论文部分内容阅读
该文提出了一种基于劋Ｔｒｏｕｓ算法的小波递归预测方法?迷げ馑惴ú捎梦蕹槿±肷⑿〔ū浠坏膭ぃ裕颍铮酰笏惴?,可以逐点把时间序列分解为与原序列长度相同的小波系数 ,适合在递归预测中应用 ,弥补了Mallat算法不能实时调整模型参数的不足。预测卫星电源母线电压数据表明 ,该方法满足预测卫星遥测数据的要求。
In this paper, a wavelet recursive prediction method based on 劋 Trous algorithm is proposed. The algorithm can decompose the time series point by point The wavelet coefficients with the same length as the original sequence are suitable for recursive prediction and make up for the shortcomings that the Mallat algorithm can not adjust the model parameters in real time. Predicting satellite power bus voltage data shows that the method meets the requirements of predicting satellite telemetry data.