摘要
在遗传算法 ( Genetic Algorithm)与误差反传 ( Back Propagation)网络结构模型相结合的基础上 ,设计了用遗传算法训练神经网络权重的新方法 ,并对吉林省梨树和德惠县的玉米进行了估产研究 ,同时与 BP算法和灰色系统理论模型进行了比较。经检验 ,计算值与实际值接近 ,并优于灰色理论模型 ,具有良好的预测效果 。
Based upon the combination of Genetic Algorithm and BP Neural Network,the estimation of maize yield in Lishu and Dehui County, Jilin Province was estimated in this paper The new approach was compared with BP Neural Network as well as Gray Model Theory The results indicate that the prediction values approximate to the real maize yield and are more accurate than BP network and gray model theory
出处
《生态学报》
CAS
CSCD
北大核心
2001年第5期716-720,共5页
Acta Ecologica Sinica
基金
国家自然科学基金重点资助项目!( 49731 0 2 0 )
关键词
作物估产
遗传算法
神经网络
玉米
estimation of crop yield
genetic algorithm
neural network