The paper presents a recursive algorithm for identification of parameters of a class of time-varying systems via Hermite polynomials. Hermite polynomials belong to the class of complete set of orthogonal polynomials. They possess very useful properties for developing recursive algorithms. These operational and recurrent properties of Hermite polynomials have been used to obtain an algorithm for estimating the parameters from input-output data. Monte-Carlo results show that consistent estimates can be obtained when the measurement noise is white. Higher variance, however, for noise in the simulation has resulted in higher bias.
展开▼
