitoring network module is used to build a ranch IoT real-time monitoring network based on the livestock breeding data set, so as to collect and transmit the livestock breeding data set in real time;

[0062] The adaptive knowledge service cloud platform module, based on the ranch IoT real-time monitoring network, can adjust the ranch environment in real time based on the real-time monitoring livestock breeding data set;

[0063] Intelligent alarm module: When the monitored livestock breeding data set exceeds the preset safety threshold, it indicates that the equipment has failed. Through the local buzzer and LED indicator light, it remotely outputs warning information and executes the emergency plan;

[0064] Remote control module: used to remotely monitor pasture environmental parameters, check livestock behavior, receive early warning information and control environmental conditioning equipment, so as to intelligently and accurately control livestock breeding environmental data and resource data.

[0065] The beneficial effects of the present invention are as follows: the present invention obtains a livestock breeding data set through real-time monitoring of pasture environmental parameters, livestock behavior and resource consumption data through multi-sensor fusion technology; constructs a pasture Internet of Things real-time monitoring network based on the livestock breeding data set to collect and transmit the livestock breeding data set in real time; based on the pasture Internet of Things real-time monitoring network, the pasture environment is regulated in real time according to the real-time monitored livestock breeding data set; when the monitored livestock breeding data set exceeds a preset safety threshold, it indicates that the equipment has failed, and through a local buzzer and LED indicator light, an early warning message is remotely output and an emergency plan is executed; by remotely monitoring pasture environmental parameters, checking livestock behavior, receiving early warning information and regulating environmental regulation equipment, livestock breeding environmental data and resource data are intelligently and accurately regulated; multi-sensor fusion technology and an Internet of Things real-time monitoring network are used to achieve real-time and accuracy of data, and automated control systems and intelligent decision support are used to improve management efficiency and accuracy, and with the help of smart water and electricity meters and an adaptive knowledge service cloud platform, resource utilization is optimized and costs are reduced, and the sustainable development of animal husbandry is promoted through precise environmental regulation and resource management. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 This is a flow chart of a livestock breeding method based on multi-sensor fusion technology and intelligent equipment provided in an embodiment of the present invention;

[0067] Figure 2 It is a schematic diagram of the module flow of the animal husbandry system based on multi-sensor fusion technology and intelligent equipment provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0068] The subject matter described herein will now be discussed with reference to exemplary embodiments. It should be understood that the discussion of these exemplary embodiments is intended only to enable those skilled in the art to better understand and implement the subject matter described herein, and that the functions and arrangements of the elements discussed may be varied without departing from the scope of this specification. Various examples may omit, substitute, or add various processes or components as needed. Furthermore, features described in some examples may be combined in other examples.

[0069] At least one embodiment of the present invention discloses a livestock farming system and method based on multi-sensor fusion technology and intelligent equipment, such as Figure 1 As shown, including:

[0070] Step 1: Real-time monitoring of pasture environmental parameters, livestock behavior, and resource consumption data, and acquisition of livestock breeding data sets through multi-sensor fusion technology.

[0071] The livestock pasture management system of the present invention consists of four parts: multi-sensor fusion technology, Internet of Things real-time monitoring network, adaptive knowledge service cloud platform and automatic control system. The system collects pasture environmental data in real time through multiple sensors, transmits it to the main controller through wireless network for processing and analysis, and automatically adjusts environmental parameters according to preset rules to ensure the suitability of the livestock growth environment. Figure 1 As shown in the figure, the overall system architecture includes the following core components:

[0072] Temperature and humidity sensor: A high-precision digital temperature and humidity sensor with a temperature accuracy of ±0.3°C and a humidity accuracy of ±2%RH is used to monitor the temperature and humidity changes in the pasture environment in real time.

[0073] Gas concentration sensors: including CO2 sensor (range 05000ppm, accuracy ±50ppm) and NH3 sensor (range 0100ppm, accuracy ±1ppm), used to monitor gas concentrations in pastures.

[0074] RGB camera: 1080P resolution, 120° field of view, installed at different locations on the pasture to monitor livestock activities and behavior patterns.

[0075] Infrared camera: Supports night vision function for continuous monitoring of livestock behavior in low-light conditions.

[0076] Smart water and electricity meter: with remote data collection function, accuracy of 0.5, and real-time monitoring of water and electricity resource consumption on the ranch.

[0077] Internet of Things real-time monitoring network module:

[0078] Wireless data transmission unit: uses ZigBee technology to build a network with a transmission distance of up to 300 meters, and has the characteristics of low power consumption and high reliability.

[0079] Main controller: It uses ARMCortexA72 architecture processor, runs Linux operating system, and is responsible for data processing, analysis and execution of control commands.

