Modal Identification of Spatial Trusses Based on Bayesian Approach

Modal identification mainly uses the input and output data of the structure to determine the natural frequency, modal damping ratio, and mode shape of the structure. The identified dynamic properties can provide the basis for structural vibration control, model updating and structural health monitoring. In the modal identification methods, Bayesian method considers the influence of system uncertainty on modal parameter identification and gives the quantified index of system uncertainty, which is an obvious advantage of Bayesian method compared with other identification methods. Space truss is usually carried on spacecraft as a support structure, and obtaining its accurate modal parameters is important for the attitude controller design of spacecraft. In this paper, a Bayesian time-domain method is proposed and used to identify the modal parameters of a numerical model of a space truss structure. First, the structure is modeled using the state equation and the mathematical relationship between the physical response of the space truss and the modal parameters is derived based on the Markov parameter matrix; secondly, the likelihood function and the posterior probability density function of the modal parameters are established based on Bayesian theory; Finally, the most probable values of the modal parameters of the space truss structure were solved using the optimization algorithm, and the posterior covariance matrix of the parameters was solved using the finite difference method to quantify the uncertainties of the parameters. The identification results and the finite element analysis results were compared to verify the accuracy of the method in this paper.