主催: The Japanese Society for Artificial Intelligence
会議名: 第34回全国大会(2020)
回次: 34
開催地: Online
開催日: 2020/06/09 - 2020/06/12
The greenhouse microclimate is an uncertain nonlinear system and affected by various physico-chemical processes, such as heat and mass transfer between plants, air, growing condition and the plastic cover, plants photosynthesis rate, cultivation methods, CO<sub>2</sub> concentration, solar radiation, atmospheric RH, temperature, etc. Growers continuously supplied hot air and CO<sub>2</sub> inside the greenhouse for cultivating eggplant during winter season. This causes ununiform distribution of CO<sub>2</sub>, RH and temperature inside the different area of the greenhouse and potentially risk against disease and pest breakout, water stress condition. Therefore, the quality and accuracy of the microclimate monitoring data for optimal control of the greenhouse microclimate has a great influence on ensuring the high crop yield with good quality throughout the year. In this experiment, we divided the greenhouse into 20 subzones and sensors were setup into three layers (top, middle, bottom). Total 72 wireless sensors (48 temperature sensors, 9 solar radiation sensors, 16 CO<sub>2</sub> and RH sensors) with 4 mobile base stations were setup for the greenhouse area of 2910m2. In this paper, we investigated the accuracy, continuation and correlation of the different parameters to optimize the number of sensors for eggplant cultivation all the year round inside a solar type greenhouse.