Input conditions for continuous-time adaptive system problems

The exponential convergence of certain algorithms used in continuous-time adaptive identification and control is investigated by deriving persistence of excitation conditions which depend on the system inputs exclusively. Stable plants and unstable plants with external stabilizing control are considered in the identification study. For the control algorithm considered, it is shown that in the absence of foreknowledge of a gain parameter, exponential stability is impossible to claim.