홈 › Papers

Prescribed-Time Newton Extremum Seeking using Delays and Time-Periodic Gains

We study prescribed-time extremum seeking (ES) for scalar maps in the presence of time delay. The problem has been solved by Yilmaz and Krstic using chirpy probing and time-varying singular gains. To alleviate the gain singularity, we present an alternative approach, employing delays with bounded time-periodic gains, for achieving prescribed-time convergence to the extremum. Our results are not extensions or refinements but a new methodological direction, even in the absence of the delay on the map. The main result we present compensates the map's delay and uses perturbation-based and the Newton (rather than gradient) approaches. The simultaneous presence of perturbation period, and two delays -- a map delay and a seeking feedback delay -- whose values are different (feedback delay must be longer than map delay), makes for an intricate situation in the design and analysis. ES can settle arbitrarily soon after four times the map delay. In the absence of a map delay, the settling time is arbitrarily short, with feedback delay chosen as one quarter of the prescribed settling time, i.e., the search settles after four times any positive feedback delay. In addition to removing the gain singularity of the Yilmaz-Krstic singular-gain prescribed-time ES, we go beyond that method's limitation to operating only up to the terminal time. With the help of averaging theorems in infinite dimension, we conduct a prescribed-time convergence analysis on a suitable perturbation-averaged \textit{target} ES system, which contains the time-periodic gains of the map and feedback delays. Since the notion of `dead-beat'' Lyapunov stabilization by time-periodic delayed feedback originates from Hale and Verduyn-Lunel (analysis, 1993) and Karafyllis (feedback design, 2006), we refer to our approach to prescribed-time ES as the `Karafyllis, Hale, Verduyn-Lunel" (KHV) PT-ES approach.

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반