Research On Irrigation Strategy Based On Grey Neural Network And Fuzzy Control For Rice In Cold Region

The shortage of water resources seriously restricts the social and economic development of our country,especially for the modernization of agriculture.According to the reports,it is about 60% of the national water supply is used in agricultural irrigation,and the utilization rate is very low,which wastes too much water resources in agricultural production.Agriculture is the foundation of people's livelihood and the foundation of our social construction.Only by solving the problem of large consumption and serious waste of irrigation could better promote the development of precision agriculture and accelerate the pace of social modernization.Heilongjiang province is the major grain producing province,but the phenomenon of food production reduction wad occurred in recent years due to the insufficient water for irrigation.For the study of water-saving irrigation technology attracted lots of agricultural experts in recent years,but there a few achievements in the field of rice irrigation in the cold region in northeast China.Irrigation water demand is mainly influenced by environmental climate factors and soil moisture.They would speed up the consumption of water to a certain extent.Only by mastering the data of these factors can better determine the actual water requirement of rice irrigation and make the best irrigation control.The paddy field environmental meteorological data were obtained from micro-meteorological stations installed in paddy fields to collect data and transmit the data,which not only improve the irrigation control efficiency,the real time and reliability of data collection,and also reduces the trouble caused by the wiring to rice harvesting and saves the cost.In order to formulate a better irrigation strategy and solve the problem of time delay,the grey neural network was structures to predict the future data based on historical and current data.The development rule of meteorological factors was studied buy grey theory and forecast the future information development.And neural networks can process some qualitative and fuzzy data by constantly updating the connection weights between neurons,without the need for strict independent variables and dependent variable coefficients.The output response of grey model was mapped to the topological structure of neural network can improve the accuracy and stability of prediction.Irrigation control is a time delay and nonlinear process,and fuzzy control is an artificial intelligence method that based on the people's experience and imitating people's way of thinking to control complex system with multiple parameters,which could effectively solving the nonlinear problem of irrigation control.MATLAB simulation was carried out to briefly introduce the design process of fuzzy controller.The irrigation strategy proposed in this paper transforms the traditional irrigation mode and turns the irrigated land into irrigated crop,which fully ensures the normal water demand of rice in the cold area,and also beneficial to promote the normal growth of the crops.This irrigation strategy could make more accurate judgment and management of actual crop water demand.The field test results on Jiu san farm in Heilongjiang province show that this study has a good water saving effect.