基于可变模糊聚类的日参考作物腾发量预报模型及应用

Daily Reference Crops Evapotranspiration Forecast Based on Variable Fuzzy Cluster

短期参考作物腾发量对作物需水量实时预报、精准灌溉的实现以及提高灌区农业水资源高效利用具有极其重要的现实意义。针对已有研究中模型输入项较多、实用性不强或实时信息考虑不够的不足,在可变模糊聚类迭代模型的基础上,建立了基于天气类型的日参考作物腾发量预报模型,通过确定晴、多云、阴、雨四种天气类型下参考作物腾发量的聚类中心,并利用聚类中心对短期相应时段天气状况下的参考作物腾发量进行预报。在郑州地区进行模拟预测,预测值的相关统计参数分别为RMESE=0.438,RE=0.123,R2=0.922,IA=0.972,预报精度令人满意,实例证明了以短期参考作物腾发量为输入的预报模型,真实地反映了田间作物短期水分变化,提出的基于天气类型的日参考作物腾发量预报模型可为实时灌溉预报提供基础数据支持。

Short-term reference crops evapotranspiration forecast has extremely important practical significance for the real-time crop water requirement prediction, implementation of precision irrigation and improving efficient utilization of agricultural water resources in irrigation area. Aimed at the deficiencies of existing research, including that the input item of model is too much, the practicability is not strong enough or considering real-time information is not enough, on the basis of the variable fuzzy cluster iteration model, the daily reference crops evapotranspiration forecast model based on the weather conditions is established in this paper to determine the clustering centers under four types of weather, just like sunny, cloudy, overcast and rain, and conduct the real-time correction to the average value for many years of the reference crops evapotranspiration by the clustering center, then the reference crops evapotrans- piration under different weather conditions in the corresponding time can be forecasted. The simulated prediction is conducted in Zhengzhou city and the relevant statistical parameters of the predictive are that RMSE= 0. 438, RE= 0. 123, R = 0. 922, RI= 0. 972. The accuracy of prediction is satisfactory, so the model in this paper can provide basic data support for real-time irrigation fore- casting.