Paper
Asymptotic Statistical Analysis of Virtual Reference Feedback Tuning Control
- Authors:
- De-zhi Tang; Xiao-yong Guo; Hong Jiang; Jian-hong Wang; Yong-hong Zhu
- Abstract
- Virtual reference feedback tuning control is a data-driven control strategy, where there is no model identification process. To compute the asymptotic covariance matrix corresponding to the unknown parameter, firstly the error using Taylor series expression is expanded. The two diagonal sub-matrices in the asymptotic covariance matrix are denoted as the asymptotic covariance matrix expression of the two unknown parameter estimation vectors in the closed-loop system. Based on this asymptotic covariance matrix, an optimal filter is constructed by solving one optimization problem on the trace operation. Finally, the efficiency and possibility of the proposed strategy are confirmed by the simulation example results.
- Keywords
- Virtual Reference Feedback Tuning Control; Asymptotic Analysis; Stochastic Optimization
- StartPage
- 1
- EndPage
- 10
- Doi
- 10.18005/JCSE0501001