摘要
该文提出一种灰色-神经网络综合预测模型。该模型由背景值构造、加权GM(1,1)模型和神经网络补偿器三部分组成。其建模机理为:首先对于原始数列进行背景值构造,然后构建加权GM(1,1)模型,同时利用神经网络补偿器获得误差补偿信号,则最终的预测值为加权GM模型的输出值加上补偿值。仿真结果验证了所提方法的有效性。
A grey neural network integrated forecasting model is proposed.The model is composed of three parts,that is:structuring method of background value,weighted grey model and neural network based error compensator.The process of the modeling is:firstly,the background value is constructed using the original sampling data,then the weighted GM(1,1)model is obtained by a new weighted modeling algorithm,at last the compensated error signal is obtained by neural network based error compensator,so the prediction value equals to the output value of the weighted GM(1,1)model plus the compensated error signal.The simulation results show that the integrated model can improve the prediction precision obviously compared to the other algorithm presented in this paper.
出处
《计算机工程与应用》
CSCD
北大核心
2004年第9期13-15,共3页
Computer Engineering and Applications
基金
国家863高技术研究发展计划
重大专项项目(编号:2003AA501100)
中国博士后科学基金项目(编号:2003034145)
关键词
预测模型
灰色系统
神经网络
forecasting model,grey system,neural network
