LIKELIHOOD ESTIMATION FOR DISTRIBUTED PARAMETER MODELS OF LARGE BEAM-LIKE STRUCTURES

In this paper, a maximum likelihood estimation for distributed parameter models of large flexible structures has been formulated. Distributed parameter models involve far fewer unknown parameters than independent modal characteristics or finite element models. The closed form solutions for the partial differential equations with corresponding boundary conditions have been derived. The closed form expressions of the sensitivity functions lead to highly efficient algorithms for analyzing ground or on-orbit test results. For illustration of this approach, experimental data of the NASA Mini-Mast truss have been used. The estimations of modal properties involve its lateral bending modes and torsional modes. The results show that distributed parameter models are promising in the parameter estimation of large flexible structures. © 1992.