2018-01-012024-05-14https://scholars.lib.ntu.edu.tw/handle/123456789/658672摘要:本年度將於使用DeLaval系統飼養規模達總頭數至少200頭的商業乳牛場,進行現場智慧化管理系統驗證。驗證可分為三個部分。第一個部分驗證「活動量異常判斷規則」方法是否真的可以應用在不同場地與牛隻的情況。第二個部分為驗證產乳量與牛隻平日的活動行為的關聯性。第三個部分則為在不同場域中,適當的調控畜養環境氣候可改變乳牛的行為活動,並增加總體的產乳量。另外,調整行為辨識方法。藉由影像可偵測乳牛飲水與進食等行為,將於較多頭牛的場域進行實地驗證並調整辨識方法,並將辨識結果作為是否發生熱緊迫之判斷依據之一。同時在臺大農 場繼續改進管理系統的介接彈性以及智慧化調控降溫設施的效率,於懷孕牛欄位佈建行為監測、影像監測以及溫濕度監測系統,收集懷孕牛在不同季節環境條件下的各種行為每日時間分配變動資料。選定部分區域架設可以接受自動調控指令的風扇與噴霧設施,在熱季進行懷孕牛於分娩前3週紓解熱緊迫的效益。<br> Abstract: The smart health management system developed in the first year will be moved to commercial dairy farm with herd size at least 200 heads to validate its functions.Rules of abnormal behavior judgement will be tested.Link between milk yield and daily behavior pattern will be examined. The effectiveness of auto-control of cooling system to relief heat stress will be tested. Automatic image analysis system for drinking and feeding behavior will be fine-tuned.At the dairy farm of National Taiwan University, advanced version of smart health management system will be further developed.Also, pregnant cow behavior patter under heat stress will be established and the effectiveness of cooling during 3 weeks before calving will be monitored.乳牛熱緊迫智慧健康管理Dairy cattleHeat stressSmart health management乳牛熱緊迫健康監測智慧化管理系統研發計畫