An algorithm for tracking rapid parameter changes in unknown MINO systems with nonlinear model uncertainties

The authors show that by adding dead zones into both recursive identification and the variables which form the detection signal, a method can be developed which will produce good results for the detection and tracking of parameter changes in unknown systems with nonlinear model uncertainties. The convergence properties in the deterministic case have been shown and simulations with a second order and two dimensional time-varying system have illustrated the power of the proposed method. The new method is shown to perform better than the recently proposed method of Chen and Norton (1987)