种猪采食行为与生长性状关联分析与实现  

Association analysis and implementation of correlation between feeding behavior and growth traits in breeding pigs

作　　者：王婷婷 黎煊 赵书红[2] 雷明刚 刘望宏 刘诗尧 WANG Tingting;LI Xuan;ZHAO Shuhong;LEI Minggang;LIU Wanghong;LIU Shiyao(Key Laboratory of Smart Farming for Agricultural Animals,Wuhan 430070,China;Center of Quality Test and Surpervison for Breeding Swine(Wuhan),MOA,Wuhan 430070,China;Shenzhen Institute of Nutrition and Health,Huazhong Agricultural University/Agricultural Genomics Institute at Shenzhen,Chinese Academy of Agricultural Sciences/Shenzhen Branch,Guangdong Laboratory for Lingnan Modern Agriculture,Shenzhen 518000,China)

机构地区：[1]农业农村部智慧养殖技术重点实验室,武汉430070 [2]农业农村部种猪质量监督检验测试中心(武汉),武汉430070 [3]华中农业大学深圳营养与健康研究院/中国农业科学院深圳农业基因组研究所/岭南现代农业科学与技术广东省实验室深圳分中心,深圳518000

出　　处：《华中农业大学学报》2022年第4期240-247,共8页Journal of Huazhong Agricultural University

基　　金：国家自然科学基金项目(31790414);中央高校基本科研业务费专项(2662020GXPY009);华中农业大学-中国农业科学院深圳农业基因组研究所合作基金项目(SZYJY2022031)。

摘　　要：为便于开展种猪采食性状研究及满足测定设备智能化发展需求,设计了一套基于物联网技术和Web技术的种猪性能测定信息管理系统。系统选取CAN总线进行数据双向传输,采用B/S(Browser/Server)前后端分离架构,前端基于Reactjs技术栈,后端采用flaskweb框架和MySQL数据库进行系统开发;2次测试结果表明,系统数据传输可靠性高,各功能模块满足预期设计要求,满足种猪性能测定实际使用需求;以杜洛克种猪为研究对象,分析采食性状与生长性状间的表型相关,并构建种猪达100 kg体质量日龄偏最小二乘回归(partial leastsquare,PLS)预测模型,交叉验证结果显示PLS测试集平均决定系数为0.68,平均MAE为3.2 d,使用试验数据对PLS模型进行测试,测试结果表明可以将采食性状与生长性状结合对种猪生长速度进行预测。In order to study feeding traits of breeding pigs and to meet the demand for intelligent development of measuring equipment,a set of information management system based on internet of things and web technology was designed.The system selects CAN bus for bi-directional data transmission,and adopts B/S(Browser/Server)front and back-end separation architecture.The front-end was based on Reactjs technology stack,and the back-end was developed with flask web framework and MySQL database.The results of two measurement tests showed that the data transmission reliability of the system is high,and functional modules meet the expected design requirements and the practical application requirements of performance measurement of breeding pigs.The phenotypic correlation between feeding traits and growth traits was analyzed in Duroc breeding pigs,and a partial least square(PLS)prediction model was constructed for breeding pigs up to 100 kg body weight.The cross-validation results showed that an average determination coefficient of PLS test set was 0.68 and the average MAE was 3.2 day.The PLS model was tested using experimental data,and the results showed that combination of feeding traits and growth traits could be used to predict the growth rate of breeding pigs in advance.

关 键 词：种猪 信息管理 采食行为 生长性能 智能化设备 精确育种 

分 类 号：S828.8[农业科学—畜牧学]