论文部分内容阅读
为了解决格型结构对噪声敏感的问题，从基于高阶累积量的均方误差（ＣＭＳＥ）准则出发，提出了一种基于高阶累积量的格型（ＣＬ）结构，并讨论了该结构具有的一些重要性质。在此基础上进一步推证了系统参数辨识的基于高阶累积量的Ｂｕｒｇ算法（ＣＢｕｒｇ），并给出了三阶ＣＢｕｒｇ算法的一种快速递归实现方案。仿真结果证明：就辨识结果的无偏性而言，这种ＣＢｕｒｇ算法的抗高斯噪声性能明显优于常规Ｂｕｒｇ算法而两种算法的运算量大体相当。
In order to solve the problem that the lattice structure is sensitive to noise, a high-order cumulant-based lattice (CL) structure is proposed based on the CMSE criterion. The structure of Some of the important properties. Based on this, the Burg algorithm (CBurg) based on high-order cumulants for system parameter identification is further validated and a fast recursive scheme for the third-order CBurg algorithm is given. The simulation results show that the anti-Gaussian noise performance of this CBurg algorithm is obviously superior to the conventional Burg algorithm in terms of the unbiased identification results, and the computations of the two algorithms are roughly equivalent.