非线性状态空间模型贝叶斯辨识的输入设计
应用统计
2013-07-25 v1 统计计算
统计方法学
摘要
我们提出了一种算法，用于设计最优输入以进行随机非线性状态空间模型的在线贝叶斯辨识。所提出的方法依赖于针对模型参数导出的后验 Cram\'er Rao 下界相对于输入序列的最小化。为了使优化问题在计算上可行，输入被参数化为输入空间中的多维马尔可夫链。通过一个仿真示例说明了所提出的方法。
引用
@article{arxiv.1307.6258,
title = {Input design for Bayesian identification of non-linear state-space models},
author = {Aditya Tulsyan and Swanand R. Khare and Biao Huang and R. Bhushan Gopaluni and J. Fraser Forbes},
journal= {arXiv preprint arXiv:1307.6258},
year = {2013}
}
备注
This article has been published in: Tulsyan, A, S.R. Khare, B. Huang, R.B. Gopaluni and J.F. Forbes (2013). Bayesian identification of non-linear state-space models: Part I- Input design. In: Proceedings of the 10th IFAC International Symposium on Dynamics and Control of Process Systems. Mumbai, India