A sow health monitoring system

Abstract

The present invention provides a sow estrus and health monitoring system including an automatic feeding station, an RFID tag, an intelligent leg ring, a LoRa gateway, a cloud computing platform, and an information terminal. The automatic feeding station and the intelligent leg ring collect sow activity data, upload it to a behavior record relay, then upload it to the LoRa gateway, and finally transmit it to the cloud computing platform. The database of the cloud computing platform records data and manages user permissions. The information terminal can obtain data from the cloud computing platform and can modify the data in the database or the behavior record relay according to user permissions. This monitoring method can provide better feeding conditions for each sow, reduce the number of non-productive days of the sow, improve the production efficiency of sow breeding, reduce the experience accumulation requirements of staff, and save human resource costs.

Description

Technical Field

[0001] The present invention relates to a monitoring system, in particular to a system for monitoring the health of sows. Background Art

[0002] In pork production, sows are responsible for breeding piglets, so their reproductive and health status are of great importance to managers. Accurately and timely determining sow estrus and health status has always been a major concern in the pig farming industry.

[0003] Accurately and promptly determining estrus during a sow's reproductive cycle and ensuring timely mating are key to improving conception rates. Currently, the following methods are commonly used to monitor estrus in sows: 1. Comprehensive observation: This involves observing behavioral changes such as restlessness, loss of appetite, and active or receptive mounting of other sows. Furthermore, changes such as swelling and moisture in the vulva and mucus discharge indicate possible estrus. 2. Boar estrus testing: This involves contact between a boar and a sow, using pheromones to stimulate the sow's estrus. During this testing, the sow will demonstrate a willingness to approach the boar, exhibit static reflexes, and accept mounting. 3. Ultrasound examination: This method uses ultrasound imaging to observe the development of the ovaries and follicles. This includes the size, shape, and texture of the ovaries, the location of the follicles, their size and elasticity, the thickness of the follicle wall, whether the follicles are ruptured, and the presence of a corpus luteum. In practice, a combination of two or three of the above methods is typically used to determine a sow's estrus status. This not only requires significant manpower and resources, but also requires a high level of experience and judgment from the staff. Given that breeding and mating personnel require practical experience, training costs are a significant expense for the pig farming industry. Large-scale pig farms are typically located in areas with very low population densities, where the demand for relevant talent is significantly out of sync with the actual supply in the talent market. Consequently, human resource issues persist in the pig farming industry.

[0004] In modern large-scale, batch-based pig farming, various intelligent technologies are beginning to be applied. For example, automated feeding stations employ RFID tags, typically attached to the pigs' ears, to identify the pigs. With the development of the Internet of Things, automated feeding stations often also include RFID tag readers to automatically record the pigs' identities during feeding. To monitor the pigs' movements, smart leg collars are installed on their bodies. While these technologies are used in pig farming, the implementation of any one technology alone does little to contribute to the development of a scientific farming environment. Summary of the Invention

[0005] This invention provides a sow estrus and health status monitoring system. It effectively monitors the sow's physical condition and determines whether it is in estrus. This system promotes the sow's healthy growth and improves conception rates, facilitating the expansion of production by breeders. Another objective of this invention is to reduce the human resource costs of breeders, enabling breeders to become qualified with minimal training.

[0006] The object of the present invention is achieved like this:

[0007] A sow estrus and health monitoring system includes an automatic feeding station, an RFID tag for pig identification, and an intelligent leg ring for collecting and receiving sow activity data. The automatic feeding station provides sow feeding and weight data, and the sow activity data collected by the intelligent leg ring is regularly uploaded to a behavior recording relay matched with the automatic feeding station. The behavior recording relay then uploads the data to a LoRa (Long Range) gateway.

[0008] The data uploaded to the LoRa gateway by the behavior record relay is transmitted to the cloud computing platform via the Internet; the cloud computing platform's database records the sow's feeding and activity data and manages user permissions.

[0009] The information terminal can push or actively pull data stored in the cloud computing platform database through the cloud computing platform, and can change the data in the database or behavior record relay according to user permissions.

[0010] The RFID tags used for pig identification are installed on the pig's ears. The smart leg rings that collect and receive sow activity data are installed on the pig's legs. The sow's feeding data provided by the automatic feeding station includes at least the sow's food intake at each time.

