Using radio frequency identification and UAV technology to realize the application of animal husbandry positioning system
chrispeng [ad_1]
introduction
Driven by factors such as sustained population growth, accelerated urbanization and increased income, the consumption of aquaculture products in China has grown rapidly, and China’s aquaculture industry has been in a period of rapid development for a relatively long period of time. The rapid and healthy development of the aquaculture industry is conducive to promoting the adjustment and optimization of the internal structure of agriculture, expanding new channels for agricultural economic growth, increasing farmers’ employment opportunities and increasing farmers’ income, and gradually forming a virtuous circle between agriculture and industry. Animal husbandry occupies a large proportion in the breeding industry. The animal husbandry industry is mainly divided into captive and free-range livestock. The free-range livestock has good meat quality and can meet people’s requirements for meat taste and nutrition. At the same time, the price of free-range livestock is high. Further increase the income of farmers.
However, due to the vast rural areas of China, especially the topography and landforms in the south, free-range farmers in mountainous and hilly areas will face the following problems:
1) It is difficult to locate livestock. In areas with large spatial areas and intricate terrain, it is difficult to obtain the specific locations of livestock.
2) Livestock is easy to lose. It is easy to be stolen or lost by yourself.
3) It is difficult to count the number of livestock. It is difficult for breeders to count the specific number of livestock.Aiming at the pain points of free-range livestock, this article focuses on the automatic cruise of drones.
Path planning algorithm, data transmission protocol, data preprocessing algorithm, design of animal husbandry positioning system based on RFID and UAV, easy to use, good applicability, can reduce the difficulty of free-range management and reduce the loss of farmers.
1 System overall scheme
1.1 System scheme design
The animal husbandry positioning system based on RFID and UAV is divided into three functional modules: data collection, data analysis and processing, and data display. The system combination uses active RFID equipment, unmanned aerial vehicles and mobile Internet technology to solve the problem of strong battery life, small size and low cost in the process of stocking livestock positioning.
The basic principle of the system is to use RFID to transmit radio frequency signals within a range of more than one hundred meters. UAVs with RFID readers and mobile phones will automatically cruise within the stocking range. Whenever RFID signals are scanned, The RFID data and signal strength are transmitted to a smart phone via Bluetooth. The phone obtains the geographic location at this time, and transmits the data to the system server through the mobile data network after data preprocessing. The server side processes and analyzes the collected location and RFID data. The mobile client App can view the latest location data of livestock, which is convenient for farmers to manage livestock.
Figure 1 Livestock positioning system
1.2 Comparison with GPS positioning solution
Breeders must meet the requirements of convenient use and low cost when positioning the stocked livestock. At the same time, livestock positioning does not require high position accuracy. Binding each livestock to a GPS device to locate it is not suitable for large-scale breeding. First, the GPS positioning device needs to obtain latitude and longitude data and send the data to the server, which consumes a lot of power. Generally, GPS devices send location information several times a day, and the battery can only last about one week. Farmers have to change batteries frequently, which is very troublesome to use. Secondly, GPS is large in size, poor in waterproof performance, and expensive, which is not conducive to large-scale use in livestock positioning. Third, because GPS actively sends location information, the situation of someone stealing livestock cannot be reported to the police in time, and the practicability is not strong.
The active RFID battery used in this system has a long battery life, up to two to three years, is small in size, and has very good waterproof performance. At the same time, the price cost is low. Calculated on the scale of 500 livestock, the GPS solution requires a total of 50,000 yuan (100 yuan for each GPS), and this system only requires 20,000 yuan (30 yuan for each RFID, 5,000 yuan for drones and card readers), and the cost is reduced by 60%. And this system has the following applications in practice:
1) Install the RFID reader and mobile phone in the livestock shed, and then count and count the livestock returned to the livestock shed.
2) RFID card readers and mobile phones are placed on the side of the main highway, once someone steals livestock, they can realize the alarm function when passing the highway.
3) The user carries the supporting RFID card reader and mobile phone, and can check the nearby RFID card through the mobile phone App, and can also find the nearby livestock.
Compared with the solution in which each livestock is bound to a GPS for positioning, this solution is more convenient to use, more practical, lower in price, and can be used in practical applications.
2 Data acquisition module design
2.1 Hardware design
1) RFID electronic tags. The system adopts 433 MHz active RFID electronic tag, which has long reading and writing distance, low power consumption and strong anti-interference ability.
