大系统的递阶辨识
Hierachical Identification of Large Scale Systems
-
摘要: 大系统的特点是维数高、待估计的参数数目多,使得辨识方法的计算量和存储量急 剧增加,以致常规辨识算法难以实现.为了减少大系统辨识的计算量,提出了计算量较小的递 阶辨识算法,并用鞅超收敛定理证明了它的收敛性.结果说明该算法可以给出大系统参数的 一致估计.Abstract: Large scale systems have so many parameters that ordinary identification algorithms have too much computational burden to realize in computer. In order to reduce the computational burden we present a hierachical identification algorithm of large scale systems and apply the martingale hyperconvergence theorem to its convergence. The results show that the proposed algorithm may give the consistent parameter estimation of large scale systems.
-
Key words:
- Large scale system /
- identification /
- parameter estimation /
- hierachical identification
计量
- 文章访问数: 2289
- HTML全文浏览量: 81
- PDF下载量: 1090
- 被引次数: 0