[0011] The behavior record relay uses an intelligent AP (Acess Point) that provides multi-interface access such as cellular network, Bluetooth, WiFi4 and WiFi5 as a behavior record relay. Whether it is a fat AP or a thin AP, any one using an ARM chipset or other chipset can be loaded into an intelligent control system, such as OpenWrt and other intelligent control systems with a Linux kernel. By editing the application that receives data through the interface that matches the automatic feeding station, the data is collected and then transmitted to the LoRa gateway. Through intelligent routing, corresponding behavior record relays can be configured according to different automatic feeding stations, thereby improving the compatibility of the system of the present invention, achieving wide support for automatic feeding stations on the market, and being upgradeable. Applications and interfaces can be changed as the automatic feeding stations are upgraded or changed, and can also be matched with existing equipment, eliminating the cost of purchasing new equipment.

[0012] The smart leg ring includes a sealed box, and a battery, a microprocessor, a data wire and a three-axis accelerometer arranged in the sealed box; the three-axis accelerometer includes a three-axis acceleration sensor, a three-axis gyroscope sensor and a three-axis magnetometer sensor, which are used to record sow activity data.

[0013] In order to improve the efficiency of large-scale farming and energy saving, the activity data collected by the smart leg ring is sent to the LoRa gateway once every hour using an interval sending method.

[0014] The behavior recording relay at the automatic feeding station communicates with the LoRa gateway via cellular networks or wireless LAN. The behavior recording relay uploads data to the LoRa gateway at a fixed rate, ranging from 0.25 to 1 time per hour. The LoRa gateway transmits activity data collected by the sow's smart leg ring via the cellular network module to a server for subsequent processing and analysis.

[0015] User rights are divided into administrator rights and general user rights. When a user with administrator rights logs in to an information terminal, he or she has the right to create and edit data through the cloud computing platform and save it to the database, or access the behavior record relay of the automatic feeding station through a point-to-point connection, change the collected data, and then upload it to the cloud computing platform database through the behavior record relay.

[0016] The cloud computing platform has a data synchronization function, which keeps the system time synchronized with the behavior record relay.

[0017] The cloud computing platform analyzes the collected sow activity data using a semi-supervised fuzzy clustering algorithm. Based on the sample characteristics and learning data, the platform determines stable cluster centers. These centers are then compared with test data to generate the final clustering results. The system categorizes sow behavior into five states: standing, lying, walking, jogging, and trotting. The system combines daily feed intake data (i.e., feed intake amount, frequency, and duration of feeding) with body weight data to create a characteristic data model for each sow. Each parameter is weighted in the logical analysis. The cloud computing platform matches the daily sow characteristic data with the previously collected sow characteristic data model and compares the magnitude of change. If the magnitude of change for all parameters is within the normal range, the data is marked and recorded. If the magnitude of change for a key parameter exceeds this range, the system uses the nine parameters combined with their weighting factors to calculate the probability of estrus or illness, making a logical judgment and sending the final result to the information terminal.

[0018] The present invention establishes a data model for each sow by monitoring and analyzing behavioral and feeding data in real time. By comparing data, it promptly identifies abnormal changes in a sow's behavior or feeding habits, thereby capturing estrus or disease characteristics. This allows breeders to gain a more objective and timely understanding of the physical condition of their sows, accurately assessing their reproductive and health status. This allows for better feeding conditions for each sow, reduces the number of non-productive days, and improves sow production efficiency. Through centralized data management, breeders or studs only need to identify data anomalies to determine the sow's health based on the data, eliminating the need for accumulated experience. Once a breeder discovers an anomaly, the data can be reported, and appropriate action can be taken based on the data. This significantly reduces the experience required and the human resources required for training. Furthermore, due to the different permissions, breeder training is only assigned general user permissions, preventing data anomalies caused by misoperation. Once a breeder or stud has gained experience, administrator permissions are assigned, and manual verification ensures that the data collected by the system accurately and completely reflects the sow's physiological and health status. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 Schematic diagram of the system topology according to an embodiment of the present invention;

[0020] Figure 2 Schematic diagram of the data upload process according to an embodiment of the present invention;

[0021] Figure 3 This is a schematic diagram of the information modification process according to an embodiment of the present invention;

[0022] Figure 4 This is a schematic diagram of the pig house layout according to an embodiment of the present invention;

[0023] Figure 5 This is a schematic diagram of the installation of the pig ear RFID tag and the smart leg ring according to an embodiment of the present invention. DETAILED DESCRIPTION

[0024] The specific embodiments of the present invention are described below.

