Simplified methods for identifying the Volterra kernels of nonlinear systems

A class of random inputs for nonlinear systems that renders the homogeneous Volterra functionals orthogonal over time is introduced. This orthogonality is exploited to derive a general input-output cyclic cross-correlation formula for identification of the individual Volterra kernels of nonlinear systems. Four specific examples of inputs in the specified class are used to obtain explicit closed forms for the general identification formula. The relationship to other methods for identification, such as that of Lee and Schetzen, based on other orthogonal functionals, such as Wiener´s, is explained