Multifunctional electronic ear tag, working method, device, equipment and storage medium

Abstract

The invention discloses a multifunctional intelligent electronic ear tag device. The ear tag adopts a head hoop and an arc-shaped ear hook structure, a connecting buckle is of a magnetic attraction type, punching is not needed, and the head hoop is made of medical silica gel, so that it is ensured that the ear tag is comfortable to wear by animals and convenient to use repeatedly. The power management module guarantees power supply, and the data monitoring module can collect multi-dimensional information such as animal body surface temperature, environment and behavior data. During working, breeding areas are divided to construct a wireless network, data are acquired and then preprocessed, and the animal growth state sensing model is utilized to analyze and upload a result. The power consumption adaptive algorithm dynamically adjusts the working mode of the ear tag according to the equipment state and the like, such as adjusting the sampling frequency and the local model complexity, so that low-power-consumption edge calculation and accurate reasoning are realized. The user terminal supports data display, interaction control and remote alarm, and helps a manager adjust a breeding strategy based on an analysis result. The system has the characteristics of high comfort, low power consumption and high precision, and provides technical support for digital transformation of animal husbandry.

Description

Technical Field

[0001] The present application relates to the technical field of intelligent electronic ear tags, and in particular to a multifunctional electronic ear tag, a working method, a device, a equipment and a storage medium. Background Art

[0002] The behavior monitoring and health status assessment of livestock in animal husbandry are of great significance to improving production efficiency and ensuring animal welfare. Traditional livestock monitoring methods mainly rely on manual observation and statistics, which not only consumes a lot of manpower and time, but is also easily affected by the subjectivity of observers and environmental conditions, making it difficult to achieve real-time and comprehensive monitoring. In recent years, with the development of Internet of Things technology and sensor technology, livestock behavior monitoring systems based on multi-sensors have gradually emerged. They collect animal behavior and physiological data through sensors such as acceleration, gyroscopes, and temperature, and then combine machine learning technology to analyze behavior patterns and assess health, greatly improving the efficiency and accuracy of monitoring.

[0003] However, traditional livestock behavior analysis systems have the following limitations in feature extraction and reasoning capabilities: ① Existing methods usually process multi-source sensor data based on a single modal feature extraction model, which cannot fully integrate the correlation information between different modal data, resulting in limited accuracy of behavior recognition; ② The system consumes high power during high-frequency sampling and high-complexity model reasoning, which makes it difficult to meet the low-power requirements of long-term monitoring scenarios; ③ The computing model on the edge device is usually a fixed structure and lacks dynamic adjustment capabilities, making it difficult to adaptively allocate computing resources according to device status and task requirements.

[0004] With the rapid development of deep learning technology, Transformer-based multimodal data processing models have been widely used in the field of behavior recognition due to their excellent performance in long- and short-range dependency modeling and cross-modal feature fusion. At the same time, technologies such as knowledge distillation, channel pruning, and reinforcement learning provide effective solutions for model optimization in edge computing environments. Many studies have begun to focus on how to use these technologies to improve the low-power reasoning capabilities of behavior monitoring equipment. For example, Banitalebi-Dehkordi proposed a knowledge distillation method for low-power object detection in "The 2021International Conference on Computer VisionWorkshops" (ICCVW 2021). By decoupling the distillation of the teacher model from the true label, the teacher model is used to provide the student model with a subset that is easier to learn, thereby reducing the demand for labeled data and reducing energy consumption. However, most of these methods focus on optimization at a single level and have not yet achieved a full range of low-power design for data sampling, model reasoning, and computing resource allocation.

[0005] In addition, the design and installation methods of traditional ear tag equipment also have certain limitations on animal comfort and management efficiency. In the existing technology, most electronic ear tag devices still use invasive punching installation, which not only causes pain and infection risks to animals, but may also induce stress reactions and affect the normal growth and behavior of animals. At the same time, the assembly and disassembly operations of many ear tag devices are complicated and inefficient, and cannot meet the farms' needs for rapid deployment and reuse. These problems limit the widespread application of existing monitoring equipment in intelligent management of animal husbandry. Summary of the invention

[0006] 1. Problem to be solved

[0007] Based on this, it is necessary to provide a multifunctional electronic ear tag, working method, device, computer equipment, computer-readable storage medium and computer program product that can be quickly deployed without punching to address the above technical problems.

[0008] 2. Technical solution

[0009] In the first aspect, the present application provides a multifunctional electronic ear tag, comprising: a headband body, an ear hook is arranged on the headband body, the ear hook is arranged in an arc shape, a connecting buckle is arranged on the ear hook, the connecting buckle includes a male mark and a female mark, the male mark is threadedly connected to the female mark, the male mark is a male mark made of soft magnetism, a magnet is embedded in the female mark, the male mark and the female mark are pressed against each other under the magnetic force of the magnet, a power management module and a data monitoring module are arranged on the headband body, the power management module supplies power to the data monitoring module, and the data monitoring module is used to collect environmental data and behavior data.

[0010] In one embodiment, the data monitoring module includes: an environment monitoring unit, a behavior monitoring unit and a positioning unit, the environment monitoring unit includes a body temperature sensor and an air quality sensor, the body temperature sensor is arranged on the inner side of the headband body, and the air quality sensors are respectively arranged on both sides of the headband body;

[0011] The behavior monitoring unit includes an acceleration sensor, a gyroscope sensor and a sound sensor, wherein the sound sensor is arranged on the ear hook, and the acceleration sensor and the gyroscope sensor are arranged on the headband body;

[0012] The positioning unit includes a Beidou satellite navigation system positioning chip, which is arranged at the center of the headband body.