[0025] Embodiment: A sow estrus and health monitoring system includes an automatic feeding station 1, a pig ear RFID tag 5 for pig identity identification, and an intelligent leg ring 4 set on the pig leg for collecting and receiving sow activity data. It is characterized in that the sow feeding data and weight data provided by the automatic feeding station 1 and the sow activity data collected and received by the intelligent leg ring 4 are regularly uploaded to the LoRa gateway connected to the behavior record relay and cloud computing platform of the automatic feeding station; the data uploaded to the LoRa gateway by the behavior record relay is transmitted to the cloud computing platform via the Internet; the database of the cloud computing platform records and stores the sow's feeding and activity data for analysis and judgment of the sow's physiological and health status; the information terminal can push or actively pull data stored in the cloud computing platform database through the cloud computing platform, and the data in the database or behavior record relay can be changed according to user permissions. In this example, the automatic feeding station 1 can use a common automatic feeding device on the market. After obtaining its data output interface, it connects to the behavior recording relay 2. This device includes an RFID reader module, an automatic feeding device, a trough weighing device, a sow weight weighing device, a pig body spraying device, and the behavior recording relay 2. The RFID reader module receives information from the pig ear RFID tag 5, and the trough weighing device records the sow's feed intake at each time. The behavior recording relay 2 receives the data reading module of the smart leg ring 4, which receives sow activity data and regularly uploads all collected data to the LoRa gateway 3.

[0026] Preferably, a smart phone is used as the information terminal, and less preferably, a personal computer is used as the information terminal. However, any device that can provide web access can be used as the information terminal.

[0027] The behavior record relay 2 uses an intelligent access point (AP) that provides access to multiple interfaces, including cellular networks, Bluetooth, Wi-Fi 4, and Wi-Fi 5. Whether it's a fat AP or a thin AP, any ARM chipset or other chipset can be integrated into an intelligent control system, such as an OpenWrt or other intelligent control system with a Linux kernel. Preferably, a cellular network module is configured to enable internet access. By editing the data access interface that matches the automatic feeding station 1, data is collected and then transmitted to the LoRa gateway 3. Through intelligent routing, the corresponding behavior record relay 2 can be configured according to different automatic feeding stations on the market, improving the compatibility of the system of the present invention.

[0028] The LoRa gateway 3 connects the behavior record relay 2 to the cloud computing platform through a cellular network and / or a wireless local area network, and transmits the data uploaded by the behavior record relay 2 to the LoRa gateway 3 to the cloud computing platform through the Internet. With the popularization of IPv6, when forming the system described in the present invention, preferably, all devices have IPv6 support enabled. Therefore, even if devices such as Bluetooth interfere with the 2.4Ghz wireless network, the cellular network can be used to transmit data to the LoRa gateway 3 through the Internet. Although the behavior record relay can also send information directly to the cloud computing platform when accessed through the Internet, in order to ensure the uniformity of the data and reduce the computing burden of the cloud computing platform, the data is uniformly uploaded through the LoRa gateway 3. Less preferably, in the absence of IPv6 access conditions, the cloud computing platform provides communication between the behavior record relay 2 and the LoRa gateway 3 through a reverse proxy. Preferably again, the behavior record relay 2 transmits the collected data from both the Internet and the local network to the LoRa gateway through the cellular network and the wireless LAN at the same time, and attaches an upload flow ID. The upload flow ID attached to the behavior record relay 2 is used to determine the ownership of the data, thereby avoiding the LoRa gateway from repeatedly recording data and uploading it to the cloud computing platform.

[0029] The cloud computing platform is used to record the sows' feeding, weight and activity data, and store them in the cloud computing platform's database for analysis and to determine the sows' reproductive and health status.

[0030] The information terminal can push or actively pull data stored in the cloud computing platform database through the cloud computing platform, and change the data in the database or behavior record relay 2 according to user permissions.

[0031] When a sow enters the automatic feeding station to eat, an RFID reader reads the sow's information and feeds her small amounts of food multiple times based on her age and gestational status. During the feeding process, the sow is weighed. When the sow finishes eating and leaves the feeding area (i.e., when the RFID tag is out of range of the RFID reader), the remaining food in the trough is weighed. After each feeding session, data on the sow's feeding time, feed intake, and weight is collected. This data is then uploaded to the Behavior Recording Relay 2.

[0032] User rights are divided into administrator rights and general user rights. When a user with administrator rights logs in to an information terminal, he or she has the right to create and edit data through the cloud computing platform and save it to the database, or access the behavior record relay 2 through a point-to-point connection, change the collected data, and then upload it to the cloud computing platform database through the behavior record relay 2.

