Abstract:Considering the temporal and spatial variation characteristics of greenhouse environment, the C++-Fluent joint optimization framework was established by constructing the greenhouse building computational fluid dynamics (CFD) model and combining the non dominated genetic algorithm with elite strategy (NSGA-Ⅱ), so as to realize the multi-objective and high-resolution optimization of greenhouse environmental factors. The CFD greenhouse model was verified by a greenhouse in Zhenjiang City, Jiangsu Province. The iterative optimization algorithm was implemented in C++ and the computing efficiency was improved by supercomputer. The optimization objectives included crop regional temperature field, carbon dioxide distribution and control of greenhouse fan energy consumption. The results showed that the CFD temperature field and velocity field were in good agreement with the experimental values at the monitoring points, and the average relative errors were 4.9% and 7.05% respectively. In order to obtain the optimal value of crop growth temperature field and carbon dioxide distribution in a certain scene and maintain low energy consumption of greenhouse fan, the greenhouse wet curtain temperature was [296.6K, 302K], and the fan outlet wind speed was [2.9m/s, 5.5m/s]. At this time, the temperature field, carbon dioxide distribution and fan energy consumption in the crop area were in the optimal range, which helped to improve crop yield and reduce greenhouse energy consumption. The computing efficiency of the optimization scheme developed under the supercomputer Linux system was significantly higher than that of the personal computer, and the computing time was shortened by about 88.09%. The strategy proposed fully considered the temporal and spatial variation characteristics of greenhouse environment and realized high-resolution and high-efficiency optimization of multiple environmental factors in greenhouse, which can provide a basis for the selection of greenhouse environmental parameters suitable for crop growth.