[0013] In a second aspect, the present application also provides a multifunctional electronic ear tag working method. The method comprises:

[0014] Divide the preset breeding areas and build a wireless network for data transmission and positioning;

[0015] Acquire multi-dimensional data of animals in the breeding area based on the monitoring module, wherein the multi-dimensional data includes at least body surface temperature, environmental parameters, movement status, sound signals, and activity tracks;

[0016] Preprocess the multidimensional data according to the data type, including converting it into numerical form using target embedding methods and normalizing and completing numerical features;

[0017] Apply the animal growth status perception model to analyze the processed data and send the analysis results to the terminal;

[0018] Adopt power consumption adaptive algorithm to dynamically adjust ear tag working mode.

[0019] In one embodiment, intelligently analyzing and processing multidimensional data based on data types includes:

[0020] For different sensor data, the target embedding method (Target Encoding) is used to convert the classification features into numerical form and calculate the conditional probability of each category corresponding to the target variable, which is expressed as:

[0021]

[0022] Among them, x i is a certain category feature value, y j is the target variable, II is the indicator function;

[0023] Z-score standardization is used to normalize numerical features and autoregressive neural network modeling is used to complete missing data.

[0024] In one embodiment, the method of applying the animal growth status perception model includes:

[0025] On the server side, build corresponding Transformer networks based on data types and fuse the extracted features;

[0026] When in the feature fusion stage, the dynamic modal attention weight is introduced, and the formula is as follows:

[0027]

[0028] Among them, q m is the query vector of modality m, k m is the key vector of mode m, cos is the cosine similarity function, α m is the dynamic modality attention weight;

[0029] The fused features are input into the cross-modal Transformer model with the MAWCL loss function added to obtain the analysis results. The loss function formula is defined as follows:

[0030]

[0031] Where M is the total number of modes, α m is the dynamic weight of mode m, L task,m is the task loss of modality m, L inconsistency,m is the consistency loss between mode m and other modes, and λ is the weight coefficient of the inconsistency loss.

[0032] In one embodiment, applying the animal growth status perception model method further includes:

[0033] The complex model trained on the server side is defined as a pre-trained animal growth status perception model A, the simplified model is defined as a local lightweight animal growth status perception model B, and the simplified model B is embedded into the ear tag;

[0034] The analysis results of the complex model on the server side are used to guide the learning of the simplified model in the form of knowledge distillation. The loss function formula is as follows:

[0035] L=(1-α)·L hard +α·T 2 ·KL(P A ||P B );

[0036] Among them, L hard is the standard cross entropy loss, KL represents KL divergence, T is the temperature coefficient, α is the weight parameter, and P A is the output distribution of model A, P B is the output distribution of model B.

[0037] In one of the embodiments, the remaining power, current power consumption and sensor temperature of the monitoring device are obtained and a state vector is formed;

[0038] The state vector is input through the policy network in the reinforcement learning framework to dynamically generate a resource allocation strategy with the following optimization objectives:

[0039]

[0040] Where π is the policy network, γ is the discount factor, and R(t) is the reward function based on power consumption and performance;

[0041]

[0042] Where Q(t) is the quality of the inference task, w1 is the weight factor for balancing power consumption P(t), w2 is the weight factor for balancing performance Q(t), and w3 is the weight factor for balancing temperature T(t);

[0043] A variable depth network is introduced to reduce the amount of inference calculation through a dynamic skip layer mechanism. The formula is as follows:

[0044]

[0045] Among them, g i (x) is the output of the i-th layer, s i Score the computational importance of the layer, τ is an adjustable threshold;

[0046] Select an appropriate sampling frequency based on the device status and the importance of the inference task. The formula is as follows:

[0047]

[0048] Among them, P sensor is the power consumption of sampling, P compute The power consumption for calculation;

[0049] When the ear tag status monitoring shows that the power consumption is within the preset threshold, the complexity of the local lightweight animal growth status perception model B is dynamically adjusted and low-frequency sampling is used to extend the ear tag working time.

[0050] In a third aspect, the present application also provides a multifunctional electronic ear tag control device. The device comprises:

[0051] A wireless network construction module, used to divide the preset breeding area and build a wireless network for data transmission and positioning;

[0052] A multi-dimensional data acquisition module, used for acquiring multi-dimensional data of animals in the breeding area based on the monitoring module, wherein the multi-dimensional data at least includes body surface temperature, environmental parameters, movement status, sound signals, and activity tracks;

[0053] Numerical preprocessing module, used to preprocess multidimensional data according to data type, including using target embedding method to convert it into numerical form and normalize and complete numerical features;

[0054] An animal growth status perception module is used to analyze the processed data using an animal growth status perception model and send the analysis results to a terminal;

[0055] The power consumption adaptive adjustment module is used to dynamically adjust the ear tag working mode by using a power consumption adaptive algorithm.

[0056] In a fourth aspect, the present application further provides a computer device. The computer device includes a memory and a processor, the memory stores a computer program, and the processor implements the following steps when executing the computer program:

[0057] Divide the preset breeding areas and build a wireless network for data transmission and positioning;

[0058] Acquire multi-dimensional data of animals in the breeding area based on the monitoring module, wherein the multi-dimensional data includes at least body surface temperature, environmental parameters, movement status, sound signals, and activity tracks;

[0059] Preprocess the multidimensional data according to the data type, including converting it into numerical form using target embedding methods and normalizing and completing numerical features;

[0060] Apply the animal growth status perception model to analyze the processed data and send the analysis results to the terminal;

[0061] Adopt power consumption adaptive algorithm to dynamically adjust ear tag working mode.

[0062] In a fifth aspect, the present application further provides a computer-readable storage medium. The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the following steps are implemented:

[0063] Divide the preset breeding areas and build a wireless network for data transmission and positioning;

[0064] Acquire multi-dimensional data of animals in the breeding area based on the monitoring module, wherein the multi-dimensional data includes at least body surface temperature, environmental parameters, movement status, sound signals, and activity tracks;

[0065] Preprocess the multidimensional data according to the data type, including converting it into numerical form using target embedding methods and normalizing and completing numerical features;

[0066] Apply the animal growth status perception model to analyze the processed data and send the analysis results to the terminal;

[0067] Adopt power consumption adaptive algorithm to dynamically adjust ear tag working mode.

