Method for measuring body temperature of livestock in original environment

Abstract

The invention discloses a method for measuring the body temperature of livestock in an original environment, and the method comprises the steps: selecting a thermosensitive device, and constructing a core model; and dynamically calibrating and measuring the body temperature of the livestock in real time. The measurement precision is greatly improved, the adaptability to the original environment is enhanced, and automation and economy are optimized.

Description

Technical Field

[0001] This invention relates to the field of Internet of Things (IoT) temperature measurement technology, and in particular to a method for measuring the body temperature of livestock in a primitive environment. Background Technology

[0002] Current implementation solutions: The most common approach for IoT sensors used in livestock body temperature monitoring is "laboratory standard calibration + fixed parameter application." For example, ear-tag temperature sensors often follow the ISO 17025 laboratory calibration standard. In a standard environment of 25℃ and 50% humidity, thermistors (such as ordinary NTC thermistors) are statically calibrated to establish a fixed correspondence between "temperature T" and "ADC voltage value" (i.e., ADC = F(T)). The collected ADC values ​​are then converted into body temperature data and uploaded to the platform via communication modules such as LoRaWAN and NB-IoT. Mainstream products on the market, such as a certain brand of cattle ear tags, use this solution. Their hardware design focuses on "meeting the accuracy requirements of standard environments" and does not specifically adapt to livestock farming scenarios.

[0003] Problems or shortcomings of existing technology: This solution has significant limitations in scenario adaptation and cannot meet the actual needs of the original breeding environment. On the one hand, dynamic changes in the breeding environment will seriously interfere with the measurement accuracy: the temperature and humidity fluctuation range in the breeding shed is large (-10℃~40℃, humidity 30%-90%), and dust and feces are easily attached to the sensor surface, resulting in changes in the heat exchange efficiency between the thermal device and the air; at the same time, individual differences in livestock (such as the difference in hair thickness between calves and adult cattle can be 2-3mm, and the amount of sebum secreted by different individuals) will change the heat conduction efficiency between the sensor and the livestock's body surface. Even for the same cow, after activity (such as rubbing against the wall causing sensor displacement), the measurement deviation will increase. On the other hand, the fixed calibration logic of existing solutions cannot be dynamically adjusted: in practical applications, the body temperature measurement error of this type of sensor often reaches ±0.5℃ or more, which is far higher than the ±0.2℃ accuracy requirement required for veterinary diagnosis; for example, when the ambient temperature rises suddenly from 10℃ to 30℃, the measurement error of a certain brand of ear tag will increase from ±0.3℃ to ±0.6℃, making it impossible to accurately determine whether cattle are in a fever state above 39.5℃, which is prone to missed or misjudged disease warnings. Summary of the Invention

[0004] In view of the above problems, the present invention is proposed to provide a method for measuring the body temperature of livestock in a primitive environment that overcomes or at least partially solves the above problems.

[0005] According to one aspect of the present invention, a method for measuring the body temperature of livestock in a primitive environment is provided, the body temperature measurement method comprising: Select a thermistor and build a core model; Dynamic calibration and real-time livestock body temperature measurement.

[0006] Optionally, the selection of the thermistor specifically includes: Thermistor selection and structural design: NTC thermistors are selected as the core sensing element. The substrate is made of 99.9% high-purity copper, with a 0.5-1μm thick gold layer on the surface and a 0.05mm thick PTFE anti-fouling coating on the outside. The device is designed as an ultra-thin patch to ensure a close fit to the skin inside the animal's ear.

[0007] Optionally, the NTC thermistor is model MF52-10K, with an accuracy of ±1% and a temperature response time ≤100ms.

[0008] Optionally, the ultra-thin patch type is specifically a structure with a thickness of 0.15mm and a diameter of 5mm.

[0009] Optionally, the construction of the core model specifically includes: core function Calibration: Under multiple environmental gradients in a simulated aquaculture scenario in the laboratory, 100 thermistor devices were subjected to cyclic testing: Under each environmental gradient, the device was attached to a standard constant temperature bath, and the temperature was increased / decreased in 0.1℃ increments. The ADC voltage value of the device was collected, and more than 10,000 sets of "temperature change ∆T - ADC change ∆ADC" data were obtained. The inverse Gaussian function model is obtained by fitting the data using the least squares method. ; The optimal parameters are a=500, b=0, c=0.3, d=100. The unit is ℃. It is a 16-bit decimal number.

