基于BP网络的膜下滴灌加工番茄墒情预报研究  

Drip Moisture Prediction of Processing Tomatoes Based on BP Networks

作　　者：陈坤[1] 雷晓云[1] 李芳松 文静[1] 

机构地区：[1]新疆农业大学水利与土木工程学院,乌鲁木齐830052 [2]新疆水利水电科学研究院,乌鲁木齐830049

出　　处：《中国农村水利水电》2012年第9期1-4,共4页China Rural Water and Hydropower

基　　金：国家自然科学基金(51060002);新疆自治区高技术研究与发展计划项目(200712111);新疆水利水电工程重点学科基金资助项目

摘　　要：针对加工蕃茄自动化滴灌系统下土壤墒情预报的问题,2011年在新疆生产建设兵团农八师国家农业科技园区,开展了土壤墒情预报的田间试验研究。通过观测得到的4个深度的土壤体积含水率以及日平均气温作为输入量,提出了一个基于BP网络的、可以对加工番茄在生育期内的墒情作出短期预报的模型。该模型分别以2d、7d为时间间隔,对自动化滴灌系统下的加工蕃茄墒情进行了预报,仿真结果显示预报效果较好。该模型用于加工番茄的土壤墒情预报是可行的,并且是一种简单、易于推广的墒情预报方法。Based on processing tomatoes under drip irrigation automation system of soil moisture forecast problem, in 2011, in the Eighth National Agricultural Science and Technology Park of Xinjiang Production and Construction Corps Agricultural Division, a field experiment was made to forecast soil moisture. Through an observation , the 4-soil volumetric water content and daily average temperature as input are obtained, a short-term prediction model is put forward. The model with two day, one week for the time interval, the automation of processing tomatoes in the drip irrigation system for soil moisture were predicted. The simulation results have shown a good prediction effect. It is feasible that the model is used for processing tomato soil moisture forecast, and it is a simple soil moisture prediction method that can be popularized.

关 键 词：加工番茄 土壤墒情预报 BP神经网络模型 

分 类 号：S152.7[农业科学—土壤学]