This paper focuses on the problem of output-only recursive identification of time-varying structures. A kernelized time-dependent autoregressive moving average (TARMA) model is proposed by expanding the time-varying model parameters onto the basis set of kernel functions in a reproducing kernel Hilbert space. A Gaussian process regression TARMA identification scheme is subsequently proposed, allowing the Gaussian process regression to operate for vector TARMA models in a recursive manner. The proposed method is employed to identify a laboratory time-varying structure consisting of a simply supported beam and a sliding mass, and is assessed against an existing recursive pseudo-linear regression TARMA method via Monte Carlo experiments. The comparison demonstrates the superior achievable accuracy of the proposed Gaussian process regression TARMA approach.
展开▼
