基于离散曲率的温室CO2优化调控模型研究  

Carbon Dioxide Optimal Control Model Based on Discrete Curvature

作　　者：胡瑾[1,2] 田紫薇 汪健康 卢有琦 辛萍萍 张海辉[1,2] HU Jin;TIAN Ziwei;WANG Jiankang;LU Youqi;XIN Pingping;ZHANG Haihui(College of Mechanical and Electronic Engineering,Northwest A&F University, Yangling, Shaanxi 712100, China;Key Laboratory of Agricultural Internet of Things,Ministry of Agriculture and Rural Affairs, Yangling,Shaanxi 712100, China;Shaanxi Key Laboratory of Agricultural InformationPerception and Intelligent Service, Yangling, Shaanxi 712100, China)

机构地区：[1]西北农林科技大学机械与电子工程学院,陕西杨凌712100 [2]农业农村部农业物联网重点实验室,陕西杨凌712100 [3]陕西省农业信息感知与智能服务重点实验室,陕西杨凌712100

出　　处：《农业机械学报》2019年第9期337-346,共10页Transactions of the Chinese Society for Agricultural Machinery

基　　金：国家自然科学基金项目(31671587、31501224);陕西省重点研发计划项目(2018TSCXL-NY-05-02);中央高校基本科研业务费专项资金项目(2452017124)

摘　　要：提出了基于离散曲率算法的温室CO2优化调控模型,通过设计嵌套试验采集温室不同温度、光照强度、CO2浓度组合下的番茄光合速率,利用支持向量机回归算法(Support vector regression algorithm,SVR)构建光合速率预测模型;以预测模型网络为目标函数,采用 L 弦长曲率算法实现CO2响应曲线离散曲率的计算,利用爬山法获得不同温度、光照强度组合条件的CO2响应曲线曲率最大点,以此作为效益最优的调控目标值,进而基于SVR构建CO2优化调控模型。结果表明,调控模型的决定系数为0.99、均方根误差为4.42 μmol/mol、平均绝对误差为 3.17 μmol/mol,拟合效果良好。与CO2饱和点目标值的调控效果对比发现,理论上CO2供需量平均下降61.81%,光合速率平均减少15.58%;验证试验中,相较饱和点调控下光合速率平均下降15.14%,CO2供需量下降57.61%,相较自然条件下光合速率升高26.70%。说明此温室CO2优化调控模型具有高效节能特点,为设施作物CO 2高效精准调控和节本增效提供了理论基础。CO2 is one of the main resources for plant photosynthesis. The slope of CO2 response curve represents the effect of CO2 concentration on photosynthetic rate. The first curvature maximum point represents the characteristic point where the effect of CO 2 concentration on photosynthetic rate becomes weak. Therefore, the acquisition of this point is the key to realize the optimal benefit control of CO2. A CO2 optimal control model based on discrete curvature algorithm was proposed. Firstly, a photosynthetic rate experiment was designed. The subject of the experiment was tomato. The experimental conditions were the different combinations of temperature, photonic flux density and CO 2 concentration. In the experiment, temperature, photon flux density and CO2 concentration gradients were set as 6,10 and20, respectively. Totally1200 sets of CO2 response data were obtained by LI 6800 portable photosynthetic rate instrument. And 80% data were used to construct photosynthetic rate prediction model based on the support vector regression, and the rest of the data were used for model verification. Then, the CO2 response curves under the nested conditions were obtained by using the established photosynthetic rate prediction model. Next, the discrete curvature value of every response curve was calculated by the L -chord discrete curvature algorithm. Using hill-climbing method, the maximum curvature value of every response curve was obtained. The CO2 concentrations corresponding to the maximum curvature values were taken as the control target values. Finally, the CO2 optimal control model was constructed based on the support vector regression. The results showed that the decision coefficient of the control model was 0.99, the mean square error was 4.42 μmol/mol, and the average absolute error was 3.17 μmol/mol. Compared with the CO2 saturation point, the CO2 demand was decreased by 61.81%, but the photosynthetic rate was decreased by15.58%. In the verification experiment, compared with the saturation point regulation, the avera

关 键 词：温室 CO2优化调控 支持向量机回归 离散曲率 爬山法 光合速率 

分 类 号：S126[农业科学—农业基础科学]