一种基于灰色模型的数据预测优化算法

摘要

近些年来,灰色模型GM(1,1)被大量应用于小样本或穷信息的预测,操作与实现步骤简单,预测精度较高.为了进一步提高GM(1,1)的预测精度,运用遗传算法动态调整GM(1,1)中的均质生成数列分辨率系数,改变通常把灰色模型的分辨率系数设置为1/2的计算模式,使得改进后的GAGM(1,1)算法针对小样本的预测具有更高的精度和鲁棒性.通过算法的数值实验,结果表明优化算法的预测精度高于传统的GM(1,1)算法及多个基于GM(1,1)的改进算法.%In recent years,grey model GM(1,1)has been widely applied to the prediction of small samples or poor information due to its simple operation and implementation step which has a high prediction accuracy. In order to further improve the prediction accuracy of GM(1,1),the genetic algorithm is applied to dynamically adjust the resolution coefficients of the homogeneous generated sequence in GM(1,1). Specifically,the calculation mode of usually setting the resolution coefficient of the grey model as 0.5 is changed to make the improved GM(1,1)algorithm has higher accuracy and robustness for small sample prediction. The result of the numerical experiments for the algorithm shows that the prediction accuracy of the optimized algorithm is higher than that of GM(1,1)algorithm and other improved algorithms.