输出误差自回归系统的分解梯度迭代算法研究

Hierarchical Gradient-Based Iterative Algorithm for Output Error Autoregressive Systems

针对输出误差自回归系统(output error autoregressive system,OEAR)辨识参数误差大,收敛速度慢的问题,本文将递阶辨识原理与梯度迭代算法(gradientbased iterative algorithm,GI)运用到输出误差自回归系统的辨识过程中,针对该系统的算法进行推导,提出了基于分解的输出误差自回归系统的梯度迭代算法。将输出误差自回归系统分解成2个子系统,通过梯度迭代算法分别对2个子系统进行辨识,最后用Matlab仿真实例进行仿真。仿真结果表明,在输入信号的作用下,系统能够更快速的收敛到比原有算法误差更小的范围内,验证了该算法的有效性。

For the problem that the output error autoregressive system has large error and slow convergence when identifying parameters,the hierarchical identification principle and the gradient based iterative algorithm are applied to the identification process of the system.Aiming at the derivation of the algorithm of this system,a gradient iterative algorithm based on decomposition-based output error autoregressive system is proposed.The basic idea is to decompose the output error autoregressive system into two subsystems,and identify the two subsystems by gradient iterative algorithm.Finally,the simulation is carried out with respect to simulation examples by using Matlab.The simulation results show that under the action of the input signal,the system can converge to a range smaller than that by using the original algorithm,and the effectiveness of the algorithm is verified.