基于智能感知和深度学习的养蚕管理系统设计

Design on Highly Efficient Silkworm Rearing Technology Based on Intelligent Perception and Processing

摘要: 新阶段，智能化、装备化养蚕成为现代蚕业高质量发展的必然趋势，但以往的养蚕仍然面临着劳力紧张、装备老旧、控制复杂、标准不高、规范不严等现实问题，急需变革养蚕方式。在养蚕环境智能监控研究的基础上，探讨研制基于智能感知和深度学习的养蚕管理系统，结合自动化、省力化养蚕实景，依托物联网智能感知技术，对养蚕环境要素和蚕体发育状态进行智能感知、识别、传输、处理，精准掌握养蚕发育进程和环境状况，运用深度残差神经网络算法(DRCNN)，对多特征数据提取和分级融合处理，及时校正养蚕数据标准，实时监控养蚕生产流程、输出技术方案，提供自动化养蚕决策，为打造智能化、装备化养蚕兼容平台提供参考。

Abstract: In the new phase, intelligent and equipped sericulture has become an inevitable trend of high-quality development of modern sericulture; however, the management of Silkworm rearing in the past still faced such practical problems as labor shortage, old equipment, complicated control, low standard and lax regulation, and so on. It is urgent to reform the mode of production of silkworm rearing. Based on the research of intelligent monitoring and control of Sericulture Environment, this paper discusses the development of sericulture management system based on intelligent perception and deep learning; combining the real scene of automatic and labor-saving sericulture production, relying on the intelligent perception technology of Internet of things, the intelligent perception, recognition, transmission and processing of the environmental factors and the developmental state of the silkworm are carried out; the development process and the environmental state of the Silkworm are accurately mastered; depth residual neural network Algorithm (DRCNN) is used; the multi-feature data are extracted and classified and fused; the data standard of sericulture should be corrected in time; the production process and output technical scheme should be monitored in real time; and the automatic decision of sericulture should be provided.