节点文献

基于嵌入式Linux的犊牛饲喂与监控系统

The Calf Feeding and Monitoring System Based on Embedded Linux

【作者】 张炎;

【导师】 彭熙伟;

【作者基本信息】 北京理工大学 , 控制工程, 2018, 硕士

【摘要】 随着人们生活水平的逐步提高,乳品市场的需求不断增加,现代畜牧业规模进一步扩大,犊牛自动饲喂设备开始受到企业的追捧。本文针对现有的犊牛饲喂设备,设计了一套基于嵌入式Linux的犊牛饲喂与监控系统,该系统不仅能够节省大量人力资源,而且能够严格保证饲喂时奶液的温度、科学的饲喂计划和良好的卫生状况,能够有效降低犊牛的发病率,对于提高养殖场的经济效益有直接的作用。首先,本文对国内外犊牛饲喂控制系统和嵌入式系统的基本概念进行了概述,分析了犊牛饲喂控制系统的基本功能和组成部分。根据犊牛的饲喂要求,并基于现有的犊牛饲喂设备,提出了本系统应具备的功能,形成了系统的总体方案。确定了主要硬件设备的选型,包括核心控制器、传感器和RFID耳标读卡器等设备。随后,搭建了嵌入式Linux系统的运行环境和开发平台。根据系统的总体方案设计了主控程序,并按照模块化的设计方案,在分析自动识别系统、饲喂控制系统、自动清洗系统、恒温控制系统和数据管理系统的工作原理的基础上,进行了软件的设计开发,实现了系统基本的自动饲喂和清洗功能。设计了犊牛精准饲喂系统,为不同品种、体况和生长状态的犊牛制定了有针对性的饲喂计划,并基于线性回归算法实现了对系统参数的准确调整,有效提高了系统的给料精确度,保证了饲喂计划的严格实施;设计了基于Haar特征的犊牛面部识别系统,使系统能够识别出犊牛面部区域,并基于犊牛鼻子的颜色特征判断其健康状况,提高了系统的智能化程度,使犊牛自动饲喂设备迈出了智能化的第一步。最后,对恒温控制系统、精确给料算法和面部识别软件进行了测试。测试结果:出奶温度能保持在39℃左右,温度变化不超过0.2℃;设备的给料精度达到99%以上;犊牛面部识别软件的识别准确率达到80%以上。测试结果均满足了设计的要求。

【Abstract】 With the continuous improvement of people’s living standards,the demand for the dairy product market continues to increase,and the scale of modern animal husbandry has further expanded.The automatic feeding equipment for calves has been sought after by companies.Aiming at the existing calf feeding equipment,a set of calf feeding and monitoring system based on embedded Linux system is design in this thesis.This system can not only save a lot of human resources,but also establish the scientific feeding plan for every calf,and strictly guarantee the temperature of milk while feeding and good sanitary conditions in the equipment.This system has effectively reduced the disease incidence of calves and has a direct effect on improving the economic benefits of the breeding farms.Firstly,the calf feeding control system at home and abroad and the basic concepts of embedded systems are summarized in this thesis,the basic functions and components of the calf feeding control system are analyzed.According to the calves’ feeding requirements,and based on the existing calf feeding equipment,the function that this system should have is proposed and the overall scheme of the system is formed.Subsequently,the selection of major hardware devices,including core controller,sensors and RFID ear tag reader are determined.Secondly,the operating environment and development platform for the embedded Linux system are built.The main control program is designed according to the overall scheme of the system,and in accordance with the modular design concept,based on the analysis of the working principles of automatic identification system,feeding control system,automatic cleaning system,constant temperature control system and data management system,the software are designed and performed.So that the basic automatic feeding and cleaning functions of the system are realized.Subsequently,a precision feeding system for calf is designed to provide targeted feeding plans for calves of different breeds,body conditions and growth status.The linear regression algorithm is used to achieve accurate adjustment of system parameters which effectively improving the accuracy of the feeding system and guarantees the strict implementation of the feeding plan.A calf facial recognition system based on the Haar feature is designed so that the system can recognize the facial area of the calf and judge the health status of the calf based on the color characteristics of the calves’ nose.The intelligent degree of the calf feeding control system has been improved and the calf automatic feeding device has taken the first step in intelligence.Finally,the thermostatic control system,precise feeding algorithm and facial recognition software are tested.Test results: the milk temperature of outlet can be maintained at about 39°C and the temperature change is not more than 0.2°C;the feeding accuracy of the equipment is over 99%;the recognition accuracy of the calf facial recognition software is over 80%.The test results meet the design requirements.

  • 【分类号】TP277;S823
  • 【被引频次】1
  • 【下载频次】109
  • 攻读期成果
节点文献中: 

本文链接的文献网络图示:

本文的引文网络