[0068] In a sixth aspect, the present application further provides a computer program product. The computer program product includes a computer program, and when the computer program is executed by a processor, the following steps are implemented:

[0069] Divide the preset breeding areas and build a wireless network for data transmission and positioning;

[0070] Acquire multi-dimensional data of animals in the breeding area based on the monitoring module, wherein the multi-dimensional data includes at least body surface temperature, environmental parameters, movement status, sound signals, and activity tracks;

[0071] Preprocess the multidimensional data according to the data type, including converting it into numerical form using target embedding methods and normalizing and completing numerical features;

[0072] Apply the animal growth status perception model to analyze the processed data and send the analysis results to the terminal;

[0073] Adopt power consumption adaptive algorithm to dynamically adjust ear tag working mode.

[0074] 3. Beneficial effects

[0075] 1. The present invention adopts a strong magnetic buckle and headband assembly design, combined with a limit mechanism and a safety structure, which not only ensures the stability of the ear tag during use, but also allows the ear tag assembly to be quickly disassembled by one-handed operation when needed. The disassembly process unlocks the buckle mechanism and releases the limit device using the built-in spring force, and then the male and female tag components can be easily rotated and removed, greatly improving the convenience of operation and the reuse rate of the equipment.

[0076] 2. In the prior art, most ear tags still need to punch holes in the ears of animals for installation, which will cause certain harm and stress reactions to the animals. The present invention adopts a head-mounted design, which does not require punching operations, avoids the pain and infection risks caused by punching, and greatly improves the comfort of the animals. The headband is made of soft and elastic medical-grade silicone. This material can fit the contour of the animal's head well and reduce the sense of oppression on the animal's head. At the same time, the place where the ear-hanging design contacts the animal's auricle also adopts a design that conforms to the physiological structure of the animal's ear, and adds soft cushioning material to further improve the wearing comfort.

[0077] 3. The present invention uses a Bluetooth module to record and transmit animal-related data, and cooperates with a swing-type power generation device to continuously power the system, ensuring the stable operation of the system. The Bluetooth module has a data redundancy storage function, which is designed to deal with possible physical damage or signal interference, thereby ensuring the continuity and integrity of the data.

[0078] 4. The present invention adopts knowledge distillation technology to effectively transfer the high-precision features of the server-side model to the lightweight local model, realizing efficient reasoning of edge devices. In addition, a power consumption adaptive algorithm is proposed, which comprehensively considers the device status, task priority and the importance of sensor data, and adjusts the computing resource allocation and sampling frequency in real time. Through adaptive optimization, the system can effectively extend the battery life of the equipment while ensuring the accuracy of data analysis, which is suitable for efficient monitoring needs in complex breeding environments.

[0079] 5. The present invention relies on deep learning algorithms, combined with multi-dimensional data collected by intelligent sensor arrays, including physiological parameters, behavioral data (activity, activity range, movement trajectory, feeding behavior, rest time, etc.) and environmental parameters, and performs comprehensive analysis through the optimized Transformer model to achieve intelligent monitoring of animal health status and behavioral patterns. At the same time, the Beidou Satellite Navigation System (BDS) is used to accurately locate animals and detect abnormal behaviors in real time. When an abnormal situation is found, the current area and location are displayed on the user terminal. The manager can quickly locate and take corresponding measures based on the terminal prompts, thereby significantly improving the timeliness and accuracy of breeding management. BRIEF DESCRIPTION OF THE DRAWINGS

[0080] Figure 1 This is a schematic diagram of the structure of a multifunctional electronic ear tag in an upright state according to an embodiment;

[0081] Figure 2 This is a schematic diagram of the structure of a multifunctional electronic ear tag in a flat state in one embodiment;

[0082] Figure 3 A diagram showing an application environment of a multifunctional electronic ear tag working method in one embodiment;

[0083] Figure 4 A flowchart of a multifunctional electronic ear tag working method in one embodiment;

[0084] Figure 5 A diagram of a multimodal data training network structure based on an optimized Transformer in one embodiment;

[0085] Figure 6 A schematic diagram of an algorithm for implementing edge computing on an ear tag device in one embodiment;

[0086] Figure 7 is a structural block diagram of a multifunctional electronic ear tag control device in one embodiment;

[0087] Figure 8 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment.

[0088] Detailed description of the drawings: 100, headband body; 200, ear hook; 300, connecting buckle; 400, power management module; 500, data monitoring module; 510, environmental monitoring unit; 5110, body temperature sensor; 5120, air quality sensor; 520, behavior monitoring unit; 5210, composite sensor; 5220, sound sensor; 530, positioning module. DETAILED DESCRIPTION

[0089] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0090] The multifunctional electronic ear tag provided in the embodiment of the present application refers to Figure 1-Figure 3 Specifically include:

[0091] A headband body 100 is provided, and an ear hook 200 is slidably connected to the headband body 100. The ear hook 200 is arranged in an arc shape to fit the surface of the animal's auricle. The ear hook 200 can adjust the distance between the headband and the ear hook 200 through a telescopic structure to ensure matching with animal heads of different sizes. A connecting buckle 300 is fixedly connected to the ear hook 200. The connecting buckle 300 includes a male mark and a female mark relatively arranged on the ear hook 200. The male mark is threadedly connected to the female mark. The male mark is a male mark made of soft magnetism. A magnet is embedded in the female mark. The male mark and the female mark are pressed against each other under the magnetic force of the magnet. The end face of the headband body 100 in contact with the animal's head is the inner wall. The inner wall of the headband body 100 is connected to a power management module 400 and a data monitoring module 500. The power management module 400 supplies power to the data monitoring module 500. The data monitoring module 500 is used to collect environmental data and behavioral data.

[0092] Among them, the power management module 400 includes a battery and a power generation device. The battery is arranged on the top of the inner wall of the headband body 100. The power generation device is arranged as a flexible piezoelectric power generation sheet. The flexible piezoelectric power generation sheet is electrically connected to the data monitoring module 500. The power management module 400 is used for battery charging management and discharge management and to provide a stable power supply for the data monitoring module 500.

