Orthoregressional-algebraic parameter identification method for linear differential equations

A combined orthoregressional-algebraic approach to the parameter identification of linear differential equations from solution measurements with additive noise is proposed. It is based on the algebraic Fliess–Sira-Ramirez method in conjunction with the orthogonal regression (TLS) method in the space of measurements transformed by the integral operators of convolution type. The consistency of the orthoregressional-algebraic method is established and a numerical comparison with the asymptotically optimal variational identification method is performed.