TY - JOUR
T1 - Asymptotic behaviour of root-loci of linear multivariable systems
AU - Kouvaritakis, B.
AU - Shaked, U.
PY - 1976/3
Y1 - 1976/3
N2 - The theory of the asymptotic behaviour of the root-loci of linear, time-invariant, multivariable, feedback systems is developed. It is shown that each of the system zeros attracts, and is a terminating point of, one of the root-loci, as the feedback gain tends to infinity. The root-loci, that are not attracted by the zeros, tend to infinity in a special pattern that is dictated by the eigen-properties of the elementary matrices of the system. To complete the geometric description of the asymptotic behaviour of the root-loci, the concept of infinite zeros and their order is introduced. Each infinite zero of order r attracts one root-locus and, together with r−1 other infinite zeros of the same order, the corresponding asymptotes form a Butterworth configuration of order r around a special point defined as a ‘ pivot ’. A detailed algorithm for the calculation of the finite and infinite zeros is given and is illustrated by examples. A synthesis technique is then proposed by which a constant feedback controller may be found such that all the root-loci asymptotes lie in the complex left-half plane and thus ensure system stability for high gains.
AB - The theory of the asymptotic behaviour of the root-loci of linear, time-invariant, multivariable, feedback systems is developed. It is shown that each of the system zeros attracts, and is a terminating point of, one of the root-loci, as the feedback gain tends to infinity. The root-loci, that are not attracted by the zeros, tend to infinity in a special pattern that is dictated by the eigen-properties of the elementary matrices of the system. To complete the geometric description of the asymptotic behaviour of the root-loci, the concept of infinite zeros and their order is introduced. Each infinite zero of order r attracts one root-locus and, together with r−1 other infinite zeros of the same order, the corresponding asymptotes form a Butterworth configuration of order r around a special point defined as a ‘ pivot ’. A detailed algorithm for the calculation of the finite and infinite zeros is given and is illustrated by examples. A synthesis technique is then proposed by which a constant feedback controller may be found such that all the root-loci asymptotes lie in the complex left-half plane and thus ensure system stability for high gains.
UR - http://www.scopus.com/inward/record.url?scp=0016926848&partnerID=8YFLogxK
U2 - 10.1080/00207177608922162
DO - 10.1080/00207177608922162
M3 - ???researchoutput.researchoutputtypes.contributiontojournal.article???
AN - SCOPUS:0016926848
SN - 0020-7179
VL - 23
SP - 297
EP - 340
JO - International Journal of Control
JF - International Journal of Control
IS - 3
ER -