[0093] For example, based on the head size and activity characteristics of the sheep, a silicone protective cover of appropriate size is selected and fixed to the front and back of the headband body 100 to provide reliable protection and limitation for the power management module 400 on the headband body 100 to prevent damage during the sheep's daily activities.

[0094] Furthermore, the headband body 100 is made of a soft and elastic medical-grade silicone material, which can better adapt to the curve of the sheep's head. For adult sheep, the circumference of the headband is generally 40-50 cm, which is more appropriate. The headband body 100 is slowly put on the sheep's head to ensure that it is centered and does not cause pressure or discomfort to the sheep's eyes, ears, etc. The memory alloy wire embedded inside the headband can quickly return to its original state when the sheep's head shakes slightly or is squeezed by external force, and maintain the fit of the headband. At the same time, the air holes opened on the headband can effectively dissipate the heat and moisture generated by the sheep's head to prevent stuffiness. Further, according to the actual size of the sheep's head, the tightness of the headband is adjusted by sliding the headband body 100 and the sliding buckle of the ear hook 200. It is generally appropriate to insert a finger at a tightness to ensure that the headband is always stable during the sheep's activities.

[0095] Specifically, a flexible piezoelectric generator is used in the power management module 400 to convert mechanical energy into electrical energy, providing stable power support for the electronic components in the ear tag and ensuring continuous operation of the device.

[0096] As a further supplement to the above, align the ear hook 200 with the sheep's auricle, and the soft cushion on the inside fits the contour of the sheep's ear, and the arc-shaped design fits tightly with the surface of the sheep's auricle. The male and female labels used in the ear hook 200 can use the strong magnetic force between the magnet and the soft magnetic material to quickly attract the male and female labels when approaching the sheep's ear to achieve preliminary positioning. Then, manually rotate the rotatable buckles on both sides of the ear hook 200 inward. The anti-slip texture of the buckle can effectively increase the friction, so that it gradually tightens during the rotation process and is firmly fixed on the sheep's auricle. The rounded edge on the inside of the buckle will not scratch the sheep's ear skin.

[0097] In the present embodiment, by means of the headband body 100 and the connecting buckle 300, there is no need to drill holes in the ears of the animal for installation, which reduces the pain and infection risk caused by drilling and greatly improves the comfort of the animal. The male label and the female label with magnets can be pressed against each other by the magnetic force, so that when the staff puts the ear tag on the animal, they can quickly press the male label and the female label against each other, pre-fix the ear hook 200 on the ear of the animal, and then thread the male label into the female label, thereby improving the deployment efficiency of the ear hook 200.

[0098] In one embodiment, referring to Figure 1-Figure 3The data monitoring module 500 includes: an environment monitoring unit 510, a behavior monitoring unit 520 and a positioning unit 530. The environment monitoring unit 510 includes a body temperature sensor 5110 and an air quality sensor 5120. The body temperature sensor 5110 is arranged at the outer top of the headband body 100. The body temperature sensor 5110 is arranged corresponding to the battery. In this embodiment, the number of air quality sensors 5120 is set to two, and the two air quality sensors 5120 are respectively arranged on the outer end surface of the headband body 100;

[0099] The behavior monitoring unit 520 includes an acceleration sensor, a gyroscope sensor and a sound sensor 5220. The sound sensor 5220 is arranged on the ear hook 200. The acceleration sensor and the gyroscope sensor are integrated into a composite structure sensor. The composite structure sensor is arranged on the inner wall of the headband body 100. The composite structure sensor is used to collect the movement data of the animal. The sound sensor 5220 is arranged on the ear hooks 200 at both ends. The eating information and emotional expression of the animal are detected through sound signals. The data collected by the composite sensor 5210 and the sound sensor 5220 are analyzed, and then the behavior pattern of the animal is judged; the positioning unit 530 includes a Beidou satellite navigation system positioning chip, which is arranged at the center of the headband body 100.

[0100] Exemplarily, the composite sensor 5210 module embedded in the headband body 100 and the ear hook 200 can comprehensively collect various data of the sheep. For example, during the sheep's feeding process, the sound sensor 5220 can capture the sound frequency and intensity of the sheep chewing grass, and combine the movement posture changes of the sheep's head and neck monitored by the acceleration sensor and the gyroscope sensor to judge the sheep's eating speed, food intake, and whether there is picky eating behavior; when the sheep is in a resting state, by monitoring the changes in muscle tension and heart rate, the sheep's sleep quality and body relaxation are analyzed. Furthermore, the body temperature sensor 5110 monitors the body temperature of the sheep in real time. If the body temperature is found to be out of the normal range, the data center will issue an early warning in time. The air quality sensor 5120 continuously detects the concentration of harmful gases such as ammonia and hydrogen sulfide and the oxygen content in the sheep house to provide data support for improving the breeding environment.

[0101] Furthermore, the multifunctional electronic ear tag uploads the collected data to the cloud data center in real time through the Bluetooth module using Bluetooth Low Energy (BLE) technology. The data redundancy storage function of low-power Bluetooth technology can ensure the integrity and continuity of the data when the signal is interfered by the metal structure or other equipment in the sheep house. Furthermore, the positioning chip based on the Beidou Satellite Navigation System (BDS) on the back of the headband body 100 can accurately track the position of the sheep in the farm. For example, when the sheep are accidentally lost or run out of the specified breeding area, the system can quickly locate and display its location information on the user terminal, making it convenient for the breeders to quickly find the sheep.

[0102] The multifunctional electronic ear tag working method provided in the embodiment of the present application can be applied to Figure 3 In the application environment shown. The terminal communicates with the cloud data center through the network, and the cloud data center is used to process the data collected by the terminal and send the analysis results back to the terminal after analysis. The terminal can be but not limited to various personal computers, laptops, smart phones, tablets, Internet of Things devices and portable wearable devices. The Internet of Things devices can be smart speakers, smart TVs, smart air conditioners, smart car-mounted devices, etc. Portable wearable devices can be smart watches, smart bracelets, head-mounted devices, etc. The cloud data center can be implemented with an independent cloud data center or a cloud data center cluster composed of multiple cloud data centers.

