This paper deals with parameter identification of LPV descriptor systems. We present a natural prediction error identification method for LPV descriptor models.We use gradient and Hessian based nonlinear optimization algorithms to minimize the cost function. The gradients and Hessians are computed using a simulation of a LPV descriptor system. The steepest-descent algorithm and the Gauss-Newton algorithm are employed together to improve the convergence and the computation speed in solving the identification problem. Numerical simulations of a magnetic levitation system, demonstrate the effectiveness of the proposed algorithm.
