Control Systems with Limited Feedback Information

ISBN-13: 9783639130720 / Angielski / Miękka / 2009 / 164 str.

Qiang Ling

Opis

This book studies control systems with limitedfeedback information. The focus is on two types oflimitations on the feedback information, dropout andquantization. By dropout, we mean that the desiredfeedback measurement is missed. By quantization, wemean the feedback measurement is only described by afinite number of bits, which introduces ``measurementerror. We consider two dropout models, independentand identically distributed (i.i.d.) processes andMarkov chains. Under i.i.d. dropouts, we provide aclosed-form expression for the outputs powerspectral density, which can be used to greatlysimplify system synthesis, such as design of theoptimal dropout compensation. Under dropouts governedby a Markov chain, we provide a necessary andsufficient stability condition, and propose a methodto compute performance measured by the outputspower. For a quantized system, we derive the minimumconstant bit rate to guarantee stability and proposea dynamic bit assignment policy(DBAP) to achieve suchminimum bit rate. It also investigates the systemperformance under DBAP. The preliminary researchconfirms the superiority of DBAP against otherquantization policies.

O autorze

Qiang Ling received the Ph.D. degree in electrical engineering
from Univ. of Notre Dame in 2005. From 2005 to 2008, he worked as
a research staff member in Seagate. He joined Univ. of Science
and Technology of China as an associate professor in 2008. His
research focuses on networked control system, signal processing,
and wireless sensor network.