[0080] Step 2: Build a ranch IoT real-time monitoring network based on the livestock breeding dataset to collect and transmit the livestock breeding dataset in real time.

[0081] Based on the data collected from current livestock ranches, we will integrate and develop intelligent equipment, build a real-time monitoring network for ranch IoT, and establish a ranch environment monitoring system. The aim is to collect and summarize ranch management information and resource information, and then build an adaptive knowledge service cloud platform for ranches.

[0082] Step 3: Based on the real-time monitoring network of the ranch IoT, the ranch environment is regulated in real time according to the real-time monitored livestock breeding data set.

[0083] Data storage and display unit: adopts distributed database storage structure to support massive data storage and fast retrieval.

[0084] Intelligent decision support unit: Based on machine learning algorithms, it provides environmental control suggestions and resource management optimization solutions.

[0085] Remote viewing and control unit: supports web and mobile access, providing real-time data display and remote control functions.

[0086] Fixed / variable frequency fan: automatically adjusts wind speed according to environmental requirements, with a maximum air volume of 12,000 m³ / h and a power range of 0.75-5.5kW.

[0087] Water curtain pump: circulation flow rate 420m³ / h, head 812m, automatically adjusts water output according to temperature and humidity conditions.

[0088] Intelligent lighting: supports dimming function, with a light intensity range of 501000 lux, and automatically adjusts the light intensity according to time and environmental requirements.

[0089] Indoor intelligent temperature control: temperature control range 1040℃, accuracy ±0.5℃, supports multi-zone independent control.

[0090] Step 4: When the monitored livestock breeding data set exceeds the preset safety threshold, it indicates that the equipment has failed, and remotely outputs warning information and executes the emergency plan through the local buzzer and LED indicator.

[0091] Intelligent alarm device: including sound and light alarm and information push system, which can remind management personnel in time under abnormal circumstances.

[0092] Step 5: Used to remotely monitor pasture environmental parameters, check livestock behavior, receive early warning information and adjust environmental regulation equipment to intelligently and accurately adjust livestock breeding environmental data and resource data.

[0093] Each step cooperates with each other through preset environmental parameter control rules to achieve comprehensive monitoring and intelligent control of the pasture environment, ensuring the suitability of the livestock growth environment and the optimization of resource utilization.

[0094] Master controller parameter deviation determination algorithm:

[0095] The main controller automatically controls the operating status of the environmental conditioning equipment according to the preset environmental parameter control rules. Among them, the parameter deviation judgment algorithm is one of the core algorithms of the system, which is used to judge whether the environmental parameters exceed the appropriate range. :

[0096] ,in, Indicates the current environmental parameter values, including but not limited to the current temperature, humidity, and CO2 concentration environmental parameters. Indicates standard parameter values, set according to livestock species and growth stage, Indicates the allowable fluctuation range of the parameter, such as the allowable fluctuation range of temperature is ±3℃. or When the corresponding control action is triggered, Indicates the threshold of the environmental suitability range.

[0097] when or (preset threshold, usually set to 15%), the parameter is judged to deviate from the appropriate range and the corresponding control action is triggered. Example: Assume that the optimal temperature of the cow house is 18℃, the allowable fluctuation range is ±3℃, and the current temperature is 23℃, ,because If the system determines that the temperature is too high, it will automatically start the cooling equipment; thus, the differences and importance of different parameters can be fully considered, and the parameters of different dimensions can be evaluated uniformly through normalization, which improves the flexibility and adaptability of the system judgment.

[0098] In order to avoid errors in single sensor measurements and maintain stability, a multi-sensor data fusion algorithm is used to process the collected data. This algorithm is based on the weighted average method, and the calculation formula is as follows:

[0099]

[0100] in, is the measurement value of the i-th sensor, is the weight coefficient of the i-th sensor, is the sum of all weight coefficients.

[0101] Weight coefficient Dynamically adjusted according to the reliability and accuracy of the sensor, the calculation method is:

[0102]

[0103] in, is the quality factor of the sensor (a value between 0 and 1), evaluated based on the sensor's historical performance. is the historical measurement standard deviation of the sensor, reflecting the stability of the sensor.

[0104] This fusion algorithm can effectively filter out the impact of outliers and improve data accuracy and reliability. For example, in pasture temperature monitoring, if five temperature sensors are deployed, this algorithm can comprehensively consider the location and performance differences of each sensor to obtain temperature data that is closer to the actual value.

