The Research Of Structural Dynamic Model Updating With Uncertainty

The model updating technology plays an important role in improving the accuracy of simulation model prediction.The traditional model updating techniques are based on deterministic framework.However,the uncertainty is inevitable,the results obtained by uncertain model updating method will be more significant.In this paper,the following uncertain model updating researches are mainly carried out on the consideration of parameter uncertainty and model form uncertainty.1.The uncertain model updating technology based on perturbation method and the interval analysis method are employed to unsymmetrical H-shaped structure for the research of correcting the complex model.The satisfactory result was achieved for both the method.And the interval analysis method is more applicable to the case of small samples.2.The uncertain model updating method based on the model deviation was proposed to correct the model of a portal frame,considering the uncertainties of the modal test data and the model form uncertainty caused by the simplification.After the correction,the model has higher prediction accuracy.3.The fractional derivative term is introduced into the system vibration equation to quantify the model form uncertainty.The damped oscillator of two-degree-of-freedom is used as the object for correction research.The precise correction results are obtained in this method,which is of great research value and broad application prospects.4.The uncertain model updating method based on neural network parameter identification is proposed for thermal model updating research.The stiffened C/SiC composite panel is selected as the research object.The research shows that the methods proposed can avoid the sensitivity analysis and can be applied to the updating of multi-field thermal structure dynamics for C/SiC composites panel.