摘要
智能可穿戴设备应用到奶牛发情期预测中能够提升奶牛的养殖经济效益。此类用于智能畜牧的可穿戴设备受到体积限制,对设备的低功耗设计提出了更高要求。传统的运动监测耳标往往需要持续采样以实现准确的奶牛计步,存在能耗较高的问题。开发了一款基于动态占空比间歇采样的奶牛耳标设备,并使用预先训练完成的支持向量机(SVM)模型对未采样区间的奶牛步数进行预测。本耳标设计还通过预测结果所体现的奶牛运动状态动态调节设备工作与休眠的占空比,进一步降低系统能耗。测试结果表明,所设计的耳标设备工作电流为3.1 mA,休眠电流为7.7μA,平均工作电流低于0.76 mA,比持续工作的耳标的电流降低约75%,大大提高了设备的工作时长。
The application of smart wearables to predict the oestrus of cows can improve the economic benefit of dairy farming.Such wearables for livestock are limited by their size and therefore place greater demands on the low power design of the devices.Traditional movement monitoring ear tags often require continuous sampling to achieve accurate cow pacing,which has the problem of high energy consumption.A dynamic duty cycle intermittent sampling based ear tag device for cows pacing is developed and a pre-trained support vector machine(SVM)model is used to predict the number of cow steps in the unsampled intervals.The ear tag design also dynamically adjusts the duty cycle of the device to operate and sleep through the cow movement state as reflected in the prediction results,further reducing the system energy consumption.Test results show that the ear tag device designed has an operating current of 3.1 mA and a sleep current of 7.7μA,with an average operating current of less than 0.76 mA,which is approximately 75%lower than the current for continuous operation,greatly improving the operating time of the device.
作者
张志鹏
符栋伟
凌明
ZHANG Zhipeng;FU Dongwei;LING Ming(Department of Electronics,Nanjing Vocational Institute of Mechatronic Technology,Nanjing Jiangsu 211306,China;School of Microelectronics,Southeast University,Nanjing Jiangsu 210096,China)
出处
《电子器件》
CAS
北大核心
2022年第6期1510-1516,共7页
Chinese Journal of Electron Devices
基金
江苏省农业科技自主创新资金项目(CX(21)3121)。
关键词
智能畜牧
间歇计步
动态占空比
步数预测
intelligent animal husbandry
intermittent step counting
dynamic duty cycle
step count prediction