[0033] The cloud computing platform possesses data synchronization capabilities, maintaining synchronization with the system time of Behavior Record Relay 2. By comparing the time points of data generation stored on the cloud computing platform and in Behavior Record Relay 2, the most recently generated time point is used as the final stored data. Preferably, because the computing power of the cloud computing platform far exceeds that of Behavior Record Relay 2, data is recorded by Behavior Record Relay 2, or data is modified by an information terminal through Behavior Record Relay 2 before being uploaded to the cloud computing platform. The cloud computing platform uses the most recent time point for the same data (i.e., transaction ID) as the basis for data generation and stores the data. Data received after collection, due to differences in collection time, is merged along the timeline.

[0034] The frequency of uploading data from the behavior record relay 2 to the LoRa gateway 3 is fixed, ranging from 0.25 times / hour to 1 time / hour.

[0035] The smart leg ring 4 includes a sealed box, and a battery, a microprocessor, a data wire and a three-axis accelerometer arranged in the sealed box; the three-axis accelerometer includes a three-axis acceleration sensor, a three-axis gyroscope sensor and a three-axis magnetometer sensor, which are used to record sow activity data.

[0036] In order to improve the efficiency of large-scale farming and energy saving, the activity data collected by the smart leg ring is sent to the LoRa gateway once every hour using an interval sending method.

[0037] Using a triaxial accelerometer installed on the sow, we can measure the sow's acceleration in the x, y, and z directions every second. Using the system's motion signature algorithm, we can categorize the sow's movements over a one-hour period into five dimensions: standing, lying, walking, jogging, and running. This classification clearly shows the time spent in each of the five postures or movement patterns within that hour, as well as the sow's movement state within a specific time period. The automatic sow feeding station 1 can capture sow feeding information over a two-hour period, including the number of times the sow eats, the duration of the meal, the amount of food consumed, and the weight.

[0038] After the administrator confirms that each sow can access automatic feeding station 1 and completes all sows' basic information, click Start Test, and the system officially begins collecting and building a characteristic data model for each sow. The system then matches a corresponding data model based on the sow's breed and weight. During the test, the system compares the data to determine the sow's reproductive and health status.

[0039] After acquiring 24 hours of data for each sow at 00:00 on the first day, the system first compares this data with the "Healthy Empty Stomach" characteristic data model for that breed's weight (data for less than 21 days). If the change in each parameter is within 20%, the data for that day is marked as "Healthy Empty Stomach" and added to the "Healthy Empty Stomach" characteristic data model. If any parameter (such as standing time, lying time, and total movement time) exceeds 20% of the Healthy Empty Stomach characteristic data model, the data is compared with the "Healthy Estrus," "Sick Empty Stomach," and "Healthy Pre-Estrous" characteristic data models. If the average degree of match for each parameter exceeds 70%, the data is marked as the corresponding characteristic data and added to that characteristic data model. If the match is less than 70%, the probability of estrus, disease, or pre-estrus is calculated based on the weight of each parameter, and a logical judgment is finally made. When the probability value of the estrus state exceeds 70%, the data of that day is marked as "healthy estrus" state; when the probability value of the disease state exceeds 70%, the data of that day is marked as "disease-free" state; when the probability value of the preestrus state exceeds 70%, the data of that day is marked as "healthy preestrus" state;

[0040] The system performs a weighted calculation of the following four states based on nine parameters in the database to determine the probability of the sow being in a certain state. The table below shows the main characteristic changes of each parameter under the four states.

[0041]

[0042] As observation data accumulates, the system's data feature models for "healthy empty-belly," "healthy estrus," and "disease empty-belly" will continue to expand. Daily data will then be compared with these expanded feature models before logical judgments are made.

[0043] If a sow's daily lying time on a given day exceeds 20% of the time in the "Healthy Empty Stomach" feature database (normally, a healthy sow's daily lying time ranges from 10-14 hours, 12 ± 2 hours. If diseased, this time can increase significantly, for example, to 16 hours, an increase of 4 hours compared to the normal 12 hours, meaning the increase is 4 / 16, exceeding 20%), the system compares all parameters for that day with all parameters in the "Disease Empty Stomach" feature database. If the match exceeds 70%, the system directly identifies the sow as "Disease Empty Stomach." If the match does not exceed 70%, the system calculates the probability of a disease state based on the weights of each parameter, and finally performs a logical judgment. If the probability calculation exceeds 70%, the sow is identified as "Disease Empty Stomach." The system classifies the data for that day as "Disease Empty Stomach" and sends a notification to the automatic feeding station 1 and the information terminal app. Automatic feeding station 1 sprays a paint mark on the sow's back when it returns to feed. Upon receiving this information, the farm manager verifies the sow's condition on-site the next day and administers treatment. If all indicators of the sow return to the normal range after treatment, the sow's data is re-entered into the database as "healthy empty."

