Design And Test Of Information Management System For Performance Determination Of Boar

Boar performance determination is the necessary way to obtain phenotypic data of boar growth performance,the material basis for boar feeding behavior research,and the key link to build the breeding system of live pigs.At present,my country’s boar performance determination station mainly relies on foreign imports,and its supporting boar performance measurement information management system has relatively simple analysis functions,and there are problems such as complicated installation and operation.Among the traits measured in breeding pigs,the age of 100 kg body weight was an important indicator to measure the growth rate of boar.This trait was obtained through analysis and calculation at the end of the determination,but there was a problem of long acquisition period.Feeding behavior of boar is closely related to growth performance.It is helpful to improve breeding efficiency and reduce breeding economy and time cost to explore the rules between feeding traits and growth traits,and to predict growth rate in advance by using early feeding traits of boar.This paper analyzs the phenotypic correlation between feeding traits and growth traits based on the growth performance measurement data of Duroc boars,build a prediction model for boars to reach the target weight day,and design a boar performance determination information management system integrating boar performance determination trait calculation,feeding trait analysis and equipment remote monitoring by using communication technology,information technology and web technology.(1)This paper analyzes the information management system of boar performance determination which is widely used in China,and analyzes the detailed functional requirements of the system in four aspects:user,real-time remote monitoring of the determination station,boar performance determination and feeding trait analysis,respectively,and determines the functions of remote monitoring,performance index calculation and feeding trait analysis of the system in combination with the actual requirements of the breeding swine quality supervision and testing Center,ministry of agriculture(wuhan).The functions of the system were determined,such as remote monitoring,performance index calculation and feeding trait analysis.(2)The phenotypic correlations between four growth traits(daily weight gain,age to100 kg body weight,feed conversion ratio and backfat)and five feeding traits(daily feed intake,number of daily feed intake,daily feed duration,single feed intake and feeding rate)were analyzed in Duroc breeding pigs.And based on the significant correlation between feeding traits and growth rate of breeding pigs,a partial least-squares prediction model was constructed for breeding pigs up to 100 kg body weight age by combining the opening weight,opening age and 30 kg body weight age of breeding pigs,and the mean coefficient of determination R~2 of the 5-fold cross-validation test set of the model was 0.68,and the mean absolute error MAE was 3.2 d.(3)This paper used B/S(Browser/Server)architecture with front and back-end separation and MVC design pattern to design and develop the information management system for breeding pig performance measurement.CAN bus is selected to complete the bidirectional transmission of data between the information management system and the measurement station,and the"confirmation","verification"and"retransmission"mechanisms are added on the basis of UDP protocol to ensure the real-time,efficient and reliable transmission of measurement data.The front-end of the information management system is based on Reactjs technology stack,Ant Design component library,Bizcharts charting library,and the server back-end of the information management system is based on flask web framework and uses My SQL for data storage.The system functions are broadly divided into four functional modules:user module,real-time monitoring module of measurement station,breeding pig growth performance measurement module and feeding trait analysis module,which realize real-time remote monitoring of measurement station,breeding pig measurement data collection,statistics and analysis,as well as feeding trait analysis.(4)After the development of the information management system was completed,two breeding trials were conducted to test the data transmission reliability and functional operational stability of the information management system.The first trial was conducted in the experimental pig farm of Huazhong Agricultural University,and the results showed no loss of feeding data.The second test was conducted in a national pig core breeding farm in Hubei Province,and the test results showed that the functional modules of the information management system operated stably and met the expected design requirements.The PLS prediction model was tested using test breeder measurement data,and the MAE between the PLS predicted age and the actual age at target weight was 5.3 d.