[0105] Environmental parameter control rules are the core basis for automatic system control. They include three scenarios: when environmental parameters exceed the appropriate range, when they fall below the appropriate range, and when some parameters exceed the limit while others fall below it. These rules are implemented using a decision tree algorithm and are formally represented as follows:

[0106] When the parameter set Existence , then execute the device control set ; That is, when the environmental parameter exceeds the threshold value of the suitable range, the first control rule is indicated;

[0107] When the parameter set Existence , then execute the device control set ; That is, when the environmental parameter is lower than the threshold value of the suitable range, the second control rule is indicated;

[0108] When the parameter set Existing simultaneously and , then execute the device control set ; That is, the control rule when some parameters exceed the limit and some are insufficient represents the third control rule,

[0109] in, Represents environmental parameters, and Represent the upper and lower thresholds of the parameters respectively, A, B and C represent the device control action sets in three different situations respectively, Indicates the The first control rule executes the device control behavior, Indicates the The second control rule executes the device control behavior, Indicates the A third control rule executes the device control behavior.

[0110] The implementation process of the decision tree algorithm is as follows:

[0111] 1. Input: current environment parameter set;

[0112] 2. Compare with the standard threshold: determine whether each parameter exceeds the threshold range;

[0113] 3. Construct a decision path: select the corresponding rule branch 4 based on the comparison result, and output: execute the corresponding device control action set.

[0114] Taking pig house environmental control as an example, some of the specific implementation rules are as follows:

[0115] Control scenarios when environmental parameters exceed the appropriate range:

[0116] Rule 1.1 (light intensity exceeds limit): When the light intensity is greater than 800 lux, the lighting device will be automatically turned off or dimmed to 50% brightness.

[0117] Rule 1.2 (CO2 concentration exceeds the limit): When CO2>2500ppm, the fixed / variable frequency fan will be automatically started and the wind speed will be set to 13m / s.

[0118] Rule 1.4 (Air temperature exceeds the limit): When the temperature is greater than 27°C, the temperature control device and fixed / variable frequency fans are automatically started, and the cooling target is set to 23°C.

[0119] Control scenarios when environmental parameters are below the appropriate range:

[0120] Rule 2.1 (Insufficient Light): When the light intensity is less than 200 lux, the lighting device will be automatically turned on to 85% brightness.

[0121] Rule 2.4 (Air temperature is too low): When the temperature is less than 16°C, start the temperature control device and set the temperature rise target to 18°C.

[0122] Rule 2.6 (temperature + insufficient light): When the light intensity is less than 200 lux and the temperature is less than 16°C, the system turns on the lighting and temperature control devices at the same time.

[0123] Control scenarios where some parameters exceed limits and some are insufficient:

[0124] Rule 3.1 (Insufficient light + Excessive temperature): When the light intensity is less than 200 lux and the temperature is greater than 27°C, the system turns on the lighting devices and starts the temperature control equipment and fixed / variable frequency fans. The application of the decision tree algorithm enables the system to quickly respond to complex environmental changes, achieve precise environmental control, and ensure the suitability of the livestock growth environment.

[0125] Multi-objective optimization model:

[0126] When multiple rules are triggered simultaneously, the system needs to decide which control actions to perform. To resolve rule conflicts, the system uses a multi-objective optimization model to determine the optimal control solution by calculating the decision priority score S:

[0127] ,in, 、 and denote the normalized scores of safety, resource conservation and efficiency, respectively. 、 、 Represent the corresponding weight coefficients respectively, and satisfy ,and , to select the rule with the highest score to execute.

[0128] In a typical pasture application scenario, the weight coefficient is set to , , , highlighting the principle of safety first.

[0129] Example: When the temperature is too high (30°C) and the CO2 concentration is too high (3000ppm), the system may trigger multiple control rules. Calculated by the multi-objective optimization model: Option 1 (only turning on the fan): = ; Option 2 (turn on the fan + water curtain system): ;

[0130] Select Score The more advanced Plan 2 is implemented, which turns on the fan and water curtain system at the same time, which can effectively reduce the temperature and reduce the CO2 concentration; the multi-objective optimization model ensures that the system can make the most reasonable control decisions when facing complex environmental changes, giving priority to livestock health and environmental safety.

[0131] Image recognition and livestock behavior analysis: Livestock behavior and health status are monitored in real time using RGB and infrared cameras, and a deep learning convolutional neural network (CNN) model is used for image recognition and analysis. The core algorithm of this model is as follows:

[0132] Feature extraction: Use convolutional layers to extract image features ;

[0133] Classification prediction: through the fully connected layer and Functions for behavioral classification Map the extracted features to different behavior category probabilities.

[0134] Behavior recognition accuracy: Evaluate recognition accuracy parameters through confusion matrix :

[0135] ;

[0136] in, represents the input image, represents the weight matrix, represents the bias vector, represents the activation function, 、 、 and Represents the number of true positive, false positive, false negative and true negative samples respectively.