[0044] If a sow's standing time in a single day exceeds 20% of the standing time in the "Healthy Empty Boobs" signature database, the system compares all parameters for that day with all parameters in the "Healthy Estrus" signature database. If the match exceeds 70%, the system determines it is in "Healthy Estrus." If the match does not exceed 70%, the probability of a disease state is calculated based on the weights of the various parameters, and a logical judgment is made. If the probability exceeds 70%, the sow is considered in "Healthy Estrus." The system classifies the data for that day as "Healthy Estrus" and sends a notification to the automatic feeding station 1 and the information terminal. The automatic feeding station 1 sprays a different paint mark on the sow's back when the sow returns to feed. Upon receiving this notification, the farm manager verifies the sow's estrus status on-site the next day and performs insemination and breeding. After breeding, the sow's breeding time can be entered directly on the automatic feeding station 1. Alternatively, the sow can be modified on a mobile phone and updated via Bluetooth connection to the system.

[0045] The data from the day of sow mating is marked as "Healthy Pregnancy Day 0." Subsequent data is first compared with the "Healthy Pregnancy Day X" feature database. If the match is greater than 70%, the system determines the sow to be in a "Healthy Pregnancy" state. If the match is less than 70%, the data is compared with the "Healthy Empty Stomach" feature database. If the match is greater than 70%, the system determines the sow to be in a "Healthy Empty Stomach" state. If the match is less than 70%, the sow continues to be classified as "Healthy Pregnancy Day X." If the sow returns to estrus around day 21, the previous estrus determination process is repeated, and the management staff will enter the new mating time after mating. The data from the day of mating is marked as "Healthy Pregnancy Day 0," and the previous process for determining "Pregnancy" or "Empty Stomach" is repeated.

[0046] If a sow's feed intake decreases by more than 20% during pregnancy compared to the "Healthy Pregnancy" signature database, the system compares all parameters for that day with all parameters in the "Disease Pregnancy" signature database. If the match exceeds 70%, the system determines it is in a "Disease Pregnancy" state. If the match does not exceed 70%, the probability of the disease state is calculated based on the weights of the various parameters, and a logical judgment is made. If the probability exceeds 70%, the sow is classified as in a "Disease Pregnancy" state. The system classifies the data for that day as "Disease Pregnancy" and sends a notification to Automatic Feeding Station 1 and the information terminal. Automatic Feeding Station 1 sprays a paint mark on the sow's back when the sow returns to feed. Upon receiving the notification, the farm manager verifies the sow's disease status on-site the next day and administers treatment. If, after treatment, all indicators of the sow return to the normal range for the sow's signature model, the sow's data for that day is re-entered into the database as "Healthy Empty."

[0047] When a sow reaches 110 days of "healthy pregnancy X days," the system sends a notification to the automatic feeding station 1 and the information terminal, prompting the farm manager to transfer the sow to the farrowing ward for farrowing. After farrowing and nursing, the sow returns to the breeding and gestation ward. After staff complete the relevant information, the process continues.

[0048] Through this system, pig farm managers can not only easily check the pregnancy and empty status of sows to better manage reproduction, but also timely understand the disease status of sows and treat them to make them recover as soon as possible.

[0049] The system described in the present invention can be deployed in large-scale breeding places in batches, and can realize equipment installation and network debugging in a pig farm with a basic sow group of 600. For example, 6 automatic feeding stations 1 can be installed in a pig house, and each feeding station can feed 30 to 40 sows. The size of a sow group is about 70. A group of sows can be statically grouped, and 2 automatic feeding stations 1 and the corresponding behavior recording relays 2 can meet the feeding requirements of a group of sows, and then the LoRa gateway will uniformly manage and upload the data in the behavior recording relays 2. Therefore, according to the actual situation of the pig house, the corresponding number of automatic feeding stations 1 and behavior recording relays 2 are configured, and then uniformly managed by the LoRa gateway, so that the system has the characteristics of flexible layout and meets the needs of large-scale breeding.

