奶牛饲喂自动机电控制系统的设计与试验

发布时间：2018-02-25 08:10

  本文关键词： 设计 控制系统 自动检测 奶牛 精准饲喂 采食行为 数据分析　出处：《农业工程学报》2017年07期 　论文类型：期刊论文

【摘要】：为开展奶牛精准饲喂及采食行为学研究,设计了一种集自动识别、饲喂、数据自动采集、数据分析与处理于一体的奶牛饲喂自动机电控制系统。该系统包括机械装置、电子识别系统、料槽称质量系统、中央控制系统、现场数据存贮及远程数据提取与分析系统等几部分。其中,机械装置包括料斗、支撑座、栏杆和阻挡单元等;电子识别系统包括阅读天线及料门启闭的气动装置;料槽称质量系统除支撑座外,还有嵌入的质量传感器及线路;中央控制系统包括微处理器、看门狗复位电路、读卡器电路、称质量数据采集电路、数据通信电路、数据收发器电路及外围驱动与稳压电路等。现场数据存贮电路接受来自各个饲喂系统的中央控制系统发送的采食行为数据,其主板结构与中央控制系统基本一致,预设可存贮记录数为14 000条,且采用堆栈数据存贮模式。远程PC端数据提取与分析系统实时管理采食行为数据,并提供多功能的数据挖掘分析。系统测试结果表明,对牛只低频RFID(134 kHz)电子耳标的识读率为100%,料及槽的计量范围为0.01~200 kg,最低称量精度10 g,实际称量相对误差≤0.15%,同时满足奶牛对最大采食量及精准饲喂对计量的需求。系统的采食行为试验表明,奶牛的日均采食次数、采食时间及采食量等采食行为均差异显著(P0.05),符合奶牛的采食行为特点。具体地,奶牛日均采食次数10~13次,日均采食时间5.38 h,而奶牛个体实际采食量与NRC(National Research Council)模型预测的采食量有-4.76%~7.83%的偏差,可能是由各种内外部因素及NRC模型的普适度造成的,有待进一步研究。总之,该系统能较好地实现奶牛个体的精细化饲喂,为研究奶牛的采食行为特点提供了在线、智能化的自动数据采集与分析平台。
[Abstract]:In order to carry out the research on the precise feeding and feeding behavior of dairy cattle, a kind of automatic mechanical and electrical control system for dairy cow feeding is designed, which integrates automatic identification, feeding, data automatic collection, data analysis and data processing. Electronic identification system, feed tank quality system, central control system, field data storage and remote data extraction and analysis system, etc. Among them, the mechanical device includes hopper, support seat, railing and blocking unit, etc. The electronic identification system includes a pneumatic device for reading the antenna and the opening and closing of the material door; a feed slot quality system includes an embedded mass sensor and a circuit in addition to the support seat; and a central control system includes a microprocessor, a watchdog reset circuit, and a card reader circuit, and a central control system includes a microprocessor, a watchdog reset circuit, and a card reader circuit. The quality data acquisition circuit, the data communication circuit, the data transceiver circuit and the peripheral drive and voltage stabilizing circuit, etc.; the field data storage circuit receives the feeding behavior data from the central control system of each feeding system. The structure of the main board is basically consistent with that of the central control system. The default number of storable records is 14 000, and the stack data storage mode is adopted. The remote PC data extraction and analysis system manages the feeding behavior data in real time. And provides multi-function data mining analysis. The system test results show that, The recognition rate of low frequency RFID(134 kHz) was 100, the measuring range of feed and trough was 0.01 ~ 200kg, the lowest weighing accuracy was 10 g, the relative error of actual weighing was 鈮，

本文编号：1533628