基于深度学习的鸡舍高温高湿环境对蛋鸡产蛋性能耦合作用机理研究

High temperature and humidity in summer in southern China aggravate the heat stress of the laying hens, which seriously affects the laying hens health and laying performance. At present, the coupling mechanism of multi-factors (air temperature, air humidity, harmful gases, etc.) in high temperature and humidity environment on laying performance of laying hens is not clear, which needs further study. To solve this problem, this project aims at the standard cascade cage laying house environment in summer in southern China, then focus on the following work, by using deep learning, computational fluid mechanics and multi-objective genetic optimization algorithms. 1) Study the numerical simulation method of rheological field of hens house, build the aerodynamic model of hens house in high temperature and humidity environment, to analyze the interaction mechanism of multi-factors in high temperature and humidity environment of hens house. 2) study on the correlation analysis method between multi-factors of high temperature and humidity environment and laying performance of laying hens, construct a comprehensive evaluation and prediction model of environmental quality of laying hens, to reveal the coupling mechanism of multi-factors of high temperature and humidity environment on laying performance of laying hens. 3) Study on the threshold searching method of high temperature and high humidity environment regulation affecting laying performance of hens house, construct the optimal threshold searching model for hens house production, to establish the optimal threshold parameters of high temperature and high humidity environment regulation. Through the above research, this study will provide a method reference for exploring the mechanism of the influence of high temperature and humidity environment on laying performance of intensive cage hens house and improve the environmental control technology of laying hen under extreme climate conditions.

南方夏季高温高湿加剧鸡群热应激，严重影响蛋鸡健康和产蛋性能。目前，鸡舍高温高湿环境多因子（空气温度、湿度、有害气体等）对蛋鸡产蛋性能耦合作用机理还不明确，有待深入研究。据此，本项目拟针对南方夏季标准化层叠式笼养蛋鸡舍，融合深度学习、计算流体力学和多目标遗传寻优算法等方法，重点开展：1）研究鸡舍流变场数值模拟方法，构建鸡舍高温高湿环境空气动力学模型，解析鸡舍高温高湿环境多因子互作机理。2）研究鸡舍高温高湿环境多因子与蛋鸡产蛋性能关联分析方法，构建鸡舍环境质量综合评价预测模型，揭示高温高湿环境多因子对蛋鸡产蛋性能耦合作用机理；3）研究影响蛋鸡产蛋性能的鸡舍高温高湿环境调控阈值搜索方法，构建蛋鸡生产最优环境调控阈值搜索模型，确立鸡舍高温高湿环境最优调控阈值参数。通过以上研究，为探索集约化笼养蛋鸡舍高温高湿环境对蛋鸡产蛋性能影响机理，进而提升极端气候条件下蛋鸡养殖环境调控技术提供方法参考。

南方夏季高温高湿加剧鸡群热应激，严重影响蛋鸡健康和产蛋性能。目前，鸡舍高温高湿环境多因子（空气温度、湿度、有害气体等）对蛋鸡产蛋性能耦合作用机理还不明确，有待深入研究。据此，本项目针对南方夏季标准化层叠式笼养蛋鸡舍，以圣迪乐生态食品有限公司安庆分公司怀宁县蛋鸡养殖基地为实验示范基地，同期选取了该基地的八层层叠式笼养蛋鸡舍和四层层叠式笼养蛋鸡舍开展实验，融合深度学习、计算流体力学和智能优化算法等方法，开展了以下研究内容：.（1）研究鸡舍高温高湿环境多因子流变场数值模拟方法，通过构建基于计算流体动力学的鸡舍环境空气动力学模型，研究分析了鸡舍不同位置蛋鸡舍空气流场、温度场和相对湿度场的分布变化规律，阐明了鸡舍高温高湿环境不同位置多因子间的关联关系。.（2）研究鸡舍高温高湿环境多因子与蛋鸡产蛋性能的非线性复杂函数表示方法，通过构建的基于粒子群优化深度置信网络算法的鸡舍高温高湿环境质量综合评价预测模型，分析蛋鸡舍环境多因子和采食量对蛋鸡产蛋性能的耦合影响机理，为鸡舍高温高湿环境精准调控提供了环境评价依据。.（3）研究构建了基于布谷鸟寻优算法的蛋鸡生产最优环境调控阈值模型，探寻鸡舍高温高湿环境最优调控阈值，解决了多环境因子关联影响下的鸡舍高温高湿环境精准调控阈值定量问题。.在此基础上，提出了南方夏季高温高湿鸡舍环境调控最优阈值范围：鸡舍内环境温度建议控制在28-29℃之间；湿度建议控制在70-80%之间；光照强度建议控制在10-20lux之间；风速建议控制在2-3m/s之间；二氧化碳浓度建议控制不高于1000ppm；氨气浓度不高于2ppm；PM2.5浓度应控制不高于1 ug/m3。并提出了有效改进示范基地设施环境质量的措施，提高了示范基地的夏季高温高湿环境下蛋鸡的生产性能。研究成果为提升极端气候条件下层叠式笼养蛋鸡舍环境调控技术提供理论依据和技术支撑，具有重要理论意义、科学价值及应用前景。