标准化生猪养殖工艺贝叶斯网络模型的构建  

Construction of Bayesian network model for standardized pig breeding technology

作　　者：王金龙 郭海[1] 袁帅 皇可 毕春光[1] WANG Jinlong;GUO Hai;YUAN Shuai;HUANG Ke;BI Chunguang(College of Information Technology,Jilin Agricultural University,Changchun 130118,China)

机构地区：[1]吉林农业大学信息技术学院,长春130118

出　　处：《黑龙江畜牧兽医》2020年第20期16-20,共5页Heilongjiang Animal Science And veterinary Medicine

基　　金：国家重点研发计划项目(〔2017〕-YFD-0502001)。

摘　　要：生猪养殖管理是标准化生猪养殖工艺的重要组成部分,对其进行研究可以有效解决生猪养殖管理中可能遇到的问题。笔者依据专家经验和贝叶斯网络理论构建生猪养殖标准化工艺模型,针对生猪养殖管理各个关键因素建立贝叶斯网络拓扑结构,采用贝叶斯估计算法对网络拓扑结构进行参数学习,确定各节点的条件概率表,并与基于专家经验建立的贝叶斯网络模型进行对比分析。结果表明:经过参数学习的贝叶斯网络模型具备有效性。说明贝叶斯网络模型是可行的,可以在一定程度上对专家经验进行代替,可对生猪养殖管理起到指导性作用,节省劳动力,提高生猪养殖管理的质量。Production management of piggery is an important part of the standardized production technology.The study of pig breeding management can effectively solve various problems that may be encountered in pig breeding management.Based on the expert knowledge and Bayesian network theory,a standardization process model of pig breeding was constructed.The Bayesian network topology was established for each key factor of piggery production management,and parameters of the network topology were learned by using the Bayesian estimation algorithm,conditional probability table of each node was determined,and Bayesian network model based on expert experience was compared and analyzed.The results showed that the Bayesian network model with parameter learning was effective,which indicates that the Bayesian network model was feasible and could replace the expert experience to a certain extent.It could play a guiding role in the production management of piggery,save labor and improve the management quality of piggery.

关 键 词：贝叶斯网络 专家经验 猪场管理 参数学习 模型 对比 

分 类 号：TP39[自动化与计算机技术—计算机应用技术]