规模化牧场自动监控设备下奶牛乳房炎影响因素与预警模型研究--基于二元Logistics回归

Study on Influencing Factors and Early Warning Model of Dairy Cow Mastitis Under Automatic Monitoring Equipment in Large-scale Pasture Based on Binary Logistics

奶牛乳房炎是危害奶牛健康的常见病之一,耗费较多人力、财力,增加了牧场的管理成本,同时严重影响动物福利。通过对比分析患病奶牛与健康奶牛的产奶量、活动量和反刍时间三项数据,采用二元Lo⁃gistics回归模型建立奶牛乳房炎预警模型,其中活动量指标在二元Logistics回归模型中拟合效果较好,模型的精确率为94%,稳定性为87%,能够证明该预警模型有效。

Dairy cow mastitis is one of the common diseases endangering the health of dairy cows.It consumes more human and financial resources,increases the management cost of pasture,and seriously affects animal welfare.By comparing and analyzing the milk yield,activity and rumination time of sick dairy cows and healthy granny cows,a dairy mastitis early warning model was established by using binary logistic regression model.The activity index fitted well in the binary logistic regression model.The accuracy of the model was 94%and the stability was 87%,which could prove that the early warning model was effective.