基于BP神经网络与模糊控制的小麦灌溉系统

Application of Wheat Water-Saving Irrigation System Based on BP Neural Network and Fuzzy Control

为了提高农业用水的利用效率,实现新疆灌区精准灌溉,在新疆阿勒泰地区福海县阔克阿尕什乡浑沃尔海,根据春小麦的生长环境和各生育时期需水量,设计了基于BP神经网络与模糊控制的小麦灌溉系统。该系统通过田间微型气象站监测、麦田传输数据,利用BP神经网络预测出小麦需水量;以需水量和土壤实际湿度的差值和差值变化率作为模糊系统的输入量,灌溉时间作为输出量。将2017年人工灌溉数据与灌溉控制系统相结合利用PYTHON做对比试验,检验灌溉系统优化的效果。结果表明,BP神经网络对春小麦需水量的预测效果较好,验证集决定系数R^(2)为0.854,相对分析误差RPD为2.014,预测结果满足春小麦实际需水标准。模糊控制系统相比于传统控制系统不会出现较大的超调量,有更好的稳定性。BP神经网络与模糊控制灌溉系统比人工灌溉节水约23.9%,说明该灌溉系统能够提高水资源的利用率,对实现精细化农业有着重要的参考意义。

In order to improve the agricultural water use efficiency and achieve precision irrigation in the Fuhai County,Quoke Agashi Township,Altay region of Xinjiang,a water-saving optimization design was conducted for the irrigation of spring wheat.Based on the water requirements of wheat during different growth stages and its environmental conditions,an irrigation system for wheat was designed using BP neural networks and fuzzy control.The system utilizes data from on-field micro-weather stations and wheat field sensors to predict the water requirements of wheat through BP neural networks.The difference and change rate of the difference between water requirements and actual soil moisture were used as input parameters for the fuzzy system,with irrigation time as the output parameter.In comparison experiments,manual irrigation data in 2017 were combined with the irrigation control system using PYTHON to achieve precise irrigation.Experimental results showed that the BP neural network had good predictive performance,with R^(2)of 0.854 and RPD of 2.014 for the validation set,meeting the actual water requirements of spring wheat.Compared to traditional control systems,fuzzy control exhibited better stability with minimal overshooting.The BP neural network and fuzzy control irrigation system saved 23.9%of the total water compared to manual irrigation,demonstrating its ability to improve water use efficiency and its significant reference value for precision agriculture.