Dynamic optimization of watering for qualitative improvement of satsuma mandarin using intelligent control techniques

To improve the quality of Satsuma mandarin, an optimal watering operation was investigated through identification and optimization using intelligent control techniques. Dynamic changes in the sugar content and the citric acid of Satsuma mandarin, as affected by rainfall and sunshine duration, was first identified using neural networks, and then an optimal operation for watering (rainfall management) that maximizes the sugar content and that minimizes the citric acid of Satsuma mandarin was sought through simulation of the identified neural-network model using genetic algorithms. An optimal solution was an operation that increases the amount of watering during the first stage and then markedly reduces it during the latter half stage of the maturing stage. This operation provided the soft texture and sweet taste of Satsuma mandarin.