[0010] Optionally, the multiple environmental gradients specifically include: temperature -10℃ to 40℃, with each 5℃ gradient; humidity 30% to 90%, with each 10% gradient.

[0011] Optionally, the standard constant temperature bath has an accuracy of ±0.01℃.

[0012] Optionally, the dynamic calibration and real-time livestock body temperature measurement specifically include: Initial dual calibration: After the sensor is fitted to the cattle, 72 hours of initial data acquisition is initiated. Individual identification: The mode of ADC values ​​over 72 hours is calculated using the Python scipy.stats.mode function and denoted as ADC_individual; Group calibration: Synchronously collect ADC data of more than 30 healthy cattle in the same group and calculate the ADC mode of the cattle group ADC_group; if the deviation between ADC_individual and ADC_group is ≤±5, then ADC_individual is confirmed as the ADC value corresponding to the normal body temperature of the cattle, and the normal body temperature T_normal is deduced in reverse. Routine Real-Time Measurement: During routine monitoring, the sensor collects the real-time ADC value (ADC_real) every minute to calculate body temperature. Calculate the change: ; Temperature conversion: via Inverse function calculation Thus, real-time body temperature can be obtained. ; Data Upload: T_real and ADC_real are uploaded to the aquaculture platform simultaneously. If T_real exceeds the range of 38.5-39.5℃, a primary warning is triggered. Anomaly Detection and Secondary Calibration: If the deviation between ADC_individual and ADC_group during initial calibration is > ±10, or if T_real exceeds the normal range three times consecutively during routine measurements, an anomaly warning will be triggered. The cattle marked as suspected of having a fever will have their normal body temperature assessment suspended. Initiate a second data collection within 72 hours. If the subsequent data still deviates from the cattle population value, coordinate with the farm's veterinarian to conduct a manual rectal temperature re-examination. If a retest confirms fever (T≥39.5℃), an abnormal body temperature record will be created in the platform to track changes in body temperature after treatment in real time; if a retest shows no abnormalities, ADC_individual will be recalculated and the normal body temperature baseline will be updated.

[0013] This invention provides a method for measuring livestock body temperature in a primitive environment. The method includes: selecting a thermistor and constructing a core model; dynamic calibration and real-time livestock body temperature measurement. The method significantly improves measurement accuracy, enhances adaptability to primitive environments, and optimizes automation and economy.

[0014] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, and in order to make the above and other objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0015] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a flowchart of a method for measuring livestock body temperature in a primitive environment, provided as an embodiment of the present invention. Detailed Implementation

[0017] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0018] The terms "comprising" and "having," and any variations thereof, in the specification, embodiments, claims, and drawings of this invention are intended to cover non-exclusive inclusion, such as including a series of steps or units.

[0019] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments.

[0020] Example 1 like Figure 1 As shown, a method for measuring livestock body temperature in a primitive environment includes: selecting a thermistor and constructing a core model; dynamic calibration and real-time livestock body temperature measurement.

[0021] This invention takes "dynamically adapting to the original environment and eliminating interference from multiple factors" as its core idea, and achieves accurate measurement through "hardware optimization + algorithm iteration". The specific implementation steps are as follows: Phase 1: Selection of Thermistor Devices and Construction of Core Model Thermistor selection and structural design: An NTC thermistor (model MF52-10K, accuracy ±1%, temperature response time ≤100ms) is selected as the core sensing element. The substrate is made of 99.9% high-purity copper (to ensure heat conduction efficiency), and the surface is plated with a 0.5-1μm thick gold layer (to reduce oxidation risk and further improve heat conduction). The exterior is coated with a 0.05mm thick PTFE anti-fouling coating (to prevent dust and feces from adhering). The overall device is designed as an ultra-thin patch type (thickness 0.15mm, diameter 5mm) to ensure a tight fit with the skin inside the animal's ear and reduce heat loss caused by air gaps.

