TY - JOUR
T1 - LOCAL STABILITY ANALYSIS FOR A CLASS OF ADAPTIVE SYSTEMS.
AU - Kosut, Robert L.
AU - Anderson, Brian D.O.
PY - 1985
Y1 - 1985
N2 - An analysis of adaptive systems is presented where a local L infinity -stability is ensured under a persistent excitation condition. The stability analysis involves establishing the exponential stability of a differential equation which arises in the study of most adaptive systems. Although the connection between exponential stability and persistent excitation is known, it is important to obtain specific formulas for the rates and gains involved. However, L infinity -stability can be obtained by using a nonlinear adaptation gain, i. e. , theta = Yh(z,e). For example, h(z,e) can arise from using a dead zone, leakage, or normalization. Such schemes can be incorporated in the general framework presented but require further analysis in order to obtain explicit signal bounds.
AB - An analysis of adaptive systems is presented where a local L infinity -stability is ensured under a persistent excitation condition. The stability analysis involves establishing the exponential stability of a differential equation which arises in the study of most adaptive systems. Although the connection between exponential stability and persistent excitation is known, it is important to obtain specific formulas for the rates and gains involved. However, L infinity -stability can be obtained by using a nonlinear adaptation gain, i. e. , theta = Yh(z,e). For example, h(z,e) can arise from using a dead zone, leakage, or normalization. Such schemes can be incorporated in the general framework presented but require further analysis in order to obtain explicit signal bounds.
UR - http://www.scopus.com/inward/record.url?scp=0022241171&partnerID=8YFLogxK
M3 - Conference article
AN - SCOPUS:0022241171
SN - 0743-1619
SP - 393
EP - 398
JO - Proceedings of the American Control Conference
JF - Proceedings of the American Control Conference
ER -