Calculating Irrigation Schedule for Rice in Temperate Regions Using a Combination of Grey Neural Network and Fuzzy Control

Zhaoyu FENG,Tianshi CUI,Zhichao ZHANG,Rui WANG,Chunli LIU,Lifeng WANG

DOI: https://doi.org/10.13522/j.cnki.ggps.2017.0386

2018-01-01

Abstract:[Objective]The purpose of this paper is to propose new models to help increase the water use efficiency in irrigation for rice in temperate regions.[Method]We used grey neural network and fuzzy control to design irrigation schedule for rice in temperate regions, considering both specific growth environment and demand of the rice for water during different growth stages. Meteorological data in rice field were measured and then transmitted from a micro meteorological station; the factors affecting the demand for irrigation strategy were estimated using the grey neural network model. The difference between the estimated demand for water and the optimal demand for water by the crop, as well as the change rate of their difference with time were taken as inputs in the fuzzy control; the duration of the irrigation was the model output. Field experiments were conducted to verify the model.[Result]The irrigation schedule calculated by the model had advantage of quick response, small overshoot and good controllability. The field experiment conducted in the rice field of Northeast Agricultural University showed 11.59％ saving in water, and the experiment in Jian Sanjiang farm of Heilongjiang Province showed 13.54％ saving in water, compared to the conventional irrigation schedule.[Conclusion]The proposed models for calculating irrigation schedule can increase water use efficiently and could play an important role in precision agriculture.