2) RFID reader. The system uses 433 MHz RFID supporting omnidirectional card reader with bluetooth module, which has the advantage of wide coverage angle, and it is convenient to transmit the card reading data to other devices such as mobile phones.
3) Mobile phone equipment. The system uses a general Android smart phone to realize the data transmission function. The Bluetooth of the mobile phone reads the data of the RFID card reader; the computing power of the mobile phone realizes the preliminary processing of the original data, and then transmits it to the server via GPRS/3G/4G, reducing the computing burden of the server.
4) UAV equipment. Based on the literature, the system uses Ardupilot’s small UAV with automatic cruise function. The Ardupilot UAV flight control system supports the setting of the fixed flight trajectory mode and the fixed flight altitude flight mode of the UAV. Due to actual usage scenarios, most of the livestock farms are uneven mountainous areas. In order to keep the UAV at a fixed height relative to the ground in the mountainous area, the system modified the UAV’s ground detection equipment and replaced the air pressure sensor with a laser ranging sensor.
2.2 Flight trajectory planning
The flight trajectory planning of UAVs is generally divided into online planning and offline planning. Reference documents, in view of the characteristics of the application scenarios of the system and the advantages of simpler and higher feasibility of offline path planning compared to offline path planning, the system adopts the offline path planning algorithm of sub-area division and spiral contraction coverage.
The steps of the spiral scanning method for sub-area segmentation are:
1) When the outer contour is more complicated, the entire area can be divided into several sub-areas. The division of sub-regions adopts the double-line scanning method, as shown in Figure 2. Use two horizontal and vertical straight lines, the horizontal straight line from top to bottom, and the vertical straight line from left to right. The two straight lines will intersect or tangent at the edge of the contour, which can be divided into several independent sub-areas.
Figure 2 Two-line scanning method to divide sub-areas
2) The sub-area adopts “spiral contraction” for full coverage. Compared with the “back-and-forth” coverage method, the former leaves a smaller uncovered area than the latter, and the former is relatively fixed at the end point, generally located near the center of gravity of the area, as shown in Figure 3.
Figure 3 Spiral covering and reciprocating covering
2.3 Data preprocessing algorithm
The drone automatically cruises in the breeding area, and the RFID reader scans the RFID tags in the area at a certain time interval T (default 1 second). At the same time, multiple RFID tags and multiple signal strengths may be scanned to form an RFID Label collection. The RFID tag collection and the latitude and longitude data obtained by the current mobile phone through GPS positioning form the original data, the format is shown in Table 1.
Due to the short collection time period, the amount of original data is generally very large. Sending directly to the server is not conducive to data analysis. Using smart phones to preprocess this raw data can greatly reduce the amount of data sent and the difficulty of data processing and analysis on the server side. As shown in Table 2, where t1 and t2 have duplicate RFIDs, they can be combined.
There are many RFID card devices that can be scanned at the same time, and the collected RFID card data needs to be deduplicated and merged. The RFID raw data scanned by the reader is a data sequence composed of metadata at the time point, namely Data={time, longitude, latitude, [{id1，rssi1｝， {id2，rssi2｝……]}, the original data is the data sequence data1, data2, data3…. The interval between reading data from the RFID reader is short (1 second), and the amount of data collected per second is relatively large. You can extend the collection time interval, increasing the 1-second time interval to 5 seconds. The method to increase the time period is:
1) Take the first second data in a 5-second time period as the initial fusion data.
2) Add the remaining data per second to the initial data. The merging method is: traverse each RFID data in the new data, if the RFID does not exist in the fusion data, then add the RFID data to the RFID list. If the RFID data already exists in the RFID list, and the current location of the metadata with the largest rssi value is set as the location of the fusion data and the rssi value.
The algorithm pseudo code is:
After data preprocessing, the data sent from the mobile phone to the server is shown in Table 3.
3 Data processing module
3.1 Data conversion
Convert metadata divided by time into data divided by RFID. The metadata format Data1= {t1, long1, lati1, {{rfid1, rssi1}, {rfid2, rssi2} collected by the data collection module. }}. Since each item of data is divided by time as the main key, it is not convenient for later data analysis and processing, and the data needs to be converted into a division with RFID number as the main key. The conversion method is to traverse each rssi and latitude and longitude data. The converted format is Data= {rfid1, {rssi1, {long1, lati1}, {rssi2, {long2, lati2},… …}}}.
