Recursive identification under scarce measurements-convergence analysis (Q5960324)

A problem of pseudo-linear recursive least squares identification of a noisy linear continuous-time SISO system from scarce input-output sampled data is considered. The input is updated periodically at a fixed rate, and the output is measured synchronously with a regular or irregular availability pattern. The convergence analysis of a recursive identification algorithm for the system parameters using scarce output data is presented for the case of regular output measurements. The unmeasured outputs are estimated by an output predictor based on the available parameter estimates and the input-output system model. The existence of wrong local equilibrium points in the limit case, when the input updating period tends to zero, and a local stability condition for the presented identification routine are established. The theoretical results are illustrated by means of simulation examples.