[0137] Trained on 30,000 labeled images, the CNN model can identify various livestock behaviors, including normal feeding, drinking, resting, and abnormal activity, with an accuracy rate of 92.5%. The system can also analyze group behavior patterns, such as feeding frequency and duration, to assess livestock health.

[0138] In actual applications, the system can automatically identify individuals with abnormal behavior in the herd, such as cattle with significantly reduced feed intake or abnormal activity patterns, and promptly alert managers to pay attention, thereby achieving early warning of livestock health status and improving the efficiency of disease prevention and treatment.

[0139] Remote viewing and control system: The adaptive knowledge service cloud platform module includes a remote viewing and control unit, which enables managers to remotely monitor pasture environmental parameters, check livestock conditions, receive early warning information, and control environmental conditioning equipment through mobile devices or computers. This unit uses a hierarchical authority control algorithm, and the authority level R is calculated as follows:

[0140] ;

[0141] in, Indicates basic authority levels 1-5, Indicates additional permission value;

[0142] Calculated based on the following factors:

[0143] ;

[0144] in, is the user role coefficient, is the time factor (to elevate privileges in an emergency), is the operation importance coefficient, 、 and Represent the weight coefficients of electricity consumption, water consumption and maintenance cost respectively.

[0145] The system sets five levels of authority management:

[0146] Level 1 (Observation Clearance): Can only view environmental data and livestock status;

[0147] Level 2 (basic control authority): can adjust non-critical equipment such as lighting and ventilation;

[0148] Level 3 (Advanced Control Authority): Can adjust key equipment such as the temperature control system and water curtain system;

[0149] Level 4 (system setting permission): can modify environmental parameter thresholds and control rules;

[0150] Level 5 (Administrator privileges): Has control authority over all system functions.

[0151] The remote control system uses 256-bit SSL encrypted transmission to ensure data security; it also supports operation log recording and auditing functions to record all remote operation behaviors and improve system security and traceability.

[0152] In actual applications, managers can view real-time ranch environmental data, equipment operating status and livestock monitoring videos through mobile apps or web pages, and remotely adjust environmental parameters or equipment operating status when abnormal situations are found, greatly improving management efficiency and response speed.

[0153] Correlation model between livestock growth status and environmental parameters

[0154] The system monitors livestock behavior patterns and feed intake, combines them with environmental parameter data for correlation analysis, and establishes a correlation model between livestock growth status and environmental parameters. This model uses a multiple regression algorithm, and the expression is:

[0155] ;

[0156] in, represents the growth condition index, T represents temperature, H represents humidity, C represents CO2 concentration, N represents NH3 concentration, L represents light intensity, β0, β1,…, β5 represent regression coefficients, and ε represents the error term, so as to obtain a quantitative analysis of the impact of environmental parameters on livestock growth conditions.

[0157] The system establishes multiple correlation models based on different livestock breeds and growth stages. For example, for fattening pigs, a model is trained using 30 days of environmental parameters and daily weight gain data. This model can predict growth performance under different environmental conditions and, through reverse calculation, determine the optimal combination of environmental parameters. Based on the correlation model, the system dynamically adjusts environmental parameter thresholds to better suit livestock growth needs, achieving personalized and precise environmental regulation. In practical application, this model has helped the farm increase the average daily weight gain of fattening pigs by 8.5% and the feed conversion rate by 5.2%.

[0158] Resource optimization management: By monitoring smart water and electricity meter data, analyzing water and electricity resource consumption, and correlating environmental control records with water and electricity consumption data, we can evaluate resource utilization efficiency. The resource optimization model is implemented through the following objective function:

[0159] ;

[0160] ;

[0161] Subject to constraints ;

[0162] in, represents the total cost, Indicates power consumption, Indicates water consumption, represents the maintenance cost, Represents a set of environmental parameters, and They represent the lower and upper thresholds of the environmental control strategy parameters respectively. The optimal control parameter values ​​are obtained by optimizing the environmental control strategy to obtain livestock growth conditions while minimizing resource consumption.

[0163] The system uses the particle swarm optimization algorithm (PSO) to solve the optimal control parameter values:

[0164] 1. Initialize particle positions and velocities, where each particle represents a set of possible environmental parameter settings;

[0165] 2. Calculate the fitness value of each particle (total cost C);

[0166] 3. Update individual optimal position and global optimal position;

[0167] 4. Update particle speed and position;

[0168] 5. Repeat step 24 until convergence.

[0169] In practical applications, the system intelligently schedules fan operating times and modes, reducing fan power consumption by approximately 15% without affecting environmental parameters. Intelligent control of the water pump and water curtain system reduces water waste and improves water utilization efficiency by 20%. Furthermore, predictive maintenance strategies reduce equipment failure rates and maintenance costs, reducing overall operating costs by approximately 18%.

