Paper
Optimal Input Signal Design for Direct Data Driven Control
- Authors:
- De-zhi Tang; Jian-hong Wang
- Abstract
- In these years, direct data driven control is widely studied to replace the classical model based control. The main advantage of direct data driven control is that the identification process of the plant does not need. Then the controller can be designed based on input-output measured data directly. In this work, the problem on how to design optimal input signal for direct data driven control is studied and one optimal correlation function of the input signal is also proposed here. The excitation signal is designed such that the expected value of the considered control is reduced. Here some results and conditions used to improve the convergence of classical prediction error method into direct data driven control are extended. Furthermore, a candidate domain of attraction for the objective function is introduced and one convergence condition which ensures that a given set is a candidate domain of attraction is derived. Generally, all results and conditions corresponding to the optimal input signal for direct data driven control are different from the classical methods. Finally, the efficiency and possibility of the proposed strategy are confirmed by the simulation example results.
- Keywords
- Direct Data Driven Control; Optimal Input Signal; Asymptotic Variance; Convergence Condition
- StartPage
- 37
- EndPage
- 47
- Doi
- 10.18005/JCSE0501004