[0103] The terminal includes a data display unit, an interactive control unit and a remote alarm unit. The data display unit presents the livestock's identity information, behavior patterns, health status and environmental data in real time through a graphical interface; the interactive control unit supports users to remotely configure monitoring parameters, adjust monitoring ranges and manage alarm thresholds through terminal devices; when the remote alarm unit detects abnormal behavior or health warnings, it sends instant reminders to users via text messages, push notifications or sound alarms to ensure that users can quickly take response measures to improve management efficiency and livestock welfare.

[0104] In one embodiment, Figure 4-Figure 6 As shown, this method is applied to Figure 3 This is achieved through the interaction between the terminal and the cloud data center. The method includes the following steps:

[0105] Step 202, divide the preset breeding area and build a wireless network for data transmission and positioning.

[0106] Before using the smart electronic ear tags, the breeding areas are first divided, and a high-precision wireless network for data transmission and positioning is built by deploying fixed positioning nodes.

[0107] Exemplarily, the farm is first divided into several breeding areas, and electronic fences are used to divide and limit the areas in different breeding areas. Further, taking a medium-sized sheep farm as an example, the farm is divided into breeding areas with different functions such as fattening area, breeding area, ewe area, etc., and electronic fences are used to accurately divide and limit the areas at the boundaries of each area. In the fattening area, according to its area of ​​about 5,000 square meters and rectangular shape, 8 Bluetooth gateways are reasonably installed around and in the center. The breeding area is relatively small at 3,000 square meters, and 6 Bluetooth gateways are installed. The ewe area is also installed with an appropriate number of Bluetooth gateways based on its layout and area, so as to build a stable and comprehensive Bluetooth networking transmission network to ensure that the electronic ear tag data on the sheep can be smoothly transmitted to the data center.

[0108] Step 204: acquiring multi-dimensional data of animals in the breeding area based on a monitoring module.

[0109] Among them, multidimensional data at least includes body surface temperature, environmental parameters, movement status, sound signals, activity trajectory

[0110] Step 206: pre-process the multi-dimensional data according to the data type.

[0111] Among them, the preprocessing operation includes using the target embedding method to convert it into numerical form and normalize and complete the numerical features.

[0112] Step 208: Apply the animal growth status perception model to analyze the processed data and send the analysis results to the terminal.

[0113] Step 210: Adopt a power consumption adaptive algorithm to dynamically adjust the ear tag working mode.

[0114] In the working method of the above-mentioned multifunctional electronic ear tag, relying on deep learning algorithms, combined with multi-dimensional data collected by intelligent sensor arrays, including physiological parameters, behavioral data (activity, activity range, movement trajectory, feeding behavior, rest time, etc.) and environmental parameters, a comprehensive analysis is performed through the optimized Transformer model to achieve intelligent monitoring of animal health status and behavior patterns. At the same time, the Beidou Satellite Navigation System (BDS) is used to accurately locate animals and detect abnormal behaviors in real time. When an abnormal situation is found, the red laser indicator light will light up on the intelligent electronic ear tag of the abnormal animal, and its current area and location will be displayed on the user terminal. Managers can quickly locate and take corresponding measures based on terminal prompts, thereby significantly improving the timeliness and accuracy of breeding management.

[0115] In one embodiment, referring to Figure 4 and Figure 5 The process of preprocessing multidimensional data based on data type specifically includes:

[0116] For different sensor data, the target embedding method is used to convert the classification features into numerical form and calculate the conditional probability of each category corresponding to the target variable, which is expressed as:

[0117]

[0118] Among them, x i is a certain category feature value, y j is the target variable, II is the indicator function;

[0119] Use Z-score standardization to normalize numerical features and use autoregressive neural network modeling to complete missing data; when multidimensional data is missing, the missing values ​​are predicted based on the autoregressive neural network. The formula is as follows:

[0120] x t =f(x t-1 , x t-2 , …, x t-k ;θ);

[0121] Among them, x t is the complement value at time t, x t-k is the previous k time step data, and θ is the model parameter.

[0122] In one embodiment, the animal growth status perception model is applied, and the method specifically includes:

[0123] On the server side, based on the data type, the corresponding feature extraction and fusion Transformer network is built to perform feature extraction and fusion operations on the corresponding modality;

[0124] When in the feature fusion stage, dynamic multimodal attention weights are introduced, and the formula is as follows:

[0125]

[0126] Among them, q m is the query vector of modality m, k m is the key vector of mode m, cos is the cosine similarity function, α m is the dynamic multimodal attention weight;

[0127] The fused features are input into the cross-modal Transformer model with the MAWCL loss function added to obtain the analysis results and push the analyzed results to the user terminal.

[0128] Among them, the cross-modal Transformer with the MAWCL loss function added uses a multi-head self-attention mechanism to model the dependencies between modalities; the encoder part uses stacked Transformer encoder blocks, each of which includes a multi-head attention mechanism and a feed-forward neural network (FFN). The decoder generates a prediction sequence through self-attention and encoder-decoder attention mechanisms. The loss function formula is defined as follows:

[0129]

[0130] Where M is the total number of modes, α m is the dynamic weight of mode m, L task,m is the task loss of modality m, L inconsistency,m is the consistency loss between mode m and other modes, and λ is the weight coefficient of the inconsistency loss.

[0131] Specifically, in terms of model training, we collected multimodal sensor data of several sheep at different time periods in the past month, as well as corresponding behavior and health status records as training data. Part of the data is the sheep's eating, resting and activity data under different weather conditions, as well as the changes in sensor data before and after the onset of disease symptoms. Through continuous training and optimization of the Transformer network, it can accurately identify various behaviors and health conditions of sheep. During the training process, the dropout rate was set to 0.2. When the accuracy of the model on the validation set did not improve for 10 consecutive epochs, the early stopping method was used to stop the training, and finally a model with good performance was obtained on the server.

