Research And Implement Of Sows’ Behavior Monitoring System Based On Wireless Sensor Network

The pig industry in China has a history of thousands of years,making China a big producer and consumer of pork.Pork is one of the most popular foods and has long been playing a significant role in animal husbandry and agriculture in China.With recent development in large-scale and intensive breeding industry,Automation technology is becoming more and more widely used in aquaculture.Determination of parturition time is a vital link in the production process of pig breeding.Traditional breeding methods like continuous observation and parturition time judgment depend mainly on the keeper and are subjective.Such methodologies have been found to be inaccurate and result in increased piglet mortality rate.Since the behavior is the basis for judging the delivery time of sows,it is important to accurately and timely monitor the prenatal behavior of sows to predict the delivery time and reduce the mortality rate.Wireless sensor network has the advantages of small size,low cost,high reliability and so on.It has been widely used in the field of animal husbandry.This study introduces a novel method design of a sow prenatal behavior monitoring system based on the wireless sensor network to provide data support model for parturition time detection.The main contents of the research include:(1)The design of hardware and software based on STM32 and CC2430 as the core of the data acquisition node and data receiving gateway of distance to head,back and tail data from the sow ultrasonic probe acquisition,and by sending and receiving wireless RF module.(2)Research on prenatal sows behavior process:distance data collected by the ultrasonic sensor,designs a filtering algorithm based on the distance of abnormal data preprocessing and classification algorithm research;sow prenatal behavior recognition with K-means clustering algorithm as the core,the judgment results are compared with the results by observing the behavior of video statistics.(3)A Web platform based on the Android platform for the comprehensive inquiry of the mobile information management terminal and the monitoring platform to facilitate the real-time query of the breeder behavior history records,and record the consistent abnormal behavior.Experiments were carried out in Jiangpu farm,Pukou District,Nanjing city,Jiangsu province.Landrace sows were experimental on;test sows at about a week before the due date were transferred to crates.The results show a stable safe system that the system can quickly identify the sow physical activities from the information and classify the sow behavior as nest,lie and standing behaviors,at an accuracy rate of 90.47%.