[0170] Intelligent alarm mechanism: When environmental parameters exceed the preset safety threshold or equipment fails, an early warning is automatically issued. The early warning level calculation formula is:

[0171]

[0172] in, Indicates the current parameter value. represents the parameter safety threshold, Indicates the normal range of parameters. represents the parameter weight coefficient, Indicates the alarm threshold, and the warning level indicates a general warning and emergency warnings There are two warning levels, and different response measures are taken respectively.

[0173] When an alarm is triggered, the warning level is divided into two levels:

[0174] General warning : Start the local LED indicator and send early warning information to the management personnel;

[0175] Emergency Warning : Start the buzzer and LED flashing alarm, send emergency notification to all management personnel, and automatically execute the emergency plan.

[0176] The emergency plan includes:

[0177] 1. Automatic start of standby equipment: such as standby generators and standby water pumps;

[0178] 2. Automatic adjustment of environmental control strategies: such as emergency ventilation and emergency cooling;

[0179] 3. Automatic allocation of emergency resources: such as prioritizing power supply and environmental control in key areas.

[0180] Taking NH3 concentration exceeding the standard as an example, when the concentration exceeds 20ppm (general warning threshold), the system activates the LED indicator light and notifies the management personnel; when it exceeds 35ppm (emergency warning threshold), the system activates the sound and light alarm, automatically adjusts the fan to the maximum air volume, turns on the emergency ventilation mode, and notifies all management personnel to take further measures.

[0181] In actual applications, the intelligent alarm mechanism effectively shortens the response time to abnormal situations from an average of 30 minutes to less than 5 minutes, greatly reducing livestock stress reactions and health risks caused by environmental abnormalities.

[0182] To verify the availability and reliability of this system, it was deployed and tested in multiple ranch environments. The following describes the actual effects of the system in combination with specific application scenarios.

[0183] Application examples in dairy farms:

[0184] At a large dairy farm in North China, the system deployed 20 sets of sensor nodes, covering five cowshed areas. Each area contains multiple sensors such as temperature and humidity, gas concentration, and light. RGB cameras and infrared cameras are also installed at key locations.

[0185] Example 1: Summer High Temperature Environment Control During an extremely hot weather event in July 2025, when the outdoor temperature reached 38°C, the system detected that the temperature inside the cowshed had risen to 32°C (exceeding the upper temperature limit of 28°C for dairy cows). The deviation D reached 133.3%, far exceeding the 15% threshold. The system immediately initiated a multi-stage cooling plan:

[0186] 1. First start the fixed frequency fan for basic ventilation;

[0187] 2. When the temperature continues to rise, the system automatically turns on the water curtain system;

[0188] 3. At the same time, adjust the lighting equipment to 30% brightness to reduce heat generation.

[0189] Through multi-objective optimization, the system selected the most energy-efficient equipment combination, maintaining a stable barn temperature within a range of 26±1°C. Compared to traditional manual intervention, this approach reduced response time by approximately 15 minutes and energy consumption by approximately 23%. During this time, CNN image recognition technology monitored the cows' feeding behavior as normal, with no signs of heat stress, ensuring their production performance.

[0190] Example 2: Abnormal gas concentration warning. During a late-night equipment operation, the system detected that the NH3 concentration in the cowshed rose rapidly from the normal 8ppm to 23ppm (exceeding the safety threshold of 20ppm), and entered the general warning state:

[0191] 1. Automatically activate the LED indicator alarm;

[0192] 2. Push abnormal notifications to the on-duty personnel’s mobile app;

[0193] 3. Automatically adjust the fan to high speed mode to increase ventilation.

[0194] After receiving the warning, the on-duty personnel checked the situation through the remote control system and discovered that the cause was a temporary malfunction in the cattle shed cleaning system. Before manual intervention, the system had already controlled NH3 concentrations to a safe level, avoiding the risk of respiratory illness in livestock. The response time for this incident was only one-sixth of that of traditional management models, significantly improving efficiency.

[0195] Pig farm application example:

[0196] At a modern pig farm in Jiangsu Province, the system covers pig areas at different growth stages and realizes intelligent control of environmental parameters throughout the entire process.

[0197] Example 1: The impact of precise environmental control on growth performance. A growth and environmental parameter correlation model established based on a multivariate regression algorithm shows that fattening pigs grow best in an environment with a temperature of 22±2°C, a humidity of 60-70%, an NH3 concentration of less than 10ppm, and a CO2 concentration of less than 1800ppm. Based on this, the system precisely controls the environment and conducts comparative tests with piggeries using traditional management methods. The results show that fattening pigs in the intelligently controlled environment:

[0198] 1. The average daily weight gain increased by 8.5%, from the original 0.82kg / day to 0.89kg / day;

[0199] 2. The feed conversion rate increased by 5.2%, from the original 2.88 to 2.733. The incidence of respiratory diseases decreased by 32%. The economic benefit analysis showed that the system investment cost can be recovered in less than 2 years, and after long-term operation, the net income per pig can be increased by about 3,545 yuan.