[0022] core function Calibration: Under simulated aquaculture conditions in the laboratory, 100 thermistors were subjected to cyclic testing under multiple environmental gradients (temperature -10℃ to 40℃, in 5℃ increments; humidity 30% to 90%, in 10% increments). At each environmental gradient, the devices were placed in a standard thermostatic bath (accuracy ±0.01℃), and the temperature was increased / decreased in 0.1℃ increments. Simultaneously, the ADC voltage values ​​of the devices were acquired (using a 16-bit ADS1115 ADC module, sampling frequency 1Hz), resulting in over 10,000 sets of temperature change data. Change "Data; by fitting the data using the least squares method, an inverse Gaussian function model is obtained." (Based on testing, the optimal parameters are a=500, b=0, c=0.3, d=100.) The unit is ℃. (The value is a 16-bit decimal number). This model can accurately describe the correspondence between "small temperature changes" and "ADC value changes" under different environments, which is different from existing fixed models.

[0023] Phase Two: Implementation of Dynamic Calibration and Real-Time Measurement Algorithms Initial dual calibration (individual + group): After the sensors are attached to the cattle, 72 hours of initial data acquisition is initiated (once every 5 minutes, for a total of 216 data points): Individual calibration: The mode of ADC values ​​over 72 hours is calculated using the Python scipy.stats.mode function and denoted as ADC_individual (initially determined as the ADC value corresponding to the normal body temperature of the cow). Group calibration: Simultaneously collect ADC data from more than 30 healthy cattle in the same herd, and calculate the herd ADC mode ADC_group; if the deviation between ADC_individual and ADC_group is ≤±5 (corresponding to a temperature deviation ≤±0.1℃), then ADC_individual is confirmed as the ADC value corresponding to the normal body temperature of that cattle, and the normal body temperature T_normal is deduced (through...). Calculate using the inverse function, or refer to the standard normal body temperature of cattle of the same breed (38.5-39.5℃).

[0024] Routine Real-Time Measurement: During routine monitoring, the sensor collects the real-time ADC value ADC_real every minute, and calculates body temperature according to the following logic: Calculate the change: ; Temperature conversion: via The inverse function (pre-stored in the sensor's local chip) is calculated. Thus, real-time body temperature can be obtained. ; Data Upload: T_real and ADC_real are uploaded to the aquaculture platform simultaneously. If T_real exceeds the range of 38.5-39.5℃, a primary warning is triggered.

[0025] Anomaly Verification and Secondary Calibration: If the deviation between ADC_individual and ADC_group during initial calibration is > ±10 (corresponding to a temperature deviation > ±0.2℃), or if T_real exceeds the normal range three consecutive times during routine measurements, an anomaly warning will be triggered. The cattle were marked as suspected cases of fever, and the determination of their normal body temperature was suspended. Initiate a second data collection within 72 hours. If the subsequent data still deviates from the cattle population value, coordinate with the farm's veterinarian to conduct a manual rectal temperature re-examination (gold standard). If a retest confirms fever (T≥39.5℃), an "abnormal body temperature record" will be created in the platform to track changes in body temperature after treatment in real time; if a retest shows no abnormalities, ADC_individual will be recalculated and the normal body temperature benchmark will be updated.

[0026] Example 2 Taking a health monitoring project for a dairy farm with 10,000 head of cattle as an example: Step 1: Hardware Preparation and Preprocessing Thermistor fabrication: Gold-plated copper substrate (99.9% pure copper, 0.1mm thick) is processed according to design parameters, NTC thermistors (MF52-10K) are attached, a 0.5μm thick gold layer is electroplated, a PTFE anti-fouling coating (0.05mm) is sprayed, and after curing, it is cut into 5mm diameter patches. Circuit integration: The thermistor, INA226 voltage amplifier chip, ADS1115ADC module, and LoRaWAN communication module (communication distance 1.5km) are integrated into a food-grade PP ear tag shell with a diameter of 25mm and a thickness of 8mm. The waterproof rating is IP67, and it has a built-in 1000mAh lithium battery (6 months of battery life). Function pre-storage: storing the inverse Gaussian function calibrated in the laboratory. And the inverse function is written to the local MCU chip of the sensor (model STM32L431).

