A stable identification structure for a class of stochastic systems

An on-line structure for identifying the parameters of a class of stochastic systems is presented and its convergence proof is established. The convergence proof, which is based on the supermartingale convergence theorem, is simple and concise and is valid under weak assumptions. It is shown that several identification schemes, some of them well known, can be derived as special cases of the proposed structure, and the global asymptotic stability of some schemes is proved.