[0132] In one embodiment, referring to Figure 4 and Figure 6 Applying the animal growth status perception model, the method specifically includes:

[0133] The complex model trained on the server side is defined as a pre-trained animal growth status perception model A, the simplified model is defined as a local lightweight animal growth status perception model B, and the simplified model B is embedded into the multifunctional electronic ear tag;

[0134] The analysis results of the complex model on the server side are used to guide the learning of the simplified model in the form of knowledge distillation. The loss function formula is as follows:

[0135] L=(1-α)·L hard +α·T 2 ·KL(P A ||P B );

[0136] Among them, L hardis the standard cross entropy loss, KL represents KL divergence, T is the temperature coefficient, α is the weight parameter, and P A is the output distribution of model A, P B is the output distribution of model B.

[0137] Specifically, according to the characteristics of the sheep data and the analysis requirements, determine the appropriate range of values ​​for the weight parameter α. For example, compare the learning effect of the simplified model B under different α values ​​and the accuracy of the sheep growth status perception through multiple experiments, and select the optimal α interval. At the same time, set the temperature coefficient T based on the distribution characteristics of the data and the comprehensive consideration of the accuracy and real-time performance of the results in the actual application scenario to ensure that the knowledge distillation process can efficiently guide the local lightweight animal growth status perception model B to learn from the pre-trained animal growth status perception model A, so that the model B embedded in the multifunctional electronic ear tag can not only use limited resources (such as computing and storage resources) to achieve fast calculations, but also achieve a more ideal sheep growth status perception effect.

[0138] In this embodiment, a teacher-student network structure is used to reduce the computational complexity of the model. Through the knowledge distillation technology, the high-precision features of the teacher network are effectively transferred to the lightweight student network to achieve efficient reasoning of edge devices. For example, when monitoring the daily activities of sheep, it is necessary to focus on key indicators such as the activity duration, activity range, and activity frequency of the sheep. These indicators play a vital role in judging the health status and growth status of the sheep. The local lightweight animal growth status perception model B can continuously track data information related to these activities by virtue of its advantage of being embedded with a multifunctional electronic ear tag to facilitate real-time data collection.

[0139] In one embodiment, referring to Figure 4 , based on the power consumption adaptive algorithm to achieve low power consumption, improve the endurance and operation stability of the device, including:

[0140] The remaining power, current power consumption and sensor temperature of the ear tag are obtained and a state vector is formed. The state vector is input through the policy network in the reinforcement learning framework to dynamically generate a resource allocation strategy. The optimization objectives are as follows:

[0141]

[0142] Where π is the policy network, γ is the discount factor, and R(t) is the reward function based on power consumption and performance;

[0143]

[0144] Among them, Q(t) is the quality of the inference task, w1 is the weight factor for balancing power consumption P(t), w2 is the weight factor for balancing performance Q(t), and w3 is the weight factor for balancing temperature T(t).

[0145] Furthermore, a variable depth network (Dynamic Depth Network) is introduced to reduce the amount of inference calculation through a dynamic layer skipping mechanism. The formula is as follows:

[0146]

[0147] Among them, g i (x) is the output of the i-th layer, s i Score the computational importance of the layer, τ is an adjustable threshold;

[0148] Select an appropriate sampling frequency based on the ear tag status and the importance of the reasoning task. The formula is as follows:

[0149]

[0150] Among them, P sensor is the power consumption of sampling, P compute The power consumption for calculation;

[0151] When the ear tag status monitoring shows that the power consumption is within the preset threshold, the complexity of the local lightweight animal growth status perception model B is dynamically adjusted and low-frequency sampling is used to extend the ear tag working time.

[0152] For example, when the sheep is in a relatively quiet resting state, some network layers that are not important for real-time monitoring can be skipped and only the layers whose importance scores exceed a threshold are calculated.

[0153] In this embodiment, a power consumption adaptive algorithm is proposed, which comprehensively considers the device status, task priority and the importance of sensor data, and adjusts the computing resource allocation and sampling frequency in real time. Through adaptive optimization, the system can effectively extend the device life while ensuring the accuracy of data analysis, which is suitable for efficient monitoring needs in complex breeding environments.

[0154] Farmers use smart terminals such as dedicated farming management login user terminal modules to display data including ear tag number, breed, date of birth, behavioral patterns including feeding time, rest cycle, activity range, health status including body temperature, heart rate, whether there is a disease warning, and sheep house environmental data including air quality, temperature, humidity, etc. to present sheep information in real time. Furthermore, the interactive control unit supports farmers to remotely configure monitoring parameters, such as adjusting the upper and lower limits of the sheep's body temperature warning, modifying the sensitivity of behavior monitoring, etc.; it can also adjust the monitoring range, such as focusing on a batch or a certain area of ​​sheep; at the same time, it can manage the alarm threshold and set different abnormal alarm conditions according to actual farming experience and the growth stage of the sheep.

[0155] Specifically, when the remote alarm unit detects abnormal behavior or health warnings in sheep, such as a sheep's body temperature continues to rise, its heart rate is abnormal, or it has not eaten for a long time, the system will immediately send an instant reminder to the breeder via SMS push notification or sound alarm. The breeder can quickly locate the problem sheep based on the detailed information displayed on the user terminal and take appropriate measures, such as isolation observation, adjusting feed formula or contacting a veterinarian for diagnosis and treatment.

[0156] In this embodiment, relying on deep learning algorithms, combined with multi-dimensional data collected by intelligent sensor arrays, including physiological parameters (body temperature), behavioral data (activity, activity range, movement trajectory, feeding behavior, rest time, etc.) and environmental parameters, a comprehensive analysis is performed through the optimized Transformer model to achieve intelligent perception and monitoring of animal health status and behavior patterns. At the same time, the Beidou Satellite Navigation System (BDS) is used to accurately locate animals and detect abnormal behaviors in real time. When an abnormal situation is found, the current area and location are displayed on the user terminal. The manager can quickly locate and take corresponding measures based on the terminal prompts, thereby significantly improving the timeliness and accuracy of breeding management.

