【摘要】 本文通过对系统辨识的设计理论分析,提出了在给出实验条件限制下,最大程度地提取数据中所含的信息内容的最优辨识方法。根据weiner—Hopf方程和?isher信息矩阵,推证出更一般性的权函数估计值H(i)的相关辨识公式,同的证明了最优输入信号和采样周期。本方法简单易行,计算机模拟计算结果表明,最优估计值H(i)比传统的方法辨识精度更高。更多还原

【Abstract】 A nover identification technique is suggested in this paper to extract the utmost information from the experimental data undfr certain copstraints. According to the weiner-hopf equation and fisher information matrix, a more general formula of the system weighing function estimation H (i) with correlation is de-veloped. The optimal input signal and sainpling cycle are obtained. The propo-sed method is simple and effective. The computer simulation results show theoptimal estimation H(i) is more accurate than that of the conventional method.更多还原