Parameter Identification of a Class of Nonlinear Systems via Piecewise Multiple Legendre Polynomials

Using a series of orthogonal functions, Piecewise multiple Legendre polynomials (PMLP), we obtain a new approach to parameter identification of a class of nonlinear systems. An algorithm is presented in the paper. The approach has the following advantages: 1) The accurate result is obtained at lower expansion order, the amount of computation is reduced; 2) The algorithm does not require initial estimation of the estimated parameters; 3) The algorithm has the function of recurrence.