摘要
针对自然时效处理ZL305A铝合金抗拉强度变化不均匀的特点,运用小波变换和灰色DGM(2,1)模型相结合,对其抗拉强度变化趋势进行预测。利用小波变换去除原始时间序列中的噪声,然后以生成的新时间序列建立DGM(2,1)模型,实现对原始时间的预测。结果表明,小波DGM(2,1)预测的平均误差仅为0.466 5%,最后一期的误差为-0.260 76%,分别比GM(1,1)模型减小了1.228%和65.066 2%。
In view of the uneven variation of tensile strength of ZL305 Aaluminum alloy in natural aging treatment,the change trend of tensile strength of ZL305 Aalloy was predicted by combining wavelet transform with grey DGM（2,1）model.The noise in the original time series was removed by the wavelet transform,and then the DGM（2,1）model was established with the new time series generated to predict the original time.The results reveal that the average prediction error is only 0.466 5%,which is decreased by 1.228% and the finished time is-0.260 76%,which is decreased by 65.066 2%.
作者
田会方
舒服华
Tian Huifang;Shu Fuhua(School of Mechanical and Electrical Engineering,Wuhan University of Technolog)
出处
《特种铸造及有色合金》
CAS
CSCD
北大核心
2018年第8期821-824,共4页
Special Casting & Nonferrous Alloys