[0200] Example 2: Resource Optimization Example: A particle swarm optimization algorithm is used to intelligently schedule pig house environmental control equipment. While ensuring appropriate environmental parameters, the equipment operation plan is optimized: 1. The fan operation adopts variable frequency technology to adjust the wind speed according to actual demand, saving 22% of electricity compared with constant-speed fans. 2. The water curtain system and temperature control equipment work together to select the optimal combination under different temperature and humidity conditions, reducing energy consumption by 18%. 3. The lighting system automatically adjusts according to the sunshine time and growth needs, reducing lighting energy consumption by 25%. One year of operation data shows that the system has helped the pig farm reduce total energy consumption by approximately 16.5%, save water resources by approximately 20%, and reduce carbon emissions by approximately 14 tons per year.

[0201] Application of cascading control rules: In actual applications, multiple parameter anomalies often occur simultaneously. The system effectively handles complex scenarios through the control rules in the claims. The following are typical scenario processing examples in actual applications:

[0202] Example 1: Insufficient light + excessive temperature scenario: On a winter morning, the system detects that the light intensity in the chicken house is only 150 lux (below the standard of 200 lux) and the temperature reaches 33°C (above the standard of 30°C). The system triggers the following control measures according to rule 3.1:

[0203] 1. Turn on the lighting device to 85% brightness to supplement the light;

[0204] 2. Start the temperature control device and set the target temperature to 27°C;

[0205] 3. Turn on the fixed-frequency fan to increase ventilation and heat dissipation. After multi-objective optimization, the system prioritizes safety (temperature control) and secondarily takes into account production performance (light supplement), ultimately restoring the environment to a suitable state within 15 minutes.

[0206] Example 2: CO2 Concentration Exceeding Limit + NH3 Concentration Exceeding Limit Scenario During a temporary ventilation system failure, the system detected that the CO2 concentration in the sheep house rose to 3200ppm (exceeding the standard of 2500ppm), and the NH3 concentration reached 18ppm (close to the warning threshold of 20ppm). The system implemented the following control measures:

[0207] 1. Start the backup fan system to increase ventilation;

[0208] 2. Turn on the water curtain pump to increase the air humidity to reduce the NH3 concentration;

[0209] 3. Trigger a level 1 alarm and notify management personnel.

[0210] Since rules 1.2 and 1.3 were met at the same time, the system selected the optimal combination through the conflict resolution mechanism, reducing the concentrations of the two gases to a safe level within 40 minutes and avoiding possible animal health risks.

[0211] To fully verify the system's stability and reliability, a six-month comprehensive test was conducted in 2025. The test environment included farms of different livestock species (cattle, pigs, sheep, and poultry), covering different climate conditions and management scales. The test focused on the following indicators:

[0212] System stability: During the six-month operation period, the system's core function availability reached 99.7%, the sensor data collection accuracy reached 98.2%, and the environmental control success rate reached 97.5%.

[0213] Intelligent decision-making capability: For 42 simulated environmental abnormalities, the system's correct recognition rate reached 95.8%, and the rationality score of the control plan reached 92.3 points (out of 100 points).

[0214] Economic benefits: Compared with traditional management methods, the system saves an average of 16.8% in energy costs, reduces feeding losses by 12.3%, and improves production efficiency by 7.6%.

[0215] User evaluation: Managers from 10 different farms rated the system with an average satisfaction score of 4.6 (out of 5), particularly praising the system's remote monitoring, intelligent early warning, and automated control functions.

[0216] The above application examples and experimental verification results show that the system of the present invention has significant application value in actual livestock breeding scenarios, and can effectively solve the shortcomings of traditional pasture management systems in real-time, accuracy, degree of automation and data management, thereby improving pasture management efficiency, reducing resource consumption, and promoting the sustainable development of animal husbandry.

[0217] The above application examples and experimental verification results show that the system of the present invention has significant application value in actual livestock breeding scenarios, and can effectively solve the shortcomings of traditional pasture management systems in real-time, accuracy, degree of automation and data management, thereby improving pasture management efficiency, reducing resource consumption, and promoting the sustainable development of animal husbandry.