3.2 Correct geographic location based on signal strength
The RFID reader mounted on the drone scans within a range of more than 100 meters
RFID card, scanning range R, the relationship between the projection distance L of the scanning distance on the ground and the flying height H is:
(1)
Assuming that the flying height of the drone is 30 meters, the reading range of the card reader is
150 meters, the calculated coverage height on the ground is 146 meters.
Due to the long reading distance of the RFID card reader, the positioning accuracy is not high. If the method of shortening the reading distance of the RFID card reader is adopted to improve the accuracy, the scanning path of the drone in the same area is required to be denser. This system uses the signal strength (RS-SI) method to correct the target position range to improve the position accuracy.
According to the formula of RSSI and distance:
(2)
but
(3)
Among them: n represents the signal propagation constant, d represents the distance from the transmitter; A represents the received signal strength at a distance of 1 m. The RSSI value will decrease as the distance increases according to formula (2). That is, the higher the RSSI value, the more accurate the collected position data. Therefore, the following corrections are made to the geographical location:
1) Select the latitude and longitude with the largest RSSI value among all the latitude and longitude data corresponding to each RFID as the collected latitude and longitude.
2) According to formula (3), the RSSI value is used to calculate the offset distance between the real longitude and latitude and the collected longitude and latitude.
The livestock location is the latitude and longitude plus the offset distance. That is, the livestock is located in the range of the circle with the latitude and longitude position as the center and the offset distance as the radius.
4 Data display module
The data display module includes mobile App client and server systems. In addition to the livestock geographic location data provided by the system’s data processing module, the App client also uses Google’s offline digital map. As shown in Figure 4, the user can view the geographic location of the stocking livestock on the map. The use of geographic location data and other information can effectively improve the daily management efficiency of livestock farmers.
The functions provided by the App client include:
1) The latest position display of each livestock.
The basic function provided by the App is to request the livestock location data stored on the server.
2) Livestock location navigation.
The App cooperates with a Bluetooth card reader to accurately locate livestock. Farmers carry mobile phone apps and RFID readers. The RFID readers receive nearby RFID tag signals and display them on the mobile phone APP via Bluetooth, thereby prompting the user of the distance of the RFID, and the user can easily find the target location according to this prompt.
Figure 4 Livestock management APP client
3) Data analysis and query services based on location data.
Provides binding and unbinding between each livestock and RFID device, and can add basic data such as the type, gender, and birthday of the corresponding livestock.
5 System experiment
5.1 Experimental method
This system was used to carry out positioning experiments on 50 sheep in a large-scale farm in Feixi County, Hefei City, Anhui Province. It is assumed that the geographic location measured using GPS is the actual location. Use multiple drone cruise measurements to compare the measured position with the position measured by the GPS device, and calculate the error values respectively, and finally get the average error.
5.2 Experimental procedure
1) Tie an RFID card reader and an Android phone to the drone, and set the phone to connect to the RFID card reader via Bluetooth.
2) Tie each RFID device to the neck of the selected sheep.
3) Tie the GPS device to the necks of the sheep with RFID devices at the same time;
Put a few sheep with RFID back into the flock, turn on the drone to perform a cruise on a predetermined track, the card reader on the drone scans to the RFID device, and transmits it to the mobile phone on the drone via Bluetooth. The mobile phone is moved via 4G Network transmission to the background server.
4) The UAV cruises several times to obtain the measured position value of the current flock to be measured.
5) Change the flock measurement equipment to another batch of flocks, and repeat the drone cruise measurement.
5.3 Experimental results and analysis
The following are the measured and actual values of a UAV flight. The data all use east longitude and north latitude.
After passing the 5 theory test, the measurement error obtained by the method in this paper is 8.24 M. This error has practical value for the positioning of livestock such as cattle and sheep.
The experimental results show that the breeding position data collected by the animal husbandry positioning system based on RFID and UAV designed in this paper is within the theoretical position estimation range, and the positioning has high reliability and accuracy.
6 Conclusion
In response to the demand for large-scale stocking of livestock, this paper designs and implements a livestock positioning system based on RFID and UAV. The focus is on the full coverage path planning of mobile robots and the geographic location correction based on RSSI. The experimental results show that the positioning system can solve the problems of difficult to count and easy to lose in the case of livestock breeding by farmers at a lower cost price, and it is easy to use.
[ad_2]