[0027] Step 2: On-site deployment and initial calibration Equipment installation: Select 300 adult dairy cows (covering different growth stages: 100 calves, 150 dairy cows, and 50 dry dairy cows), and attach an ear tag sensor to the inside of each cow's left ear to ensure that the device is in close contact with the skin; install a LoRaWAN gateway (coverage radius 1.5km) in the center of the farm and connect it to the farm's Alibaba Cloud IoT platform; Data collection: The platform starts collecting data for the first 72 hours, recording ADC data every 5 minutes, and also recording the temperature and humidity inside the breeding shed (once per hour). Dual-calibration implementation: For each dairy cow, calculate the 72-hour ADC mode value (e.g., calf A's ADC_individual=2320, dairy cow B's ADC_individual=2350); calculate the ADC group value of 300 dairy cows ADC_group=2345; if calf A's ADC_individual deviates from ADC_group by -25 (corresponding to a temperature deviation of -0.5℃), it is marked as a suspected abnormality, and the remaining dairy cows with deviations ≤±5 are confirmed to have normal ADC values ​​(e.g., dairy cow B is confirmed to have ADC_individual=2350, and the normal body temperature T_normal=39.0℃).

[0028] Step 3: Daily Monitoring and Anomaly Handling Real-time measurement: In daily monitoring, the sensor collects ADC_real data once every minute and calculates... The inverse function is used to obtain T_real and then uploaded to the platform; for example, the milk-producing cow B. ,calculate (Within the normal range); Abnormal handling: After a second data collection at 72 hours, calf A's ADC_individual value was still 2320 (deviation from the average value of -25), triggering a red alert on the platform and pushing the information to the veterinary terminal; the veterinarian manually re-examined calf A's rectal temperature and found it to be 40.3℃ (confirming fever), established a treatment record on the platform, and tracked the temperature changes daily until T_real dropped below 39.0℃, at which point its normal ADC value was recalibrated; Results verification: During a 30-day monitoring period, 50 dairy cows were randomly selected for artificial rectal temperature comparison. The deviation between the measured values ​​and the rectal temperature values ​​of this method was ≤ ±0.2℃, with an accuracy rate of 98.5%. The sensor did not malfunction due to dust adhesion, the maintenance cycle was extended to 3 months, and the labor cost was reduced by 220,000 yuan / year compared with the existing technology.

[0029] Beneficial effects: Measurement accuracy has been greatly improved: through " The "dynamic model" eliminates interference from environmental temperature and humidity, and the "individual + group dual calibration" adapts to individual differences in livestock. In practical applications, the body temperature measurement error can be controlled within ±0.2℃, meeting the accuracy requirements of veterinary diagnosis (e.g., the standard for judging fever in cattle is T≥39.5℃, and accurate data can avoid missed / false diagnoses). Compared with the existing technology's error of ±0.5℃, this solution improves the accuracy of disease early warning by more than 16.5% (a large-scale farm with 10,000 head of livestock can reduce 10-15 false alarms per month, reducing the cost of manual re-examination by veterinarians).

[0030] Enhanced adaptability to the original environment: On the hardware level, the PTFE anti-fouling coating extends the sensor maintenance cycle from 1 month in the existing technology to 3 months, reducing the cost of manual cleaning in farms; on the algorithm level, it does not rely on laboratory calibration and can work stably in complex environments of -10℃ to 40℃ and humidity of 30% to 90%, adapting to farming scenarios in cold winters in northern regions and humid summers in southern regions, without the need to build additional constant temperature protection facilities.

[0031] Automation and economic optimization: It enables 72-hour automatic calibration, real-time monitoring, and anomaly warning without the need for manual calibration. Compared with the existing technology that requires monthly manual sampling and calibration (5 minutes per cow), a farm with 10,000 head of cattle can save about 200,000 to 250,000 yuan in labor costs per year. At the same time, the hardware design can be adapted to existing ear tag sensor molds, and the algorithm can be integrated into existing livestock IoT platforms. Enterprises do not need to reinvest in production lines or platform development, and upgrade costs are reduced by more than 40%.

