一种基于衰减系数的混沌时间序列预测法

为提高加权一阶局域模型的预测精度,提出一种改进型混沌时间序列预测方法.该方法用衰减系数和时间延迟修正向量距离公式,调节邻近点与中心点的相关性,同时,只用邻近点中与预测值相关性最大的分量进行线性拟合.利用该方法对Henon混沌时间序列进行预测的结果表明,衰减系数取最佳值时,相对于现有算法,该方法可以更精确地预测混沌时间序...

A prediction method of chaotic time series based on attenuation coefficient

A modified prediction method of chaotic time series is introduced for improving the prediction accuracy of adding-weight one-rank local-region model. An attenuation coefficient and the delay time are applied to amend the vector distance formula, so different relevance of each adjacent point and the center point can be adjusted. And the linear fitting is only applied to the component of adjacent points which is most relevant to the predicted value. The Henon chaotic time series are forecasted using the improved method, and simulation results show that the prediction accuracy is improved with the optimal attenuation coefficient in the proposed method compared to the original one.