摘要
导出了 GM( 1 ,1 )模型微分方程的自记忆计算格式 ,建立了一种新的对历史多个时刻观测数据具有记忆功能的灰色自记忆模型 ,模型克服了 GM( 1 ,1 )模型“对初值比较敏感”,预见期短和序列数据非负的局限 ,并有效提高了模型的计算精度 .大量数值计算结果充分表明了该模型具有精度高 ,稳定性好的优点 .
The self-memory numerical method for solving GM(1,1) model is derived and a novel grey model is made。This novel model can recollect last server data and get over GM(1,1)'s shortages。 some numerical tests for this method are made。The numerical examples show that this method has high accurate results and good stability.
出处
《系统工程理论与实践》
EI
CSCD
北大核心
2003年第8期114-117,129,共5页
Systems Engineering-Theory & Practice