Claims

1. A sow health monitoring system, including a smart leg ring, pig ear RFID tags, an automatic feeding station, a LoRa gateway, a cloud computing platform, and an information terminal, characterized by: The sow feeding and weight data collected by the automatic feeding station and the sow activity data collected by the smart leg ring are regularly uploaded to the behavior record relay that matches the automatic feeding station, and then uploaded to the LoRa gateway by the behavior record relay; Sow activity data includes standing, lying, walking, jogging, and running. Sow feeding information includes feeding frequency, feeding time, feed intake, and weight. The LoRa gateway connects the behavior record relay to the cloud computing platform via the cellular network and / or wireless LAN. The data uploaded by the behavior record relay to the LoRa gateway is transmitted to the cloud computing platform via the Internet. The cloud computing platform's database records the sow's feeding and activity data and manages user permissions. The smart leg ring is equipped with a sealed box, which includes a battery, a microprocessor, a data cable, and a three-axis accelerometer. The three-axis accelerometer includes a three-axis acceleration sensor, a three-axis gyroscope sensor, and a three-axis magnetometer sensor to record sow activity data. The information terminal pushes or actively pulls data stored in the cloud computing platform database through the cloud computing platform, and modifies the data in the database or behavior record relay according to user permissions; After the manager confirms that each sow can enter the automatic feeding station to eat and completes the basic information of all sows, click to start the test, and the system officially begins to collect and establish a characteristic data model for each sow. The characteristic data is the sow's activity data and the sow's feeding information. At this time, the system matches a corresponding data model according to the sow's breed and weight. During the test, the data is compared to determine the sow's reproductive and health status. The health status analyzed by the sow activity data is "healthy empty", "healthy estrus", "sick empty", "healthy preestrus", "healthy pregnancy", and "sick pregnancy". After collecting the sow's data within 24 hours every day, it is compared with the "healthy empty" model corresponding to the breed and weight. When the change range of each parameter is within 20%, the data is marked as "healthy empty stomach" and added to the "healthy empty stomach" feature data model; When a parameter exceeds 20% of the characteristic data of the healthy empty-breasted model, the data of that day will be further compared with the characteristic data models of "healthy estrus", "sick empty-breasted", and "healthy pre-estrus". When the average matching degree of each parameter exceeds 70%, the data of that day will be directly marked as the corresponding characteristic data and included in the characteristic data model; When the average matching degree of each parameter is less than 70%, the probability of estrus, disease and pre-estrus is calculated according to the weight of each parameter, and finally a logical judgment is made. The system performs a weighted calculation of the four states of pre-estrus, estrus, disease and pregnancy based on the data of standing time, lying time, slow walking time, jogging time, sprinting time, as well as feeding frequency, feeding time, feed intake and weight in the database to obtain the probability value of the sow being in a certain state. When the probability value of estrus exceeds 70%, the data of that day is marked as "healthy estrus"; When the probability value of the disease state exceeds 70%, the data of that day is marked as "disease-free" state; When the probability value of the proestrus state exceeds 70%, the data for that day is marked as "healthy proestrus" state.

2. The sow health monitoring system according to claim 1, characterized in that: User rights are divided into administrator rights and general user rights. When users with administrator rights log in to the information terminal, they have the right to create and edit data through the cloud computing platform and save it to the database; Or access the behavior record relay through a point-to-point connection, modify the collected data, and then upload it to the cloud computing platform database through the behavior record relay; General user permissions can only query data on the cloud computing platform or behavior record relay.

3. The sow health monitoring system according to claim 1, characterized in that: The LoRa gateway includes a cellular network module and a wireless module. The behavior record relay communicates with the LoRa gateway through the cellular network and / or wireless LAN, and then transmits the data to the cloud computing platform through the LoRa gateway.

4. The sow health monitoring system according to claim 3, characterized in that: When the pig ear RFID tag exceeds the reading distance, the automatic feeding station starts recording data and transmits it to the behavior recording relay. The behavior recording relay temporarily stores the data sent by the automatic feeding station.

5. The sow health monitoring system according to claim 4, characterized in that: The frequency of uploading behavior record relay data to the LoRa gateway is fixed, ranging from 0.25 times / hour to 1 time / hour.

6. The sow health monitoring system according to claim 5, characterized in that: The behavior record relay has a data synchronization function and keeps synchronized with the system time of the cloud computing platform.

Images