基于人工神经网络的田间信息插值方法研究  

Interpolation method of field information based on the artificial neural network

作　　者：何勇[1] 张淑娟[1] 方慧[1] 

机构地区：[1]浙江大学生物系统工程与食品科学学院,杭州310029

出　　处：《农业工程学报》2004年第3期120-123,共4页Transactions of the Chinese Society of Agricultural Engineering

基　　金：国家自然科学基金项目(30270773);浙江省自然科学基金(301270);教育部高等学校优秀青年教师教学科研奖励计划资助(02411);浙江省自然科学基金人才基金资助项目(RC02067)

摘　　要：提出了一种基于人工神经网络的田间信息插值新方法,并利用ArcView3.2软件绘制碱解氮的BP神经网络插值空间分布图和球状插值分布图,并对BP神经网络插值方法和克立格球状插值方法的结果进行了误差分析。结果表明,BP神经网络的插值方法优于克立格球状插值法,该方法有利于田间信息空间分布特性准确、直观的表达,有利于农田精确施肥、灌溉、播种等精细农业生产管理。In this paper, based on the principle for function approaching of BP neural network, the method of BP artificial neural network for field information interpolation was studied, and soil available N spatial distribution maps of BP artificial neural network interpolation and Kriging interpolation were drawn using ArcView3.2, the errors for interpolation results of the two methods were analyzed. The result showed that interpolation method of BP artificial neural network excels Kriging interpolation method, and the neural network interpolation method can be used to express the field information spatial distribution visually and exactly, which can be used as a reference for precision fertilization, precision irrigation and other precision management in the field.

关 键 词：人工神经网络 土壤性质 插值方法 空间分布图 误差分析 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程] S131.3[自动化与计算机技术—控制科学与工程]