基于非线性多项式的自适应重构算法

研究实现了一种基于非线性多项式的重构模型,并将该模型与递推最小二乘法(RLS)相结合,实现对跳频序列的动力学系统进行自适应重构。从仿真实验可以看出该重构算法对一般的跳频序列具有较高的重构精度和准确重构率。针对RLS算法复杂度较高的特点,提出了利用多项式参数的数学特性及采用稀疏自适应滤波算法来降低算法复杂度。通过选择适当的门限值,可以在不损失重构精度的前提下尽可能地降低算法复杂度。

An Adaptive Reconstruction Algorithm Based on Nonlinear Polynomial

An adaptive reconstruction model based on nonlinear polynomial is researched and realized in this paper.Combining it with Recursive Least Squares（RLS） algorithm,we can get the Dynamics of frequency hopping series adaptively.Simulation experiments have demonstrated that the reconstruction algorithm can provide high reconstruction and satisfactory percentage of reconstruction for some typical frequency hopping series.For the characteristic of high computation complexity of RLS algorithm,a new adaptive smoothing algorithm is proposed.Which based on the mathematical features of Bernstein polynomial parameters and a sparse adaptive filtering algorithm.Choosing the appropriate threshold value,we can simplify the complicacy as much as possible with high reconstructing accuracy.