[0218] like Figure 2 As shown, the livestock breeding system based on multi-sensor fusion technology and intelligent equipment includes:

[0219] Multi-sensor acquisition module, used to monitor pasture environmental parameters, livestock behavior and resource consumption data in real time, and obtain livestock breeding data sets through multi-sensor fusion technology;

[0220] An IoT real-time monitoring network module is used to build a ranch IoT real-time monitoring network based on the livestock breeding data set, so as to collect and transmit the livestock breeding data set in real time;

[0221] The adaptive knowledge service cloud platform module, based on the ranch IoT real-time monitoring network, can adjust the ranch environment in real time based on the real-time monitoring livestock breeding data set;

[0222] Intelligent alarm module: When the monitored livestock breeding data set exceeds the preset safety threshold, it indicates that the equipment has failed. Through the local buzzer and LED indicator, the warning information is remotely output and the emergency plan is executed;

[0223] Remote control module: By remotely monitoring pasture environmental parameters, checking livestock behavior, receiving early warning information and adjusting environmental conditioning equipment, it can intelligently and accurately control livestock breeding environmental data and resource data.

[0224] The above describes an embodiment of the present invention, but this embodiment is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Ordinary technicians in this field can also make more forms of equivalent embodiments based on the inspiration of this embodiment, all of which are protected by this embodiment.

Claims

1. A livestock breeding method based on multi-sensor fusion technology and intelligent equipment, characterized in that: include: Real-time monitoring of pasture environmental parameters, livestock behavior, and resource consumption data, and acquisition of livestock breeding data sets through multi-sensor fusion technology; Building a ranch IoT real-time monitoring network based on the livestock breeding dataset to collect and transmit the livestock breeding dataset in real time; Based on the real-time monitoring network of the ranch Internet of Things, the ranch environment is regulated in real time according to the real-time monitoring livestock breeding data set; When the monitored livestock breeding data set exceeds the preset safety threshold, it indicates that the equipment has failed. Through the local buzzer and LED indicator, the warning information is remotely output and the emergency plan is executed; By remotely monitoring pasture environmental parameters, checking livestock behavior, receiving early warning information and adjusting environmental conditioning equipment, livestock breeding environmental data and resource data can be intelligently and accurately controlled.

2. The livestock breeding method based on multi-sensor fusion technology and intelligent equipment according to claim 1, characterized in that: include: Collect pasture environmental data through multi-sensor fusion technology, including light intensity, CO2 concentration, NH3 concentration, air temperature, air humidity parameters and livestock behavior data; The collected data is transmitted to the main controller in real time through the wireless network in the real-time monitoring network of the Internet of Things, realizing real-time transmission and centralized processing of data; The main controller determines whether the environmental parameters exceed the appropriate range according to the preset environmental parameter control rules. When the environmental parameters exceed the threshold of the environmental suitable range, the main controller automatically controls the corresponding environmental adjustment equipment to adjust, realizing intelligent control of the environment; Upload environmental data and control records to the adaptive knowledge service cloud platform module for storage and analysis.

3. The livestock breeding method based on multi-sensor fusion technology and intelligent equipment according to claim 2, characterized in that: include: The main controller automatically controls the operating status of the environmental conditioning equipment according to the preset environmental parameter control rules; Environmental parameter control rules set environmental suitable range thresholds for light intensity, CO2 concentration, NH3 concentration, air temperature and air humidity parameters. When one or more parameters exceed or fall below the environmental suitable range threshold, the main controller automatically starts or shuts down the corresponding environmental adjustment device to enable precise control of environmental parameters. The main controller uses the parameter deviation determination algorithm to determine whether the environmental parameters exceed the appropriate range threshold. : ,in, Indicates the current environment parameter value. Indicates the standard parameter value, Indicates the parameter's allowable fluctuation range. or When the corresponding control action is triggered, Indicates the threshold of the environmental suitability range.

4. The livestock breeding method based on multi-sensor fusion technology and intelligent equipment according to claim 3, characterized in that: The environmental parameter control rules include: When the parameter set Existence , then execute the device control set ; That is, when the environmental parameter exceeds the threshold value of the suitable range, the first control rule is indicated; When the parameter set Existence , then execute the device control set ; That is, when the environmental parameter is lower than the threshold value of the suitable range, the second control rule is indicated; When the parameter set Existing simultaneously and , then execute the device control set ; That is, the control rule when some parameters exceed the limit and some are insufficient represents the third control rule, in, Represents environmental parameters, and Represent the upper and lower thresholds of the parameters respectively, A, B and C represent the device control action sets in three different situations respectively, Indicates the The first control rule executes the device control behavior, Indicates the The second control rule executes the device control behavior, Indicates the A third control rule executes the device control behavior.

