摘要
在综合考虑生态系统中水华发生的机理特点基础上,采用改进的BP神经网络实现了对叶绿素最高点的非线性预测;利用灰色WPGM(1,1)模型的累加生成运算(AGO)对叶绿素最高值对应的时刻进行推算,从而预测水华的爆发时间点。经检验,神经网络预测结合灰色WPGM(1,1)预测模型相对误差在10%左右,能够对水华的发生进行判断和预报,有利于综合整治方案的优化和统筹。
Considering comprehensively the characteristics of water bloom occurrence mechanism in ecological system,the nonlinear prediction of the chlorophyll highest point is implemented based on improved BP neural network;and the time of maximum chlorophyll value is calculated based on the Accumulated Generating Operation（AGO） of WPGM（1,1） model.By inspection,the relative error of chlorophyll highest point model is controlled about 10%.The model is able to judge and predict the water bloom outbreak,and is beneficial to optimize and orchestrate the comprehensive regulation scheme
出处
《计算机工程与应用》
CSCD
北大核心
2011年第13期231-233,共3页
Computer Engineering and Applications
基金
北京市教委科技项目(No.KM200810011003)
关键词
水华预测
叶绿素尖点
BP神经网络
灰色拓扑预测
water bloom prediction
the highest point of chlorophyll
Back Propagation（BP） neural network
grey topologicalprediction
