摘要
针对温室温度、光照、水肥控制不协调导致的能耗高、水肥利用率低的问题,提出了基于深度强化学习的温室环境协调控制方法。以能耗、光合速率为优化目标,采用深度强化学习算法训练模型,对温度、光照调控目标值进行优化;通过分析不同营养液灌溉量对作物长势的影响,确定灌溉量动态调整方法;开发了基于深度强化学习的温室环境协调控制系统软硬件。实验结果表明:该方法能够协调控制温室温度、光照和水肥环境因子,与传统控制方法相比,环境调控能耗降低8.1%,营养液灌溉量降低7.9%,光合速率提升2.7%,能够为温室环境高效控制提供决策支持。
In response to the issues of high energy consumption and low nutrient utilization caused by the lack of coordination in greenhouse temperature,light,and water-fertilizer control,proposed a deep reinforcement learning-based method for coordinated control of the greenhouse environment.With energy consumption and photosynthetic rate as the optimization objectives,a deep reinforcement learning algorithm was used to train the model to optimize the target values of temperature and light regulation.The impact of different nutrient irrigation levels on crop growth was analyzed to determine a dynamic adjustment method for irrigation volume.The software and hardware of the coordinated control system of greenhouse environment based on deep reinforcement learning were developed.The experimental results showed that the method can coordinate the control of greenhouse temperature,light and water and fertilizer environmental factors,and compared with the traditional control method,the energy consumption of environmental regulation was reduced by 8.1%,the amount of nutrient solution irrigation was reduced by 7.9%,and the photosynthetic rate was increased by 2.7%,which can provide decision support for the efficient control of greenhouse environment.
作者
左志宇
牟晋东
毛罕平
韩绿化
胡建平
张晓东
金文帅
Zuo Zhiyu;Mu Jindong;Mao Hanping;Han Lyuhua;Hu Jianping;Zhang Xiaodong;Jin Wenshuai(Key Laboratory of Modern Agricultural Equipment and Technology,Ministry of Education,Jiangsu University,Zhenjiang 212013,China)
出处
《农机化研究》
北大核心
2025年第5期22-27,共6页
Journal of Agricultural Mechanization Research
基金
国家大宗蔬菜产业技术体系岗位专家任务项目(CARS-23-D-03)
江苏省农业科技自主创新项目(CX(20)1005)。
关键词
温室
深度强化学习
协调控制
光合速率
能耗
greenhouse
deep reinforcement learning
coordinated control
photosynthetic rate
energy consumption
