摘要

It is well known in the literature that delay-adaptive control schemes for general linear systems with input delay, induced from predictor feedback, typically incorporate distributed terms in both the control law and the update law for the delay estimator, leading to the infinite-dimension characteristic of the schemes. Furthermore, when the full actuator state is not employed, only local regulation has been achieved in the sense that the initial state has to be sufficiently small and, at the same time, the initial guess of the delay has to be sufficiently close to its actual value. In this paper, by restricting our attention to systems without exponentially unstable poles, we avoid implementational issues arising from typical infinite-dimensional control schemes by developing a finite-dimensional scheme inspired by the truncated predictor feedback (TPF) design technique. Our scheme achieves global regulation as long as a sufficiently small neighborhood of the actual delay is identified. An adaptation-free version of the robust TPF, which is inferred from the delay-adaptive control scheme, is also included to reveal a second level of robustness to the uncertainty in the value of the delay that appears in the state transition matrix of the TPF.

  • 出版日期2018-5-25