[0032] The above specific embodiments further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for measuring livestock body temperature in a primitive environment, characterized in that, The body temperature measurement method includes: Select a thermistor and build a core model; Dynamic calibration and real-time livestock body temperature measurement.

2. The method for measuring livestock body temperature in a primitive environment according to claim 1, characterized in that, The selection of the thermal device specifically includes: Thermistor selection and structural design: NTC thermistors are selected as the core sensing element. The substrate is made of 99.9% high-purity copper, with a 0.5-1μm thick gold layer on the surface and a 0.05mm thick PTFE anti-fouling coating on the outside. The device is designed as an ultra-thin patch to ensure a close fit to the skin inside the animal's ear.

3. The method for measuring livestock body temperature in a primitive environment according to claim 1, characterized in that, The NTC thermistor is model MF52-10K, with an accuracy of ±1% and a temperature response time of ≤100ms.

4. The method for measuring livestock body temperature in a primitive environment according to claim 1, characterized in that, The ultra-thin patch type specifically has a thickness of 0.15mm and a diameter of 5mm.

5. The method for measuring livestock body temperature in a primitive environment according to claim 1, characterized in that, The construction of the core model specifically includes: core function ； Calibration: Under multiple environmental gradients in a simulated aquaculture scenario in the laboratory, 100 thermistor devices were subjected to cyclic testing: Under each environmental gradient, the device was attached to a standard constant temperature bath, and the temperature was increased / decreased in 0.1℃ increments. The ADC voltage values ​​of the device were collected, and more than 10,000 sets of "temperature change ∆T - ADC change ∆ADC" data were obtained. The inverse Gaussian function model is obtained by fitting the data using the least squares method. ； The optimal parameters are a=500, b=0, c=0.3, d=100. The unit is ℃. It is a 16-bit decimal number.

6. The method for measuring livestock body temperature in a primitive environment according to claim 5, characterized in that, The multiple environmental gradients specifically include: temperature -10℃ to 40℃, with each 5℃ gradient; humidity 30% to 90%, with each 10% gradient.

7. The method for measuring livestock body temperature in a primitive environment according to claim 5, characterized in that, The standard constant temperature bath has an accuracy of ±0.01℃.

8. The method for measuring livestock body temperature in a primitive environment according to claim 1, characterized in that, The dynamic calibration and real-time livestock body temperature measurement specifically include: Initial dual calibration: After the sensor is fitted to the cattle, 72 hours of initial data acquisition is initiated. Individual identification: The mode of ADC values ​​over 72 hours is calculated using the Python scipy.stats.mode function and denoted as ADC_individual; Group calibration: Synchronously collect ADC data of more than 30 healthy cattle in the same group and calculate the ADC mode of the cattle group ADC_group; if the deviation between ADC_individual and ADC_group is ≤±5, then ADC_individual is confirmed as the ADC value corresponding to the normal body temperature of the cattle, and the normal body temperature T_normal is deduced in reverse. Routine Real-Time Measurement: During routine monitoring, the sensor collects the real-time ADC value (ADC_real) every minute to calculate body temperature. Calculate the change: ； Temperature conversion: via Inverse function calculation Thus, real-time body temperature can be obtained. ; Data upload: and Simultaneously uploaded to the aquaculture platform, if Temperatures exceeding the 38.5-39.5℃ range trigger a basic warning. Anomaly Detection and Secondary Calibration: If the deviation between ADC_individual and ADC_group during initial calibration is > ±10, or if T_real exceeds the normal range three times consecutively during routine measurements, an anomaly warning will be triggered. The cattle marked as suspected of having a fever will have their normal body temperature assessment suspended. Initiate a second data collection within 72 hours. If the subsequent data still deviates from the cattle population value, coordinate with the farm's veterinarian to conduct a manual rectal temperature re-examination. If a retest confirms fever (T≥39.5℃), an abnormal body temperature record will be created in the platform to track changes in body temperature after treatment in real time; if a retest shows no abnormalities, ADC_individual will be recalculated and the normal body temperature baseline will be updated.

Images