基于过程神经网络的农作物生长预测算法研究  

Research of Crop Growing Prediction Algorithms Based on Process Neural Network

作　　者：仝志民[1] 荣丽红[1] 孟耀华[1,2] 张品秀[1] 刘英楠[1] 

机构地区：[1]黑龙江八一农垦大学信息技术学院,大庆163319 [2]哈尔滨工业大学通信技术研究所

出　　处：《黑龙江八一农垦大学学报》2013年第3期85-87,101,共4页journal of heilongjiang bayi agricultural university

基　　金：黑龙江省教育厅科研项目(12511354);黑龙江八一农垦大学博士启动基金项目(B2010-10)

摘　　要：针对农作物基本发育过程应用过程神经网络模型,建立了农作物生长预测算法。该算法可以描述并预测农作物的生长如光合作用和呼吸作用,并可通过反馈式训练学习实现自动修正动态输入函数因子。实验结果表明,所提出的农作物生长预测算法可用于描述和预估农作物生长环境条件对农作物生产潜力的影响,其对农业生产具有一定的现实指导意义。According to the basic process of crops growth, the crops growth forecast algorithm was established based on process neural network mode. The growth of crops such as photosynthesis and respiration could be described by the algorithm, and automatic revision dynamic input function factor could be input through the feedback type of training and learning. The experimental results showed that crops growth forecast algorithm could be used to describe and estimate the degree of influence of environment on the production of crops, which was very important for the production of crop.

关 键 词：过程神经网络 训练学习 特征函数 生长预测 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]