(J. of the Society of Agricultural Structures, Japan, 31(3), 163-170, 2000)
Studies on Control Systems of Greenhouse Environment (Part 4)
- Prediction and automatic watering system for soil moisture using multiple regression analysis -
Kazuhiro
NAKANO, Takuo KAWAKAMI, Yasuo OHTSUKA and Shinobu MIZUNO
Summary
The objective of this study is to develop the automatic watering system which can predict the changes of soil moisture; pF value, in a greenhouse and control pF value appropriately. The soil moisture control system and the environmental supervisory system were constructed in this study. Changes of pF values were determined as the dependent variable in the multiple regression, and temperature, humidity and solar radiation were used for the independent variables. The cultivated results using the above multiple regression method were compared with farmer's manual watering.
For the cucumber cultivation, the amount of watering, the yield, the quality of vegetables and the income were measured. Differences in the results between the multiple regression method and the farmer's manual watering were almost the same in the above factors. For the tomato cultivation, the total yield of vegetables in the multiple regression method was 10% fewer than that by the farmer's manual watering. But the sugar content, the quality of the vegetables and the total income in the multiple regression method were almost the same result in the farmer's manual watering.
The automatic watering system using multiple regression method developed in this study is useful for the prediction of soil moisture and appropriate control of pF value in the greenhouse.