[0157] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0158] Based on the same inventive concept, the embodiment of the present application also provides a multifunctional electronic ear tag and a control device for realizing the multifunctional electronic ear tag and the working method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more multifunctional electronic ear tags and control device embodiments provided below can refer to the limitations of the multifunctional electronic ear tag and the working method above, and will not be repeated here.

[0159] In one embodiment, Figure 7 As shown, a multifunctional electronic ear tag control device is provided, including: a wireless network construction module, a multidimensional data acquisition module, a numerical preprocessing module, an animal growth status perception module and a power consumption adaptive adjustment module, wherein:

[0160] A wireless network construction module, used to divide the preset breeding area and build a wireless network for data transmission and positioning;

[0161] A multi-dimensional data acquisition module, used for acquiring multi-dimensional data of animals in the breeding area based on the monitoring module, wherein the multi-dimensional data at least includes body surface temperature, environmental parameters, movement status, sound signals, and activity tracks;

[0162] Numerical preprocessing module, used to preprocess multidimensional data according to data type, including using target embedding method to convert it into numerical form and normalize and complete numerical features;

[0163] An animal growth status perception module is used to analyze the processed data using an animal growth status perception model and send the analysis results to a terminal;

[0164] The power consumption adaptive adjustment module is used to dynamically adjust the ear tag working mode by using a power consumption adaptive algorithm.

[0165] In one embodiment, the numerical preprocessing module is further used to: for different sensor data, use the target embedding method to convert the classification features into numerical form, and calculate the conditional probability of each category corresponding to the target variable, which is expressed as: Among them, x i is a certain category feature value, y j is the target variable, II is the indicator function; Z-score standardization is used to normalize the numerical features and autoregressive neural network modeling is used to complete the missing data.

[0166] In one embodiment, the animal growth status perception module is also used to: build corresponding feature extraction and fusion Transformer networks based on data types on the server side, perform feature extraction of corresponding modalities and perform fusion operations; when in the feature fusion stage, introduce dynamic multimodal attention weights, the formula is as follows: Among them, q m is the query vector of modality m, k m is the key vector of mode m, cos is the cosine similarity function, α m is the dynamic multimodal attention weight; the fused features are input into the cross-modal Transformer model with the MAWCL loss function added to obtain the analysis results. The loss function formula is defined as follows: Where M is the total number of modes, α m is the dynamic weight of mode m, L task，m is the task loss of modality m, L inconsistency，mis the consistency loss between mode m and other modes, and λ is the weight coefficient of the inconsistency loss.

[0167] In one embodiment, the animal growth status perception module is also used to define the complex model trained on the server side as a pre-trained animal growth status perception model A, define the simplified model as a local lightweight animal growth status perception model B, and embed the simplified model B into the multifunctional electronic ear tag; guide the simplified model learning by the analysis results of the complex model on the server side in the form of knowledge distillation, and the loss function formula is as follows: L = (1-α)·L hard +α·T 2 ·KL(P A ||P B ), where L hard is the standard cross entropy loss, KL represents KL divergence, T is the temperature coefficient, α is the weight parameter, and P A is the output distribution of model A, P B is the output distribution of model B.

[0168] In one embodiment, the power consumption adaptive adjustment module is further used to: obtain the remaining power, current power consumption and sensor temperature of the monitoring device and form a state vector; the state vector is input through the policy network in the reinforcement learning framework to dynamically generate a resource allocation strategy, and the optimization objectives are as follows: Where π is the policy network, γ is the discount factor, and R(t) is the reward function based on power consumption and performance; Among them, Q(t) is the quality of the inference task, w1 is the weight factor for balancing power consumption P(t), w2 is the weight factor for balancing performance Q(t), and w3 is the weight factor for balancing temperature T(t). A variable depth network is introduced to reduce the amount of inference calculation through a dynamic layer skipping mechanism. The formula is as follows: Among them, g i (x) is the output of the i-th layer, s i Score the computational importance of the layer, τ is an adjustable threshold; select an appropriate sampling frequency based on the device state and the importance of the inference task, the formula is as follows: Among them, P sensor is the power consumption of sampling, P compute The power consumption is calculated; when the device status monitoring shows that the power consumption is within the preset threshold, the device uses a low-frequency sampling frequency to upload data in real time.

[0169] Each module in the above-mentioned multifunctional electronic ear tag and control device can be implemented in whole or in part by software, hardware or a combination thereof. Each of the above-mentioned modules can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute operations corresponding to each of the above modules.

[0170] In one embodiment, a computer device is provided. The computer device may be a cloud data center, and its internal structure diagram may be as follows: Figure 8 As shown. The computer device includes a processor, a memory and a network interface connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a multifunctional electronic ear tag and a working method are implemented.

[0171] In one embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as follows: Figure 8 As shown. The computer device includes a processor, a memory, a communication interface, a display screen and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, a mobile cellular network, NFC (near field communication) or other technologies. When the computer program is executed by the processor, a multifunctional electronic ear tag and a working method are implemented. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covered on the display screen, or a button, a trackball or a touchpad set on the computer device housing, or an external keyboard, touchpad or mouse, etc.

[0172] Those skilled in the art will understand that Figure 8 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0173] In one embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above-mentioned method embodiments when executing the computer program.

[0174] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0175] In one embodiment, a computer program product is provided, including a computer program, which implements the steps in the above method embodiments when executed by a processor.

[0176] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0177] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited to this.