5. The livestock breeding method based on multi-sensor fusion technology and intelligent equipment according to claim 4, characterized in that: include: When multiple rules are triggered simultaneously, control actions are executed in the order of security first, resource conservation second, and efficiency last, to obtain reasonable control decisions; A multi-objective optimization model is used to handle rule conflicts, and the decision priority score S is calculated as follows: ,in, 、 and denote the normalized scores of safety, resource conservation and efficiency, respectively. 、 、 Represent the corresponding weight coefficients respectively, and satisfy ,and , to select the rule with the highest score to execute.

6. The livestock breeding method based on multi-sensor fusion technology and intelligent equipment according to claim 1, characterized in that: include: Based on the collected image data, image recognition technology is used to analyze livestock feed intake, activity patterns, and abnormal behaviors to obtain data on livestock health and behavior. Among them, a deep learning convolutional neural network (CNN) model is used for image recognition and analysis. The model can represent: Feature extraction: ; Classification prediction: ; Behavior recognition accuracy: ; in, represents the input image, represents the weight matrix, represents the bias vector, represents the activation function, 、 、 and represent the number of true positive, false positive, false negative, and true negative samples, respectively.

7. The livestock breeding method based on multi-sensor fusion technology and intelligent equipment according to claim 1, characterized in that: Obtain permission levels based on user role, operation time, and operation importance Sure: ,in, Indicates basic authority level 15, Indicates additional permission value; The correlation model between livestock growth status and environmental parameters is expressed as follows: ； in, represents the growth status index, T represents temperature, H represents humidity, C represents CO2 concentration, N represents NH3 concentration, L represents light intensity, β0, β1,…, β5 represent regression coefficients, and ε represents the error term, in order to obtain a quantitative analysis of the impact of environmental parameters on livestock growth status; Adjust environmental parameter thresholds according to permission level dynamics and association models to personalize and precisely control environmental parameters.

8. The livestock breeding method based on multi-sensor fusion technology and intelligent equipment according to claim 1, characterized in that: include: Monitor smart water and electricity meter data and analyze water and electricity resource consumption; Correlate and analyze environmental control records with water and electricity consumption data to assess resource utilization efficiency; Optimize environmental control strategies to achieve dual optimization of economic and environmental benefits of pasture production; The optimized environmental control strategy: ； ； Subject to constraints ; in, represents the total cost, Indicates power consumption, Indicates water consumption, represents the maintenance cost, 、 and Represent the weight coefficients of electricity consumption, water consumption and maintenance cost respectively, Represents a set of environmental parameters, and They represent the lower and upper thresholds of the environmental control strategy parameters respectively. The optimal control parameter values ​​are obtained by optimizing the environmental control strategy to obtain livestock growth conditions while minimizing resource consumption.

9. The livestock breeding method based on multi-sensor fusion technology and intelligent equipment according to claim 1, characterized in that: include: Intelligent alarm mechanism, when environmental parameters exceed preset safety thresholds or equipment failure occurs, early warning is issued in the following ways: Start local buzzer and LED indicator to give on-site alarm; Send early warning information to managers' mobile devices through the adaptive knowledge service cloud platform; Automatically record alarm events and generate suggestions for handling abnormal situations to provide support for management decisions; In an emergency, the system automatically executes emergency plans, starts backup equipment or adjusts environmental control strategies. The warning level calculation formula is as follows: ,when When the alarm is triggered; in, Indicates the current parameter value. represents the parameter safety threshold, Indicates the normal range of parameters. represents the parameter weight coefficient, Indicates the alarm threshold, and the warning level indicates a general warning and emergency warnings There are two warning levels, and different response measures are taken respectively.

10. A livestock breeding system based on multi-sensor fusion technology and intelligent equipment, used to implement the livestock breeding method based on multi-sensor fusion technology and intelligent equipment according to any one of claims 19, characterized in that: include: Multi-sensor acquisition module, used to monitor pasture environmental parameters, livestock behavior and resource consumption data in real time, and obtain livestock breeding data sets through multi-sensor fusion technology; An IoT real-time monitoring network module is used to build a ranch IoT real-time monitoring network based on the livestock breeding data set, so as to collect and transmit the livestock breeding data set in real time; The adaptive knowledge service cloud platform module, based on the ranch IoT real-time monitoring network, can adjust the ranch environment in real time based on the real-time monitoring livestock breeding data set; Intelligent alarm module: When the monitored livestock breeding data set exceeds the preset safety threshold, it indicates that the equipment has failed. Through the local buzzer and LED indicator light, it remotely outputs warning information and executes the emergency plan; Remote control module: used to remotely monitor pasture environmental parameters, check livestock behavior, receive early warning information and control environmental conditioning equipment, so as to intelligently and accurately control livestock breeding environmental data and resource data.

Images