INTELLIGENT ANIMAL FEEDING SYSTEM BASED ON MULTISOURCE INFORMATION ACQUISITION

基于多源信息采集的动物智能饲喂系统

DOI : https://doi.org/10.35633/inmateh-79-65

Authors

Linwei LI

College of Information Science and Engineering, Shanxi Agricultural University, Taigu 030801, Shanxi/China;Key Laboratory of Equipment and Informatization in Environment Controlled Agriculture, Ministry of Agriculture and Rural Affairs, Hangzhou 310058, Zhejiang/China;

Jie BAI

Department of Big Data and Intelligent Engineering, Shanxi Institute of Technology, Yangquan 045000, Shanxi/china;

Xinyu ZHAO

College of Information Science and Engineering, Shanxi Agricultural University, Taigu 030801, Shanxi/China;

Guanzhen LI

College of Computer Science and Software Engineering, Hohai University, Nanjin 211100, Jiangsu/China;

Yuanrong TIAN

Faculty of Software Technologies, Shanxi Agricultural University, Taigu 030801, Shanxi/China;

Peng CHENG

Combat Support Academy, Rocket Force University of Engineering, Xi’an 710025, Shanxi/China;

Dongyao LIU

Department of Mechanical Engineering,Shanxi Institute of Technology, Yangquan 045000, Shanxi/china;

Zhenyu LIU

Dryland Farm Machinery Key Technology and Equipment Key Laboratory of Shanxi Province, Taigu 030801, Shanxi/China;College of Agricultural Engineering, Shanxi Agricultural University, Taigu 030801, Shanxi/China;

Senhao YANG

College of Information Science and Engineering, Shanxi Agricultural University, Taigu 030801, Shanxi/China;

Tao ZHAO

Department of Big Data and Intelligent Engineering, Shanxi Institute of Technology, Yangquan 045000, Shanxi/china;

(*) Xiaoping JI

Faculty of Software Technologies, Shanxi Agricultural University, Taigu 030801, Shanxi/China;

(*) Corresponding authors:

jixiaoping@sxau.edu.cn |

Xiaoping JI

Abstract

With the increasing refinement of livestock feeding management, conventional manual feeding and record-keeping methods often fail to ensure reliable confirmation of individual animal arrival and accurate identity matching. This may result in ineffective feeding, feed misdelivery, missed feeding, and deviations from the prescribed feed amount. To address these limitations, an intelligent animal feeding system based on multisource information acquisition was developed. The system integrates camera-based monitoring with microphone-array sound-source localization to support target positioning and combines radio-frequency identification (RFID)-based identity verification with weight-based arrival confirmation to enable individualized feeding while minimizing animal–target mismatches. The system also collects body weight and environmental data to generate traceable individual records and supports data transmission and command delivery through an ESP8266 module using the Message Queuing Telemetry Transport (MQTT) protocol. A WeChat mini-program is integrated into the platform to provide remote control, feeding-plan configuration, and data visualization, as well as AI-assisted recognition and question-answering functions. Experimental results demonstrated that the proposed system successfully integrates intelligent feeding with multisource information acquisition. Under the combined audio–visual configuration, the target-positioning success rate reached 92.67%, while the combined RFID–weight configuration achieved a trigger accuracy of 93.33%. These results indicate that the system can support refined feeding management and improve process traceability.

Abstract in Chinese

随着饲喂管理逐步向精细化发展，传统人工饲喂与记录的方式难以保证个体到位确认与身份匹配，易出现无效投喂、误投漏投与饲喂量偏差等问题。为此，本文构建基于多源信息采集的动物智能饲喂系统。系统通过摄像头监测与麦克风阵列声源方位数据辅助定位，并结合Radio Frequency Identification（RFID）身份确认与称重到位判定，实现个体化饲喂，避免对象错配；系统同步采集体重与环境参数，形成可追溯的个体记录，通过 ESP8266 与 Message Queuing Telemetry Transport（MQTT）实现数据上报与指令下发；微信小程序与平台对接，提供远端控制、方案配置与数据可视化，并集成 AI 辅助识别与问答服务。实验结果表明，本系统能够实现智能饲喂与信息采集的一体化运行，在音频与视频联合配置下，指向成功率为 92.67%；在 RFID 与称重联合配置下，触发准确率为 93.33%。上述结果可为精细化饲喂管理与过程溯源提供参考依据。

Indexed in