[0178] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0179] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A multifunctional electronic ear tag, characterized in that: include: A headband body (100), wherein an ear hook (200) is provided on the headband body (100), wherein the ear hook (200) is arranged in an arc shape, wherein a connecting buckle (300) is provided on the ear hook (200), wherein the connecting buckle (300) comprises a male mark and a female mark, wherein the male mark is threadedly connected to the female mark, wherein the male mark is a male mark made of soft magnetic material, wherein a magnet is embedded in the female mark, wherein the male mark and the female mark are pressed against each other under the magnetic force of the magnet, wherein a power management module (400) and a data monitoring module (500) are provided on the headband body (100), wherein the power management module (400) supplies power to the data monitoring module (500), and wherein the data monitoring module (500) is used to collect environmental data and behavioral data.

2. The method according to claim 1, characterized in that The data monitoring module (500) comprises: an environment monitoring unit (510), a behavior monitoring unit (520) and a positioning unit (530); the environment monitoring unit (510) comprises a body temperature sensor (5110) and an air quality sensor (5120); the body temperature sensor (5110) is arranged on the inner side of the headband body (100); and the air quality sensor (5120) is arranged on both sides of the headband body (100); The behavior monitoring unit (520) comprises an acceleration sensor, a gyroscope sensor and a sound sensor (5220), the sound sensor (5220) is arranged on the ear hook (200), and the acceleration sensor and the gyroscope sensor are arranged on the headband body (100); The positioning unit (530) comprises a Beidou satellite navigation system positioning chip and is arranged at the center of the headband body (100).

3. A multifunctional electronic ear tag working method, applied to any of the multifunctional electronic ear tags as claimed in claims 1-2, characterized in that: The method comprises: Divide the preset breeding areas and build a wireless network for data transmission and positioning; Acquire multi-dimensional data of animals in the breeding area based on the monitoring module, wherein the multi-dimensional data includes at least body surface temperature, environmental parameters, movement status, sound signals, and activity tracks; Preprocess the multidimensional data according to the data type, including converting it into numerical form using target embedding methods and normalizing and completing numerical features; Apply the animal growth status perception model to analyze the processed data and send the analysis results to the terminal; Adopt power consumption adaptive algorithm to dynamically adjust ear tag working mode.

4. The method according to claim 3, characterized in that The process of preprocessing multidimensional data based on data types includes: For different sensor data, the target embedding method is used to convert the classification features into numerical form and calculate the conditional probability of each category corresponding to the target variable, which is expressed as: Among them, x i is a certain category feature value, y j is the target variable, II is the indicator function; Z-score standardization is used to normalize numerical features and autoregressive neural network modeling is used to complete missing data.

5. The method according to claim 3, characterized in that: The application of the animal growth status perception model comprises: On the server side, based on the data type, the corresponding feature extraction and fusion Transformer network is built to perform feature extraction and fusion operations on the corresponding modality; When in the feature fusion stage, dynamic multimodal attention weights are introduced, and the formula is as follows: Among them, q m is the query vector of modality m, k m is the key vector of mode m, cos is the cosine similarity function, α m is the dynamic multimodal attention weight; The fused features are input into the cross-modal Transformer model with the MAWCL loss function added to obtain the analysis results. The loss function formula is defined as follows: Where M is the total number of modes, α m is the dynamic weight of mode m, L task,m is the task loss of modality m, L inconsistency,m is the consistency loss between mode m and other modes, and λ is the weight coefficient of the inconsistency loss.

6. The method according to claim 3, characterized in that The application of the animal growth status perception model comprises: The complex model trained on the server side is defined as a pre-trained animal growth status perception model A, the simplified model is defined as a local lightweight animal growth status perception model B, and the simplified model B is embedded into the multifunctional electronic ear tag; The analysis results of the complex model on the server side are used to guide the simplified model learning in the form of knowledge distillation. The loss function formula is as follows: L=(1-α)·L hard +α·T 2 ·KL(P A ||P B ); Among them, L hard is the standard cross entropy loss, KL represents KL divergence, T is the temperature coefficient, α is the weight parameter, and P A is the output distribution of model A, P B is the output distribution of model B.

7. The method according to claim 3, characterized in that The power consumption adaptive algorithm also includes: Obtain the remaining power, current power consumption and sensor temperature of the ear tag and form a state vector; The state vector is input through the policy network in the reinforcement learning framework to dynamically generate a resource allocation strategy with the following optimization objectives: Where π is the policy network, γ is the discount factor, and R(t) is the reward function based on power consumption and performance; Where Q(t) is the quality of the inference task, w1 is the weight factor for balancing power consumption P(t), w2 is the weight factor for balancing performance Q(t), and w3 is the weight factor for balancing temperature T(t); A variable depth network is introduced to reduce the amount of inference calculation through a dynamic skip layer mechanism. The formula is as follows: Among them, g i (x) is the output of the i-th layer, s i Score the computational importance of the layer, τ is an adjustable threshold; Select an appropriate sampling frequency based on the ear tag status and the importance of the reasoning task. The formula is as follows: Among them, P sensor is the power consumption of sampling, P compute The power consumption for calculation; When the ear tag status monitoring shows that the power consumption is within the preset threshold, the complexity of the local lightweight animal growth status perception model B is dynamically adjusted and low-frequency sampling is used to extend the ear tag working time.

8. A multifunctional electronic ear tag control device, characterized in that: The device comprises: A wireless network construction module, used to divide the preset breeding area and build a wireless network for data transmission and positioning; A multi-dimensional data acquisition module, used for acquiring multi-dimensional data of animals in the breeding area based on the monitoring module, wherein the multi-dimensional data at least includes body surface temperature, environmental parameters, movement status, sound signals, and activity tracks; Numerical preprocessing module, used to preprocess multidimensional data according to data type, including using target embedding method to convert it into numerical form and normalize and complete numerical features; An animal growth status perception module is used to analyze the processed data using an animal growth status perception model and send the analysis results to a terminal; The power consumption adaptive adjustment module is used to dynamically adjust the ear tag working mode by using a power consumption adaptive algorithm.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 3 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 3 to 7 are implemented.

Images